Datasets:

ChicagoHAI

/

CaseSumm

like 11

Follow

ChicagoHAI 19

587.US.2018_17-778

When petitioner Jamar Quarles pled guilty to being a felon in possession of a firearm in violation of 18 U. S. C. §922(g)(1), he also appeared to qualify for enhanced sentencing under the Armed Career Criminal Act because he had at least three prior “violent felony” convictions, §924(e). He claimed, however, that a 2002 Michigan conviction for third-degree home invasion did not qualify, even though §924(e) defines “violent felony” to include “burglary,” and the generic statutory term “burglary” means “unlawful or unprivileged entry into, or remaining in, a building or structure, with intent to commit a crime,” Taylor v. United States, 495 U.S. 575, 599 (emphasis added). Quarles argued that Michigan’s third-degree home invasion statute—which applies when a person “breaks and enters a dwelling or enters a dwelling without permission and, at any time while he or she is entering, present in, or exiting the dwelling, commits a misdemeanor,” Mich. Comp. Laws Ann. §750.110a(4)(a) (emphasis added)—swept too broadly. Specifically, he claimed, it encompassed situations where the defendant forms the intent to commit a crime at any time while unlawfully remaining in a dwelling, while generic remaining-in burglary occurs only when the defendant has the intent to commit a crime at the exact moment when he or she first unlawfully remains in a building or structure. The District Court rejected that argument, and the Sixth Circuit affirmed. Held: 1. Generic remaining-in burglary occurs under §924(e) when the defendant forms the intent to commit a crime at any time while unlawfully remaining in a building or structure. In ordinary usage, “remaining-in” refers to a continuous activity, and this Court has followed that ordinary meaning in analogous legal contexts, see, e.g., United States v. Cores, 356 U.S. 405, 408. Those contexts thus inform the interpretation of “remaining-in” burglary in §924(e): The common understanding of “remaining in” as a continuous event means that burglary occurs for purposes of §924(e) if the defendant forms the intent to commit a crime at any time during the continuous event of unlawfully remaining in a building or structure. The intent to commit a crime must be contemporaneous with unlawful entry or remaining, but the defendant’s intent is contemporaneous with the unlawful remaining so long as the defendant forms the intent at any time while unlawfully remaining. That conclusion is supported by the body of state law as of 1986, when Congress enacted §924(e). Quarles’ narrow interpretation makes little sense in light of Congress’ rationale for specifying burglary as a violent felony. Congress “singled out burglary” because of its “inherent potential for harm to persons,” Taylor, 495 U. S., at 588, and the possibility of a violent confrontation does not depend on the exact moment when the burglar forms the intent to commit a crime while unlawfully present in a building or structure. Quarles’ interpretation would also thwart the stated goals of the Armed Career Criminal Act by presumably eliminating many States’ burglary statutes as predicate offenses under §924(e). Pp. 3–9. 2. For the Court’s purposes here, the Michigan home-invasion statute substantially corresponds to or is narrower than generic burglary. The conclusion that generic remaining-in burglary occurs when the defendant forms the intent to commit a crime at any time while unlawfully remaining in a building or structure resolves this case. When deciding whether a state law is broader than generic burglary, the state law’s “exact definition or label” does not control. Taylor, 495 U. S., at 599. So long as the state law in question “substantially corresponds” to (or is narrower than) generic burglary, the conviction qualifies. Ibid. Pp. 9–10. 850 F.3d 836, affirmed. Kavanaugh, J., delivered the opinion for a unanimous Court. Thomas, J., filed a concurring opinion.

Section 924(e) of Title 18, also known as the Armed Career Criminal Act, mandates a minimum 15-year prison sentence for a felon who unlawfully possesses a firearm and has three prior convictions for a “serious drug offense” or “violent felony.” Section 924(e) defines “violent felony” to include “burglary.” Under this Court’s 1990 decision in Taylor v. United States, 495 U.S. 575, the generic statu- tory term “burglary” means “unlawful or unprivileged entry into, or remaining in, a building or structure, with intent to commit a crime.” Id., at 599 (emphasis added). The exceedingly narrow question in this case concerns remaining-in burglary. The question is whether remaining-in burglary (i) occurs only if a person has the intent to commit a crime at the exact moment when he or she first unlawfully remains in a building or structure, or (ii) more broadly, occurs when a person forms the intent to commit a crime at any time while unlawfully remaining in a building or structure. For purposes of §924(e), we conclude that remaining-in burglary occurs when the defendant forms the intent to commit a crime at any time while unlawfully remaining in a building or structure. We affirm the judgment of the U. S. Court of Appeals for the Sixth Circuit. I On August 24, 2013, police officers in Grand Rapids, Michigan, responded to a 911 call. When the officers arrived at the scene, the caller, Chasity Warren, told the officers that she had just escaped from her boyfriend, Jamar Quarles. Warren said that Quarles had threatened her at gunpoint and also hit her. While the police officers were speaking with Warren, Quarles drove by. The officers then arrested Quarles and later searched his house. Inside they found a semiautomatic pistol. Quarles pled guilty to being a felon in possession of a firearm in violation of 18 U. S. C. §922(g)(1). Quarles had at least three prior convictions that appeared to qualify as violent felonies under the Armed Career Criminal Act, 18 U. S. C. §924(e). Those three convictions were: (1) a 2002 Michigan conviction for third-degree home invasion stemming from an attempt to chase down an ex-girlfriend who had sought refuge in a nearby apartment; (2) a 2004 Michigan conviction for assault with a dangerous weapon based on an incident where Quarles held a gun to the head of another ex-girlfriend and threatened to kill her; and (3) a 2008 Michigan conviction for assault with a dangerous weapon arising from an altercation with another man and that same ex-girlfriend in which Quarles shot at the man. In the sentencing proceedings for his federal felon-in-possession offense, Quarles argued that his 2002 Michigan conviction for third-degree home invasion did not qualify as a burglary under §924(e). Under this Court’s precedents, the District Court had to decide whether the Michigan statute under which Quarles was convicted in 2002 was broader than the generic definition of burglary set forth in Taylor (in which case the conviction would not qualify as a prior conviction under §924(e)) or, instead, whether the Michigan statute “substantially correspond[ed]” to or was narrower than the generic definition of burglary set forth in Taylor. 495 U. S., at 602. To reiterate, Taylor interpreted burglary under §924(e) to mean “unlawful or unprivileged entry into, or remaining in, a building or structure, with intent to commit a crime.” Id., at 599 (emphasis added). Under the Michigan law at issue here, a person commits third-degree home invasion if he or she “breaks and enters a dwelling or enters a dwelling without permission and, at any time while he or she is entering, present in, or exiting the dwelling, commits a misdemeanor.” Mich. Comp. Laws Ann. §750.110a(4)(a) (West 2004) (emphasis added). Quarles argued to the District Court that the Michigan third-degree home invasion statute swept too broadly to qualify as burglary under §924(e) because the Michigan statute encompassed situations where the defendant forms the intent to commit a crime at any time while unlawfully remaining in a dwelling, not at the exact moment when the defendant is first unlawfully present in a dwelling. The District Court rejected that argument and sentenced Quarles to 17 years in prison. The Sixth Circuit affirmed. 850 F.3d 836, 840 (2017). We granted certio- rari in light of a Circuit split on the question of how to assess state remaining-in burglary statutes for purposes of §924(e). 586 U. S. ___ (2019). II Section 924(e) lists “burglary” as a qualifying predicate offense for purposes of the Armed Career Criminal Act. But §924(e) does not define “burglary.” The question here is how to define “burglary” under §924(e). We do not write on a clean slate. See Taylor, 495 U. S., at 599. At common law, burglary was confined to unlawful breaking and entering a dwelling at night with the intent to commit a felony. See, e.g., 4 W. Blackstone, Commentaries on the Laws of England 224 (1769). But by the time Congress passed and President Reagan signed the current version of §924(e) in 1986, state burglary statutes had long since departed from the common-law formulation. See Taylor, 495 U. S., at 593–596. In addition to casting off relics like the requirement that there be a breaking, or that the unlawful entry occur at night, a majority of States by 1986 prohibited unlawfully “remaining in” a building or structure with intent to commit a crime. Those remaining-in statutes closed a loophole in some States’ laws by extending burglary to cover situations where a per- son enters a structure lawfully but stays unlawfully—for example, by remaining in a store after closing time without permission to do so. In the 1990 Taylor decision, this Court interpreted the term “burglary” in §924(e) in accord with the more expansive understanding of burglary that had become common by 1986: “We believe that Congress meant by ‘burglary’ the generic sense in which the term is now used in the criminal codes of most States.” 495 U. S., at 598. The Court concluded that generic burglary under §924(e) means “unlawful or unprivileged entry into, or remaining in, a building or structure, with intent to commit a crime.” Id., at 599 (emphasis added). A defendant’s prior conviction under a state statute qualifies as a predicate burglary under §924(e) if the state statute—regardless of its “exact definition or label”—“substantially corresponds” to or is narrower than the generic definition of burglary. Id., at 599, 602. In this case, we must determine the scope of generic remaining-in burglary under Taylor—in particular, the timing of the intent requirement. Quarles argues that remaining-in burglary occurs only when the defendant has the intent to commit a crime at the exact moment when he or she first unlawfully remains in a building or structure. The Government argues for a broader definition of remaining-in burglary. According to the Government, remaining-in burglary occurs when the defendant forms the intent to commit a crime at any time while unlawfully present in a building or structure. We agree with the Government. As noted, Taylor interpreted generic burglary under §924(e) to include remaining-in burglary. Id., at 599. In ordinary usage, “remaining in” refers to a continuous activity. See United States v. Cores, 356 U.S. 405, 408 (1958); see also Webster’s New International Dictionary 2106 (2d ed. 1949); 8 Oxford English Dictionary 418 (1933). This Court has followed that ordinary meaning in analogous legal contexts. For example, when interpreting a federal criminal statute punishing any “ ‘alien crewman who willfully remains in the United States in excess of the number of days allowed,’ ” the Court stated that “the crucial word ‘remains’ permits no connotation other than continuing presence.” Cores, 356 U. S., at 408. The law of trespass likewise proscribes remaining on the land of another without permission. In that context, the term “remain” refers to “a continuing trespass for the entire time during which the actor wrongfully remains.” Restatement (Second) of Torts §158, Comment m, p. 280 (1965). Those interpretations of “remaining in” in analogous areas of the law inform our interpretation of “remaining-in” burglary in §924(e). In particular, the common understanding of “remaining in” as a continuous event means that burglary occurs for purposes of §924(e) if the defendant forms the intent to commit a crime at any time during the continuous event of unlawfully remaining in a building or structure. To put it in conventional criminal law terms: Because the actus reus is a continuous event, the mens rea matches the actus reus so long as the burglar forms the intent to commit a crime at any time while unlawfully present in the building or structure. Quarles insists, however, that to constitute a burglary under §924(e), the intent to commit a crime must be contemporaneous with unlawful entry or remaining. That is true. But the defendant’s intent is contemporaneous with the unlawful remaining so long as the defendant forms the intent at any time while unlawfully remaining. Put sim- ply, for burglary predicated on unlawful entry, the defendant must have the intent to commit a crime at the time of entry. For burglary predicated on unlawful remaining, the defendant must have the intent to commit a crime at the time of remaining, which is any time during which the defendant unlawfully remains. That conclusion is supported by the States’ laws as of 1986 when Congress enacted §924(e). As of 1986, a majority of States proscribed remaining-in burglary. At that time, there was not much case law addressing the precise timing of the intent requirement for remaining-in bur- glary. That is presumably because in most remaining-in burglaries, the defendant has the intent to commit a crime when he or she first unlawfully remains in a building or structure. The timing issue arises only in the rarer cases where the defendant forms the intent to commit a crime only after unlawfully remaining in the building or structure for a while. In any event, for present purposes, the important point is that all of the state appellate courts that had definitively addressed this issue as of 1986 had interpreted remaining-in burglary to occur when the defendant forms the intent to commit a crime at any time while unlawfully present in the building or structure. See Gratton v. State, 456 So. 2d 865, 872 (Ala. Crim. App. 1984); State v. Embree, 130 Ariz. 64, 66, 633 P.2d 1057, 1059 (App. 1981); Keith v. State, 138 Ga. App. 239, 225 S.E.2d 719, 720 (1976); State v. Mogenson, 10 Kan. App. 2d 470, 472–476, 701 P.2d 1339, 1343–1345 (1985); State v. Papineau, 53 Ore. App. 33, 38, 630 P.2d 904, 906–907 (1981).[1] Especially in light of the body of state law as of 1986, it is not likely that Congress intended generic burglary under §924(e) to include (i) a burglar who intends to commit a crime at the exact moment when he or she first unlawfully remains in a building or structure, but to exclude (ii) a burglar who forms the intent to commit a crime at any time while unlawfully remaining in a building or structure. Indeed, excluding that latter category of burglaries from generic burglary under §924(e) would make little sense in light of Congress’ rationale for specifying burglary as a violent felony. As the Court recognized in Taylor, Congress “singled out burglary” because of its “inherent potential for harm to persons.” 495 U. S., at 588. Burglary is dangerous because it “creates the possibility of a violent confrontation between the offender and an occupant, caretaker, or some other person who comes to investigate.” Ibid.; see also United States v. Stitt, 586 U. S. ___, ___ (2018) (slip op., at 6). With respect to remaining-in burglary, the possibility of a violent confrontation does not depend on the exact moment when the burglar forms the intent to commit a crime while unlawfully present in a building or structure. Once an intruder is both unlawfully present inside a building or structure and has the requisite intent to commit a crime, all of the reasons that led Congress to include burglary as a §924(e) predicate fully apply. The dangers of remaining-in burglary are not tied to the esoteric question of precisely when the defendant forms the intent to commit a crime. That point underscores that Congress, when enacting §924(e) in 1986, would not have understood the meaning of burglary to hinge on exactly when the defendant forms the intent to commit a crime while unlawfully present in a building or structure. Moreover, to interpret remaining-in burglary narrowly, as Quarles advocates, would thwart the stated goals of the Armed Career Criminal Act. After all, most burglaries involve unlawful entry, not unlawful remaining in. Yet if we were to narrowly interpret the remaining-in category of generic burglary so as to require that the defendant have the intent to commit a crime at the exact moment he or she first unlawfully remains, then many States’ bur- glary statutes would be broader than generic burglary. As a result, under our precedents, many States’ burglary statutes would presumably be eliminated as predicate offenses under §924(e). That result not only would defy common sense, but also would defeat Congress’ stated objective of imposing enhanced punishment on armed career criminals who have three prior convictions for burglary or other violent felonies. We should not lightly conclude that Congress enacted a self-defeating statute. See, e.g., Stokeling v. United States, 586 U. S. ___, ___ (2019) (slip op., at 8); Taylor, 495 U. S., at 594. To sum up: The Armed Career Criminal Act does not define the term “burglary.” In Taylor, the Court explained that “Congress did not wish to specify an exact formulation that an offense must meet in order to count as ‘burglary’ for enhancement purposes.” Id., at 599. And the Court recognized that the definitions of burglary “vary” among the States. Id., at 598. The Taylor Court therefore interpreted the generic term “burglary” in §924(e) in light of: the ordinary understanding of burglary as of 1986; the States’ laws at that time; Congress’ recognition of the dangers of burglary; and Congress’ stated objective of imposing increased punishment on armed career criminals who had committed prior burglaries. Looking at those sources, the Taylor Court interpreted generic burglary under §924(e) to encompass remaining-in burglary. Looking at those same sources, we interpret remaining-in burglary under §924(e) to occur when the defendant forms the intent to commit a crime at any time while unlawfully present in a building or structure. III In light of our conclusion that generic remaining-in burglary occurs when the defendant forms the intent to commit a crime at any time while unlawfully remaining in a building or structure, Quarles’ case is easily resolved. The question in Quarles’ case is whether the Michigan home-invasion statute under which he was convicted in 2002 is broader than generic burglary or, instead, “substantially corresponds” to or is narrower than generic burglary. Id., at 602. Regarding that inquiry, the Taylor Court cautioned courts against seizing on modest state-law deviations from the generic definition of burglary. A state law’s “exact definition or label” does not control. Id., at 599. As the Court stated in Taylor, so long as the state law in question “substantially corresponds” to (or is narrower than) generic burglary, the conviction qualifies under §924(e). Id., at 602. As stated above, generic remaining-in burglary occurs under §924(e) when the defendant forms the intent to commit a crime at any time while unlawfully remaining in a building or structure. For the Court’s purposes here, the Michigan statute substantially corresponds to or is narrower than generic burglary.[2] * * * We affirm the judgment of the U. S. Court of Appeals for the Sixth Circuit. It is so ordered. Notes 1 The consensus position has not changed. Today, of the States that have addressed the question, at least 18 have adopted the “at any time” interpretation of remaining-in burglary, and only 3 appear to have adopted the narrower interpretation. Of those 18 States, some have adopted the broader “at any time” interpretation by statute. See Colo. Rev. Stat. §18–4–201(3) (2018); Del. Code Ann., Tit. 11, §829(e) (2015); Haw. Rev. Stat. Ann. §708–812.5 (2014); Mich. Comp. Laws Ann. §750.110a(4)(a) (West 2004); Minn. Stat. §§609.581(4), 609.582(3) (2016); Mont. Code Ann. §45–6–204(1) (2017); Tenn. Code Ann. §39–14–402(a)(3) (2018); Tex. Penal Code Ann. §30.02(a)(3) (West 2019). And in addition to the five pre-1986 state-court decisions identified in the text above, at least five post-1986 state-court decisions have adopted the “at any time” interpretation of “remaining in.” See Braddy v. State, 111 So. 3d 810, 844 (Fla. 2012) (per curiam); State v. Walker, 600 N.W.2d 606, 609 (Iowa 1999); State v. DeNoyer, 541 N.W.2d 725, 732 (S. D. 1995); State v. Rudolph, 970 P.2d 1221, 1228–1229 (Utah 1998); State v. Allen, 127 Wash. App. 125, 135, 110 P.3d 849, 853–855 (2005). By contrast, three state courts appear to have adopted the narrower interpretation. Shetters v. State, 751 P.2d 31, 36, n. 2 (Alaska App. 1988); People v. Gaines, 74 N.Y.2d 358, 361–363, 546 N.E.2d 913, 915–916 (1989); In re J. N. S., 258 Ore. App. 310, 318–319, 308 P.3d 1112, 1117–1118 (2013). 2 In his brief, Quarles alternatively suggests that Michigan’s home-invasion statute actually does not require that the defendant have any intent to commit a crime at any time while unlawfully present in a dwelling. Brief for Petitioner 9. Quarles offers no support for his suggestion that there is no mens rea requirement. In any event, Quarles did not preserve that argument, and we do not address it.

588.US.2018_17-9560

Petitioner Rehaif entered the United States on a nonimmigrant student visa to attend university but was dismissed for poor grades. He subsequently shot two firearms at a firing range. The Government prosecuted him under 18 U. S. C. §922(g), which makes it unlawful for certain persons, including aliens illegally in the country, to possess firearms, and §924(a)(2), which provides that anyone who “knowingly violates” the first provision can be imprisoned for up to 10 years. The jury at Rehaif’s trial was instructed that the Government was not required to prove that he knew that he was unlawfully in the country. It returned a guilty verdict. The Eleventh Circuit affirmed. Held: In a prosecution under §922(g) and §924(a)(2), the Government must prove both that the defendant knew he possessed a firearm and that he knew he belonged to the relevant category of persons barred from possessing a firearm. Pp. 3–12. (a) Whether a criminal statute requires the Government to prove that the defendant acted knowingly is a question of congressional intent. This inquiry starts from a longstanding presumption that Congress intends to require a defendant to possess a culpable mental state regarding “each of the statutory elements that criminalize otherwise innocent conduct,” United States v. X-Citement Video, Inc., 513 U.S. 64, 72, normally characterized as a presumption in favor of “scienter.” There is no convincing reason to depart from this presumption here. The statutory text supports the presumption. It specifies that a defendant commits a crime if he “knowingly” violates §922(g), which makes possession of a firearm unlawful when the following elements are satisfied: (1) a status element (here “being an alien . . . illegally or unlawfully in the United States”); (2) a possession element (to “possess”); (3) a jurisdictional element (“in or affecting commerce”); and (4) a firearm element (a “firearm or ammunition”). Aside from the jurisdictional element, which is not subject to the presumption in favor of scienter, §922(g)’s text simply lists the elements that make a defendant’s behavior criminal. The term “knowingly” is normally read “as applying to all the subsequently listed elements of the crime.” Flores-Figueroa v. United States, 556 U.S. 646, 650. And the “knowingly” requirement clearly applies to §922(g)’s possession element, which follows the status element in the statutory text. There is no basis for interpreting “knowingly” as applying to the second §922(g) element but not the first. This reading of the statute is also consistent with a basic principle underlying the criminal law: the importance of showing what Blackstone called “a vicious will.” Scienter requirements advance this principle by helping to separate wrongful from innocent acts. That is the case here. Possessing a gun can be entirely innocent. It is the defendant’s status, not his conduct alone, that makes the difference. Without knowledge of that status, a defendant may lack the intent needed to make his behavior wrongful. Pp. 3–7. (b) The Government’s arguments to the contrary are unpersuasive. In claiming that Congress does not normally require defendants to know their own status, it points to statutes where the defendant’s status is the “crucial element” separating innocent from wrongful conduct. X-Citement Video, supra, at 73. Those statutes are quite different from the provisions at issue here, where the defendant’s status separates innocent from wrongful conduct. The Government also argues that whether an alien is “illegally or unlawfully in the United States” is a question of law, not fact, and thus appeals to the maxim that “ignorance of the law” is no excuse. But that maxim normally applies where a defendant possesses the requisite mental state in respect to the elements of the crime but claims to be unaware of a law forbidding his conduct. That maxim does not normally apply where a defendant’s mistaken impression about a collateral legal question causes him to misunderstand his conduct’s significance, thereby negating an element of the offense. Rehaif’s status as an alien “illegally or unlawfully in the United States” refers to what commentators call a “collateral” question of law, and a mistake regarding that status negates an element of the offense. Finally, the statutory and legislative history on which the Government relies is at best inclusive. Pp. 7–11. 888 F.3d 1138, reversed and remanded. Breyer, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Alito, J., filed a dissenting opinion, in which Thomas, J., joined.

A federal statute, 18 U. S. C. §922(g), provides that “[i]t shall be unlawful” for certain individuals to possess firearms. The provision lists nine categories of individuals subject to the prohibition, including felons and aliens who are “illegally or unlawfully in the United States.” Ibid. A separate provision, §924(a)(2), adds that anyone who “knowingly violates” the first provision shall be fined or imprisoned for up to 10 years. (Emphasis added.) The question here concerns the scope of the word “knowingly.” Does it mean that the Government must prove that a defendant knew both that he engaged in the relevant conduct (that he possessed a firearm) and also that he fell within the relevant status (that he was a felon, an alien unlawfully in this country, or the like)? We hold that the word “knowingly” applies both to the defendant’s conduct and to the defendant’s status. To convict a defendant, the Government therefore must show that the defendant knew he possessed a firearm and also that he knew he had the relevant status when he possessed it. I Petitioner Hamid Rehaif entered the United States on a nonimmigrant student visa to attend university. After he received poor grades, the university dismissed him and told him that his “ ‘immigration status’ ” would be terminated unless he transferred to a different university or left the country. App. to Pet. for Cert. 3a. Rehaif did neither. Rehaif subsequently visited a firing range, where he shot two firearms. The Government learned about his target practice and prosecuted him for possessing firearms as an alien unlawfully in the United States, in violation of §922(g) and §924(a)(2). At the close of Rehaif’s trial, the judge instructed the jury (over Rehaif’s objection) that the “United States is not required to prove” that Rehaif “knew that he was illegally or unlawfully in the United States.” App. to Pet. for Cert. 4a (internal quotation marks omitted). The jury returned a guilty verdict, and Rehaif was sentenced to 18 months’ imprisonment. Rehaif appealed. He argued that the judge erred in instructing the jury that it did not need to find that he knew he was in the country unlawfully. The Court of Appeals for the Eleventh Circuit, however, concluded that the jury instruction was correct, and it affirmed Rehaif’s conviction. See 888 F.3d 1138, 1148 (2018). The Court of Appeals believed that the criminal law generally does not require a defendant to know his own status, and further observed that no court of appeals had required the Government to establish a defendant’s knowledge of his status in the analogous context of felon-in-possession prosecutions. Id., at 1145–1146. We granted certiorari to consider whether, in prosecutions under §922(g) and §924(a)(2), the Government must prove that a defendant knows of his status as a person barred from possessing a firearm. We now reverse. II Whether a criminal statute requires the Government to prove that the defendant acted knowingly is a question of congressional intent. See Staples v. United States, 511 U.S. 600, 605 (1994). In determining Congress’ intent, we start from a longstanding presumption, traceable to the common law, that Congress intends to require a defendant to possess a culpable mental state regarding “each of the statutory elements that criminalize otherwise innocent conduct.” United States v. X-Citement Video, Inc., 513 U.S. 64, 72 (1994); see also Morissette v. United States, 342 U.S. 246, 256–258 (1952). We normally characterize this interpretive maxim as a presumption in favor of “scienter,” by which we mean a presumption that criminal statutes require the degree of knowledge sufficient to “mak[e] a person legally responsible for the consequences of his or her act or omission.” Black’s Law Dictionary 1547 (10th ed. 2014). We apply the presumption in favor of scienter even when Congress does not specify any scienter in the statutory text. See Staples, 511 U. S., at 606. But the presumption applies with equal or greater force when Congress includes a general scienter provision in the statute itself. See ALI, Model Penal Code §2.02(4), p. 22 (1985) (when a statute “prescribes the kind of culpability that is sufficient for the commission of an offense, without distinguishing among the material elements thereof, such provision shall apply to all the material elements of the offense, unless a contrary purpose plainly appears”). A Here we can find no convincing reason to depart from the ordinary presumption in favor of scienter. The statutory text supports the presumption. The text of §924(a)(2) says that “[w]hoever knowingly violates” certain subsections of §922, including §922(g), “shall be” subject to penalties of up to 10 years’ imprisonment. The text of §922(g) in turn provides that it “shall be unlawful for any person . . . , being an alien . . . illegally or unlawfully in the United States,” to “possess in or affecting commerce, any firearm or ammunition.” The term “knowingly” in §924(a)(2) modifies the verb “violates” and its direct object, which in this case is §922(g). The proper interpretation of the statute thus turns on what it means for a defendant to know that he has “violate[d]” §922(g). With some here-irrelevant omissions, §922(g) makes possession of a firearm or ammunition unlawful when the following elements are satisfied: (1) a status element (in this case, “being an alien . . . illegally or unlawfully in the United States”); (2) a possession element (to “possess”); (3) a jurisdictional element (“in or affecting commerce”); and (4) a firearm element (a “firearm or ammunition”). No one here claims that the word “knowingly” modifies the statute’s jurisdictional element. Jurisdictional elements do not describe the “evil Congress seeks to prevent,” but instead simply ensure that the Federal Government has the constitutional authority to regulate the defendant’s conduct (normally, as here, through its Commerce Clause power). Luna Torres v. Lynch, 578 U. S. ___, ___–___ (2016) (slip op., at 15–16). Because jurisdictional elements normally have nothing to do with the wrongfulness of the defendant’s conduct, such elements are not subject to the presumption in favor of scienter. See id., at ___ (slip op., at 16). Jurisdictional element aside, however, the text of §922(g) simply lists the elements that make a defendant’s behavior criminal. As “a matter of ordinary English grammar,” we normally read the statutory term “ ‘knowingly’ as applying to all the subsequently listed elements of the crime.” Flores-Figueroa v. United States, 556 U.S. 646, 650 (2009); see also id., at 652 (we “ordinarily read a phrase in a criminal statute that introduces the elements of a crime with the word ‘knowingly’ as applying that word to each element”). This is notably not a case where the modifier “knowingly” introduces a long statutory phrase, such that questions may reasonably arise about how far into the statute the modifier extends. See id., at 659 (Alito, J., concurring in part). And everyone agrees that the word “knowingly” applies to §922(g)’s possession element, which is situated after the status element. We see no basis to interpret “knowingly” as applying to the second §922(g) element but not the first. See United States v. Games-Perez, 667 F.3d 1136, 1143 (CA10 2012) (Gorsuch, J., concurring). To the contrary, we think that by specifying that a defendant may be convicted only if he “knowingly violates” §922(g), Congress intended to require the Government to establish that the defendant knew he violated the material elements of §922(g). B Beyond the text, our reading of §922(g) and §924(a)(2) is consistent with a basic principle that underlies the criminal law, namely, the importance of showing what Blackstone called “a vicious will.” 4 W. Blackstone, Commentaries on the Laws of England 21 (1769). As this Court has explained, the understanding that an injury is criminal only if inflicted knowingly “is as universal and persistent in mature systems of law as belief in freedom of the human will and a consequent ability and duty of the normal individual to choose between good and evil.” Morissette, 342 U. S., at 250. Scienter requirements advance this basic principle of criminal law by helping to “separate those who understand the wrongful nature of their act from those who do not.” X-Citement Video, 513 U. S., at 72–73, n. 3. The cases in which we have emphasized scienter’s importance in separating wrongful from innocent acts are legion. See, e.g., id., at 70; Staples, 511 U. S., at 610; Liparota v. United States, 471 U.S. 419, 425 (1985); United States v. Bailey, 444 U.S. 394, 406, n. 6 (1980); United States v. United States Gypsum Co., 438 U.S. 422, 436 (1978); Morissette, 342 U. S., at 250–251. We have interpreted statutes to include a scienter requirement even where the statutory text is silent on the question. See Staples, 511 U. S., at 605. And we have interpreted statutes to include a scienter requirement even where “the most grammatical reading of the statute” does not support one. X-Citement Video, 513 U. S., at 70. Applying the word “knowingly” to the defendant’s status in §922(g) helps advance the purpose of scienter, for it helps to separate wrongful from innocent acts. Assuming compliance with ordinary licensing requirements, the possession of a gun can be entirely innocent. See Staples, 511 U. S., at 611. It is therefore the defendant’s status, and not his conduct alone, that makes the difference. Without knowledge of that status, the defendant may well lack the intent needed to make his behavior wrongful. His behavior may instead be an innocent mistake to which criminal sanctions normally do not attach. Cf. O. Holmes, The Common Law 3 (1881) (“even a dog distinguishes between being stumbled over and being kicked”). We have sometimes declined to read a scienter requirement into criminal statutes. See United States v. Balint, 258 U.S. 250, 254 (1922). But we have typically declined to apply the presumption in favor of scienter in cases involving statutory provisions that form part of a “regulatory” or “public welfare” program and carry only minor penalties. See Staples, 511 U. S., at 606; Morissette, 342 U. S., at 255–259. The firearms provisions before us are not part of a regulatory or public welfare program, and they carry a potential penalty of 10 years in prison that we have previously described as “harsh.” X-Citement Video, 513 U. S., at 72. Hence, this exception to the presumption in favor of scienter does not apply. III The Government’s arguments to the contrary do not convince us that Congress sought to depart from the normal presumption in favor of scienter. The Government argues that Congress does not normally require defendants to know their own status. But the Government supports this claim primarily by referring to statutes that differ significantly from the provisions at issue here. One of these statutes prohibits “an officer, employee, contractor, or consultant of the United States” from misappropriating classified information. 18 U. S. C. §1924(a). Another statute applies to anyone “at least eighteen years of age” who solicits a minor to help avoid detection for certain federal crimes. 21 U. S. C. §861(a)(2). A third applies to a “parent [or] legal guardian” who allows his child to be used for child pornography. 18 U. S. C. §2251(b). We need not decide whether we agree or disagree with the Government’s interpretation of these statutes. In the provisions at issue here, the defendant’s status is the “crucial element” separating innocent from wrongful conduct. X-Citement Video, 513 U. S., at 73. But in the statutes cited by the Government, the conduct prohibited—misappropriating classified information, seeking to evade detection for certain federal crimes, and facilitating child pornography—would be wrongful irrespective of the defendant’s status. This difference assures us that the presumption in favor of scienter applies here even assuming the Government is right that these other statutes do not require knowledge of status. Nor do we believe that Congress would have expected defendants under §922(g) and §924(a)(2) to know their own statuses. If the provisions before us were construed to require no knowledge of status, they might well apply to an alien who was brought into the United States unlawfully as a small child and was therefore unaware of his un- lawful status. Or these provisions might apply to a person who was convicted of a prior crime but sentenced only to probation, who does not know that the crime is “punish- able by imprisonment for a term exceeding one year.” §922(g)(1) (emphasis added); see also Games-Perez, 667 F. 3d, at 1138 (defendant held strictly liable regarding his status as a felon even though the trial judge had told him repeatedly—but incorrectly—that he would “leave this courtroom not convicted of a felony”). As we have said, we normally presume that Congress did not intend to impose criminal liability on persons who, due to lack of knowledge, did not have a wrongful mental state. And we doubt that the obligation to prove a defendant’s knowledge of his status will be as burdensome as the Government suggests. See Staples, 511 U. S., at 615, n. 11 (“knowledge can be inferred from circumstantial evidence”). The Government also argues that whether an alien is “illegally or unlawfully in the United States” is a question of law, not fact, and thus appeals to the well-known maxim that “ignorance of the law” (or a “mistake of law”) is no excuse. Cheek v. United States, 498 U.S. 192, 199 (1991). This maxim, however, normally applies where a defendant has the requisite mental state in respect to the elements of the crime but claims to be “unaware of the existence of a statute proscribing his conduct.” 1 W. LaFave & A. Scott, Substantive Criminal Law §5.1(a), p. 575 (1986). In contrast, the maxim does not normally apply where a defendant “has a mistaken impression concerning the legal effect of some collateral matter and that mistake results in his misunderstanding the full significance of his conduct,” thereby negating an element of the offense. Ibid.; see also Model Penal Code §2.04, at 27 (a mistake of law is a defense if the mistake negates the “knowledge . . . required to establish a material element of the offense”). Much of the confusion surrounding the ignorance-of-the-law maxim stems from “the failure to distinguish [these] two quite different situations.” LaFave, Substantive Criminal Law §5.1(d), at 585. We applied this distinction in Liparota, where we considered a statute that imposed criminal liability on “whoever knowingly uses, transfers, acquires, alters, or possesses” food stamps “in any manner not authorized by the statute or the regulations.” 471 U. S., at 420 (quotation altered). We held that the statute required scienter not only in respect to the defendant’s use of food stamps, but also in respect to whether the food stamps were used in a “manner not authorized by the statute or regulations.” Id., at 425, n. 9. We therefore required the Government to prove that the defendant knew that his use of food stamps was unlawful—even though that was a question of law. See ibid. This case is similar. The defendant’s status as an alien “illegally or unlawfully in the United States” refers to a legal matter, but this legal matter is what the commentators refer to as a “collateral” question of law. A defendant who does not know that he is an alien “illegally or unlawfully in the United States” does not have the guilty state of mind that the statute’s language and purposes require. The Government finally turns for support to the statu- tory and legislative history. Congress first enacted a criminal statute prohibiting particular categories of persons from possessing firearms in 1938. See Federal Firearms Act, 52Stat. 1250. In 1968, Congress added new cate- gories of persons subject to the prohibition. See Omnibus Crime Control and Safe Streets Act, 82Stat. 197. Then, in 1986, Congress passed the statute at issue here, the Firearms Owners’ Protection Act, 100Stat. 449, note following 18 U. S. C. §921, which reorganized the prohibition on firearm possession and added the language providing that only those who violate the prohibition “knowingly” may be held criminally liable. The Government says that, prior to 1986, the courts had reached a consensus that the law did not require the Government to prove scienter regarding a defendant’s status. And the Government relies on the interpretive canon providing that when particular statutory language has received a settled judicial construction, and Congress subsequently reenacts that “same language,” courts should presume that Congress intended to ratify the judicial consensus. Helsinn Healthcare S. A. v. Teva Pharmaceuticals USA, Inc., 586 U. S. ___, ___ (2019) (slip op., at 7). Prior to 1986, however, there was no definitive judicial consensus that knowledge of status was not needed. This Court had not considered the matter. As the Government says, most lower courts had concluded that the statute did not require knowledge of status. See, e.g., United States v. Pruner, 606 F.2d 871, 874 (CA9 1979). But the Sixth Circuit had held to the contrary, specifically citing the risk that a defendant “may not be aware of the fact” that barred him from possessing a firearm. United States v. Renner, 496 F.2d 922, 926 (1974). And the Fourth Circuit had found that knowledge of a defendant’s status was not needed because the statute “[b]y its terms” did not require knowledge of status. United States v. Williams, 588 F.2d 92 (1978) (per curiam). This last-mentioned circumstance is important. Any pre-1986 consensus involved the statute as it read prior to 1986—without any explicit scienter provision. But Congress in 1986 added a provision clarifying that a defendant could be convicted only if he violated the prohibition on firearm possession “knowingly.” This addition, which would serve no apparent purpose under the Government’s view, makes it all but impossible to draw any inference that Congress intended to ratify a pre-existing consensus when, in 1986, it amended the statute. The Government points to the House Report on the legislation, which says that the 1986 statute would require the Government to prove “that the defendant’s conduct was knowing.” H. R. Rep. No. 99–495, p. 10 (1986) (emphasis added). Although this statement speaks of “conduct” rather than “status,” context suggests that the Report may have meant the former to include the latter. In any event, other statements suggest that the word “knowingly” was intended to apply to both conduct and status. The Senate Report, for example, says that the proposed amendments sought to exclude “individuals who lack all criminal intent and knowledge,” without distinguishing between conduct and status. S. Rep. No. 97–476, p. 15 (1982). And one Senate sponsor of the bill pointed out that the absence of a scienter requirement in the prior statutes had resulted in “severe penalties for unintentional missteps.” 132 Cong. Rec. 9590 (1986) (statement of Sen. Hatch). Thus, assuming without deciding that statutory or legislative history could overcome the longstanding presumption in favor of scienter, that history here is at best inconclusive. * * * The Government asks us to hold that any error in the jury instructions in this case was harmless. But the lower courts did not address that question. We therefore leave the question for those courts to decide on remand. See Thacker v. TVA, 587 U. S. ___, ___ (2019) (slip op., at 10) (citing Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7 (2005)). We conclude that in a prosecution under 18 U. S. C. §922(g) and §924(a)(2), the Government must prove both that the defendant knew he possessed a firearm and that he knew he belonged to the relevant category of persons barred from possessing a firearm. We express no view, however, about what precisely the Government must prove to establish a defendant’s knowledge of status in respect to other §922(g) provisions not at issue here. See post, at 13–15 (Alito, J., dissenting) (discussing other statuses listed in §922(g) not at issue here). We accordingly reverse the judgment of the Court of Appeals and re- mand the case for further proceedings consistent with this opinion. It is so ordered. APPENDIX 18 U. S. C. §924(a)(2) “Whoever knowingly violates subsection (a)(6), (d), (g), (h), (i), (j), or (o) of section 922 shall be fined as provided in this title, imprisoned not more than 10 years, or both.” 18 U. S. C. §922(g) “It shall be unlawful for any person— “(1) who has been convicted in any court of, a crime punishable by imprisonment for a term exceeding one year; “(2) who is a fugitive from justice; “(3) who is an unlawful user of or addicted to any controlled substance . . . ; “(4) who has been adjudicated as a mental defective or who has been committed to a mental institution; “(5) who, being an alien—(A) is illegally or unlawfully in the United States; or (B) . . . has been admitted to the United States under a nonimmigrant visa (as that term is defined in section 101(a)(26) of the Immigration and Nationality Act (8 U.S.C. 1101(a)(26))); “(6) who has been discharged from the Armed Forces under dishonorable conditions; “(7) who, having been a citizen of the United States, has renounced his citizenship; “(8) who is subject to a court order that—(A) was issued after a hearing of which such person received actual notice, and at which such person had an opportunity to participate; (B) restrains such person from harassing, stalking, or threatening an intimate partner of such person or child of such intimate partner or person, or engaging in other conduct that would place an intimate partner in reasonable fear of bodily injury to the partner or child; and (C)(i) includes a finding that such person represents a credible threat to the physical safety of such intimate partner or child; or (ii) by its terms explicitly prohibits the use, attempted use, or threatened use of physical force against such intimate partner or child that would reasonably be expected to cause bodily injury; or “(9) who has been convicted in any court of a misdemeanor crime of domestic violence, to ship or transport in interstate or foreign commerce, or possess in or affecting commerce, any firearm or ammunition; or to receive any firearm or ammunition which has been shipped or transported in interstate or foreign commerce.”

587.US.2018_16-1094

The Foreign Sovereign Immunities Act of 1976 (FSIA) generally immunizes foreign states from suit in this country unless one of several enumerated exceptions to immunity applies. 28 U. S. C. §§1604, 1605–1607. If an exception applies, the FSIA provides subject-matter jurisdiction in federal district court, §1330(a), and personal jurisdiction “where service has been made under section 1608,” §1330(b). Section 1608(a) provides four methods of serving civil process, including, as relevant here, service “by any form of mail requiring a signed receipt, to be addressed and dispatched . . . to the head of the ministry of foreign affairs of the foreign state concerned,” §1608(a)(3). Respondents, victims of the bombing of the USS Cole and their family members, sued the Republic of Sudan under the FSIA, alleging that Sudan provided material support to al Qaeda for the bombing. The court clerk, at respondents’ request, addressed the service packet to Sudan’s Minister of Foreign Affairs at the Sudanese Embassy in the United States and later certified that a signed receipt had been returned. After Sudan failed to appear in the litigation, the District Court entered a default judgment for respondents and subsequently issued three orders requiring banks to turn over Sudanese assets to pay the judgment. Sudan challenged those orders, arguing that the judgment was invalid for lack of personal jurisdiction, because §1608(a)(3) required that the service packet be sent to its foreign minister at his principal office in Sudan, not to the Sudanese Embassy in the United States. The Second Circuit affirmed, reasoning that the statute was silent on where the mailing must be sent and that the method chosen was consistent with the statute’s language and could be reasonably expected to result in delivery to the foreign minister. Held: Most naturally read, §1608(a)(3) requires a mailing to be sent directly to the foreign minister’s office in the foreign state. Pp. 5–17. (a) A letter or package is “addressed” to an intended recipient when his or her name and address are placed on the outside. The noun “address” means “a residence or place of business.” Webster’s Third New International Dictionary 25. A foreign nation’s embassy in the United States is neither the residence nor the usual place of business of that nation’s foreign minister. Similarly, to “dispatch” a letter to an addressee connotes sending it directly. It is also significant that service under §1608(a)(3) requires a signed returned receipt to ensure delivery to the addressee. Pp. 5–9. (b) Several related provisions in §1608 support this reading. Section 1608(b)(3)(B) contains similar “addressed and dispatched” language, but also says that service by its method is permissible “if reasonably calculated to give actual notice.” Respondents’ suggestion that §1608(a)(3) embodies a similar standard runs up against well-settled principles of statutory interpretation. See Department of Homeland Security v. MacLean, 574 U. S. ___, ___, and Mackey v. Lanier Collection Agency & Service, Inc., 486 U.S. 825, 837. Section 1608(b)(2) expressly allows service on an agent, specifies the particular individuals who are permitted to be served as agents of the recipient, and makes clear that service on the agent may occur in the United States. Congress could have included similar terms in §1608(a)(3) had it intended the provision to operate in this manner. Section 1608(c) deems service to have occurred under all methods only when there is a strong basis for concluding that the service packet will very shortly thereafter come into the hands of a foreign official who will know what needs to be done. Under §1608(a)(3), that occurs when the person who receives it from the carrier signs for it. Interpreting §1608(a)(3) to require that a service packet be sent to a foreign minister’s own office rather than to a mailroom employee in a foreign embassy better harmonizes the rules for determining when service occurs. Pp. 9–13. (c) This reading of §1608(a)(3) avoids potential tension with the Federal Rules of Civil Procedure and the Vienna Convention on Diplomatic Relations. If mailing a service packet to a foreign state’s embassy in the United States were sufficient, then it would appear to be easier to serve the foreign state than to serve a person in that foreign state under Rule 4. The natural reading of §1608(a)(3) also avoids the potential international implications arising from the State Department’s position that the Convention’s principle of inviolability precludes serving a foreign state by mailing process to the foreign state’s embassy in the United States. Pp. 13–15. (d) Respondents’ remaining arguments are unavailing. First, their suggestion that §1608(a)(3) demands that service be sent “to a location that is likely to have a direct line of communication to the foreign minister” creates difficult line-drawing problems that counsel in favor of maintaining a clear, administrable rule. Second, their claim that §1608(a)(4)—which requires that process be sent to the Secretary of State in “Washington, District of Columbia”—shows that Congress did not intend §1608(a)(3) to have a similar locational requirement is outweighed by the countervailing arguments already noted. Finally, they contend that it would be unfair to throw out their judgment based on petitioner’s highly technical and belatedly raised argument. But in cases with sensitive diplomatic implications, the rule of law demands adherence to strict rules, even when the equities seem to point in the opposite direction. Pp. 15–17. 802 F.3d 399, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Sotomayor, Kagan, Gorsuch, and Kav- anaugh, JJ., joined. Thomas, J., filed a dissenting opinion.

This case concerns the requirements applicable to a particular method of serving civil process on a foreign state. Under the Foreign Sovereign Immunities Act of 1976 (FSIA), a foreign state may be served by means of a mailing that is “addressed and dispatched . . . to the head of the ministry of foreign affairs of the foreign state concerned.” 28 U. S. C. §1608(a)(3). The question now before us is whether this provision is satisfied when a service packet that names the foreign minister is mailed to the foreign state’s embassy in the United States. We hold that it is not. Most naturally read, §1608(a)(3) requires that a mailing be sent directly to the foreign minister’s office in the minister’s home country. I A Under the FSIA, a foreign state is immune from the jurisdiction of courts in this country unless one of several enumerated exceptions to immunity applies. 28 U. S. C. §§1604, 1605–1607. If a suit falls within one of these exceptions, the FSIA provides subject-matter jurisdiction in federal district courts. §1330(a). The FSIA also provides for personal jurisdiction “where service has been made under section 1608.” §1330(b). Section 1608(a) governs service of process on “a foreign state or political subdivision of a foreign state.” §1608(a); Fed. Rule Civ. Proc. 4(j)(1). In particular, it sets out in hierarchical order the following four methods by which “[s]ervice . . . shall be made.” 28 U. S. C. §1608(a). The first method is by delivery of a copy of the summons and complaint “in accordance with any special arrangement for service between the plaintiff and the foreign state or political subdivision.” §1608(a)(1). “[I]f no special arrangement exists,” service may be made by the second method, namely, delivery of a copy of the summons and complaint “in accordance with an applicable international convention on service of judicial documents.” §1608(a)(2). If service is not possible under either of the first two methods, the third method, which is the one at issue in this case, may be used. This method calls for “sending a copy of the summons and complaint and a notice of suit, together with a translation of each into the official language of the foreign state, by any form of mail requiring a signed receipt, to be addressed and dispatched by the clerk of the court to the head of the ministry of foreign affairs of the foreign state concerned.” §1608(a)(3) (emphasis added). Finally, if service cannot be made within 30 days under §1608(a)(3), service may be effected by sending the service packet “by any form of mail requiring a signed receipt, to be addressed and dispatched by the clerk of the court to the Secretary of State in Washington, District of Columbia,” for transmittal “through diplomatic channels to the foreign state.” §1608(a)(4). Once served, a foreign state or political subdivision has 60 days to file a responsive pleading. §1608(d). If the foreign state or political subdivision does not do this, it runs the risk of incurring a default judgment. See §1608(e). A copy of any such default judgment must be “sent to the foreign state or political subdivision in the [same] manner prescribed for service.” Ibid. B On October 12, 2000, the USS Cole, a United States Navy guided-missile destroyer, entered the harbor of Aden, Yemen, for what was intended to be a brief refueling stop. While refueling was underway, a small boat drew along the side of the Cole, and the occupants of the boat detonated explosives that tore a hole in the side of the Cole. Seventeen crewmembers were killed, and dozens more were injured. Al Qaeda later claimed responsibility for the attack. Respondents in this case are victims of the USS Cole bombing and their family members. In 2010, respondents sued petitioner, the Republic of Sudan, alleging that Sudan had provided material support to al Qaeda for the bombing. See 28 U. S. C. §§1605A(a)(1), (c). Because respondents brought suit under the FSIA, they were required to serve Sudan with process under §1608(a). It is undisputed that service could not be made under §1608(a)(1) or §1608(a)(2), and respondents therefore turned to §1608(a)(3). At respondents’ request, the clerk of the court sent the service packet, return receipt requested, to: “Republic of Sudan, Deng Alor Koul, Minister of Foreign Affairs, Embassy of the Republic of Sudan, 2210 Massachusetts Avenue NW, Washington, DC 20008.” App. 172. The clerk certified that the service packet had been sent and, a few days later, certified that a signed receipt had been returned.[1] After Sudan failed to appear in the litigation, the District Court for the District of Columbia held an evidentiary hearing and entered a $314 million default judgment against Sudan. Again at respondents’ request, the clerk of the court mailed a copy of the default judgment in the same manner that the clerk had previously used. See §1608(e). With their default judgment in hand, respondents turned to the District Court for the Southern District of New York, where they sought to register the judgment and satisfy it through orders requiring several banks to turn over Sudanese assets. See 28 U. S. C. §1963 (providing for registration of judgments for enforcement in other districts). Pursuant to §1610(c), the District Court entered an order confirming that a sufficient period of time had elapsed following the entry and notice of the default judgment, and the court then issued three turnover orders. At this point, Sudan made an appearance for the purpose of contesting jurisdiction. It filed a notice of appeal from each of the three turnover orders and contended on appeal that the default judgment was invalid for lack of personal jurisdiction. In particular, Sudan maintained that §1608(a)(3) required that the service packet be sent to its foreign minister at his principal office in Khartoum, the capital of Sudan, and not to the Sudanese Embassy in the United States. The Court of Appeals for the Second Circuit rejected this argument and affirmed the orders of the District Court. 802 F.3d 399 (2015). The Second Circuit reasoned that, although §1608(a)(3) requires that a service packet be mailed “to the head of the ministry of foreign affairs of the foreign state concerned,” the statute “is silent as to a specific location where the mailing is to be addressed.” Id., at 404. In light of this, the court concluded that “the method chosen by plaintiffs—a mailing addressed to the minister of foreign affairs at the embassy—was consistent with the language of the statute and could reasonably be expected to result in delivery to the intended person.” Ibid. Sudan filed a petition for rehearing, and the United States filed an amicus curiae brief in support of Sudan’s petition. The panel ordered supplemental briefing and heard additional oral argument, but it once again affirmed, reiterating its view that §1608(a)(3) “does not specify that the mailing be sent to the head of the ministry of foreign affairs in the foreign country.” 838 F.3d 86, 91 (CA2 2016). The court thereafter denied Sudan’s petition for rehearing en banc. Subsequent to the Second Circuit’s decision, the Court of Appeals for the Fourth Circuit held in a similar case that §1608(a)(3) “does not authorize delivery of service to a foreign state’s embassy even if it correctly identifies the intended recipient as the head of the ministry of foreign affairs.” Kumar v. Republic of Sudan, 880 F.3d 144, 158 (2018), cert. pending, No. 17–1269. We granted certiorari to resolve this conflict. 585 U. S. ___ (2018) II A The question before us concerns the meaning of §1608(a)(3), and in interpreting that provision, “[w]e begin ‘where all such inquiries must begin: with the language of the statute itself.’ ” Caraco Pharmaceutical Laboratories, Ltd. v. Novo Nordisk A/S, 566 U.S. 399, 412 (2012) (quoting United States v. Ron Pair Enterprises, Inc., 489 U.S. 235, 241 (1989)). As noted, §1608(a)(3) requires that service be sent “by any form of mail requiring a signed receipt, to be addressed and dispatched by the clerk of the court to the head of the ministry of foreign affairs of the foreign state concerned.” The most natural reading of this language is that service must be mailed directly to the foreign minister’s office in the foreign state. Although this is not, we grant, the only plausible reading of the statutory text, it is the most natural one. See, e.g., United States v. Hohri, 482 U.S. 64, 69–71 (1987) (choosing the “more natural” reading of a statute); ICC v. Texas, 479 U.S. 450, 456–457 (1987) (same); see also Florida Dept. of Revenue v. Piccadilly Cafeterias, Inc., 554 U.S. 33, 41 (2008) (similar). A key term in §1608(a)(3) is the past participle “addressed.” A letter or package is “addressed” to an intended recipient when his or her name and “address” is placed on the outside of the item to be sent. And the noun “address,” in the sense relevant here, means “the designation of a place (as a residence or place of business) where a person or organization may be found or communicated with.” Webster’s Third New International Dictionary 25 (1971) (Webster’s Third); see also Webster’s Second New International Dictionary 30 (1957) (“the name or description of a place of residence, business, etc., where a person may be found or communicated with”); Random House Dictionary of the English Language 17 (1966) (“the place or the name of the place where a person, organization, or the like is located or may be reached”); American Heritage Dictionary 15 (1969) (“[t]he location at which a particular organization or person may be found or reached”); Oxford English Dictionary 106 (1933) (OED) (“the name of the place to which any one’s letters are directed”). Since a foreign nation’s embassy in the United States is neither the residence nor the usual place of business of that nation’s foreign minister and is not a place where the minister can customarily be found, the most common understanding of the minister’s “address” is inconsistent with the interpretation of §1608(a)(3) adopted by the court below and advanced by respondents. We acknowledge that there are circumstances in which a mailing may be “addressed” to the intended recipient at a place other than the individual’s residence or usual place of business. For example, if the person sending the mailing does not know the intended recipient’s current home or business address, the sender might use the intended recipient’s last known address in the hope that the mailing will be forwarded. Or a sender might send a mailing to a third party who is thought to be in a position to ensure that the mailing is ultimately received by the intended recipient. But in the great majority of cases, addressing a mailing to X means placing on the outside of the mailing both X’s name and the address of X’s residence or customary place of work. Section 1608(a)(3)’s use of the term “dispatched” points in the same direction. To “dispatch” a communication means “to send [it] off or away (as to a special destination) with promptness or speed often as a matter of official business.” Webster’s Third 653; see also OED 478 (“To send off post-haste or with expedition or promptitude (a messenger, message, etc., having an express destination)”). A person who wishes to “dispatch” a letter to X will generally send it directly to X at a place where X is customarily found. The sender will not “dispatch” the letter in a roundabout way, such as by directing it to a third party who, it is hoped, will then send it on to the intended recipient. A few examples illustrate this point. Suppose that a person is instructed to “address” a letter to the Attorney General of the United States and “dispatch” the letter (i.e., to “send [it] off post-haste”) to the Attorney General. The person giving these instructions would likely be disappointed and probably annoyed to learn that the letter had been sent to, let us say, the office of the United States Attorney for the District of Idaho. And this would be so even though a U. S. Attorney’s office is part of the Department headed by the Attorney General and even though such an office would very probably forward the letter to the Attorney General’s office in Washington. Similarly, a person who instructs a subordinate to dispatch a letter to the CEO of a big corporation that owns retail outlets throughout the country would probably be irritated to learn that the letter had been mailed to one of those stores instead of corporate headquarters. To “dispatch” a letter to an addressee connotes sending it directly. A similar understanding underlies the venerable “mailbox rule.” As first-year law students learn in their course on contracts, there is a presumption that a mailed acceptance of an offer is deemed operative when “dispatched” if it is “properly addressed.” Restatement (Second) of Contracts § 66, p. 161 (1979) (Restatement); Rosenthal v. Walker, 111 U.S. 185, 193 (1884). But no acceptance would be deemed properly addressed and dispatched if it lacked, and thus was not sent to, the offer- or’s address (or an address that the offeror held out as the place for receipt of an acceptance). See Restatement § 66, Comment b. It is also significant that service under §1608(a)(3) requires a signed returned receipt, a standard method for ensuring delivery to the addressee. Cf. Black’s Law Dictionary 1096 (10th ed. 2014) (defining “certified mail” as “[m]ail for which the sender requests proof of delivery in the form of a receipt signed by the addressee”). We assume that certified mail sent to a foreign minister will generally be signed for by a subordinate, but the person who signs for the minister’s certified mail in the foreign ministry itself presumably has authority to receive mail on the minister’s behalf and has been instructed on how that mail is to be handled. The same is much less likely to be true for an employee in the mailroom of an embassy. For all these reasons, we think that the most natural reading of §1608(a)(3) is that the service packet must bear the foreign minister’s name and customary address and that it be sent to the minister in a direct and expeditious way. And the minister’s customary office is the place where he or she generally works, not a farflung outpost that the minister may at most occasionally visit. B Several related provisions in §1608 support this reading. See Davis v. Michigan Dept. of Treasury, 489 U.S. 803, 809 (1989) (“It is a fundamental canon of statutory construction that the words of a statute must be read in their context and with a view to their place in the overall statutory scheme”). 1 One such provision is §1608(b)(3)(B). Section 1608(b) governs service on “an agency or instrumentality of a foreign state.” And like §1608(a)(3), §1608(b)(3)(B) requires delivery of a service packet to the intended recipient “by any form of mail requiring a signed receipt, to be addressed and dispatched by the clerk of the court.” But §1608(b)(3)(B), unlike §1608(a)(3), contains prefatory language saying that service by this method is permissible “if reasonably calculated to give actual notice.” Respondents read §1608(a)(3) as embodying a similar requirement. See Brief for Respondents 34. At oral argument, respondents’ counsel stressed this point, arguing that respondents’ interpretation of §1608(a)(3) “gives effect” to the “familiar” due process standard articulated in Mullane v. Central Hanover Bank & Trust Co., 339 U.S. 306 (1950), which is “the notion that [service] must be reasonably calculated to give notice.” Tr. of Oral Arg. 37–38. This argument runs up against two well-settled principles of statutory interpretation. First, “Congress generally acts intentionally when it uses particular language in one section of a statute but omits it in another.” Department of Homeland Security v. MacLean, 574 U. S. ___, ___ (2015) (slip op., at 7). Because Congress included the “reasonably calculated to give actual notice” language only in §1608(b), and not in §1608(a), we resist the suggestion to read that language into §1608(a). Second, “we are hesitant to adopt an interpretation of a congressional enactment which renders superfluous another portion of that same law.” Mackey v. Lanier Collection Agency & Service, Inc., 486 U.S. 825, 837 (1988). Here, respondents encounter a superfluity problem when they argue that the “addressed and dispatched” clause in §1608(a)(3) gives effect to the Mullane due process standard. They fail to account for the fact that §1608(b)(3)(B) contains both the “addressed and dispatched” and “reasonably calculated to give actual notice” requirements. If respondents were correct that “addressed and dispatched” means “reason- ably calculated to give notice,” then the phrase “reasonably calculated to give actual notice” in §1608(b)(3) would be superfluous. Thus, as the dissent agrees, §1608(a)(3) “does not deem a foreign state properly served solely because the service method is reasonably calculated to provide actual notice.” Post, at 2 (opinion of Thomas, J.). 2 Section 1608(b)(2) similarly supports our interpretation of §1608(a)(3). Section 1608(b)(2) provides for delivery of a service packet to an officer or a managing or general agent of the agency or instrumentality of a foreign state or “to any other agent authorized by appointment or by law to receive service of process in the United States.” This language is significant for three reasons. First, it expressly allows service on an agent. Second, it specifies the particular individuals who are permitted to be served as agents of the recipient. Third, it makes clear that service on the agent may occur in the United States if an agent here falls within the provision’s terms. If Congress had contemplated anything similar under §1608(a)(3), there is no apparent reason why it would not have included in that provision terms similar to those in §1608(b)(2). Respondents would have us believe that Congress was content to have the courts read such terms into §1608(a)(3). In view of §1608(b)(2), this seems un- likely.[2] See also post, at 2 (“Nor does the FSIA authorize service on a foreign state by utilizing an agent designated to receive process for the state”). 3 Section 1608(c) further buttresses our reading of §1608(a)(3). Section 1608(c) sets out the rules for determining when service “shall be deemed to have been made.” For the first three methods of service under §1608(a), service is deemed to have occurred on the date indicated on “the certification, signed and returned postal receipt, or other proof of service applicable to the method of service employed.” §1608(c)(2). The sole exception is service under §1608(a)(4), which requires the Secretary of State to transmit a service packet to the foreign state through diplomatic channels. Under this method, once the Secretary has transmitted the packet, the Secretary must send to the clerk of the court “a certified copy of the diplomatic note indicating when the papers were transmitted.” §1608(a)(4). And when service is effected in this way, service is regarded as having occurred on the transmittal date shown on the certified copy of the diplomatic note. §1608(c)(1). Under all these methods, service is deemed to have occurred only when there is a strong basis for concluding that the service packet will very shortly thereafter come into the hands of a foreign official who will know what needs to be done. Under §1608(a)(4), where service is transmitted by the Secretary of State through diplomatic channels, there is presumably good reason to believe that the service packet will quickly come to the attention of a high-level foreign official, and thus service is regarded as having been completed on the date of transmittal. And under §§1608(a)(1), (2), and (3), where service is deemed to have occurred on the date shown on a document signed by the person who received it from the carrier, Congress presumably thought that the individuals who signed for the service packet could be trusted to ensure that the service packet is handled properly and expeditiously. It is easy to see why Congress could take that view with respect to a person designated for the receipt of process in a “special arrangement for service between the plaintiff and the foreign state or political subdivision,” §1608(a)(1), and a person so designated under “an applicable international convention,” §1608(a)(2). But what about §1608(a)(3), the provision now before us? Who is more comparable to those who sign for mail under §§1608(a)(1) and (2)? A person who works in the office of the foreign minister in the minister’s home country and is authorized to receive and process the minister’s mail? Or a mailroom employee in a foreign embassy? We think the answer is obvious, and therefore interpreting §1608(a)(3) to require that a service packet be sent to a foreign minister’s own office better harmonizes the rules for determining when service is deemed to have been made. Respondents seek to soften the blow of an untimely delivery to the minister by noting that the foreign state can try to vacate a default judgment under Federal Rule of Civil Procedure 55(c). Brief for Respondents 27. But that is a poor substitute for sure and timely receipt of service, since a foreign state would have to show “good cause” to vacate the judgment under that Rule. Here, as with the previously mentioned provisions in §1608, giving §1608(a)(3) its ordinary meaning better harmonizes the various provisions in §1608 and avoids the oddities that respondents’ interpretation would create. C The ordinary meaning of the “addressed and dispatched” requirement in §1608(a)(3) also has the virtue of avoiding potential tension with the Federal Rules of Civil Procedure and the Vienna Convention on Diplomatic Relations. 1 Take the Federal Rules of Civil Procedure first. At the time of the FSIA’s enactment, Rule 4(i), entitled “Alternative provisions for service in a foreign-country,” set out certain permissible methods of service on “part[ies] in a foreign country.” Fed. Rule Civ. Proc. 4(i)(1) (1976). One such method was “by any form of mail, requiring a signed receipt, to be addressed and dispatched by the clerk of the court to the party to be served.” Rule 4(i)(1)(D) (emphasis added). Rule 4(i)(2) further provided that “proof of service” pursuant to that method “shall include a receipt signed by the addressee or other evidence of delivery to the addressee satisfactory to the court.” (Emphasis added.) The current version of Rule 4 is similar. See Rules 4(f )(2)(C)(ii), 4(l)(2)(B). The virtually identical methods of service outlined in Rule 4 and §1608(a)(3) pose a problem for respondents’ position: If mailing a service packet to a foreign state’s embassy in the United States were sufficient for purposes of §1608(a)(3), then it would appear to be easier to serve the foreign state than to serve a person in that foreign state. This is so because a receipt signed by an embassy employee would not necessarily satisfy Rule 4 since such a receipt would not bear the signature of the foreign minister and might not constitute evidence that is sufficient to show that the service packet had actually been delivered to the minister. It would be an odd state of affairs for a foreign state’s inhabitants to enjoy more protections in federal courts than the foreign state itself, particularly given that the foreign state’s immunity from suit is at stake. The natural reading of §1608(a)(3) avoids that oddity. 2 Our interpretation of §1608(a)(3) avoids concerns regarding the United States’ obligations under the Vienna Convention on Diplomatic Relations. We have previously noted that the State Department “helped to draft the FSIA’s language,” and we therefore pay “special attention” to the Department’s views on sovereign immunity. Bolivarian Republic of Venezuela v. Helmerich & Payne Int’l Drilling Co., 581 U. S. ___, ___ (2017) (slip op., at 9). It is also “well settled that the Executive Branch’s interpretation of a treaty ‘is entitled to great weight.’ ” Abbott v. Abbott, 560 U.S. 1, 15 (2010) (quoting Sumitomo Shoji America, Inc. v. Avagliano, 457 U.S. 176, 185 (1982)). Article 22(1) of the Vienna Convention provides: “The premises of the mission shall be inviolable. The agents of the receiving State may not enter them, except with the consent of the head of the mission.” Vienna Convention on Diplomatic Relations, Apr. 18, 1961, 23 U. S. T. 3237, T. I. A. S. No. 7502. Since at least 1974, the State Department has taken the position that Article 22(1)’s principle of inviolability precludes serving a foreign state by mailing process to the foreign state’s embassy in the United States. See Service of Legal Process by Mail on Foreign Governments in the United States, 71 Dept. State Bull. 458–459 (1974). In this case, the State Department has reiterated this view in amicus curiae briefs filed in this Court and in the Second Circuit. The Government also informs us that United States embassies do not accept service of process when the United States is sued in a foreign court, and the Government expresses concern that accepting respondents’ interpretation of §1608 might imperil this practice. Brief for United States as Amicus Curiae 25–26. Contending that the State Department held a different view of Article 22(1) before 1974, respondents argue that the Department’s interpretation of the Vienna Convention is wrong, but we need not decide this question. By giving §1608(a)(3) its most natural reading, we avoid the potential international implications of a contrary interpretation. III Respondents’ remaining arguments do not alter our conclusion. First, respondents contend that §1608(a)(3) says nothing about where the service packet must be sent. See Brief for Respondents 22 (“the statute is silent as to the location where the service packet should be sent”). But while it is true that §1608(a)(3) does not expressly provide where service must be sent, it is common ground that this provision must implicitly impose some requirement. Respondents acknowledge this when they argue that the provision demands that service be sent “to a location that is likely to have a direct line of communication to the foreign minister.” Id., at 34; cf. post, at 6 (stating that sending a letter to a Washington-based embassy “with a direct line of communication” to the foreign minister seems as efficient as sending it to the minister’s office in the foreign state). The question, then, is precisely what §1608(a)(3) implicitly requires. Respondents assure us that a packet sent to “an embassy plainly would qualify,” while a packet sent to “a tourism office plainly would not.” Brief for Respondents 34. But if the test is whether “a location . . . is likely to have a direct line of communication to the foreign minister,” ibid., it is not at all clear why service could not be sent to places in the United States other than a foreign state’s embassy. Why not allow the packet to be sent, for example, to a consulate? The residence of the foreign state’s ambassador? The foreign state’s mission to the United Nations? Would the answer depend on the size or presumed expertise of the staff at the delivery location? The difficult line-drawing problems that flow from respondents’ interpretation of §1608(a)(3) counsel in favor of maintaining a clear, administrable rule: The service packet must be mailed directly to the foreign minister at the minister’s office in the foreign state. Second, respondents (and the dissent, see post, at 5–6) contrast the language of §1608(a)(3) with that of §1608(a)(4), which says that service by this method requires that process be sent to the Secretary of State in “Washington, District of Columbia.” If Congress wanted to require that process under §1608(a)(3) be sent to a foreign minister’s office in the minister’s home country, respondents ask, why didn’t Congress use a formulation similar to that in §1608(a)(4)? This is respondents’ strongest argument, and in the end, we see no entirely satisfactory response other than that §1608(a) does not represent an example of perfect draftsmanship. We grant that the argument based on the contrasting language in §1608(a)(4) cuts in respondents’ favor, but it is outweighed in our judgment by the countervailing arguments already noted. Finally, respondents contend that it would be “the height of unfairness to throw out [their] judgment” based on the highly technical argument belatedly raised by petitioner. See Brief for Respondents 35. We understand respondents’ exasperation and recognize that enforcing compliance with §1608(a)(3) may seem like an empty formality in this particular case, which involves highly publicized litigation of which the Government of Sudan may have been aware prior to entry of default judgment. But there are circumstances in which the rule of law demands adherence to strict requirements even when the equities of a particular case may seem to point in the opposite direction. The service rules set out in §1608(a)(3), which apply to a category of cases with sensitive diplo- matic implications, clearly fall into this category. Under those rules, all cases must be treated the same. Moreover, as respondents’ counsel acknowledged at oral argument, holding that Sudan was not properly served under §1608(a)(3) is not the end of the road. Tr. of Oral Arg. 56. Respondents may attempt service once again under §1608(a)(3), and if that attempt fails, they may turn to §1608(a)(4). When asked at argument to provide examples of any problems with service under §1608(a)(4), respondents’ counsel stated that he was unaware of any cases where such service failed. Id., at 59–62. * * * We interpret §1608(a)(3) as it is most naturally understood: A service packet must be addressed and dispatched to the foreign minister at the minister’s office in the foreign state. We therefore reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 Sudan questions whether respondents named the correct foreign minister and whether the Sudanese Embassy received the service packet. Because we find the service deficient in any event, we assume for the sake of argument that the correct name was used and that the Embassy did receive the packet. 2 Notably, the idea of treating someone at a foreign state’s embassy as an agent for purposes of service on the foreign state was not unfamiliar to Congress. An earlier proposed version of the FSIA would have permitted service on a foreign state by sending the service packet “to the ambassador or chief of mission of the foreign state.” See S. 566, 93d Cong., 1st Sess., §1608, p. 6 (1973).

587.US.2018_17-1594

The Leahy-Smith America Invents Act (AIA) of 2011 created the Patent Trial and Appeal Board, 35 U. S. C. §6(c), and established three types of administrative review proceedings before the Board that enable a “person” other than the patent owner to challenge the validity of a patent post-issuance: (1) “inter partes review,” §311; (2) “post-grant review,” §321; and (3) “covered-business-method review” (CBM review), note following §321. After an adjudicatory proceeding, the Board either confirms the patent claims or cancels some or all of them, §§318(b), 328(b). Any “dissatisfied” party may then seek judicial review in the Federal Circuit, §§319, 329. In addition to AIA review proceedings, a patent can be reexamined either in federal court during a defense to an infringement suit, §282(b), or in an ex parte reexamination by the Patent Office, §§301(a), 302(a). Return Mail, Inc., owns a patent that claims a method for processing undeliverable mail. The Postal Service subsequently introduced an enhanced address-change service to process undeliverable mail, which Return Mail asserted infringed its patent. The Postal Service petitioned for ex parte reexamination of the patent, but the Patent Office confirmed the patent’s validity. Return Mail then sued the Postal Service in the Court of Federal Claims, seeking compensation for the unauthorized use of its invention. While that suit was pending, the Postal Service petitioned for CBM review. The Patent Board concluded that the subject matter of Return Mail’s claims was ineligible to be patented and thus canceled the claims underlying its patent. The Federal Circuit affirmed, concluding, as relevant here, that the Government is a “person” eligible to petition for CBM review. Held: The Government is not a “person” capable of instituting the three AIA review proceedings. Pp. 6–18. (a) In the absence of an express definition of the term “person” in the patent statutes, the Court applies a “longstanding interpretive presumption that ‘person’ does not include the sovereign,” and thus excludes a federal agency like the Postal Service. Vermont Agency of Natural Resources v. United States ex rel. Stevens, 529 U.S. 765, 780–781. This presumption reflects “common usage,” United States v. Mine Workers, 330 U.S. 258, 275, as well as an express directive from Congress in the Dictionary Act, 1 U. S. C. §1. The Dictionary Act does not include the Federal Government among the persons listed in the definition of “person” that courts use “[i]n determining the meaning of any Act of Congress, unless the context indicates otherwise,” §1. Contrary to the Postal Service’s contention otherwise, this Court has, in several instances, applied the presumption against treating the Government as a statutory person even when, as here, doing so would exclude the Government or one of its agencies from accessing a benefit or favorable procedural device. See, e.g., United States v. Cooper Corp., 312 U.S. 600, 604–605, 614. Thus, the Court here proceeds from the presumption that the Government is not a “person” authorized to initiate these proceedings absent an affirmative showing to the contrary. Pp. 7–9. (b) The Postal Service must point to some indication in the AIA’s text or context affirmatively showing that Congress intended to include the Government as a “person,” but its arguments are unpersuasive. Pp. 9–17. (1) The Postal Service first argues that the AIA’s reference to a “person” in the context of post-issuance review proceedings must include the Government because other references to persons in the patent statutes appear to do so. The consistent-usage principle—i.e., when Congress uses a word to mean one thing in one part of the statute, it will mean the same thing elsewhere in the statute—however, “ ‘ readily yields to context, ’ ” especially when a statutory term is used throughout a statute and takes on “distinct characters” in distinct statutory provisions. Utility Air Regulatory Group v. EPA, 573 U.S. 302, 320. Here, where there are at least 18 references to “person[s]” in the Patent Act and the AIA, no clear trend is shown: Sometimes “person” plainly includes or excludes the Government, but sometimes, as here, it might be read either way. The mere existence of some Government-inclusive references cannot make the “affirmative showing,” Stevens, 529 U. S., at 781, required to overcome the presumption that the Government is not a “person” eligible to petition for AIA review proceedings. Pp. 9–12. (2) The Postal Service next points to the Federal Government’s longstanding history with the patent system, arguing that because federal officers have been able to apply for patents in the name of the United States since 1883, Congress must have intended to allow the Government access to AIA review proceedings. But the Government’s ability to obtain a patent does not speak to whether Congress meant for the Government to participate as a third-party challenger in AIA proceedings established only eight years ago. Moreover, even assuming that the Government may petition for ex parte reexamination of an issued patent, as a 1981 Patent Office Manual of Patent Examining Procedure (MPEP) indicates, an ex parte reexamination process is fundamentally different from an AIA review proceeding. The former process is internal, and the party challenging the patent may not participate. By contrast, adversarial, adjudicatory AIA review proceedings are between the “person” who petitioned for review and the patent owner; they include briefing, a hearing, discovery, and the presentation of evidence; and the losing party has appeal rights. Congress may have had good reason to authorize the Government to initiate a hands-off ex parte reexamination but not to become a party to the AIA’s full-blown adversarial proceeding. Nothing suggests that Congress had the 1981 MPEP statement about ex parte reexamination in mind when it created the AIA review proceedings. And because there is no “settled” meaning of the term “person” with respect to the newly established AIA review proceedings, see Bragdon v. Abbott, 524 U.S. 624, 645, the MPEP does not justify putting aside the presumptive meaning of “person.” Pp. 13–15. (3) Finally, the Postal Service argues that it must be a “person” who may petition for AIA review proceedings because, like other potential infringers, it is subject to civil liability and can assert a defense of patent invalidity. It would thus be anomalous, the Postal Service posits, to deny it a benefit afforded to other infringers—namely, the ability to challenge a patent de novo before the Patent Office, rather than only with clear and convincing evidence in defense to an infringement suit. Federal agencies, however, face lower and more calculable risks than nongovernmental actors, so it is reasonable for Congress to have treated them differently. Excluding federal agencies from AIA review proceedings also avoids the awkward situation of having a civilian patent owner defend the patentability of her invention in an adversarial, adjudicatory proceeding initiated by one federal agency and overseen by a different federal agency. Pp. 15–17. 868 F.3d 1350, reversed and remanded. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Alito, Gorsuch, and Kavanaugh, JJ., joined. Breyer, J., filed a dissenting opinion, in which Ginsburg and Kagan, JJ., joined.

In the Leahy-Smith America Invents Act of 2011, 35 U. S. C. §100 et seq., Congress created the Patent Trial and Appeal Board and established three new types of administrative proceedings before the Board that allow a “person” other than the patent owner to challenge the validity of a patent post-issuance. The question presented in this case is whether a federal agency is a “person” able to seek such review under the statute. We conclude that it is not. I A The Constitution empowers Congress “[t]o promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective . . . Discoveries.” Art. I, §8, cl. 8. Pursuant to that authority, Congress established the United States Patent and Trademark Office (Patent Office) and tasked it with “the granting and issuing of patents.” 35 U. S. C. §§1, 2(a)(1). To obtain a patent, an inventor submits an application describing the proposed patent claims to the Patent Office. See §§111(a)(1), 112. A patent examiner then reviews the application and prior art (the information available to the public at the time of the application) to determine whether the claims satisfy the statutory requirements for patent- ability, including that the claimed invention is useful, novel, nonobvious, and contains eligible subject matter. See §§101, 102, 103. If the Patent Office accepts the claim and issues a patent, the patent owner generally obtains exclusive rights to the patented invention throughout the United States for 20 years. §§154(a)(1), (2). After a patent issues, there are several avenues by which its validity can be revisited. The first is through a defense in an infringement action. Generally, one who intrudes upon a patent without authorization “infringes the patent” and becomes subject to civil suit in the federal district courts, where the patent owner may demand a jury trial and seek monetary damages and injunctive relief. §§271(a), 281–284. If, however, the Federal Government is the alleged patent infringer, the patent owner must sue the Government in the United States Court of Federal Claims and may recover only “reasonable and entire compensation” for the unauthorized use. 28 U. S. C. §1498(a). Once sued, an accused infringer can attempt to prove by clear and convincing evidence “that the patent never should have issued in the first place.” Microsoft Corp. v. i4i L. P., 564 U.S. 91, 96–97 (2011); see 35 U. S. C. §282(b). If a defendant succeeds in showing that the claimed invention falls short of one or more patentability requirements, the court may deem the patent invalid and absolve the defendant of liability. The Patent Office may also reconsider the validity of issued patents. Since 1980, the Patent Act has empowered the Patent Office “to reexamine—and perhaps cancel—a patent claim that it had previously allowed.” Cuozzo Speed Technologies, LLC v. Lee, 579 U. S. ___, ___ (2016) (slip op., at 3). This procedure is known as ex parte reexamination. “Any person at any time” may cite to the Patent Office certain prior art that may “bea[r] on the patentability of any claim of a particular patent”; and the person may additionally request that the Patent Office reexamine the claim on that basis. 35 U. S. C. §§301(a), 302(a). If the Patent Office concludes that the prior art raises “a substantial new question of patentability,” the agency may reexamine the patent and, if warranted, cancel the patent or some of its claims. §§303(a), 304–307. The Director of the Patent Office may also, on her “own initiative,” initiate such a proceeding. §303(a). In 1999 and 2002, Congress added an “inter partes reexamination” procedure, which similarly invited “[a]ny person at any time” to seek reexamination of a patent on the basis of prior art and allowed the challenger to participate in the administrative proceedings and any subsequent appeal. See §311(a) (2000 ed.); §§314(a), (b) (2006 ed.); Cuozzo Speed Technologies, 579 U. S., at ___ (slip op., at 3). B In 2011, Congress overhauled the patent system by enacting the America Invents Act (AIA), which created the Patent Trial and Appeal Board and phased out inter partes reexamination. See 35 U. S. C. §6; H. R. Rep. No. 112–98, pt. 1, pp. 46–47. In its stead, the AIA tasked the Board with overseeing three new types of post-issuance review proceedings. First, the “inter partes review” provision permits “a person” other than the patent owner to petition for the review and cancellation of a patent on the grounds that the invention lacks novelty or nonobviousness in light of “patents or printed publications” existing at the time of the patent application. §311. Second, the “post-grant review” provision permits “a person who is not the owner of a patent” to petition for review and cancellation of a patent on any ground of pat- entability. §321; see §§282(b)(2), (b)(3). Such proceedings must be brought within nine months of the patent’s issuance. §321. Third, the “covered-business-method review” (CBM review) provision provides for changes to a patent that claims a method for performing data processing or other operations used in the practice or management of a financial product or service. AIA §§18(a)(1), (d)(1), 125Stat. 329, note following 35 U. S. C. §321, p. 1442. CBM review tracks the “standards and procedures of” post-grant review with two notable exceptions: CBM review is not limited to the nine months following issuance of a patent, and “[a] person” may file for CBM review only as a defense against a charge or suit for infringement. §18(a)(1)(B), 125Stat. 330.[1] The AIA’s three post-issuance review proceedings are adjudicatory in nature. Review is conducted by a three-member panel of the Patent Trial and Appeal Board, 35 U. S. C. §6(c), and the patent owner and challenger may seek discovery, file affidavits and other written memo- randa, and request an oral hearing, see §§316, 326; AIA §18(a)(1), 125Stat. 329; Oil States Energy Services, LLC v. Greene’s Energy Group, LLC, 584 U. S. ___, ___–___ (2018) (slip op., at 3–4). The petitioner has the burden of proving unpatentability by a preponderance of the evidence. §§282, 316(e), 326(e). The Board then either confirms the patent claims or cancels some or all of the claims. §§318(b), 328(b). Any party “dissatisfied” with the Board’s final decision may seek judicial review in the Court of Appeals for the Federal Circuit, §§319, 329; see §141(c), and the Director of the Patent Office may intervene, §143. In sum, in the post-AIA world, a patent can be reexamined either in federal court during a defense to an infringement action, in an ex parte reexamination by the Patent Office, or in the suite of three post-issuance review proceedings before the Patent Trial and Appeal Board. The central question in this case is whether the Federal Government can avail itself of the three post-issuance review proceedings, including CBM review. C Return Mail, Inc., owns U. S. Patent No. 6,826,548 (’548 patent), which claims a method for processing mail that is undeliverable. Beginning in 2003, the United States Postal Service allegedly began exploring the possibility of licensing Return Mail’s invention for use in handling the country’s undelivered mail. But the parties never reached an agreement. In 2006, the Postal Service introduced an enhanced address-change service to process undeliverable mail. Return Mail’s representatives asserted that the new service infringed the ’548 patent, and the company renewed its offer to license the claimed invention to the Postal Service. In response, the Postal Service petitioned for ex parte reexamination of the ’548 patent. The Patent Office canceled the original claims but issued several new ones, confirming the validity of the ’548 patent. Return Mail then sued the Postal Service in the Court of Federal Claims, seeking compensation for the Postal Service’s unauthorized use of its invention, as reissued by the Pat- ent Office. While the lawsuit was pending, the Postal Service again petitioned the Patent Office to review the ’548 patent, this time seeking CBM review. The Patent Board instituted review. The Board agreed with the Postal Service that Return Mail’s patent claims subject matter that was ineligible to be patented, and it canceled the claims underlying the ’548 patent. A divided panel of the Court of Appeals for the Federal Circuit affirmed. See 868 F.3d 1350 (2017). As relevant here, the Federal Circuit held, over a dissent, that the Government is a “person” eligible to petition for CBM review. Id., at 1366; see AIA §18(a)(1)(B), 125Stat. 330 (only a qualifying “person” may petition for CBM review). The court then affirmed the Patent Board’s decision on the merits, invalidating Return Mail’s patent claims. We granted certiorari to determine whether a federal agency is a “person” capable of petitioning for post-issuance review under the AIA.[2] 586 U. S. ___ (2018). II The AIA provides that only “a person” other than the patent owner may file with the Office a petition to institute a post-grant review or inter partes review of an issued patent. 35 U. S. C. §§311(a), 321(a). The statute likewise provides that a “person” eligible to seek CBM review may not do so “unless the person or the person’s real party in interest or privy has been sued for infringement.” AIA §18(a)(1)(B), 125Stat. 330. The question in this case is whether the Government is a “person” capable of instituting the three AIA review proceedings. A The patent statutes do not define the term “person.” In the absence of an express statutory definition, the Court applies a “longstanding interpretive presumption that ‘person’ does not include the sovereign,” and thus excludes a federal agency like the Postal Service. Vermont Agency of Natural Resources v. United States ex rel. Stevens, 529 U.S. 765, 780–781 (2000); see United States v. Mine Workers, 330 U.S. 258, 275 (1947); United States v. Cooper Corp., 312 U.S. 600, 603–605 (1941); United States v. Fox, 94 U.S. 315, 321 (1877). This presumption reflects “common usage.” Mine Workers, 330 U. S., at 275. It is also an express directive from Congress: The Dictionary Act has since 1947 provided the definition of “ ‘person’ ” that courts use “[i]n determining the meaning of any Act of Congress, unless the context indicates otherwise.” 1 U. S. C. §1; see Rowland v. California Men’s Colony, Unit II Men’s Advisory Council, 506 U.S. 194, 199–200 (1993). The Act provides that the word “ ‘person’ . . . include[s] corporations, companies, associations, firms, partnerships, societies, and joint stock companies, as well as individuals.” §1. Notably absent from the list of “person[s]” is the Federal Government. See Mine Workers, 330 U. S., at 275 (reasoning that Congress’ express inclusion of partnerships and corporations in §1 implies that Congress did not intend to include the Government). Thus, although the presumption is not a “hard and fast rule of exclusion,” Cooper, 312 U. S., at 604–605, “it may be disregarded only upon some affirmative showing of statutory intent to the contrary,” Stevens, 529 U. S., at 781. The Postal Service contends that the presumption is strongest where interpreting the word “person” to include the Government imposes liability on the Government, and is weakest where (as here) interpreting “person” in that way benefits the Government. In support of this argument, the Postal Service points to a different interpretive canon: that Congress must unequivocally express any waiver of sovereign immunity for that waiver to be effective. See FAA v. Cooper, 566 U.S. 284, 290 (2012). That clear-statement rule inherently applies only when a party seeks to hold the Government liable for its actions; otherwise immunity is generally irrelevant. In the Postal Service’s view, the presumption against treating the Government as a statutory person works in tandem with the clear-statement rule regarding immunity, such that both apply only when a statute would subject the Government to liability. Our precedents teach otherwise. In several instances, this Court has applied the presumption against treating the Government as a statutory person when there was no question of immunity, and doing so would instead exclude the Federal Government or one of its agencies from accessing a benefit or favorable procedural device. In Cooper, 312 U. S., at 604–605, 614, for example, the Court held that the Federal Government was not “ ‘[a]ny person’ ” who could sue for treble damages under §7 of the Sherman Anti-Trust Act. Accord, International Primate Protection League v. Administrators of Tulane Ed. Fund, 500 U.S. 72, 82–84 (1991) (concluding that the National Institutes of Health was not authorized to remove an action as a “ ‘person acting under [a federal]’ officer” pursuant to 28 U. S. C. §1442(a)(1)); Davis v. Pringle, 268 U.S. 315, 317–318 (1925) (reasoning that “normal usages of speech” indicated that the Government was not a “person” entitled to priority under the Bankruptcy Act); Fox, 94 U. S., at 321 (holding that the Federal Government was not a “ ‘person capable by law of holding real estate,’ ” absent “an express definition to that effect”). Thus, although the presumption against treating the Government as a statutory person is “ ‘particularly applicable where it is claimed that Congress has subjected the [sovereign] to liability to which they had not been subject before,’ ” Stevens, 529 U. S., at 781, it is hardly confined to such cases. Here, too, we proceed from the presumption that the Government is not a “person” authorized to initiate these proceedings absent an affirmative showing to the contrary. B Given the presumption that a statutory reference to a “person” does not include the Government, the Postal Service must show that the AIA’s context indicates otherwise. Although the Postal Service need not cite to “an express contrary definition,” Rowland, 506 U. S., at 200, it must point to some indication in the text or context of the statute that affirmatively shows Congress intended to include the Government. See Cooper, 312 U. S., at 605. The Postal Service makes three arguments for displacing the presumption. First, the Postal Service argues that the statutory text and context offer sufficient evidence that the Government is a “person” with the power to petition for AIA review proceedings. Second, the Postal Service contends that federal agencies’ long history of participation in the patent system suggests that Congress intended for the Government to participate in AIA review proceedings as well. Third, the Postal Service maintains that the statute must permit it to petition for AIA review because §1498 subjects the Government to liability for infringement. None delivers. 1 The Postal Service first argues that the AIA’s reference to a “person” in the context of post-issuance review proceedings must include the Government because other references to persons in the patent statutes appear to do so. Indeed, it is often true that when Congress uses a word to mean one thing in one part of the statute, it will mean the same thing elsewhere in the statute. See Merrill Lynch, Pierce, Fenner & Smith Inc. v. Dabit, 547 U.S. 71, 86 (2006). This principle, however, “readily yields to context,” especially when a statutory term is used throughout a statute and takes on “distinct characters” in distinct statutory provisions. See Utility Air Regulatory Group v. EPA, 573 U.S. 302, 320 (2014) (internal quotation marks omitted). That is the case here. The Patent Act and the AIA refer to “person[s]” in at least 18 different places, and there is no clear trend: Sometimes “person” plainly includes the Government,[3] sometimes it plainly excludes the Government,[4] and sometimes—as here—it might be read either way. Looking on the bright side, the Postal Service and the dissent, see post, at 2, focus on §207(a)(1), which authorizes “[e]ach [f]ederal agency” to “apply for, obtain, and maintain patents or other forms of protection . . . on inventions in which the Federal Government owns a right, title, or interest.” It follows from §207(a)(1)’s express inclusion of federal agencies among those eligible to apply for patents that the statute’s references to “person[s]” in the subsections governing the patent-application process and questions of patentability (§§102(a), 118, and 119) must also include federal agencies.[5] In other words, the right described in §207(a)(1) provides a sufficient contextual clue that the word “person”—when used in the other provisions governing the application process §207(a)(1) makes available to federal agencies—includes the Government. But §207(a)(1) provides no such clue as to the interpretation of the AIA review provisions because it implies nothing about what a federal agency may or may not do following the issuance of someone else’s patent. Conversely, reading the review provisions to exclude the Govern- ment has no bearing on a federal agency’s right to obtain a patent under §207(a)(1). An agency may still apply for and obtain patents whether or not it may petition for a review proceeding under the AIA seeking cancellation of a patent it does not own. There is thus no reason to think that “person” must mean the same thing in these two different parts of the statute. See Utility Air, 573 U. S., at 320.[6] The Postal Service cites other provisions that may refer to the Government—namely, the “intervening rights” provisions that offer certain protections for “any person” who is lawfully making or using an invention when the Patent Office modifies an existing patent claim in a way that deems the person’s (previously lawful) use to be infringement. See §§252, 307(b), 318(c), 328(c). The Postal Service argues that the Government must be among those protected by these provisions and from there deduces that it must also be permitted to petition for AIA review proceedings because the review provisions and the intervening-rights provisions were all added to the Patent Act by the AIA at the same time. See Powerex Corp. v. Reliant Energy Services, Inc., 551 U.S. 224, 232 (2007) (invoking the consistent-usage canon where the same term was used in related provisions enacted at the same time). But regardless of whether the intervening-rights provisions apply to the Government (a separate interpretive question that we have no occasion to answer here), the Postal Service’s chain of inferences overlooks a confounding link: The consistent-usage canon breaks down where Congress uses the same word in a statute in multiple conflicting ways. As noted, that is the case here. In the face of such inconsistency, the mere existence of some Government-inclusive references cannot make the “affirmative showing,” Stevens, 529 U. S., at 781, required to overcome the presumption that Congress did not intend to include the Government among those “person[s]” eligible to petition for AIA review proceedings.[7] 2 The Postal Service next points to the Federal Government’s longstanding history with the patent system. It reminds us that federal officers have been able to apply for patents in the name of the United States since 1883, see Act of Mar. 3, 1883, 22Stat. 625—which, in the Postal Service’s view, suggests that Congress intended to allow the Government access to AIA review proceedings as well. But, as already explained, the Government’s ability to obtain a patent under §207(a)(1) does not speak to whether Congress meant for the Government to participate as a third-party challenger in AIA review proceedings. As to those proceedings, there is no longstanding practice: The AIA was enacted just eight years ago.[8] More pertinently, the Postal Service and the dissent both note that the Patent Office since 1981 has treated federal agencies as “persons” who may cite prior art to the agency or request an ex parte reexamination of an issued patent. See post, at 5. Recall that §301(a) provides that “[a]ny person at any time may cite to the Office in writing . . . prior art . . . which that person believes to have a bearing on the patentability of any claim of a particular patent.” As memorialized in the Patent Office’s Manual of Patent Examining Procedure (MPEP), the agency has understood §301’s reference to “any person” to include “governmental entit[ies].” Dept. of Commerce, Patent and Trademark Office, MPEP §§2203, 2212 (4th rev. ed., July 1981). We might take account of this “executive interpretation” if we were determining whether Congress meant to include the Government as a “person” for purposes of the ex parte reexamination procedures themselves. See, e.g., United States v. Hermanos y Compañia, 209 U.S. 337, 339 (1908). Here, however, the Patent Office’s statement in the 1981 MPEP has no direct relevance. Even assuming that the Government may petition for ex parte reexamination, ex parte reexamination is a fundamentally different process than an AIA post-issuance review proceeding.[9] Both share the common purpose of allowing non-patent owners to bring questions of patent validity to the Patent Office’s attention, but they do so in meaningfully different ways. In an ex parte reexamination, the third party sends information to the Patent Office that the party believes bears on the patent’s validity, and the Patent Office decides whether to reexamine the patent. If it decides to do so, the reexamination process is internal; the challenger is not permitted to participate in the Patent Office’s process. See 35 U. S. C. §§302, 303. By contrast, the AIA post-issuance review proceedings are adversarial, adjudicatory proceedings between the “person” who petitioned for review and the patent owner: There is briefing, a hearing, discovery, and the presentation of evidence, and the losing party has appeal rights. See supra, at 4–5. Thus, there are good reasons Congress might have authorized the Government to initiate a hands-off ex parte reexamination but not to become a party to a full-blown adversarial proceeding before the Patent Office and any subsequent appeal. After all, the Government is already in a unique position among alleged infringers given that 28 U. S. C. §1498 limits patent owners to bench trials before the Court of Federal Claims and monetary damages, whereas 35 U. S. C. §271 permits patent owners to demand jury trials in the federal district courts and seek other types of relief. Thus, there is nothing to suggest that Congress had the 1981 MPEP statement in mind when it enacted the AIA. It is true that this Court has often said, “[w]hen administrative and judicial interpretations have settled the meaning of an existing statutory provision, repetition of the same language in a new statute indicates, as a general matter, the intent to incorporate its administrative and judicial interpretations as well.” Bragdon v. Abbott, 524 U.S. 624, 645 (1998). But there is no “settled” meaning of the term “person” with respect to the newly established AIA review proceedings. Accordingly, the MPEP does not justify putting aside the presumptive meaning of “person” here. 3 Finally, the Postal Service argues that it must be a “person” who may petition for AIA review proceedings because, like other potential infringers, it is subject to civil liability and can assert a defense of patent invalidity. See §§282(b)(2)–(3). In the Postal Service’s view, it is anomalous to deny it a benefit afforded to other infringers—the ability to challenge a patent de novo before the Patent Office, rather than only as an infringement defense that must be proved by clear and convincing evidence. See ibid.; Microsoft Corp., 564 U. S., at 95 (holding that §282’s presumption of validity in litigation imposes a clear and convincing evidence standard on defendants seeking to prove invalidity). The Postal Service overstates the asymmetry. Agencies retain the ability under §282 to assert defenses to infringement. Once sued, an agency may, like any other accused infringer, argue that the patent is invalid, and the agency faces the same burden of proof as a defendant in any other infringement suit. The Postal Service lacks only the additional tool of petitioning for the initiation of an administrative proceeding before the Patent Office under the AIA, a process separate from defending an infringement suit. We see no oddity, however, in Congress’ affording nongovernmental actors an expedient route that the Government does not also enjoy for heading off potential infringement suits. Those other actors face greater and more uncertain risks if they misjudge their right to use technology that is subject to potentially invalid patents. Most notably, §1498 restricts a patent owner who sues the Government to her “reasonable and entire compensation” for the Government’s infringing use; she cannot seek an injunction, demand a jury trial, or ask for punitive damages, all of which are available in infringement suits against nongovernmental actors under §271(e)(4). Thus, although federal agencies remain subject to damages for impermissible uses, they do not face the threat of preliminary injunctive relief that could suddenly halt their use of a patented invention, and they enjoy a degree of certainty about the extent of their potential liability that ordinary accused infringers do not. Because federal agencies face lower risks, it is reasonable for Congress to have treated them differently.[10] Finally, excluding federal agencies from the AIA review proceedings avoids the awkward situation that might result from forcing a civilian patent owner (such as Return Mail) to defend the patentability of her invention in an adversarial, adjudicatory proceeding initiated by one federal agency (such as the Postal Service) and overseen by a different federal agency (the Patent Office). We are therefore unpersuaded that the Government’s exclusion from the AIA review proceedings is sufficiently anomalous to overcome the presumption that the Government is not a “person” under the Act.[11] III For the foregoing reasons, we hold that a federal agency is not a “person” who may petition for post-issuance re- view under the AIA. The judgment of the United States Court of Appeals for the Federal Circuit is therefore reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 The CBM review program will stop accepting new claims in 2020. See AIA §18(a)(3)(A), 125Stat. 330; 77 Fed. Reg. 48687 (2012). 2 The Federal Circuit rejected Return Mail’s argument that the Postal Service cannot petition for CBM review for the independent reason that a suit against the Government under 28 U. S. C. §1498 is not a suit for infringement. 868 F.3d 1350, 1366 (2017). We denied Return Mail’s petition for certiorari on this question and therefore have no occasion to resolve it in this case. Accordingly, we assume that a §1498 suit is one for infringement and refer to it as the same. 3 For example, the statute expressly includes the Government as a “person” in §296(a), which, as enacted, provided that States “shall not be immune . . . from suit in Federal court by any person, including any governmental or nongovernmental entity, for infringement of a patent under section 271.” 35 U. S. C. §296(a) (1988 ed., Supp. IV) (ruled unconstitutional by Florida Prepaid Postsecondary Ed. Expense Bd. v. College Savings Bank, 527 U.S. 627, 630 (1999)). 4 For example, in §6(a), the Patent Act provides that the administrative patent judges comprising the Board must be “persons of competent legal knowledge and scientific ability.” Likewise, §257(e) requires the Patent Office Director to treat as confidential any referral to the Attorney General of suspected fraud in the patent process unless the United States charges “a person” with a criminal offense in connection with the fraud. See also §2(b)(11) (authorizing the Patent Office to cover the expenses of “persons” other than federal employees attending programs on intellectual-property protection); §100(h) (defining a “ ‘joint research agreement’ ” as a written agreement between “2 or more persons or entities”). Some of these provisions (§§2(b)(11), 6(a), and 100(h)) were enacted as part of the AIA, alongside the AIA review proceedings. See 125Stat. 285, 313, 335. 5 Section 102(a) provides that “[a] person shall be entitled to a patent” as long as the patent is novel. Section 118 states that “[a] person to whom the inventor has assigned” an invention may file a patent application. Section 119 discusses the effect of a patent application filed in a foreign country “by any person” on the patent-application process in the United States. 6 Likewise, we are not persuaded by the dissent’s suggestion that §207(a)(3)—which authorizes federal agencies “to protect and administer rights” to federally owned inventions—provides a statutory basis for the Postal Service’s initiation of AIA review proceedings. See post, at 5–6. The statute explains how a federal agency is to “protect” those rights: “either directly or through contract,” such as by “acquiring rights for and administering royalties” or “licensing.” §207(a)(3). TheAIA review proceedings, which a “person” may initiate regardless of ownership, do not fall clearly within the ambit of §207(a)(3). 7 The dissent responds that we should set aside the statutory references to “person[s]” that naturally exclude the Government and instead count only those references that expressly or impliedly include the Government. See post, at 3–4. But the point of the canon the PostalService invokes is to ascertain the meaning of a statutory term from its consistent usage in other parts of the statute, not to pick sides among differing uses. 8 Moreover, for those of us who consider legislative history, there is none that suggests Congress considered whether the Federal Government or its agencies would have access to the AIA review proceedings. 9 As discussed above, see supra, at 2–4, ex parte reexamination is not one of the three new proceedings added by the AIA, and therefore the question whether its reference to a “person” includes the Government is beyond the scope of the question presented. Moreover, neither party contests that a federal agency may cite prior art to the Patent Office and ask for ex parte reexamination. 10 If the Government were a “person” under the AIA, yet another anomaly might arise under the statute’s estoppel provisions. Those provisions generally preclude a party from relitigating issues in any subsequent proceedings in federal district court, before the International Trade Commission, and (for inter partes review and post-grant review) before the Patent Office. See 35 U. S. C. §§315(e), 325(e); AIA§18(a)(1)(D), 125Stat. 330. Because infringement suits against the Government must be brought in the Court of Federal Claims—which is not named in the estoppel provisions—the Government might not be precluded by statute from relitigating claims raised before the Patent Office if it were able to institute post-issuance review under the AIA. See 28 U. S. C. §1498(a). Although Return Mail cites this asymmetry in support of its interpretation, we need not rely on it, because Return Mail already prevails for the reasons given above. At any rate, the practical effect of the estoppel provisions’ potential inapplicability to the Government is uncertain given that this Court has not decided whether common-law estoppel applies in §1498 suits. 11 Nor do we find persuasive the dissent’s argument that the Postal Service should be allowed to petition for post-issuance review proceedings because its participation would further the purpose of the AIA: to provide a cost-effective and efficient alternative to litigation in the courts. See post, at 5; H. R. Rep. No. 112–98, pt. 1, pp. 47–48 (2001). Statutes rarely embrace every possible measure that would further their general aims, and, absent other contextual indicators of Congress’ intent to include the Government in a statutory provision referring to a “person,” the mere furtherance of the statute’s broad purpose does not overcome the presumption in this case. See Cooper, 312 U. S., at 605 (“[I]t is not our function to engraft on a statute additions which we think the legislature logically might or should have made”).

586.US.2018_17-1625

A jury awarded Oracle damages after finding that Rimini Street had infringed various Oracle copyrights. After judgment, the District Court also awarded Oracle fees and costs, including $12.8 million for litigation expenses such as expert witnesses, e-discovery, and jury consulting. In affirming the $12.8 million award, the Ninth Circuit acknowledged that it covered expenses not included within the six categories of costs that the general federal statute authorizing district courts to award costs, 28 U. S. C. §§1821 and 1920, provides may be awarded against a losing party. The court nonetheless held that the award was appropriate because the Copyright Act gives federal district courts discretion to award “full costs” to a party in copyright litigation, 17 U. S. C. §505. Held: The term “full costs” in §505 of the Copyright Act means the costs specified in the general costs statute codified at §§1821 and 1920. Pp. 3–12. (a) Sections 1821 and 1920 define what the term “costs” encompasses in subject-specific federal statutes such as §505. Congress may authorize awards of expenses beyond the six categories specified in the general costs statute, but courts may not award litigation expenses that are not specified in §§1821 and 1920 absent explicit authority. This Court’s precedents have consistently adhered to that approach. See Crawford Fitting Co. v. J. T. Gibbons, Inc., 482 U.S. 437; West Virginia Univ. Hospitals, Inc. v. Casey, 499 U.S. 83; Arlington Central School Dist. Bd. of Ed. v. Murphy, 548 U.S. 291. The Copyright Act does not explicitly authorize the award of litigation expenses beyond the six categories specified in §§1821 and 1920, which do not authorize an award for expenses such as expert witness fees, e-discovery expenses, and jury consultant fees. Pp. 3–6. (b) Oracle’s counterarguments are not persuasive. First, Oracle argues that the word “full” authorizes courts to award expenses beyond the costs specified in §§1821 and 1920. The term “full” is an adjective that means the complete measure of the noun it modifies. It does not, therefore, alter the meaning of the word “costs” in §505. Rather, “full costs” are all the “costs” otherwise available under the relevant law. Second, Oracle maintains that the term “full costs” in the Copyright Act is a historical term of art that encompasses more than the “costs” listed in §§1821 and 1920. Oracle argues that Congress imported the meaning of the term “full costs” from the English copyright statutes into the Copyright Act in 1831. It contends that the 1831 meaning of “full costs” allows the transfer of all expenses of litigation, beyond those specified in any costs schedule, and overrides anything that Congress enacted in the Fee Act of 1853 or any subsequent costs statute. Courts need not, however, undertake extensive historical excavation to determine the meaning of costs statutes. See Crawford Fitting Co., 482 U. S., at 445. In any event, Oracle has not shown that the phrase “full costs” had an established meaning in English or American law that covered more than the full amount of the costs listed in the applicable costs schedule. Case law since 1831 also refutes Oracle’s historical argument. Third, Oracle advances a variety of surplusage arguments. According to Oracle, after Congress made the costs award discretionary in 1976, district courts could award any amount of costs up to 100 percent, and so Rimini’s reading of the word “full” now adds nothing to “costs.” Because Congress would not have intended “full” to be surplusage, Oracle contends, Congress must have employed the term “full” to mean expenses beyond the costs specified in §§1821 and 1920. But even if the term “full” lacked any continuing significance after 1976, the meaning of “costs” did not change. Oracle’s interpretation would also create its own redundancy problem by rendering the second sentence of §505—which covers attorney’s fees—largely redundant because §505’s first sentence presumably would already cover those fees. Finally, Oracle’s argument, even if correct, overstates the significance of statutory surplusage and redundancy. See, e.g., Marx v. General Revenue Corp., 568 U.S. 371, 385. Pp. 6–11. 879 F.3d 948, reversed in part and remanded. Kavanaugh, J., delivered the opinion for a unanimous Court.

The Copyright Act gives federal district courts discretion to award “full costs” to a party in copyright litigation. 17 U. S. C. §505. In the general statute governing awards of costs, Congress has specified six categories of litigation expenses that qualify as “costs.” See 28 U. S. C. §§1821, 1920. The question presented in this case is whether the Copyright Act’s reference to “full costs” authorizes a court to award litigation expenses beyond the six categories of “costs” specified by Congress in the general costs statute. The statutory text and our precedents establish that the answer is no. The term “full” is a term of quantity or amount; it does not expand the categories or kinds of expenses that may be awarded as “costs” under the general costs statute. In copyright cases, §505’s authorization for the award of “full costs” therefore covers only the six categories specified in the general costs statute, codified at §§1821 and 1920. We reverse in relevant part the judgment of the U. S. Court of Appeals for the Ninth Circuit, and we remand the case for further proceedings consistent with this opinion. I Oracle develops and licenses software programs that manage data and operations for businesses and non-profit organizations. Oracle also offers its customers software maintenance services. Rimini Street sells third-party software maintenance services to Oracle customers. In doing so, Rimini competes with Oracle’s software maintenance services. Oracle sued Rimini and its CEO in Federal District Court in Nevada, asserting claims under the Copyright Act and various other federal and state laws. Oracle alleged that Rimini, in the course of providing software support services to Oracle customers, copied Oracle’s software without licensing it. A jury found that Rimini had infringed various Oracle copyrights and that both Rimini and its CEO had violated California and Nevada computer access statutes. The jury awarded Oracle $35.6 million in damages for copyright infringement and $14.4 million in damages for violations of the state computer access statutes. After judgment, the District Court ordered the defendants to pay Oracle an additional $28.5 million in attorney’s fees and $4.95 million in costs; the Court of Appeals reduced the latter award to $3.4 million. The District Court also ordered the defendants to pay Oracle $12.8 million for litigation expenses such as expert witnesses, e-discovery, and jury consulting. That $12.8 million award is the subject of the dispute in this case. As relevant here, the U. S. Court of Appeals for the Ninth Circuit affirmed the District Court’s $12.8 million award. The Court of Appeals recognized that the general federal statute authorizing district courts to award costs, 28 U. S. C. §§1821 and 1920, lists only six categories of costs that may be awarded against the losing party. And the Court of Appeals acknowledged that the $12.8 million award covered expenses not included within those six categories. But the Court of Appeals, relying on Circuit precedent, held that the District Court’s $12.8 million award for additional expenses was still appropriate because §505 permits the award of “full costs,” a term that the Ninth Circuit said was not confined to the six categories identified in §§1821 and 1920. 879 F.3d 948, 965−966 (2018). We granted certiorari to resolve disagreement in the Courts of Appeals over whether the term “full costs” in §505 authorizes awards of expenses other than those costs identified in §§1821 and 1920. 585 U. S. ___ (2018). Compare 879 F. 3d, at 965–966; Twentieth Century Fox Film Corp. v. Entertainment Distributing, 429 F.3d 869 (CA9 2005), with Artisan Contractors Assn. of Am., Inc. v. Frontier Ins. Co., 275 F.3d 1038 (CA11 2001); Pinkham v. Camex, Inc., 84 F.3d 292 (CA8 1996). II A Congress has enacted more than 200 subject-specific federal statutes that explicitly authorize the award of costs to prevailing parties in litigation. The Copyright Act is one of those statutes. That Act provides that a district court in a copyright case “in its discretion may allow the recovery of full costs by or against any party other than the United States or an officer thereof.” 17 U. S. C. §505. In the general “costs” statute, codified at §§1821 and 1920 of Title 28, Congress has specified six categories of litigation expenses that a federal court may award as “costs,”[1] and Congress has detailed how to calculate the amount of certain costs. Sections 1821 and 1920 in essence define what the term “costs” encompasses in the subject-specific federal statutes that provide for an award of costs. Sections 1821 and 1920 create a default rule and establish a clear baseline against which Congress may legislate. Consistent with that default rule, some federal statutes simply refer to “costs.” In those cases, federal courts are limited to awarding the costs specified in §§1821 and 1920. If, for particular kinds of cases, Congress wants to authorize awards of expenses beyond the six categories specified in the general costs statute, Congress may do so. For example, some federal statutes go beyond §§1821 and 1920 to expressly provide for the award of expert witness fees or attorney’s fees. See West Virginia Univ. Hospitals, Inc. v. Casey, 499 U.S. 83, 89, n. 4 (1991). Indeed, the Copyright Act expressly provides for awards of attorney’s fees as well as costs. 17 U. S. C. §505. And the same Congress that enacted amendments to the Copyright Act in 1976 enacted several other statutes that expressly authorized awards of expert witness fees. See Casey, 499 U. S., at 88. But absent such express authority, courts may not award litigation expenses that are not specified in §§1821 and 1920. Our precedents have consistently adhered to that approach. Three cases illustrate the point. In Crawford Fitting Co. v. J. T. Gibbons, Inc., the question was whether courts could award expert witness fees under Rule 54(d) of the Federal Rules of Civil Procedure. Rule 54(d) authorizes an award of “costs” but does not expressly refer to expert witness fees. 482 U.S. 437, 441 (1987). In defining what expenses qualify as “costs,” §§1821 and 1920 likewise do not include expert witness fees. We therefore held that the prevailing party could not obtain expert witness fees: When “a prevailing party seeks reimbursement for fees paid to its own expert witnesses, a federal court is bound by the limit of §1821(b), absent contract or explicit statutory authority to the contrary.” Id., at 439. In Casey, we interpreted 42 U. S. C. §1988, the federal statute authorizing an award of “costs” in civil rights litigation. We described Crawford Fitting as holding that §§1821 and 1920 “define the full extent of a federal court’s power to shift litigation costs absent express statutory authority to go further.” 499 U. S., at 86. In accord with Crawford Fitting, we concluded that §1988 does not authorize awards of expert witness fees because §1988 supplies no “ ‘explicit statutory authority’ ” to award expert witness fees. 499 U. S., at 87 (quoting Crawford Fitting, 482 U. S., at 439). In Arlington Central School Dist. Bd. of Ed. v. Murphy, we considered the Individuals with Disabilities Education Act, which authorized an award of costs. The question was whether that Act’s reference to “costs” encompassed expert witness fees. We again explained that “costs” is “ ‘a term of art that generally does not include expert fees.’ ” 548 U.S. 291, 297 (2006); see also Taniguchi v. Kan Pa- cific Saipan, Ltd., 566 U.S. 560, 573 (2012). We stated: “[N]o statute will be construed as authorizing the taxation of witness fees as costs unless the statute ‘refer[s] explic- itly to witness fees.’ ” Murphy, 548 U. S., at 301 (quoting Crawford Fitting, 482 U. S., at 445). Our cases, in sum, establish a clear rule: A statute awarding “costs” will not be construed as authorizing an award of litigation expenses beyond the six categories listed in §§1821 and 1920, absent an explicit statutory instruction to that effect. See Murphy, 548 U. S., at 301 (requiring “ ‘explici[t]’ ” authority); Casey, 499 U. S., at 86 (requiring “ ‘explicit’ ” authority); Crawford Fitting, 482 U. S., at 439 (requiring “explicit statutory authority”). Here, the Copyright Act does not explicitly authorize the award of litigation expenses beyond the six categories specified in §§1821 and 1920. And §§1821 and 1920 in turn do not authorize an award for expenses such as expert witness fees, e-discovery expenses, and jury consultant fees, which were expenses encompassed by the District Court’s $12.8 million award to Oracle here. Rimini argues that the $12.8 million award therefore cannot stand. B To sustain its $12.8 million award, Oracle advances three substantial arguments. But we ultimately do not find those arguments persuasive. First, although Oracle concedes that it would lose this case if the Copyright Act referred only to “costs,” Oracle stresses that the Copyright Act uses the word “full” before “costs.” Oracle argues that the word “full” authorizes courts to award expenses beyond the costs specified in §§1821 and 1920. We disagree. “Full” is a term of quantity or amount. It is an adjective that means the complete measure of the noun it modifies. See American Heritage Dictionary 709 (5th ed. 2011); Oxford English Dictionary 247 (2d ed. 1989). As we said earlier this Term: “Adjectives modify nouns—they pick out a subset of a category that possesses a certain quality.” Weyerhaeuser Co. v. United States Fish and Wildlife Serv., 586 U. S. ___, ___ (2018) (slip op., at 8). The adjective “full” in §505 therefore does not alter the meaning of the word “costs.” Rather, “full costs” are all the “costs” otherwise available under law. The word “full” operates in the phrase “full costs” just as it operates in other common phrases: A “full moon” means the moon, not Mars. A “full breakfast” means breakfast, not lunch. A “full season ticket plan” means tickets, not hot dogs. So too, the term “full costs” means costs, not other expenses. The dispute here, therefore, turns on the meaning of the word “costs.” And as we have explained, the term “costs” refers to the costs generally available under the federal costs statute—§§1821 and 1920. “Full costs” are all the costs generally available under that statute. Second, Oracle maintains that the term “full costs” in the Copyright Act is a historical term of art that encompasses more than the “costs” listed in the relevant costs statute—here, §§1821 and 1920. We again disagree. Some general background: From 1789 to 1853, federal courts awarded costs and fees according to the relevant state law of the forum State. See Crawford Fitting, 482 U. S., at 439−440; Alyeska Pipeline Service Co. v. Wilderness Society, 421 U.S. 240, 247−250 (1975). In 1853, Congress departed from that state-focused approach. That year, Congress passed and President Fillmore signed a comprehensive federal statute establishing a federal schedule for the award of costs in federal court. Crawford Fitting, 482 U. S., at 440; 10Stat. 161. Known as the Fee Act of 1853, that 1853 statute has “carried forward to today” in §§1821 and 1920 “ ‘without any apparent intent to change the controlling rules.’ ” Crawford Fitting, 482 U. S., at 440. As we have said, §§1821 and 1920 provide a comprehensive schedule of costs for proceedings in federal court. Now some copyright law background: The term “full [c]osts” appeared in the first copyright statute in England, the Statute of Anne. 8 Anne c.19, §8 (1710). In the United States, the Federal Copyright Act of 1831 borrowed the phrasing of English copyright law and used the same term, “full costs.” Act of Feb. 3, 1831, §12, 4Stat. 438–439. That term has appeared in subsequent revisions of the Copyright Act, through the Act’s most recent substantive alterations in 1976. See Act of July 8, 1870, §108, 16Stat. 215; Copyright Act of 1909, §40, 35Stat. 1084; Copyright Act of 1976, §505, 90Stat. 2586. Oracle argues that English copyright statutes awarding “full costs” allowed the transfer of all expenses of litigation, beyond what was specified in any costs schedule. According to Oracle, Congress necessarily imported that meaning of the term “full costs” into the Copyright Act in 1831. And according to Oracle, that 1831 meaning overrides anything that Congress enacted in any costs statute in 1853 or later. To begin with, our decision in Crawford Fitting explained that courts should not undertake extensive historical excavation to determine the meaning of costs statutes. We said that §§1821 and 1920 apply regardless of when individual subject-specific costs statutes were enacted. 482 U. S., at 445. The Crawford Fitting principle eliminates the need for that kind of historical analysis and confirms that the Copyright Act’s reference to “full costs” must be interpreted by reference to §§1821 and 1920. In any event, Oracle’s historical argument fails even on its own terms. Oracle has not persuasively demonstrated that as of 1831, the phrase “full costs” had an established meaning in English or American law that covered more than the full amount of the costs listed in the applicable costs schedule. On the contrary, the federal courts as of 1831 awarded costs in accord with the costs schedule of the relevant state law. See id., at 439−440; Alyeska Pipeline, 421 U. S., at 250. And state laws at the time tended to use the term “full costs” to refer to, among other things, full cost awards as distinguished from the half, double, or treble cost awards that were also commonly available under state law at the time.[2] That usage accorded with the ordinary meaning of the term. At the time, the word “full” conveyed the same meaning that it does today: “Complete; entire; not defective or partial.” 1 N. Webster, An American Dictionary of the English Language 89 (1828); see also 1 S. Johnson, A Dictionary of the English Language 817 (1773) (“Complete, such as that nothing further is desired or wanted; Complete without abatement; at the utmost degree”). Full costs did not encompass expenses beyond those costs that otherwise could be awarded under the applicable state law. The case law since 1831 also refutes Oracle’s historical argument. If Oracle’s account of the history were correct, federal courts starting in 1831 presumably would have interpreted the term “full costs” in the Copyright Act to allow awards of litigation expenses that were not ordinarily available as costs under the applicable costs schedule. But Rimini points out that none of the more than 800 available copyright decisions awarding costs from 1831 to 1976—that is, from the year the term “full costs” first appeared in the Copyright Act until the year that the Act was last significantly amended—awarded expenses other than those specified by the applicable state or federal law. Tr. of Oral Arg. 7. Oracle has not refuted Rimini’s argument on that point. Oracle cites no §505 cases where federal courts awarded expert witness fees or other litigation expenses of the kind at issue here until the Ninth Circuit’s 2005 decision adopting the interpretation of §505 that the Ninth Circuit followed in this case. See Twentieth Century Fox, 429 F.3d 869. In light of the commonly understood meaning of the term “full costs” as of 1831 and the case law since 1831, Oracle’s historical argument falls short. The best interpretation is that the term “full costs” meant in 1831 what it means now: the full amount of the costs specified by the applicable costs schedule. Third, Oracle advances a variety of surplusage arguments. Oracle contends, for example, that the word “full” would be unnecessary surplusage if Rimini’s argument were correct. We disagree. The award of costs in copyright cases was mandatory from 1831 to 1976. See §40, 35Stat. 1084; §12, 4Stat. 438–439. During that period, the term “full” fixed both a floor and a ceiling for the amount of “costs” that could be awarded. In other words, the term “full costs” required an award of 100 percent of the costs available under the applicable costs schedule. Oracle says that even if that interpretation of “full costs” made sense before 1976, the meaning of the term “full costs” changed in 1976. That year, Congress amended the Copyright Act to make the award of costs discretionary rather than mandatory. See §505, 90Stat. 2586. According to Oracle, after Congress made the costs award discretionary, district courts could award any amount of costs up to 100 percent and so Rimini’s reading of the word “full” now adds nothing to “costs.” If we assume that Congress in 1976 did not intend “full” to be surplusage, Oracle argues that Congress must have employed the term “full” to mean expenses beyond the costs specified in §§1821 and 1920. For several reasons, that argument does not persuade us. To begin with, even if the term “full” lacked any continuing significance after 1976, the meaning of “costs” did not change. The term “costs” still means those costs specified in §§1821 and 1920. It makes little sense to think that Congress in 1976, when it made the award of full costs discretionary rather than mandatory, silently expanded the kinds of expenses that a court may otherwise award as costs in copyright suits.[3] Moreover, Oracle’s interpretation would create its own redundancy problem by rendering the second sentence of §505 largely redundant. That second sentence provides: “Except as otherwise provided by this title, the court may also award a reasonable attorney’s fee to the prevailing party as part of the costs.” 17 U. S. C. §505. If Oracle were right that “full costs” covers all of a party’s litigation expenditures, then the first sentence of §505 would presumably already cover attorney’s fees and the second sentence would be largely unnecessary. In order to avoid some redundancy, Oracle’s interpretation would create other redundancy. Finally, even if Oracle is correct that the term “full” has become unnecessary or redundant as a result of the 1976 amendment, Oracle overstates the significance of statutory surplusage or redundancy. Redundancy is not a silver bullet. We have recognized that some “redundancy is ‘hardly unusual’ in statutes addressing costs.” Marx v. General Revenue Corp., 568 U.S. 371, 385 (2013). If one possible interpretation of a statute would cause some redundancy and another interpretation would avoid redundancy, that difference in the two interpretations can supply a clue as to the better interpretation of a statute. But only a clue. Sometimes the better overall reading of the statute contains some redundancy. * * * The Copyright Act authorizes federal district courts to award “full costs” to a party in copyright litigation. That term means the costs specified in the general costs statute, §§1821 and 1920. We reverse in relevant part the judgment of the Court of Appeals, and we remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 The six categories that a federal court may award as costs are: “(1) Fees of the clerk and marshal; “(2) Fees for printed or electronically recorded transcripts necessarily obtained for use in the case; “(3) Fees and disbursements for printing and witnesses; “(4) Fees for exemplification and the costs of making copies of any materials where the copies are necessarily obtained for use in the case; “(5) Docket fees under section 1923 of this title; “(6) Compensation of court appointed experts, compensation of interpreters, and salaries, fees, expenses, and costs of special interpretation services under section 1828 of this title.” 28 U. S. C. §1920. In addition, §1821 provides particular reimbursement rates for witnesses’ “[p]er diem and mileage” expenses. 2 See, e.g., 1 Laws of Pa., ch. DCXLV, pp. 371, 373 (1810) (“full costs” and “double costs”); 2 Rev. Stat. N. Y., pt. III, ch. X, Tit. 1, §§16, 25 (1836) (“full,” “double,” and “treble” costs); Rev. Stat. Mass., pt. III, Tit. VI, ch. 121, §§4, 7, 8, 11, 18 (1836) (“one quarter,” “full,” “double,” and “treble” costs). 3 Rimini further suggests that “full” still has meaning after 1976 because the statute gives the district court discretion to award either full costs or no costs, unlike statutes that refer only to “costs,” which allow courts to award any amount of costs up to full costs. In light of our disposition of the case, we need not and do not consider that argument.

588.US.2018_18-422

Voters and other plaintiffs in North Carolina and Maryland filed suits challenging their States’ congressional districting maps as unconstitutional partisan gerrymanders. The North Carolina plaintiffs claimed that the State’s districting plan discriminated against Democrats, while the Maryland plaintiffs claimed that their State’s plan discriminated against Republicans. The plaintiffs alleged violations of the First Amendment, the Equal Protection Clause of the Fourteenth Amendment, the Elections Clause, and Article I, §2. The District Courts in both cases ruled in favor of the plaintiffs, and the defendants appealed directly to this Court. Held: Partisan gerrymandering claims present political questions beyond the reach of the federal courts. Pp. 6–34. (a) In these cases, the Court is asked to decide an important question of constitutional law. Before it does so, the Court “must find that the question is presented in a ‘case’ or ‘controversy’ that is . . . ‘of a Judiciary Nature.’ ” DaimlerChrysler Corp. v. Cuno, 547 U.S. 332, 342. While it is “the province and duty of the judicial department to say what the law is,” Marbury v. Madison, 1 Cranch 137, 177, sometimes the law is that the Judiciary cannot entertain a claim because it presents a nonjusticiable “political question,” Baker v. Carr, 369 U.S. 186, 217. Among the political question cases this Court has identified are those that lack “judicially discoverable and manageable standards for resolving [them].” Ibid. This Court’s partisan gerrymandering cases have left unresolved the question whether such claims are claims of legal right, resolvable according to legal principles, or political questions that must find their resolution elsewhere. See Gill v. Whitford, 585 U. S. ___, ___. Partisan gerrymandering was known in the Colonies prior to Independence, and the Framers were familiar with it at the time of the drafting and ratification of the Constitution. They addressed the election of Representatives to Congress in the Elections Clause, Art. I, §4, cl. 1, assigning to state legislatures the power to prescribe the “Times, Places and Manner of holding Elections” for Members of Congress, while giving Congress the power to “make or alter” any such regulations. Congress has regularly exercised its Elections Clause power, including to address partisan gerrymandering. But the Framers did not set aside all electoral issues as questions that only Congress can resolve. In two areas—one-person, one-vote and racial gerrymandering—this Court has held that there is a role for the courts with respect to at least some issues that could arise from a State’s drawing of congressional districts. But the history of partisan gerrymandering is not irrelevant. Aware of electoral districting problems, the Framers chose a characteristic approach, assigning the issue to the state legislatures, expressly checked and balanced by the Federal Congress, with no suggestion that the federal courts had a role to play. Courts have nonetheless been called upon to resolve a variety of questions surrounding districting. The claim of population inequality among districts in Baker v. Carr, for example, could be decided under basic equal protection principles. 369 U. S., at 226. Racial discrimination in districting also raises constitutional issues that can be addressed by the federal courts. See Gomillion v. Lightfoot, 364 U.S. 339, 340. Partisan gerrymandering claims have proved far more difficult to adjudicate, in part because “a jurisdiction may engage in constitutional political gerrymandering.” Hunt v. Cromartie, 526 U.S. 541, 551. To hold that legislators cannot take their partisan interests into account when drawing district lines would essentially countermand the Framers’ decision to entrust districting to political entities. The “central problem” is “determining when political gerrymandering has gone too far.” Vieth v. Jubelirer, 541 U.S. 267, 296 (plurality opinion). Despite considerable efforts in Gaffney v. Cummings, 412 U.S. 735, 753; Davis v. Bandemer, 478 U.S. 109, 116–117; Vieth, 541 U. S., at 272–273; and League of United Latin American Citizens v. Perry, 548 U.S. 399, 414 (LULAC), this Court’s prior cases have left “unresolved whether . . . claims [of legal right] may be brought in cases involving allegations of partisan gerrymandering,” Gill, 585 U. S., at ___. Two “threshold questions” remained: standing, which was addressed in Gill, and “whether [such] claims are justiciable.” Ibid. Pp. 6–14. (b) Any standard for resolving partisan gerrymandering claims must be grounded in a “limited and precise rationale” and be “clear, manageable, and politically neutral.” Vieth, 541 U. S., at 306–308 (Kennedy, J., concurring in judgment). The question is one of degree: How to “provid[e] a standard for deciding how much partisan dominance is too much.” LULAC, 548 U. S., at 420 (opinion of Kennedy, J.). Partisan gerrymandering claims rest on an instinct that groups with a certain level of political support should enjoy a commensurate level of political power and influence. Such claims invariably sound in a desire for proportional representation, but the Constitution does not require proportional representation, and federal courts are neither equipped nor authorized to apportion political power as a matter of fairness. It is not even clear what fairness looks like in this context. It may mean achieving a greater number of competitive districts by undoing packing and cracking so that supporters of the disadvantaged party have a better shot at electing their preferred candidates. But it could mean engaging in cracking and packing to ensure each party its “appropriate” share of “safe” seats. Or perhaps it should be measured by adherence to “traditional” districting criteria. Deciding among those different visions of fairness poses basic questions that are political, not legal. There are no legal standards discernible in the Constitution for making such judgments. And it is only after determining how to define fairness that one can even begin to answer the determinative question: “How much is too much?” The fact that the Court can adjudicate one-person, one-vote claims does not mean that partisan gerrymandering claims are justiciable. This Court’s one-person, one-vote cases recognize that each person is entitled to an equal say in the election of representatives. It hardly follows from that principle that a person is entitled to have his political party achieve representation commensurate to its share of statewide support. Vote dilution in the one-person, one-vote cases refers to the idea that each vote must carry equal weight. That requirement does not extend to political parties; it does not mean that each party must be influential in proportion to the number of its supporters. The racial gerrymandering cases are also inapposite: They call for the elimination of a racial classification, but a partisan gerrymandering claim cannot ask for the elimination of partisanship. Pp. 15–21. (c) None of the proposed “tests” for evaluating partisan gerrymandering claims meets the need for a limited and precise standard that is judicially discernible and manageable. Pp. 22–30. (1) The Common Cause District Court concluded that all but one of the districts in North Carolina’s 2016 Plan violated the Equal Protection Clause by intentionally diluting the voting strength of Democrats. It applied a three-part test, examining intent, effects, and causation. The District Court’s “predominant intent” prong is borrowed from the test used in racial gerrymandering cases. However, unlike race-based decisionmaking, which is “inherently suspect,” Miller v. Johnson, 515 U.S. 900, 915, districting for some level of partisan advantage is not unconstitutional. Determining that lines were drawn on the basis of partisanship does not indicate that districting was constitutionally impermissible. The Common Cause District Court also required the plaintiffs to show that vote dilution is “likely to persist” to such a degree that the elected representatives will feel free to ignore the concerns of the supporters of the minority party. Experience proves that accurately predicting electoral outcomes is not simple, and asking judges to predict how a particular districting map will perform in future elections risks basing constitutional holdings on unstable ground outside judicial expertise. The District Court’s third prong—which gave the defendants an opportunity to show that discriminatory effects were due to a “legitimate redistricting objective”—just restates the question asked at the “predominant intent” prong. Pp. 22–25. (2) The District Courts also found partisan gerrymandering claims justiciable under the First Amendment, coalescing around a basic three-part test: proof of intent to burden individuals based on their voting history or party affiliation, an actual burden on political speech or associational rights, and a causal link between the invidious intent and actual burden. But their analysis offers no “clear” and “manageable” way of distinguishing permissible from impermissible partisan motivation. Pp. 25–27. (3) Using a State’s own districting criteria as a baseline from which to measure how extreme a partisan gerrymander is would be indeterminate and arbitrary. Doing so would still leave open the question of how much political motivation and effect is too much. Pp. 27–29. (4) The North Carolina District Court further held that the 2016 Plan violated Article I, §2, and the Elections Clause, Art. I, §4, cl. 1. But the Vieth plurality concluded—without objection from any other Justice—that neither §2 nor §4 “provides a judicially enforceable limit on the political considerations that the States and Congress may take into account when districting.” 541 U. S., at 305. Any assertion that partisan gerrymanders violate the core right of voters to choose their representatives is an objection more likely grounded in the Guarantee Clause of Article IV, §4, which “guarantee[s] to every State in [the] Union a Republican Form of Government.” This Court has several times concluded that the Guarantee Clause does not provide the basis for a justiciable claim. See, e.g., Pacific States Telephone & Telegraph Co. v. Oregon, 223 U.S. 118. Pp. 29–30. (d) The conclusion that partisan gerrymandering claims are not justiciable neither condones excessive partisan gerrymandering nor condemns complaints about districting to echo into a void. Numerous States are actively addressing the issue through state constitutional amendments and legislation placing power to draw electoral districts in the hands of independent commissions, mandating particular districting criteria for their mapmakers, or prohibiting drawing district lines for partisan advantage. The Framers also gave Congress the power to do something about partisan gerrymandering in the Elections Clause. That avenue for reform established by the Framers, and used by Congress in the past, remains open. Pp. 30–34. 318 F. Supp. 3d 777 and 348 F. Supp. 3d 493, vacated and remanded. Roberts, C. J., delivered the opinion of the Court, in which Thomas, Alito, Gorsuch, and Kavanaugh, JJ., joined. Kagan, J., filed a dissenting opinion, in which Ginsburg, Breyer, and Sotomayor, JJ., joined. Notes 1 Together with No. 18–726, Lamone et al. v. Benisek et al., on appeal from the United States District Court for the District of Maryland.

Voters and other plaintiffs in North Carolina and Maryland challenged their States’ congressional districting maps as unconstitutional partisan gerrymanders. The North Carolina plaintiffs complained that the State’s districting plan discriminated against Democrats; the Maryland plaintiffs complained that their State’s plan discriminated against Republicans. The plaintiffs alleged that the gerrymandering violated the First Amendment, the Equal Protection Clause of the Fourteenth Amendment, the Elections Clause, and Article I, §2, of the Constitution. The District Courts in both cases ruled in favor of the plaintiffs, and the defendants appealed directly to this Court. These cases require us to consider once again whether claims of excessive partisanship in districting are “justiciable”—that is, properly suited for resolution by the federal courts. This Court has not previously struck down a districting plan as an unconstitutional partisan gerrymander, and has struggled without success over the past several decades to discern judicially manageable standards for deciding such claims. The districting plans at issue here are highly partisan, by any measure. The question is whether the courts below appropriately exercised judicial power when they found them unconstitutional as well. I A The first case involves a challenge to the congressional redistricting plan enacted by the Republican-controlled North Carolina General Assembly in 2016. Rucho v. Common Cause, No. 18–422. The Republican legislators leading the redistricting effort instructed their mapmaker to use political data to draw a map that would produce a congressional delegation of ten Republicans and three Democrats. 318 F. Supp. 3d 777, 807–808 (MDNC 2018). As one of the two Republicans chairing the redistricting committee stated, “I think electing Republicans is better than electing Democrats. So I drew this map to help foster what I think is better for the country.” Id., at 809. He further explained that the map was drawn with the aim of electing ten Republicans and three Democrats because he did “not believe it [would be] possible to draw a map with 11 Republicans and 2 Democrats.” Id., at 808. One Demo- cratic state senator objected that entrenching the 10–3 advantage for Republicans was not “fair, reasonable, [or] balanced” because, as recently as 2012, “Democratic congressional candidates had received more votes on a statewide basis than Republican candidates.” Ibid. The General Assembly was not swayed by that objection and approved the 2016 Plan by a party-line vote. Id., at 809. In November 2016, North Carolina conducted congressional elections using the 2016 Plan, and Republican candidates won 10 of the 13 congressional districts. Id., at 810. In the 2018 elections, Republican candidates won nine congressional districts, while Democratic candidates won three. The Republican candidate narrowly prevailed in the remaining district, but the State Board of Elections called a new election after allegations of fraud. This litigation began in August 2016, when the North Carolina Democratic Party, Common Cause (a nonprofit organization), and 14 individual North Carolina voters sued the two lawmakers who had led the redistricting effort and other state defendants in Federal District Court. Shortly thereafter, the League of Women Voters of North Carolina and a dozen additional North Carolina voters filed a similar complaint. The two cases were consolidated. The plaintiffs challenged the 2016 Plan on multiple constitutional grounds. First, they alleged that the Plan violated the Equal Protection Clause of the Fourteenth Amendment by intentionally diluting the electoral strength of Democratic voters. Second, they claimed that the Plan violated their First Amendment rights by retaliating against supporters of Democratic candidates on the basis of their political beliefs. Third, they asserted that the Plan usurped the right of “the People” to elect their preferred candidates for Congress, in violation of the requirement in Article I, §2, of the Constitution that Members of the House of Representatives be chosen “by the People of the several States.” Finally, they alleged that the Plan violated the Elections Clause by exceeding the State’s delegated authority to prescribe the “Times, Places and Manner of holding Elections” for Members of Congress. After a four-day trial, the three-judge District Court unanimously concluded that the 2016 Plan violated the Equal Protection Clause and Article I of the Constitution. The court further held, with Judge Osteen dissenting, that the Plan violated the First Amendment. Common Cause v. Rucho, 279 F. Supp. 3d 587 (MDNC 2018). The defendants appealed directly to this Court under 28 U. S. C. §1253. While that appeal was pending, we decided Gill v. Whitford, 585 U. S. ___ (2018), a partisan gerrymandering case out of Wisconsin. In that case, we held that a plaintiff asserting a partisan gerrymandering claim based on a theory of vote dilution must establish standing by showing he lives in an allegedly “cracked” or “packed” district. Id., at ___ (slip op., at 17). A “cracked” district is one in which a party’s supporters are divided among multiple districts, so that they fall short of a majority in each; a “packed” district is one in which a party’s supporters are highly concentrated, so they win that district by a large margin, “wasting” many votes that would improve their chances in others. Id., at ___–___ (slip op., at 3–4). After deciding Gill, we remanded the present case for further consideration by the District Court. 585 U. S. ___ (2018). On remand, the District Court again struck down the 2016 Plan. 318 F. Supp. 3d 777. It found standing and concluded that the case was appropriate for judicial resolution. On the merits, the court found that “the General Assembly’s predominant intent was to discriminate against voters who supported or were likely to support non-Republican candidates,” and to “entrench Republican candidates” through widespread cracking and packing of Democratic voters. Id., at 883–884. The court rejected the defendants’ arguments that the distribution of Republican and Democratic voters throughout North Carolina and the interest in protecting incumbents neutrally explained the 2016 Plan’s discriminatory effects. Id., at 896–899. In the end, the District Court held that 12 of the 13 districts constituted partisan gerrymanders that violated the Equal Protection Clause. Id., at 923. The court also agreed with the plaintiffs that the 2016 Plan discriminated against them because of their political speech and association, in violation of the First Amendment. Id., at 935. Judge Osteen dissented with respect to that ruling. Id., at 954–955. Finally, the District Court concluded that the 2016 Plan violated the Elections Clause and Article I, §2. Id., at 935–941. The District Court enjoined the State from using the 2016 Plan in any election after the November 2018 general election. Id., at 942. The defendants again appealed to this Court, and we postponed jurisdiction. 586 U. S. ___ (2019). B The second case before us is Lamone v. Benisek, No. 18–726. In 2011, the Maryland Legislature—dominated by Democrats—undertook to redraw the lines of that State’s eight congressional districts. The Governor at the time, Democrat Martin O’Malley, led the process. He appointed a redistricting committee to help redraw the map, and asked Congressman Steny Hoyer, who has described himself as a “serial gerrymanderer,” to advise the committee. 348 F. Supp. 3d 493, 502 (Md. 2018). The Governor later testified that his aim was to “use the redistricting process to change the overall composition of Maryland’s congressional delegation to 7 Democrats and 1 Republican by flipping” one district. Ibid. “[A] decision was made to go for the Sixth,” ibid., which had been held by a Republican for nearly two decades. To achieve the required equal population among districts, only about 10,000 residents needed to be removed from that district. Id., at 498. The 2011 Plan accomplished that by moving roughly 360,000 voters out of the Sixth District and moving 350,000 new voters in. Overall, the Plan reduced the number of registered Republicans in the Sixth District by about 66,000 and increased the number of registered Democrats by about 24,000. Id., at 499–501. The map was adopted by a party-line vote. Id., at 506. It was used in the 2012 election and succeeded in flipping the Sixth District. A Democrat has held the seat ever since. In November 2013, three Maryland voters filed this lawsuit. They alleged that the 2011 Plan violated the First Amendment, the Elections Clause, and Article I, §2, of the Constitution. After considerable procedural skirmishing and litigation over preliminary relief, the District Court entered summary judgment for the plaintiffs. 348 F. Supp. 3d 493. It concluded that the plaintiffs’ claims were justiciable, and that the Plan violated the First Amendment by diminishing their “ability to elect their candidate of choice” because of their party affiliation and voting history, and by burdening their associational rights. Id., at 498. On the latter point, the court relied upon findings that Republicans in the Sixth District “were burdened in fundraising, attracting volunteers, campaigning, and generating interest in voting in an atmosphere of general confusion and apathy.” Id., at 524. The District Court permanently enjoined the State from using the 2011 Plan and ordered it to promptly adopt a new plan for the 2020 election. Id., at 525. The defendants appealed directly to this Court under 28 U. S. C. §1253. We postponed jurisdiction. 586 U. S. ___ (2019). II A Article III of the Constitution limits federal courts to deciding “Cases” and “Controversies.” We have understood that limitation to mean that federal courts can address only questions “historically viewed as capable of resolution through the judicial process.” Flast v. Cohen, 392 U.S. 83, 95 (1968). In these cases we are asked to decide an important question of constitutional law. “But before we do so, we must find that the question is presented in a ‘case’ or ‘controversy’ that is, in James Madison’s words, ‘of a Judiciary Nature.’ ” DaimlerChrysler Corp. v. Cuno, 547 U.S. 332, 342 (2006) (quoting 2 Records of the Federal Convention of 1787, p. 430 (M. Farrand ed. 1966)). Chief Justice Marshall famously wrote that it is “the province and duty of the judicial department to say what the law is.” Marbury v. Madison, 1 Cranch 137, 177 (1803). Sometimes, however, “the law is that the judicial department has no business entertaining the claim of unlawfulness—because the question is entrusted to one of the political branches or involves no judicially enforceable rights.” Vieth v. Jubelirer, 541 U.S. 267, 277 (2004) (plurality opinion). In such a case the claim is said to present a “political question” and to be nonjusticiable—outside the courts’ competence and therefore beyond the courts’ jurisdiction. Baker v. Carr, 369 U.S. 186, 217 (1962). Among the political question cases the Court has identified are those that lack “judicially discoverable and manageable standards for resolving [them].” Ibid. Last Term in Gill v. Whitford, we reviewed our partisan gerrymandering cases and concluded that those cases “leave unresolved whether such claims may be brought.” 585 U. S., at ___ (slip op., at 13). This Court’s authority to act, as we said in Gill, is “grounded in and limited by the necessity of resolving, according to legal principles, a plaintiff’s particular claim of legal right.” Ibid. The question here is whether there is an “appropriate role for the Federal Judiciary” in remedying the problem of partisan gerrymandering—whether such claims are claims of legal right, resolvable according to legal principles, or political questions that must find their resolution elsewhere. Id., at ___ (slip op., at 8). B Partisan gerrymandering is nothing new. Nor is frustration with it. The practice was known in the Colonies prior to Independence, and the Framers were familiar with it at the time of the drafting and ratification of the Constitution. See Vieth, 541 U. S., at 274 (plurality opinion). During the very first congressional elections, George Washington and his Federalist allies accused Patrick Henry of trying to gerrymander Virginia’s districts against their candidates—in particular James Madison, who ultimately prevailed over fellow future President James Monroe. Hunter, The First Gerrymander? 9 Early Am. Studies 792–794, 811 (2011). See 5 Writings of Thomas Jefferson 71 (P. Ford ed. 1895) (Letter to W. Short (Feb. 9, 1789)) (“Henry has so modelled the districts for representatives as to tack Orange [county] to counties where he himself has great influence that Madison may not be elected into the lower federal house”). In 1812, Governor of Massachusetts and future Vice President Elbridge Gerry notoriously approved congressional districts that the legislature had drawn to aid the Democratic-Republican Party. The moniker “gerrymander” was born when an outraged Federalist newspaper observed that one of the misshapen districts resembled a salamander. See Vieth, 541 U. S., at 274 (plurality opinion); E. Griffith, The Rise and Development of the Gerrymander 17–19 (1907). “By 1840, the gerrymander was a recognized force in party politics and was generally attempted in all legislation enacted for the formation of election districts. It was generally conceded that each party would attempt to gain power which was not proportionate to its numerical strength.” Id., at 123. The Framers addressed the election of Representatives to Congress in the Elections Clause. Art. I, §4, cl. 1. That provision assigns to state legislatures the power to prescribe the “Times, Places and Manner of holding Elections” for Members of Congress, while giving Congress the power to “make or alter” any such regulations. Whether to give that supervisory authority to the National Government was debated at the Constitutional Convention. When those opposed to such congressional oversight moved to strike the relevant language, Madison came to its defense: “[T]he State Legislatures will sometimes fail or refuse to consult the common interest at the expense of their local coveniency or prejudices. . . . Whenever the State Legislatures had a favorite measure to carry, they would take care so to mould their regulations as to favor the candidates they wished to succeed.” 2 Records of the Federal Convention of 1787, at 240–241. During the subsequent fight for ratification, the provision remained a subject of debate. Antifederalists predicted that Congress’s power under the Elections Clause would allow Congress to make itself “omnipotent,” setting the “time” of elections as never or the “place” in difficult to reach corners of the State. Federalists responded that, among other justifications, the revisionary power was necessary to counter state legislatures set on undermining fair representation, including through malapportionment. M. Klarman, The Framers’ Coup: The Making of the United States Constitution 340–342 (2016). The Federalists were, for example, concerned that newly developing population centers would be deprived of their proper electoral weight, as some cities had been in Great Britain. See 6 The Documentary History of the Ratification of the Constitution: Massachusetts 1278–1279 (J. Kaminski & G. Saladino eds. 2000). Congress has regularly exercised its Elections Clause power, including to address partisan gerrymandering. The Apportionment Act of 1842, which required single-member districts for the first time, specified that those districts be “composed of contiguous territory,” Act of June 25, 1842, ch. 47, 5Stat. 491, in “an attempt to forbid the practice of the gerrymander,” Griffith, supra, at 12. Later statutes added requirements of compactness and equality of population. Act of Jan. 16, 1901, ch. 93, §3, 31Stat. 733; Act of Feb. 2, 1872, ch. 11, §2, 17Stat. 28. (Only the single member district requirement remains in place today. 2 U. S. C. §2c.) See Vieth, 541 U. S., at 276 (plurality opinion). Congress also used its Elections Clause power in 1870, enacting the first comprehensive federal statute dealing with elections as a way to enforce the Fifteenth Amendment. Force Act of 1870, ch. 114, 16Stat. 140. Starting in the 1950s, Congress enacted a series of laws to protect the right to vote through measures such as the suspension of literacy tests and the prohibition of English-only elections. See, e.g., 52 U. S. C. §10101 et seq. Appellants suggest that, through the Elections Clause, the Framers set aside electoral issues such as the one before us as questions that only Congress can resolve. See Baker, 369 U. S., at 217. We do not agree. In two areas—one-person, one-vote and racial gerrymandering—our cases have held that there is a role for the courts with respect to at least some issues that could arise from a State’s drawing of congressional districts. See Wesberry v. Sanders, 376 U.S. 1 (1964); Shaw v. Reno, 509 U.S. 630 (1993) (Shaw I ). But the history is not irrelevant. The Framers were aware of electoral districting problems and considered what to do about them. They settled on a characteristic approach, assigning the issue to the state legislatures, expressly checked and balanced by the Federal Congress. As Alexander Hamilton explained, “it will . . . not be denied that a discretionary power over elections ought to exist somewhere. It will, I presume, be as readily conceded that there were only three ways in which this power could have been reasonably modified and disposed: that it must either have been lodged wholly in the national legislature, or wholly in the State legislatures, or primarily in the latter, and ultimately in the former.” The Federalist No. 59, p. 362 (C. Rossiter ed. 1961). At no point was there a suggestion that the federal courts had a role to play. Nor was there any indication that the Framers had ever heard of courts doing such a thing. C Courts have nevertheless been called upon to resolve a variety of questions surrounding districting. Early on, doubts were raised about the competence of the federal courts to resolve those questions. See Wood v. Broom, 287 U.S. 1 (1932); Colegrove v. Green, 328 U.S. 549 (1946). In the leading case of Baker v. Carr, voters in Tennessee complained that the State’s districting plan for state representatives “debase[d]” their votes, because the plan was predicated on a 60-year-old census that no longer reflected the distribution of population in the State. The plaintiffs argued that votes of people in overpopulated districts held less value than those of people in less-populated districts, and that this inequality violated the Equal Protection Clause of the Fourteenth Amendment. The District Court dismissed the action on the ground that the claim was not justiciable, relying on this Court’s precedents, including Colegrove. Baker v. Carr, 179 F. Supp. 824, 825, 826 (MD Tenn. 1959). This Court reversed. It identified various considerations relevant to determining whether a claim is a nonjusticiable political question, including whether there is “a lack of judicially discover- able and manageable standards for resolving it.” 369 U. S., at 217. The Court concluded that the claim of population inequality among districts did not fall into that category, because such a claim could be decided under basic equal protection principles. Id., at 226. In Wesberry v. Sanders, the Court extended its ruling to malapportionment of congressional districts, holding that Article I, §2, required that “one man’s vote in a congressional election is to be worth as much as another’s.” 376 U. S., at 8. Another line of challenges to districting plans has focused on race. Laws that explicitly discriminate on the basis of race, as well as those that are race neutral on their face but are unexplainable on grounds other than race, are of course presumptively invalid. The Court applied those principles to electoral boundaries in Gomillion v. Lightfoot, concluding that a challenge to an “uncouth twenty-eight sided” municipal boundary line that excluded black voters from city elections stated a constitutional claim. 364 U.S. 339, 340 (1960). In Wright v. Rockefeller, 376 U.S. 52 (1964), the Court extended the reasoning of Gomillion to congressional districting. See Shaw I, 509 U. S., at 645. Partisan gerrymandering claims have proved far more difficult to adjudicate. The basic reason is that, while it is illegal for a jurisdiction to depart from the one-person, one-vote rule, or to engage in racial discrimination in districting, “a jurisdiction may engage in constitutional political gerrymandering.” Hunt v. Cromartie, 526 U.S. 541, 551 (1999) (citing Bush v. Vera, 517 U.S. 952, 968 (1996); Shaw v. Hunt, 517 U.S. 899, 905 (1996) (Shaw II); Miller v. Johnson, 515 U.S. 900, 916 (1995); Shaw I, 509 U. S., at 646). See also Gaffney v. Cummings, 412 U.S. 735, 753 (1973) (recognizing that “[p]olitics and political considerations are inseparable from districting and apportionment”). To hold that legislators cannot take partisan interests into account when drawing district lines would essentially countermand the Framers’ decision to entrust districting to political entities. The “central problem” is not determining whether a jurisdiction has engaged in partisan gerrymandering. It is “determining when political gerrymandering has gone too far.” Vieth, 541 U. S., at 296 (plurality opinion). See League of United Latin American Citizens v. Perry, 548 U.S. 399, 420 (2006) (LULAC) (opinion of Kennedy, J.) (difficulty is “providing a standard for deciding how much partisan dominance is too much”). We first considered a partisan gerrymandering claim in Gaffney v. Cummings in 1973. There we rejected an equal protection challenge to Connecticut’s redistricting plan, which “aimed at a rough scheme of proportional representation of the two major political parties” by “wiggl[ing] and joggl[ing] boundary lines” to create the appropriate number of safe seats for each party. 412 U. S., at 738, 752, n. 18 (internal quotation marks omitted). In upholding the State’s plan, we reasoned that districting “inevitably has and is intended to have substantial political consequences.” Id., at 753. Thirteen years later, in Davis v. Bandemer, we addressed a claim that Indiana Republicans had cracked and packed Democrats in violation of the Equal Protection Clause. 478 U.S. 109, 116–117 (1986) (plurality opinion). A majority of the Court agreed that the case was justiciable, but the Court splintered over the proper standard to apply. Four Justices would have required proof of “intentional discrimination against an identifiable political group and an actual discriminatory effect on that group.” Id., at 127. Two Justices would have focused on “whether the boundaries of the voting districts have been distorted deliberately and arbitrarily to achieve illegitimate ends.” Id., at 165 (Powell, J., concurring in part and dissenting in part). Three Justices, meanwhile, would have held that the Equal Protection Clause simply “does not supply judicially manageable standards for resolving purely political gerrymandering claims.” Id., at 147 (O’Connor, J., concurring in judgment). At the end of the day, there was “no ‘Court’ for a standard that properly should be applied in determining whether a challenged redistricting plan is an unconstitutional partisan political gerrymander.” Id., at 185, n. 25 (opinion of Powell, J.). In any event, the Court held that the plaintiffs had failed to show that the plan violated the Constitution. Eighteen years later, in Vieth, the plaintiffs complained that Pennsylvania’s legislature “ignored all traditional redistricting criteria, including the preservation of local government boundaries,” in order to benefit Republican congressional candidates. 541 U. S., at 272–273 (plurality opinion) (brackets omitted). Justice Scalia wrote for a four-Justice plurality. He would have held that the plaintiffs’ claims were nonjusticiable because there was no “judicially discernible and manageable standard” for deciding them. Id., at 306. Justice Kennedy, concurring in the judgment, noted “the lack of comprehensive and neutral principles for drawing electoral boundaries [and] the absence of rules to limit and confine judicial intervention.” Id., at 306–307. He nonetheless left open the possibility that “in another case a standard might emerge.” Id., at 312. Four Justices dissented. In LULAC, the plaintiffs challenged a mid-decade redistricting map approved by the Texas Legislature. Once again a majority of the Court could not find a justiciable standard for resolving the plaintiffs’ partisan gerrymandering claims. See 548 U. S., at 414 (noting that the “disagreement over what substantive standard to apply” that was evident in Bandemer “persists”). As we summed up last Term in Gill, our “considerable efforts in Gaffney, Bandemer, Vieth, and LULAC leave unresolved whether . . . claims [of legal right] may be brought in cases involving allegations of partisan gerrymandering.” 585 U. S., at ___ (slip op., at 13). Two “threshold questions” remained: standing, which we addressed in Gill, and “whether [such] claims are justiciable.” Ibid. III A In considering whether partisan gerrymandering claims are justiciable, we are mindful of Justice Kennedy’s counsel in Vieth: Any standard for resolving such claims must be grounded in a “limited and precise rationale” and be “clear, manageable, and politically neutral.” 541 U. S., at 306–308 (opinion concurring in judgment). An important reason for those careful constraints is that, as a Justice with extensive experience in state and local politics put it, “[t]he opportunity to control the drawing of electoral boundaries through the legislative process of apportionment is a critical and traditional part of politics in the United States.” Bandemer, 478 U. S., at 145 (opinion of O’Connor, J.). See Gaffney, 412 U. S., at 749 (observing that districting implicates “fundamental ‘choices about the nature of representation’ ” (quoting Burns v. Richardson, 384 U.S. 73, 92 (1966))). An expansive standard requiring “the correction of all election district lines drawn for partisan reasons would commit federal and state courts to unprecedented intervention in the American political process,” Vieth, 541 U. S., at 306 (opinion of Kennedy, J.). As noted, the question is one of degree: How to “provid[e] a standard for deciding how much partisan dominance is too much.” LULAC, 548 U. S., at 420 (opinion of Kennedy, J.). And it is vital in such circumstances that the Court act only in accord with especially clear standards: “With uncertain limits, intervening courts—even when proceeding with best intentions—would risk assuming political, not legal, responsibility for a process that often produces ill will and distrust.” Vieth, 541 U. S., at 307 (opinion of Kennedy, J.). If federal courts are to “inject [themselves] into the most heated partisan issues” by adjudicating partisan gerrymandering claims, Bandemer, 478 U. S., at 145 (opinion of O’Connor, J.), they must be armed with a standard that can reliably differentiate unconstitutional from “constitutional political gerrymandering.” Cromartie, 526 U. S., at 551. B Partisan gerrymandering claims rest on an instinct that groups with a certain level of political support should enjoy a commensurate level of political power and influence. Explicitly or implicitly, a districting map is alleged to be unconstitutional because it makes it too difficult for one party to translate statewide support into seats in the legislature. But such a claim is based on a “norm that does not exist” in our electoral system—“statewide elections for representatives along party lines.” Bandemer, 478 U. S., at 159 (opinion of O’Connor, J.). Partisan gerrymandering claims invariably sound in a desire for proportional representation. As Justice O’Connor put it, such claims are based on “a conviction that the greater the departure from proportionality, the more suspect an apportionment plan becomes.” Ibid. “Our cases, however, clearly foreclose any claim that the Constitution requires proportional representation or that legislatures in reapportioning must draw district lines to come as near as possible to allocating seats to the contending parties in proportion to what their anticipated statewide vote will be.” Id., at 130 (plurality opinion). See Mobile v. Bolden, 446 U.S. 55, 75–76 (1980) (plurality opinion) (“The Equal Protection Clause of the Fourteenth Amendment does not require proportional representation as an imperative of political organization.”). The Founders certainly did not think proportional representation was required. For more than 50 years after ratification of the Constitution, many States elected their congressional representatives through at-large or “general ticket” elections. Such States typically sent single-party delegations to Congress. See E. Engstrom, Partisan Gerry- mandering and the Construction of American Democracy 43–51 (2013). That meant that a party could garner nearly half of the vote statewide and wind up without any seats in the congressional delegation. The Whigs in Alabama suffered that fate in 1840: “their party garnered 43 percent of the statewide vote, yet did not receive a single seat.” Id., at 48. When Congress required single-member districts in the Apportionment Act of 1842, it was not out of a general sense of fairness, but instead a (mis)calculation by the Whigs that such a change would improve their electoral prospects. Id., at 43–44. Unable to claim that the Constitution requires proportional representation outright, plaintiffs inevitably ask the courts to make their own political judgment about how much representation particular political parties deserve—based on the votes of their supporters—and to rearrange the challenged districts to achieve that end. But federal courts are not equipped to apportion political power as a matter of fairness, nor is there any basis for concluding that they were authorized to do so. As Justice Scalia put it for the plurality in Vieth: “ ‘Fairness’ does not seem to us a judicially manage- able standard. . . . Some criterion more solid and more demonstrably met than that seems to us necessary to enable the state legislatures to discern the limits of their districting discretion, to meaningfully constrain the discretion of the courts, and to win public acceptance for the courts’ intrusion into a process that is the very foundation of democratic decisionmaking.” 541 U. S., at 291. The initial difficulty in settling on a “clear, manageable and politically neutral” test for fairness is that it is not even clear what fairness looks like in this context. There is a large measure of “unfairness” in any winner-take-all system. Fairness may mean a greater number of competitive districts. Such a claim seeks to undo packing and cracking so that supporters of the disadvantaged party have a better shot at electing their preferred candidates. But making as many districts as possible more competitive could be a recipe for disaster for the disadvantaged party. As Justice White has pointed out, “[i]f all or most of the districts are competitive . . . even a narrow statewide preference for either party would produce an overwhelming majority for the winning party in the state legislature.” Bandemer, 478 U. S., at 130 (plurality opinion). On the other hand, perhaps the ultimate objective of a “fairer” share of seats in the congressional delegation is most readily achieved by yielding to the gravitational pull of proportionality and engaging in cracking and packing, to ensure each party its “appropriate” share of “safe” seats. See id., at 130–131 (“To draw district lines to maximize the representation of each major party would require creating as many safe seats for each party as the demographic and predicted political characteristics of the State would permit.”); Gaffney, 412 U. S., at 735–738. Such an approach, however, comes at the expense of competitive districts and of individuals in districts allocated to the opposing party. Or perhaps fairness should be measured by adherence to “traditional” districting criteria, such as maintaining political subdivisions, keeping communities of interest together, and protecting incumbents. See Brief for Bipartisan Group of Current and Former Members of the House of Representatives as Amici Curiae; Brief for Professor Wesley Pegden et al. as Amici Curiae in No. 18–422. But protecting incumbents, for example, enshrines a particular partisan distribution. And the “natural political geography” of a State—such as the fact that urban electoral districts are often dominated by one political party—can itself lead to inherently packed districts. As Justice Kennedy has explained, traditional criteria such as compactness and contiguity “cannot promise political neutrality when used as the basis for relief. Instead, it seems, a decision under these standards would unavoidably have significant political effect, whether intended or not.” Vieth, 541 U. S., at 308–309 (opinion concurring in judgment). See id., at 298 (plurality opinion) (“[P]acking and cracking, whether intentional or no, are quite consistent with adherence to compactness and respect for political subdivision lines”). Deciding among just these different visions of fairness (you can imagine many others) poses basic questions that are political, not legal. There are no legal standards discernible in the Constitution for making such judgments, let alone limited and precise standards that are clear, manageable, and politically neutral. Any judicial decision on what is “fair” in this context would be an “unmoored determination” of the sort characteristic of a political question beyond the competence of the federal courts. Zivotofsky v. Clinton, 566 U.S. 189, 196 (2012). And it is only after determining how to define fairness that you can even begin to answer the determinative question: “How much is too much?” At what point does permissible partisanship become unconstitutional? If compliance with traditional districting criteria is the fairness touchstone, for example, how much deviation from those criteria is constitutionally acceptable and how should mapdrawers prioritize competing criteria? Should a court “reverse gerrymander” other parts of a State to counteract “natural” gerrymandering caused, for example, by the urban concentration of one party? If a districting plan protected half of the incumbents but redistricted the rest into head to head races, would that be constitutional? A court would have to rank the relative importance of those traditional criteria and weigh how much deviation from each to allow. If a court instead focused on the respective number of seats in the legislature, it would have to decide the ideal number of seats for each party and determine at what point deviation from that balance went too far. If a 5–3 allocation corresponds most closely to statewide vote totals, is a 6–2 allocation permissible, given that legislatures have the authority to engage in a certain degree of partisan gerrymandering? Which seats should be packed and which cracked? Or if the goal is as many competitive districts as possible, how close does the split need to be for the district to be considered competitive? Presumably not all districts could qualify, so how to choose? Even assuming the court knew which version of fairness to be looking for, there are no discernible and manageable standards for deciding whether there has been a violation. The questions are “unguided and ill suited to the development of judicial standards,” Vieth, 541 U. S., at 296 (plurality opinion), and “results from one gerrymandering case to the next would likely be disparate and inconsistent,” id., at 308 (opinion of Kennedy, J.). Appellees contend that if we can adjudicate one-person, one-vote claims, we can also assess partisan gerrymandering claims. But the one-person, one-vote rule is relatively easy to administer as a matter of math. The same cannot be said of partisan gerrymandering claims, because the Constitution supplies no objective measure for assessing whether a districting map treats a political party fairly. It hardly follows from the principle that each person must have an equal say in the election of representatives that a person is entitled to have his political party achieve representation in some way commensurate to its share of statewide support. More fundamentally, “vote dilution” in the one-person, one-vote cases refers to the idea that each vote must carry equal weight. In other words, each representative must be accountable to (approximately) the same number of constituents. That requirement does not extend to political parties. It does not mean that each party must be influential in proportion to its number of supporters. As we stated unanimously in Gill, “this Court is not responsible for vindicating generalized partisan preferences. The Court’s constitutionally prescribed role is to vindicate the individual rights of the people appearing before it.” 585 U. S., at ___ (slip op., at 21). See also Bandemer, 478 U. S., at 150 (opinion of O’Connor, J.) (“[T]he Court has not accepted the argument that an ‘asserted entitlement to group representation’ . . . can be traced to the one person, one vote principle.” (quoting Bolden, 446 U. S., at 77)).[1] Nor do our racial gerrymandering cases provide an appropriate standard for assessing partisan gerrymandering. “[N]othing in our case law compels the conclusion that racial and political gerrymanders are subject to precisely the same constitutional scrutiny. In fact, our country’s long and persistent history of racial discrimination in voting—as well as our Fourteenth Amendment jurisprudence, which always has reserved the strictest scrutiny for discrimination on the basis of race—would seem to compel the opposite conclusion.” Shaw I, 509 U. S., at 650 (citation omitted). Unlike partisan gerrymandering claims, a racial gerrymandering claim does not ask for a fair share of political power and influence, with all the justiciability conundrums that entails. It asks instead for the elimination of a racial classification. A partisan gerrymandering claim cannot ask for the elimination of partisanship. IV Appellees and the dissent propose a number of “tests” for evaluating partisan gerrymandering claims, but none meets the need for a limited and precise standard that is judicially discernible and manageable. And none provides a solid grounding for judges to take the extraordinary step of reallocating power and influence between political parties. A The Common Cause District Court concluded that all but one of the districts in North Carolina’s 2016 Plan violated the Equal Protection Clause by intentionally diluting the voting strength of Democrats. 318 F. Supp. 3d, at 923. In reaching that result the court first required the plaintiffs to prove “that a legislative mapdrawer’s predominant purpose in drawing the lines of a particular district was to ‘subordinate adherents of one political party and entrench a rival party in power.’ ” Id., at 865 (quoting Arizona State Legislature v. Arizona Independent Redistricting Comm’n, 576 U. S. ___, ___ (2015) (slip op., at 1)). The District Court next required a showing “that the dilution of the votes of supporters of a disfavored party in a particular district—by virtue of cracking or packing—is likely to persist in subsequent elections such that an elected representative from the favored party in the district will not feel a need to be responsive to constituents who support the disfavored party.” 318 F. Supp. 3d, at 867. Finally, after a prima facie showing of partisan vote dilution, the District Court shifted the burden to the defendants to prove that the discriminatory effects are “attributable to a legitimate state interest or other neutral explanation.” Id., at 868. The District Court’s “predominant intent” prong is borrowed from the racial gerrymandering context. In racial gerrymandering cases, we rely on a “predominant intent” inquiry to determine whether race was, in fact, the reason particular district boundaries were drawn the way they were. If district lines were drawn for the purpose of separating racial groups, then they are subject to strict scrutiny because “race-based decisionmaking is inherently suspect.” Miller, 515 U. S., at 915. See Bush, 517 U. S., at 959 (principal opinion). But determining that lines were drawn on the basis of partisanship does not indicate that the districting was improper. A permissible intent—securing partisan advantage—does not become constitutionally impermissible, like racial discrimination, when that permissible intent “predominates.” The District Court tried to limit the reach of its test by requiring plaintiffs to show, in addition to predominant partisan intent, that vote dilution “is likely to persist” to such a degree that the elected representative will feel free to ignore the concerns of the supporters of the minority party. 318 F. Supp. 3d, at 867. But “[t]o allow district courts to strike down apportionment plans on the basis of their prognostications as to the outcome of future elections . . . invites ‘findings’ on matters as to which neither judges nor anyone else can have any confidence.” Bandemer, 478 U. S., at 160 (opinion of O’Connor, J.). See LULAC, 548 U. S., at 420 (opinion of Kennedy, J.) (“[W]e are wary of adopting a constitutional standard that invalidates a map based on unfair results that would occur in a hypothetical state of affairs.”). And the test adopted by the Common Cause court requires a far more nuanced prediction than simply who would prevail in future political contests. Judges must forecast with unspecified certainty whether a prospective winner will have a margin of victory sufficient to permit him to ignore the supporters of his defeated opponent (whoever that may turn out to be). Judges not only have to pick the winner—they have to beat the point spread. The appellees assure us that “the persistence of a party’s advantage may be shown through sensitivity testing: probing how a plan would perform under other plausible electoral conditions.” Brief for Appellees League of Women Voters of North Carolina et al. in No. 18–422, p. 55. See also 318 F. Supp. 3d, at 885. Experience proves that accurately predicting electoral outcomes is not so simple, either because the plans are based on flawed assumptions about voter preferences and behavior or because demographics and priorities change over time. In our two leading partisan gerrymandering cases themselves, the predictions of durability proved to be dramatically wrong. In 1981, Republicans controlled both houses of the Indiana Legislature as well as the governorship. Democrats challenged the state legislature districting map enacted by the Republicans. This Court in Bandemer rejected that challenge, and just months later the Democrats increased their share of House seats in the 1986 elections. Two years later the House was split 50–50 between Democrats and Republicans, and the Democrats took control of the chamber in 1990. Democrats also challenged the Pennsylvania congressional districting plan at issue in Vieth. Two years after that challenge failed, they gained four seats in the delegation, going from a 12–7 minority to an 11–8 majority. At the next election, they flipped another Republican seat. Even the most sophisticated districting maps cannot reliably account for some of the reasons voters prefer one candidate over another, or why their preferences may change. Voters elect individual candidates in individual districts, and their selections depend on the issues that matter to them, the quality of the candidates, the tone of the candidates’ campaigns, the performance of an incumbent, national events or local issues that drive voter turnout, and other considerations. Many voters split their tickets. Others never register with a political party, and vote for candidates from both major parties at different points during their lifetimes. For all of those reasons, asking judges to predict how a particular districting map will perform in future elections risks basing constitutional holdings on unstable ground outside judicial expertise. It is hard to see what the District Court’s third prong—providing the defendant an opportunity to show that the discriminatory effects were due to a “legitimate redistricting objective”—adds to the inquiry. 318 F. Supp. 3d, at 861. The first prong already requires the plaintiff to prove that partisan advantage predominates. Asking whether a legitimate purpose other than partisanship was the motivation for a particular districting map just restates the question. B The District Courts also found partisan gerrymandering claims justiciable under the First Amendment, coalescing around a basic three-part test: proof of intent to burden individuals based on their voting history or party affiliation; an actual burden on political speech or associational rights; and a causal link between the invidious intent and actual burden. See Common Cause, 318 F. Supp. 3d, at 929; Benisek, 348 F. Supp. 3d, at 522. Both District Courts concluded that the districting plans at issue violated the plaintiffs’ First Amendment right to association. The District Court in North Carolina relied on testimony that, after the 2016 Plan was put in place, the plaintiffs faced “difficulty raising money, attracting candidates, and mobilizing voters to support the political causes and issues such Plaintiffs sought to advance.” 318 F. Supp. 3d, at 932. Similarly, the District Court in Maryland examined testimony that “revealed a lack of enthusiasm, indifference to voting, a sense of disenfranchisement, a sense of disconnection, and confusion,” and concluded that Republicans in the Sixth District “were burdened in fundraising, attracting volunteers, campaigning, and generating interest in voting.” 348 F. Supp. 3d, at 523–524. To begin, there are no restrictions on speech, association, or any other First Amendment activities in the districting plans at issue. The plaintiffs are free to engage in those activities no matter what the effect of a plan may be on their district. The plaintiffs’ argument is that partisanship in districting should be regarded as simple discrimination against supporters of the opposing party on the basis of political viewpoint. Under that theory, any level of partisanship in districting would constitute an infringement of their First Amendment rights. But as the Court has explained, “[i]t would be idle . . . to contend that any political consideration taken into account in fashioning a reapportionment plan is sufficient to invalidate it.” Gaffney, 412 U. S., at 752. The First Amendment test simply describes the act of districting for partisan advantage. It provides no standard for determining when partisan activity goes too far. As for actual burden, the slight anecdotal evidence found sufficient by the District Courts in these cases shows that this too is not a serious standard for separating constitutional from unconstitutional partisan gerrymandering. The District Courts relied on testimony about difficulty drumming up volunteers and enthusiasm. How much of a decline in voter engagement is enough to constitute a First Amendment burden? How many door knocks must go unanswered? How many petitions unsigned? How many calls for volunteers unheeded? The Common Cause District Court held that a partisan gerrymander places an unconstitutional burden on speech if it has more than a “de minimis” “chilling effect or adverse impact” on any First Amendment activity. 318 F. Supp. 3d, at 930. The court went on to rule that there would be an adverse effect “even if the speech of [the plaintiffs] was not in fact chilled”; it was enough that the districting plan “makes it easier for supporters of Republican candidates to translate their votes into seats,” thereby “enhanc[ing] the[ir] relative voice.” Id., at 933 (internal quotation marks omitted). These cases involve blatant examples of partisanship driving districting decisions. But the First Amendment analysis below offers no “clear” and “manageable” way of distinguishing permissible from impermissible partisan motivation. The Common Cause court embraced that conclusion, observing that “a judicially manageable framework for evaluating partisan gerrymandering claims need not distinguish an ‘acceptable’ level of partisan gerrymandering from ‘excessive’ partisan gerrymandering” because “the Constitution does not authorize state redistricting bodies to engage in such partisan gerrymandering.” Id., at 851. The decisions below prove the prediction of the Vieth plurality that “a First Amendment claim, if it were sustained, would render unlawful all consideration of political affiliation in districting,” 541 U. S., at 294, contrary to our established precedent. C The dissent proposes using a State’s own districting criteria as a neutral baseline from which to measure how extreme a partisan gerrymander is. The dissent would have us line up all the possible maps drawn using those criteria according to the partisan distribution they would produce. Distance from the “median” map would indicate whether a particular districting plan harms supporters of one party to an unconstitutional extent. Post, at 18–19, 25 (opinion of Kagan, J.). As an initial matter, it does not make sense to use criteria that will vary from State to State and year to year as the baseline for determining whether a gerrymander violates the Federal Constitution. The degree of partisan advantage that the Constitution tolerates should not turn on criteria offered by the gerrymanderers themselves. It is easy to imagine how different criteria could move the median map toward different partisan distributions. As a result, the same map could be constitutional or not depending solely on what the mapmakers said they set out to do. That possibility illustrates that the dissent’s proposed constitutional test is indeterminate and arbitrary. Even if we were to accept the dissent’s proposed baseline, it would return us to “the original unanswerable question (How much political motivation and effect is too much?).” Vieth, 541 U. S., at 296–297 (plurality opinion). Would twenty percent away from the median map be okay? Forty percent? Sixty percent? Why or why not? (We appreciate that the dissent finds all the unanswerable questions annoying, see post, at 22, but it seems a useful way to make the point.) The dissent’s answer says it all: “This much is too much.” Post, at 25–26. That is not even trying to articulate a standard or rule. The dissent argues that there are other instances in law where matters of degree are left to the courts. See post, at 27. True enough. But those instances typically involve constitutional or statutory provisions or common law confining and guiding the exercise of judicial discretion. For example, the dissent cites the need to determine “substantial anticompetitive effect[s]” in antitrust law. Post, at 27 (citing Ohio v. American Express Co., 585 U. S. ___ (2018)). That language, however, grew out of the Sherman Act, understood from the beginning to have its “origin in the common law” and to be “familiar in the law of this country prior to and at the time of the adoption of the [A]ct.” Standard Oil Co. of N. J. v. United States, 221 U.S. 1, 51 (1911). Judges began with a significant body of law about what constituted a legal violation. In other cases, the pertinent statutory terms draw meaning from related provisions or statutory context. Here, on the other hand, the Constitution provides no basis whatever to guide the exercise of judicial discretion. Common experience gives content to terms such as “substantial risk” or “substantial harm,” but the same cannot be said of substantial deviation from a median map. There is no way to tell whether the prohibited deviation from that map should kick in at 25 percent or 75 percent or some other point. The only provision in the Constitution that specifically addresses the matter assigns it to the political branches. See Art. I, §4, cl. 1. D The North Carolina District Court further concluded that the 2016 Plan violated the Elections Clause and Article I, §2. We are unconvinced by that novel approach. Article I, §2, provides that “[t]he House of Representatives shall be composed of Members chosen every second Year by the People of the several States.” The Elections Clause provides that “[t]he Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations, except as to the Places of chusing Senators.” Art. I, §4, cl. 1. The District Court concluded that the 2016 Plan exceeded the North Carolina General Assembly’s Elections Clause authority because, among other reasons, “the Elections Clause did not empower State legislatures to disfavor the interests of supporters of a particular candidate or party in drawing congressional districts.” 318 F. Supp. 3d, at 937. The court further held that partisan gerrymandering infringes the right of “the People” to select their representatives. Id., at 938–940. Before the District Court’s decision, no court had reached a similar conclusion. In fact, the plurality in Vieth concluded—without objection from any other Justice—that neither §2 nor §4 of Article I “provides a judicially enforceable limit on the political considerations that the States and Congress may take into account when districting.” 541 U. S., at 305. The District Court nevertheless asserted that partisan gerrymanders violate “the core principle of [our] republican government” preserved in Art. I, §2, “namely, that the voters should choose their representatives, not the other way around.” 318 F. Supp. 3d, at 940 (quoting Arizona State Legislature, 576 U. S., at ___ (slip op., at 35); internal quotation marks omitted; alteration in original). That seems like an objection more properly grounded in the Guarantee Clause of Article IV, §4, which “guarantee[s] to every State in [the] Union a Republican Form of Government.” This Court has several times concluded, however, that the Guarantee Clause does not provide the basis for a justiciable claim. See, e.g., Pacific States Telephone & Telegraph Co. v. Oregon, 223 U.S. 118 (1912). V Excessive partisanship in districting leads to results that reasonably seem unjust. But the fact that such gerrymandering is “incompatible with democratic principles,” Arizona State Legislature, 576 U. S., at ___ (slip op., at 1), does not mean that the solution lies with the federal judiciary. We conclude that partisan gerrymandering claims present political questions beyond the reach of the federal courts. Federal judges have no license to reallocate political power between the two major political parties, with no plausible grant of authority in the Constitution, and no legal standards to limit and direct their decisions. “[J]udicial action must be governed by standard, by rule,” and must be “principled, rational, and based upon reasoned distinctions” found in the Constitution or laws. Vieth, 541 U. S., at 278, 279 (plurality opinion). Judicial review of partisan gerrymandering does not meet those basic requirements. Today the dissent essentially embraces the argument that the Court unanimously rejected in Gill: “this Court can address the problem of partisan gerrymandering because it must.” 585 U. S., at ___ (slip op., at 12). That is not the test of our authority under the Constitution; that document instead “confines the federal courts to a properly judicial role.” Town of Chester v. Laroe Estates, Inc., 581 U. S. ___, ___ (2017) (slip op., at 4). What the appellees and dissent seek is an unprecedented expansion of judicial power. We have never struck down a partisan gerrymander as unconstitutional—despite various requests over the past 45 years. The expansion of judicial authority would not be into just any area of controversy, but into one of the most intensely partisan aspects of American political life. That intervention would be unlimited in scope and duration—it would recur over and over again around the country with each new round of districting, for state as well as federal representatives. Consideration of the impact of today’s ruling on democratic principles cannot ignore the effect of the unelected and politically unaccountable branch of the Federal Government assuming such an extraordinary and unprecedented role. See post, at 32–33. Our conclusion does not condone excessive partisan gerrymandering. Nor does our conclusion condemn complaints about districting to echo into a void. The States, for example, are actively addressing the issue on a number of fronts. In 2015, the Supreme Court of Florida struck down that State’s congressional districting plan as a violation of the Fair Districts Amendment to the Florida Constitution. League of Women Voters of Florida v. Detzner, 172 So. 3d 363 (2015). The dissent wonders why we can’t do the same. See post, at 31. The answer is that there is no “Fair Districts Amendment” to the Federal Constitution. Provisions in state statutes and state constitutions can provide standards and guidance for state courts to apply. (We do not understand how the dissent can maintain that a provision saying that no districting plan “shall be drawn with the intent to favor or disfavor a political party” provides little guidance on the question. See post, at 31, n. 6.) Indeed, numerous other States are restricting partisan considerations in districting through legislation. One way they are doing so is by placing power to draw electoral districts in the hands of independent commissions. For example, in November 2018, voters in Colorado and Michigan approved constitutional amendments creating multimember commissions that will be responsible in whole or in part for creating and approving district maps for congressional and state legislative districts. See Colo. Const., Art. V, §§44, 46; Mich. Const., Art. IV, §6. Missouri is trying a different tack. Voters there overwhelmingly approved the creation of a new position—state demographer—to draw state legislative district lines. Mo. Const., Art. III, §3. Other States have mandated at least some of the traditional districting criteria for their mapmakers. Some have outright prohibited partisan favoritism in redistricting. See Fla. Const., Art. III, §20(a) (“No apportionment plan or individual district shall be drawn with the intent to favor or disfavor a political party or an incumbent.”); Mo. Const., Art. III, §3 (“Districts shall be designed in a manner that achieves both partisan fairness and, secondarily, competitiveness. ‘Partisan fairness’ means that parties shall be able to translate their popular support into legislative representation with approximately equal efficiency.”); Iowa Code §42.4(5) (2016) (“No district shall be drawn for the purpose of favoring a political party, incumbent legislator or member of Congress, or other person or group.”); Del. Code Ann., Tit. xxix, §804 (2017) (providing that in determining district boundaries for the state legislature, no district shall “be created so as to unduly favor any person or political party”). As noted, the Framers gave Congress the power to do something about partisan gerrymandering in the Elections Clause. The first bill introduced in the 116th Congress would require States to create 15-member independent commissions to draw congressional districts and would establish certain redistricting criteria, including protection for communities of interest, and ban partisan gerrymandering. H. R. 1, 116th Cong., 1st Sess., §§2401, 2411 (2019). Dozens of other bills have been introduced to limit reliance on political considerations in redistricting. In 2010, H. R. 6250 would have required States to follow standards of compactness, contiguity, and respect for political subdivisions in redistricting. It also would have prohibited the establishment of congressional districts “with the major purpose of diluting the voting strength of any person, or group, including any political party,” except when necessary to comply with the Voting Rights Act of 1965. H. R. 6250, 111th Cong., 2d Sess., §2 (referred to committee). Another example is the Fairness and Independence in Redistricting Act, which was introduced in 2005 and has been reintroduced in every Congress since. That bill would require every State to establish an independent commission to adopt redistricting plans. The bill also set forth criteria for the independent commissions to use, such as compactness, contiguity, and population equality. It would prohibit consideration of voting history, political party affiliation, or incumbent Representative’s residence. H. R. 2642, 109th Cong., 1st Sess., §4 (referred to subcommittee). We express no view on any of these pending proposals. We simply note that the avenue for reform established by the Framers, and used by Congress in the past, remains open. * * * No one can accuse this Court of having a crabbed view of the reach of its competence. But we have no commission to allocate political power and influence in the absence of a constitutional directive or legal standards to guide us in the exercise of such authority. “It is emphatically the province and duty of the judicial department to say what the law is.” Marbury v. Madison, 1 Cranch, at 177. In this rare circumstance, that means our duty is to say “this is not law.” The judgments of the United States District Court for the Middle District of North Carolina and the United States District Court for the District of Maryland are vacated, and the cases are remanded with instructions to dismiss for lack of jurisdiction. It is so ordered. Notes 1 The dissent’s observation that the Framers viewed political parties “with deep suspicion, as fomenters of factionalism and symptoms of disease in the body politic” post, at 9, n. 1 (opinion of Kagan, J.) (internal quotation marks and alteration omitted), is exactly right. Its inference from that fact is exactly wrong. The Framers would have been amazed at a constitutional theory that guarantees a certain degree of representation to political parties.

587.US.2018_17-1606

The Social Security Act permits judicial review of “any final decision . . . after a hearing” by the Social Security Administration (SSA). 42 U. S. C. §405(g). Claimants for, as relevant here, supplemental security income disability benefits under Title XVI of the Act must generally proceed through a four-step administrative process in order to obtain federal-court review: (1) seek an initial determination of eligibility; (2) seek reconsideration of that determination; (3) request a hearing before an administrative law judge (ALJ); and (4) seek review of the ALJ’s decision by the SSA’s Appeals Council. See 20 CFR §416.1400. A request for Appeals Council review generally must be made within 60 days of receiving the ALJ’s ruling, §416.1468; if the claimant misses the deadline and cannot show good cause for doing so, the Appeals Council dismisses the request, §416.1471. Petitioner Ricky Lee Smith’s claim for disability benefits under Title XVI was denied at the initial-determination stage, upon reconsideration, and on the merits after a hearing before an ALJ. The Appeals Council later dismissed Smith’s request for review as untimely. Smith sought judicial review of the dismissal in a Federal District Court, which held that it lacked jurisdiction to hear the suit. The Sixth Circuit affirmed, maintaining that the Appeals Council’s dismissal of an untimely petition is not a “final decision” subject to federal-court review. Held: An Appeals Council dismissal on timeliness grounds after a claimant has had an ALJ hearing on the merits qualifies as a “final decision . . . made after a hearing” for purposes of allowing judicial review under §405(g). Pp. 5–16. (a) The statute’s text supports this reading. In the first clause (“any final decision”), the phrase “final decision” clearly denotes some kind of terminal event, and Congress’ use of “any” suggests an intent to use that term “expansive[ly],” Ali v. Federal Bureau of Prisons, 552 U.S. 214, 218–219. The Appeals Council’s dismissal of Smith’s claim fits that language: The SSA’s regulations make it the final stage of review. See 20 CFR §416.1472. As for the second clause (“made after a hearing”), Smith obtained the kind of hearing that §405(g) most naturally suggests: an ALJ hearing on the merits. This case differs from Califano v. Sanders, 430 U.S. 99, where the Court found that the SSA’s denial of a claimant’s petition to reopen a prior denial of his claim for benefits—a second look that the agency had made available to claimants as a matter of grace—was not a final decision under §405(g). Here, by contrast, the SSA’s “final decision” is much more closely tethered to the relevant “hearing.” A primary application for benefits may not be denied without an ALJ hearing (if requested), §405(b)(1), and a claimant’s access to this first bite at the apple is a matter of legislative right rather than agency grace. There is also no danger here of thwarting Congress’ own deadline, where the only potential untimeliness concerns Smith’s request for Appeals Council review, not his request for judicial review following the agency’s ultimate determination. Pp. 6–9. (b) The statutory context also weighs in Smith’s favor. Appeals from SSA determinations are, by their nature, appeals from the action of a federal agency. In the separate administrative-law context of Administrative Procedure Act (APA) review, an action is “final” if it both (1) “mark[s] the ‘consummation’ of the agency’s decisionmaking process” and (2) is “one by which ‘rights or obligations have been determined,’ or from which ‘legal consequences will flow.’ ” Bennett v. Spear, 520 U.S. 154, 177–178. Both conditions are satisfied when a Social Security claimant has reached the final step of the SSA’s four-step process and has had his request for review dismissed as untimely. While the administrative-exhaustion requirement “should be applied with regard for the particular administrative scheme at issue,” Weinberger v. Salfi, 422 U.S. 749, 765, the differences between the two Acts here suggest that Congress wanted more oversight by the courts rather than less under §405(g) and that “Congress designed [the statute as a whole] to be ‘unusually protective’ of claimants,” Bowen v. City of New York, 476 U.S. 467, 480. SSA is also a massive enterprise and mistakes will occur; Congress did not suggest that it intended for this claimant-protective statute to leave a claimant with no recourse to the courts if a mistake does happen. Pp. 9–10. (c) Smith’s entitlement to judicial review is confirmed by “the strong presumption that Congress intends judicial review of administrative action.” Bowen v. Michigan Academy of Family Physicians, 476 U.S. 667, 670. The heavy burden for rebutting this presumption is not met here. Congress left it to the SSA to define the procedures that claimants like Smith must first pass through, but it has not suggested that it intended for the SSA to be the unreviewable arbiter of whether claimants have complied with those procedures. Pp. 10–11. (d) The arguments of amicus in support of the judgment do not alter this conclusion. Amicus first argues that the phrase “final decision . . . made after a hearing” refers to a conclusive disposition, after exhaustion, of a benefits claim on the merits. However, this Court’s precedents do not support that reading; the Appeals Council’s dismissal is not merely collateral but an end to a proceeding in which a substantial factual record has already been developed and on which considerable resources have already been expended; and Smith’s case is distinct from Sanders. Amicus also claims that permitting greater judicial review could risk a flood of litigation, given the large volume of claims handled by the SSA, but that result is unlikely, because the number of Appeals Council untimeliness dismissals is comparatively small, and because data from the Eleventh Circuit, which follows the interpretation adopted here, do not bear out amicus’ warning. Third, amicus flags related contexts that could be informed by this ruling, but those issues are not before the Court. Finally, amicus argues that §405(g) is ambiguous and that the SSA’s longstanding interpretation of its meaning—prior to a change of position in this case—is entitled to deference under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, but this is not the kind of question on which courts defer to agencies. Pp. 11–14. (e) A reviewing court that disagrees with the procedural ground for the Appeals Council dismissal should in the ordinary case remand the case to allow the agency to address substantive issues in the first place. While there would be jurisdiction for a court to reach the merits, this general rule comports with fundamental administrative-law principles and is confirmed by the Court’s cases discussing exhaustion in the Social Security context, see City of New York, 476 U. S., at 485. Pp. 14–16. 880 F.3d 813, reversed and remanded. Sotomayor, J., delivered the opinion for a unanimous Court.

The Social Security Act allows for judicial review of “any final decision . . . made after a hearing” by the Social Security Administration (SSA). 42 U. S. C. §405(g). Petitioner Ricky Lee Smith was denied Social Security benefits after a hearing by an administrative law judge (ALJ) and later had his appeal from that denial dismissed as untimely by the SSA’s Appeals Council—the agency’s final decisionmaker. This case asks whether the Appeals Council’s dismissal of Smith’s claim is a “final decision . . . made after a hearing” so as to allow judicial review under §405(g). We hold that it is. I A Congress enacted the Social Security Act in 1935, responding to the crisis of the Great Depression. 49Stat. 620; F. Bloch, Social Security Law and Practice 13 (2012). In its early days, the program was administered by a body called the Social Security Board; that role has since passed on to the Board’s successor, the SSA.[1] In 1939, Congress amended the Act, adding various provisions that—subject to changes not at issue here—continue to govern cases like this one. See Social Security Act Amendments of 1939, ch. 666, 53Stat. 1360. First, Congress gave the agency “full power and authority to make rules and regulations and to establish procedures . . . necessary or appropriate to carry out” the Act. §405(a). Second, Congress directed the agency “to make findings of fac[t] and decisions as to the rights of any individual applying for a payment” and to provide all eligible claimants—that is, people seeking benefits—with an “opportunity for a hearing with respect to such decision[s].” §405(b)(1). Third, and most centrally, Congress provided for judicial review of “any final decision of the [agency] made after a hearing.” §405(g). At the same time, Congress made clear that review would be available only “as herein provided”—that is, only under the terms of §405(g). §405(h); see Heckler v. Ringer, 466 U.S. 602, 614–615 (1984). In 1940, the Social Security Board created the Appeals Council, giving it responsibility for overseeing and reviewing the decisions of the agency’s hearing officers (who, today, are ALJs).[2] Though the Appeals Council originally had just three members, its ranks have since swelled to include over 100 individuals serving as either judges or officers.[3] The Appeals Council remains a creature of regulatory rather than statutory creation. Today, the Social Security Act provides disability benefits under two programs, known by their statutory headings as Title II and Title XVI. See §401 et seq. (Title II); §1381 et seq. (Title XVI). Title II “provides old-age, survivor, and disability benefits to insured individuals irrespective of financial need.” Bowen v. Galbreath, 485 U.S. 74, 75 (1988). Title XVI provides supplemental security income benefits “to financially needy individuals who are aged, blind, or disabled regardless of their insured status.” Ibid. The regulations that govern the two programs are, for today’s purposes, equivalent. See Sims v. Apfel, 530 U.S. 103, 107, n. 2 (2000).[4] Likewise, §405(g) sets the terms of judicial review for each. See §1383(c)(3). Modern-day claimants must generally proceed through a four-step process before they can obtain review from a federal court. First, the claimant must seek an initial determination as to his eligibility. Second, the claimant must seek reconsideration of the initial determination. Third, the claimant must request a hearing, which is conducted by an ALJ. Fourth, the claimant must seek review of the ALJ’s decision by the Appeals Council. See 20 CFR §416.1400. If a claimant has proceeded through all four steps on the merits, all agree, §405(g) entitles him to judicial review in federal district court.[5] The tension in this case stems from the deadlines that SSA regulations impose for seeking each successive stage of review. A party who seeks Appeals Council review, as relevant here, must file his request within 60 days of receiving the ALJ’s ruling, unless he can show “good cause for missing the deadline.” §416.1468. The Appeals Council’s review is discretionary: It may deny even a timely request without issuing a decision. See §416.1481. If a claimant misses the deadline and cannot show good cause, however, the Appeals Council does not deny the request but rather dismisses it. §416.1471. Dismissals are “binding and not subject to further review” by the SSA. §416.1472. The question here is whether a dismissal for untimeliness, after the claimant has had an ALJ hearing, is a “final decision . . . made after a hearing” for purposes of allowing judicial review under §405(g). B Petitioner Ricky Lee Smith applied for disability benefits under Title XVI in 2012. Smith’s claim was denied at the initial-determination stage and upon reconsideration. Smith then requested an ALJ hearing, which the ALJ held in February 2014 before issuing a decision denying Smith’s claim on the merits in March 2014. The parties dispute what happened next. Smith’s attorney says that he sent a letter requesting Appeals Council review in April 2014, well within the 60-day deadline. The SSA says that it has no record of receiving any such letter. In late September 2014, Smith’s attorney sent a copy of the letter that he assertedly had mailed in April. The SSA, noting that it had no record of prior receipt, counted the date of the request as the day that it received the copy. The Appeals Council accordingly determined that Smith’s submission was untimely, concluded that Smith lacked good cause for missing the deadline, and dismissed Smith’s request for review. Smith sought judicial review of that dismissal in the U. S. District Court for the Eastern District of Kentucky. The District Court held that it lacked jurisdiction to hear his suit. The U. S. Court of Appeals for the Sixth Circuit affirmed, maintaining that “an Appeals Council decision to refrain from considering an untimely petition for review is not a ‘final decision’ subject to judicial review in federal court.’ ” Smith v. Commissioner of Social Security, 880 F.3d 813, 814 (2018). Smith petitioned this Court for certiorari. Responding to Smith’s petition, the Government stated that while the Sixth Circuit’s decision accorded with the SSA’s longstanding position, the Government had “reexamined the question and concluded that its prior position was incorrect.” Brief for Respondent on Pet. for Cert. 15. We granted certiorari to resolve a conflict among the Courts of Appeals. 586 U. S. ___ (2018).[6] Because the Government agrees with Smith that the Appeals Council’s dismissal meets §405(g)’s terms, we appointed Deepak Gupta as amicus curiae to defend the judgment below. 586 U. S. ___ (2018). He has ably discharged his duties. II Section 405(g), as noted above, provides for judicial review of “any final decision . . . made after a hearing.” This provision, the Court has explained, contains two separate elements: first, a “jurisdictional” requirement that claims be presented to the agency, and second, a “waivable . . . requirement that the administrative remedies prescribed by the Secretary be exhausted.” Mathews v. Eldridge, 424 U.S. 319, 328 (1976). This case involves the latter, nonjurisdictional element of administrative exhaustion. While §405(g) delegates to the SSA the authority to dictate which steps are generally required, see Sims, 530 U. S., at 106, exhaustion of those steps may not only be waived by the agency, see Weinberger v. Salfi, 422 U.S. 749, 767 (1975), but also excused by the courts, see Bowen v. City of New York, 476 U.S. 467, 484 (1986); Eldridge, 424 U. S., at 330.[7] The question here is whether a dismissal by the Appeals Council on timeliness grounds after a claimant has received an ALJ hearing on the merits qualifies as a “final decision . . . made after a hearing” for purposes of allowing judicial review under §405(g). In light of the text, the context, and the presumption in favor of the reviewability of agency action, we conclude that it does. A We begin with the text. Taking the first clause (“any final decision”) first, we note that the phrase “final decision” clearly denotes some kind of terminal event,[8] and Congress’ use of the word “any” suggests an intent to use that term “expansive[ly],” see Ali v. Federal Bureau of Prisons, 552 U.S. 214, 218–219 (2008). The Appeals Council’s dismissal of Smith’s claim fits that language: Under the SSA’s own regulations, it was the final stage of review. See 20 CFR §416.1472. Turning to the second clause (“made after a hearing”), we note that this phrase has been the subject of some confusion over the years. On the one hand, the statute elsewhere repeatedly uses the word “hearing” to signify an ALJ hearing,[9] which suggests that, in the ordinary case, the phrase here too denotes an ALJ hearing. See, e.g., IBP, Inc. v. Alvarez, 546 U.S. 21, 34 (2005) (noting “the normal rule of statutory interpretation that identical words used in different parts of the same statute are generally presumed to have the same meaning”). On the other hand, the Court’s precedents make clear that an ALJ hearing is not an ironclad prerequisite for judicial review. See, e.g., City of New York, 476 U. S., at 484 (emphasizing the Court’s “ ‘intensely practical’ ” approach to the applicability of the exhaustion requirement and disapproving “mechanical application” of a set of factors). There is no need today to give §405(g) a definition for all seasons, because, in any event, this is a mine-run case and Smith obtained the kind of hearing that §405(g) most naturally suggests: an ALJ hearing on the merits.[10] In other words, even giving §405(g) a relatively strict reading, Smith appears to satisfy its terms.[11] Smith cannot, however, satisfy §405(g)’s “after a hearing” requirement as a matter of mere chronology.[12] In Califano v. Sanders, 430 U.S. 99 (1977), the Court considered whether the SSA’s denial of a claimant’s petition to reopen a prior denial of his claim for benefits qualified as a final decision under §405(g). Id., at 102–103, 107–109. The Court concluded that it did not, reasoning that a petition to reopen was a matter of agency grace that could be denied without a hearing altogether and that allowing judicial review would thwart Congress’ own deadline for seeking such review. See id., at 108–109. That the SSA’s denial of the petition to reopen (1) was conclusive and (2) postdated an ALJ hearing did not, alone, bring it within the meaning of §405(g). Here, by contrast, the SSA’s “final decision” is much more closely tethered to the relevant “hearing.” Unlike a petition to reopen, a primary application for benefits may not be denied without an ALJ hearing (assuming the claimant timely requests one, as Smith did). §405(b)(1). Moreover, the claimant’s access to this first bite at the apple is indeed a matter of legislative right rather than agency grace. See id., at 108. And, again unlike the situation in Sanders, there is no danger here of thwarting Congress’ own deadline, given that the only potential untimeliness here concerns Smith’s request for Appeals Council review—not his request for judicial review following the agency’s ultimate determination. B The statutory context weighs in Smith’s favor as well. Appeals from SSA determinations are, by their nature, appeals from the action of a federal agency, and in the separate administrative-law context of the Administrative Procedure Act (APA), an action is “final” if it both (1) “mark[s] the ‘consummation’ of the agency’s decisionmaking process” and (2) is “one by which ‘rights or obligations have been determined,’ or from which ‘legal consequences will flow.’ ” Bennett v. Spear, 520 U.S. 154, 177–178 (1997). Both conditions are satisfied when a Social Secu- rity claimant has reached the fourth and final step of the SSA’s four-step process and has had his request for review dismissed as untimely. It is consistent to treat the Appeals Council’s dismissal of Smith’s claim as a final decision as well. To be clear, “the doctrine of administrative exhaustion should be applied with a regard for the particular administrative scheme at issue,” Salfi, 422 U. S., at 765, and we leave this axiom undisturbed today. The Social Security Act and the APA are different statutes, and courts must remain sensitive to their differences. See, e.g., Sullivan v. Hudson, 490 U.S. 877, 885 (1989) (observing that “[a]s provisions for judicial review of agency action go, §405(g) is somewhat unusual” in that its “detailed provisions . . . suggest a degree of direct interaction between a federal court and an administrative agency alien to” APA review). But at least some of these differences suggest that Congress wanted more oversight by the courts in this context rather than less, see ibid.,[13] and the statute as a whole is one that “Congress designed to be ‘unusually protective’ of claimants,” City of New York, 476 U. S., at 480. We note further that the SSA is a massive enterprise,[14] and mistakes will occur. See Brief for National Organization of Social Security Claimants’ Representatives as Amicus Curiae 13 (collecting examples).[15] The four steps preceding judicial review, meanwhile, can drag on for years.[16] While mistakes by the agency may be admirably rare, we do not presume that Congress intended for this claimant-protective statute, see City of New York, 476 U. S., at 480, to leave a claimant without recourse to the courts when such a mistake does occur—least of all when the claimant may have already expended a significant amount of likely limited resources in a lengthy proceeding. C Smith’s entitlement to judicial review is confirmed by “the strong presumption that Congress intends judicial review of administrative action.” Bowen v. Michigan Academy of Family Physicians, 476 U.S. 667, 670 (1986). “That presumption,” of course, “is rebuttable: It fails when a statute’s language or structure demonstrates that Congress wanted an agency to police its own conduct.” Mach Mining, LLC v. EEOC, 575 U.S. 480, ___–___ (2015) (slip op., at 4–5). But the burden for rebutting it is “ ‘heavy,’ ” id., at ___ (slip op., at 5), and that burden is not met here. While Congress left it to the SSA to define the procedures that claimants like Smith must first pass through, see Sims, 530 U. S., at 106, Congress has not suggested that it intended for the SSA to be the unreviewable arbiter of whether claimants have complied with those procedures. Where, as here, a claimant has received a claim-ending timeliness determination from the agency’s last-in-line decisionmaker after bringing his claim past the key procedural post (a hearing) mentioned in §405(g), there has been a “final decision . . . made after a hearing” under §405(g).[17] III Amicus’ arguments to the contrary have aided our consideration of this case, but they have not dissuaded us from concluding that the Appeals Council’s dismissal of Smith’s claim satisfied §405(g). Amicus first argues that the phrase “final decision . . . made after a hearing” refers to a conclusive disposition, after exhaustion, of a benefits claim on the merits—that is, on a basis for which the Social Security Act entitles a claimant to a hearing. This reading follows, amicus argues, from the Court’s observations that §405(g) generally requires exhaustion, and moreover from Sanders’ suggestion, see 430 U. S., at 108, that review is not called for where a claimant loses on an agency-determined procedural ground that is divorced from the substantive matters for which a hearing is required. Even if Smith did receive a hearing on the merits, amicus argues, the conclusive determination was not on that basis, and “[i]t would be unnatural to read the statute as throwing open the gates to judicial review of any final decision, no matter how collateral,” just because such a hearing occurred. Brief for Court-Appointed Amicus Curiae 34. We disagree. First, as noted above, the Court’s precedents do not make exhaustion a pure necessity, indicating instead that while the SSA is empowered to define the steps claimants must generally take, the SSA is not also the unreviewable arbiter of whether a claimant has sufficiently complied with those steps. See supra, at 5–6, and n. 7. Second, the Appeals Council’s dismissal is not merely collateral; such a dismissal calls an end to a proceeding in which a substantial factual record has already been developed and on which considerable resources have already been expended. See supra, at 10, and n. 16. Accepting amicus’ argument would mean that a claimant could make it to the end of the SSA’s process and then have judicial review precluded simply because the Appeals Council stamped “untimely” on the request, even if that designation were patently inaccurate. While there may be contexts in which the law is so unforgiving, this is not one. See supra, at 9–11. Smith’s case, as noted above, is also distinct from Sanders. See supra, at 8. Sanders, after all, involved the SSA’s denial of a petition for reopening—a second look that the agency had made available to claimants as a matter of grace. See 430 U. S., at 101–102, 107–108. But Smith is not seeking a second look at an already-final denial; he argues that he was wrongly prevented from continuing to pursue his primary claim for benefits. That primary claim, meanwhile, is indeed a matter of statutory entitlement. See §405(b). Amicus also emphasizes that the SSA handles a large volume of claims, such that a decision providing for greater judicial review could risk a flood of litigation. That result seems unlikely for a few reasons. First, the number of Appeals Council untimeliness dismissals is comparatively small—something on the order of 2,500 dismissals out of 160,000 dispositions per year.[18] Second, the interpretation that Smith and the Government urge has been the law since 1983 in the Eleventh Circuit, and the data there do not bear out amicus’ warning. See Reply Brief for Respondent 14–15 (collecting statistics). Third, while amicus flags related contexts that could be informed by today’s ruling, see Brief for Court-Appointed Amicus Curiae 36–40, those issues are not before us. We therefore do not address them other than to reinforce that such questions must be considered in the light of “the particular administrative scheme at issue.” See Salfi, 422 U. S., at 765. Today’s decision, therefore, hardly knocks loose a line of dominoes. Finally, amicus argues that the meaning of §405(g) is ambiguous and that the SSA’s longstanding interpretation of §405(g)—prior to its changed position during the pendency of this case—is entitled to deference under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). The Government and Smith maintain that the statute unambiguously supports the Government’s new position, and Smith further asserts that deference is inappropriate where the Government itself has rejected the interpretation in question in its filings. We need not decide whether the statute is unambiguous or what to do with the curious situation of an amicus curiae seeking deference for an interpretation that the Government’s briefing rejects. Chevron deference “ ‘is premised on the theory that a statute’s ambiguity constitutes an implicit delegation from Congress to the agency to fill in the statutory gaps.’ ” King v. Burwell, 576 U. S. ___, ___ (2015) (slip op., at 8). The scope of judicial review, meanwhile, is hardly the kind of question that the Court presumes that Congress implicitly delegated to an agency. Indeed, roughly six years after Chevron was decided, the Court declined to give Chevron deference to the Secretary of Labor’s interpretation of a federal statute that would have foreclosed private rights of action under certain circumstances. See Adams Fruit Co. v. Barrett, 494 U.S. 638, 649–650 (1990). As the Court explained, Congress’ having created “a role for the Department of Labor in administering the statute” did “not empower the Secretary to regulate the scope of the judicial power vested by the statute.” Id., at 650. Rather, “[a]lthough agency determinations within the scope of delegated authority are entitled to deference, it is fundamental ‘that an agency may not bootstrap itself into an area in which it has no jurisdiction.’ ” Ibid. Here, too, while Congress has empowered the SSA to create a scheme of administrative exhaustion, see Sims, 530 U. S., at 106, Congress did not delegate to the SSA the power to determine “the scope of the judicial power vested by” §405(g) or to determine conclusively when its dictates are satisfied. Adams Fruit Co., 494 U. S., at 650. Consequently, having concluded that Smith and the Government have the better reading of §405(g), we need go no further. IV Although they agree that §405(g) permits judicial review of the Appeals Council’s dismissal in this case, Smith and the Government disagree somewhat about the scope of review on remand.[19] Smith argues that if a reviewing court disagrees with the procedural ground for dismissal, it can then proceed directly to the merits, while the Government argues that the proper step in such a case would be to remand. We largely agree with the Government. To be sure, there would be jurisdiction for a federal court to proceed to the merits in the way that Smith avers. For one, as noted above, exhaustion itself is not a jurisdictional prerequisite. See supra, at 5–6. Moreover, §405(g) states that a reviewing “court shall have power to enter, upon the pleadings and transcript of the record, a judgment affirming, modifying, or reversing the decision of the Commissioner of Social Security, with or without remanding the cause for a rehearing”—a broad grant of authority that reflects the high “degree of direct interaction between a federal court and an administrative agency” envisioned by §405(g). Hudson, 490 U. S., at 885. In short, there is no jurisdictional bar to a court’s reaching the merits. Fundamental principles of administrative law, however, teach that a federal court generally goes astray if it decides a question that has been delegated to an agency if that agency has not first had a chance to address the question. See, e.g., INS v. Orlando Ventura, 537 U.S. 12, 16, 18 (2002) (per curiam); ICC v. Locomotive Engineers, 482 U.S. 270, 283 (1987); cf. SEC v. Chenery Corp., 318 U.S. 80, 88 (1943) (“For purposes of affirming no less than reversing its orders, an appellate court cannot intrude upon the domain which Congress has exclusively entrusted to an administrative agency”). The Court’s cases discussing exhaustion in the Social Security context confirm the prudence of applying this general principle here, where the agency’s final decisionmaker has not had a chance to address the merits at all.[20] See City of New York, 476 U. S., at 485 (“Because of the agency’s expertise in administering its own regulations, the agency ordinarily should be given the opportunity to review application of those regulations to a particular factual context”); Salfi, 422 U. S., at 765 (explaining that exhaustion serves to “preven[t] premature interference with agency processes” and to give the agency “an opportunity to correct its own errors,” “to afford the parties and the courts the benefit of its experience and expertise,” and to produce “a record which is adequate for judicial review”). Accordingly, in an ordinary case, a court should restrict its review to the procedural ground that was the basis for the Appeals Council dismissal and (if necessary) allow the agency to address any residual substantive questions in the first instance.[21] V We hold that where the SSA’s Appeals Council has dismissed a request for review as untimely after a claimant has obtained a hearing from an ALJ on the merits, that dismissal qualifies as a “final decision . . . made after a hearing” within the meaning of §405(g). The judgment of the United States Court of Appeals for the Sixth Circuit is therefore reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 See Koch & Koplow, The Fourth Bite at the Apple: A Study of the Operation and Utility of the Social Security Administration’s Appeals Council, 17 Fla. St. U. L. Rev. 199, 234–235 (1990) (Koch & Koplow). 2 See id., at 235. 3 SSA, Brief History and Current Information About the Appeals Council, https://www.ssa.gov/appeals/about_ac.html (all Internet mate-rials as last visited May 22, 2019). 4 Because Smith seeks benefits under Title XVI, we cite to the regulations that govern Title XVI, which are located at 20 CFR pt. 416 (2018). The regulations that govern Title II are located at 20 CFR pt. 404. 5 Of course, if the result at any of the four preceding stages is fully favorable, there is generally no need to proceed further. 6 Seven Courts of Appeals have held that there is no judicial review under these circumstances, while two have held that there is. Compare Brandtner v. Department of Health & Human Servs., 150 F.3d 1306, 1307 (CA10 1998); Bacon v. Sullivan, 969 F.2d 1517, 1520 (CA3 1992); Matlock v. Sullivan, 908 F.2d 492, 494 (CA9 1990); Harper v. Bowen, 813 F.2d 737, 743 (CA5 1987); Adams v. Heckler, 799 F.2d 131, 133 (CA4 1986); Smith v. Heckler, 761 F.2d 516, 518 (CA8 1985); Dietsch v. Schweiker, 700 F.2d 865, 867 (CA2 1983), with Casey v. Berryhill, 853 F.3d 322, 326 (CA7 2017); Bloodsworth v. Heckler, 703 F.2d 1233, 1239 (CA11 1983). 7 While Califano v. Sanders, 430 U.S. 99 (1977), can be read to cabin Eldridge and Salfi to only constitutional claims, the Court’s subsequent decision in City of New York demonstrates that this understanding of §405(g) can extend to cases lacking Eldridge’s and Salfi’s constitutional character. See City of New York, 476 U. S., at 474–475, and n. 5, 482–484; see also City of New York v. Heckler, 578 F. Supp. 1109, 1124–1125 (EDNY 1984) (ruling that the agency’s actions violated the Social Security Act and its own regulations and thus declining to reach the plaintiffs’ constitutional argument). 8 See 5 Oxford English Dictionary 920 (2d ed. 1989) (Final: “Marking the last stage of a process; leaving nothing to be looked for or expected; ultimate”); 4 Oxford English Dictionary 222 (1933) (same); see also Webster’s New World College Dictionary 542 (5th ed. 2016) (Final: “leaving no further chance for action, discussion, or change; deciding; conclusive”); Merriam-Webster’s Collegiate Dictionary 469 (11th ed. 2011) (Final: “coming at the end: being the last in a series, process, or progress”). 9 See 42 U. S. C. §405(b)(1) (entitling claimants to a hearing on the merits); §405(b)(2) (discussing “reconsideration” of certain findings “before any hearing under paragraph (1) on the issue of such entitlement”); §405(g) (discussing factual findings and evidence resulting from such a “hearing”); §405(h) (discussing binding effect of decision “after a hearing”); see also §§1383(c)(1)(A), (3) (similar). 10 We note as well that the “hearing” referred to in §405(g) cannot be a hearing before the Appeals Council. Congress provided for a hearing in §405(b) and for judicial review “after a hearing” in §405(g) before the Appeals Council even existed. See supra, at 2. Moreover, the Appeals Council makes many decisions without a hearing—e.g., denying a petition for review without giving reasons—that are nevertheless plainly reviewable. See 20 CFR §§ 416.1400(a)(5), 416.1467, 416.1481. Accordingly, the fact that there was no Appeals Council hearing—much like the fact that there was no reasoned Appeals Council decision on the merits—does not bar review. 11 We return below to the possibility, suggested by amicus, that “final decision . . . made after a hearing” could signify a final decision “on a matter on which the Act requires a hearing.” Brief for Court-Appointed Amicus Curiae 13; see infra, at 11–12. Here, we note only that while Congress certainly could have written something like “final decision on the merits . . . made after a hearing,” it did not. 12 The alternative risks untenable breadth. The Battle of Yorktown predates our ruling today, but no one would describe today’s opinion as a “decision made after the Battle of Yorktown.” As we explain, how-ever, the dismissal of Smith’s claim is tethered to Smith’s hearing in a way that more distant events are not. 13 The noteworthy counterpoint is §405(h), which withdraws federal-court jurisdiction under 28 U. S. C. §§ 1331, 1346. While that provision clearly serves “to route review through” §405(g), see Sanders, 430 U. S., at 103, n. 3; see also Heckler v. Ringer, 466 U.S. 602, 614–615 (1984), that routing choice does not simultaneously constrict the route that Congress did provide. 14 For example, the agency receives roughly 2.5 million new disability claims per year. See SSA, Annual Performance Report Fiscal Years 2017–2019, p. 32 (Feb. 12, 2018), https://www.ssa.gov/budget/FY19Files/2019APR.pdf. 15 See also Koch & Koplow 257 (noting that each Appeals Council member “typically spends only ten to fifteen minutes reviewing an average case” given “the pressures of the caseload”). 16 See SSA, FY 2020 Congressional Justification 9 (Mar. 2019) (estimating 2019 average processing time for the first three steps at 113 days, 105 days, and 515 days, respectively), https://www.ssa.gov/budget/FY20Files/FY20-JEAC.pdf; Brief for National Organization of Social Security Claimants’ Representatives as Amicus Curiae 11. 17 A different question would be presented by a claimant who assertedly faltered at an earlier step—e.g., whose request for an ALJ hearing was dismissed as untimely and who then appealed that determination to the Appeals Council before seeking judicial review. While such a claimant would not have received a “hearing” at all, the Court’s precedents also make clear that a hearing is not always required. See supra, at 5–6. Because such a situation is not before us, we do not address it. 18 See Brief for Respondent 43, n. 17 (number of timeliness dismissals); SSA, Annual Statistical Supplement 2018 (Table 2.F11) (number of dispositions), https://www.ssa.gov/policy/docs/statcomps/supplement/2018/2f8-2f11.pdf. 19 The parties agree, as do we, on the standard of review: abuse of discretion as to the overall conclusion, and “substantial evidence” “as to any fact.” See §405(g); see also Brief for Respondent 43–44; Tr. of Oral Arg. 5; cf. Bowen v. City of New York, 476 U.S. 467, 483 (1986) (“Ordinarily, the Secretary has discretion to decide when to waive the exhaustion requirement”). 20 We make no statement, by contrast, regarding the applicability of this line of cases to situations in which the Appeals Council has had a chance to address the merits. Cf. Sims v. Apfel, 530 U.S. 103, 110–112 (2000) (plurality opinion) (discussing why the inquisitorial nature of SSA proceedings counsels against imposing an issue-exhaustion requirement). 21 By the same token, remand may be forgone in rarer cases, such as where the Government joins the claimant in asking the court to reach the merits or where remand would serve no meaningful purpose.

586.US.2018_17-5554

Petitioner Stokeling pleaded guilty to possessing a firearm and ammunition after having been convicted of a felony, in violation of 18 U. S. C. §922(g)(1). Based on Stokeling’s prior criminal history, the probation office recommended the mandatory minimum 15-year prison term that the Armed Career Criminal Act (ACCA) provides for §922(g) violators who have three previous convictions “for a violent felony,” §924(e). As relevant here, Stokeling objected that his prior Florida robbery conviction was not a “violent felony,” which ACCA defines, in relevant part, as “any crime punishable by imprisonment for a term exceeding one year” that “has as an element the use, attempted use, or threatened use of physical force against the person of another,” §924(e)(2)(B)(i). The District Court held that Stokeling’s actions during the robbery did not justify an ACCA sentence enhancement, but the Eleventh Circuit reversed. Held: 1. ACCA’s elements clause encompasses a robbery offense that requires the defendant to overcome the victim’s resistance. Pp. 3–12. (a) As originally enacted, ACCA prescribed a sentence enhancement for certain individuals with three prior convictions “for robbery or burglary,” 18 U. S. C. App. §1202(a) (1982 ed., Supp. II), and defined robbery as an unlawful taking “by force or violence,” §1202(c)(8)—a clear reference to common-law robbery, which required a level of “force” or “violence” sufficient to overcome the resistance of the victim, however slight. When Congress amended ACCA two years later, it replaced the enumerated crimes with the elements clause, an expanded enumerated offenses clause, and the now-defunct residual clause. The new elements clause extended ACCA to cover any offense that has as an element “the use, attempted use, or threatened use of physical force,” §924(e)(2)(B)(i) (emphasis added). By replacing robbery with a clause that has “force” as its touchstone, Congress retained the same common-law definition that undergirded the definition of robbery in the original ACCA. This understanding is buttressed by the then widely accepted definitions of robbery among the States, a significant majority of which defined nonaggravated robbery as requiring a degree of force sufficient only to overcome a victim’s resistance. Under Stokeling’s reading, many of those state robbery statutes would not qualify as ACCA predicates. But federal criminal statutes should not be construed in ways that would render them inapplicable in many States. Pp. 3–8. (b) This understanding of “physical force” comports with Johnson v. United States, 559 U. S. 133. The force necessary for misdemeanor battery addressed in Johnson does not require resistance or even physical aversion on the part of the victim. Rather, the “slightest offensive touching” would qualify. Id., at 139. It is thus different in kind from the force necessary to overcome resistance by a victim, which is inherently “violent” in the sense contemplated by Johnson and “suggest[s] a degree of power that would not be satisfied by the merest touching.” Ibid. Johnson did not purport, as Stokeling suggests, to establish a force threshold so high as to exclude even robbery from ACCA’s scope. Pp. 8–10. (c) Stokeling’s suggested definition of “physical force”—force “reasonably expected to cause pain or injury”—is inconsistent with the degree of force necessary to commit robbery at common law. Moreover, the Court declined to adopt this standard in Johnson. Stokeling’s proposal would prove exceedingly difficult to apply, would impose yet another indeterminable line-drawing exercise on the lower courts, and is not supported by United States v. Castleman, 572 U. S. 157. Pp. 10–12. 2. Robbery under Florida law qualifies as an ACCA-predicate offense under the elements clause. The term “physical force” in ACCA encompasses the degree of force necessary to commit common-law robbery. And the Florida Supreme Court has made clear that the robbery statute requires “resistance by the victim that is overcome by the physical force of the offender.” Robinson v. State, 692 So. 2d 883, 886. Pp. 12–13. 684 Fed. Appx. 870, affirmed. Thomas, J., delivered the opinion of the Court, in which Breyer, Alito, Gorsuch, and Kavanaugh, JJ., joined. Sotomayor, J., filed a dissenting opinion, in which Roberts, C. J., and Ginsburg and Kagan, JJ., joined.

This case requires us to decide whether a robbery offense that has as an element the use of force sufficient to overcome a victim’s resistance necessitates the use of “physical force” within the meaning of the Armed Career Criminal Act (ACCA), 18 U. S. C. §924(e)(2)(B)(i). We conclude that it does. I In the early hours of July 27, 2015, two people burgled the Tongue & Cheek restaurant in Miami Beach, Florida. Petitioner Denard Stokeling was an employee of the restaurant, and the Miami Beach Police identified him as a suspect based on surveillance video from the burglary and witness statements. After conducting a criminal background check, police learned that Stokeling had previously been convicted of three felonies—home invasion, kidnaping, and robbery. When confronted, Stokeling admitted that he had a gun in his backpack. The detectives opened the backpack and discovered a 9-mm semiautomatic firearm, a magazine, and 12 rounds of ammunition. Stokeling pleaded guilty in federal court to possessing a firearm and ammunition after having been convicted of a felony, in violation of 18 U. S. C. §922(g)(1). The probation office recommended that Stokeling be sentenced as an armed career criminal under ACCA, which provides that a person who violates §922(g) and who has three previous convictions for a “violent felony” shall be imprisoned for a minimum of 15 years. §924(e). ACCA defines “violent felony” as “any crime punishable by imprisonment for a term exceeding one year” that “(i) has as an element the use, attempted use, or threatened use of physical force against the person of another; or “(ii) is burglary, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.” §924(e)(2)(B). As relevant here, Stokeling objected that his 1997 Florida robbery conviction was not a predicate offense under ACCA. This conviction, he argued, did not qualify under the first clause—the “elements clause”—because Florida robbery does not have “as an element the use, attempted use, or threatened use of physical force.” [1]* Under Florida law, robbery is defined as “the taking of money or other property . . . from the person or custody of another, . . . when in the course of the taking there is the use of force, violence, assault, or putting in fear.” Fla. Stat. §812.13(1) (1995). The Florida Supreme Court has explained that the “use of force” necessary to commit robbery requires “resistance by the victim that is overcome by the physical force of the offender.” Robinson v. State, 692 So. 2d 883, 886 (1997). Instead of applying a categorical approach to the elements clause, the District Court evaluated whether the facts of Stokeling’s robbery conviction were serious enough to warrant an enhancement. The court concluded that, although Stokeling “ ‘grabbed [the victim] by the neck and tried to remove her necklaces’ ” as she “ ‘held onto’ ” them, his actions did not “justify an enhancement.” Sentencing Hearing in 15–cv–20815 (SD Fla.), Doc. 45, pp. 10–11. The court then sentenced Stokeling to less than half of the mandatory minimum 15-year term of imprisonment provided by ACCA. The Eleventh Circuit reversed. 684 Fed. Appx. 870 (2017). It held that the District Court erred in making its own factual determination about the level of violence involved in Stokeling’s particular robbery offense. Id., at 871. The court also rejected Stokeling’s argument that Florida robbery does not categorically require sufficient force to constitute a violent felony under ACCA’s elements clause. Id., at 871–872. We granted certiorari to address whether the “force” required to commit robbery under Florida law qualifies as “physical force” for purposes of the elements clause. 584 U. S. ___ (2018). We now affirm. II Construing the language of the elements clause in light of the history of ACCA and our opinion in Johnson v. United States, 559 U. S. 133 (2010), we conclude that the elements clause encompasses robbery offenses that require the criminal to overcome the victim’s resistance. A As originally enacted, ACCA prescribed a 15-year minimum sentence for any person who received, possessed, or transported a firearm following three prior convictions “for robbery or burglary.” 18 U. S. C. App. §1202(a) (1982 ed., Supp. II). Robbery was defined in relevant part as “any felony consisting of the taking of the property of another from the person or presence of another by force or violence.” §1202(c)(8) (1982 ed., Supp. II) (emphasis added). The statute’s definition mirrored the elements of the common-law crime of robbery, which has long required force or violence. At common law, an unlawful taking was merely larceny unless the crime involved “violence.” 2 J. Bishop, Criminal Law §1156, p. 860 (J. Zane & C. Zollman eds., 9th ed. 1923). And “violence” was “committed if sufficient force [was] exerted to overcome the resistance encountered.” Id., at 861. A few examples illustrate the point. Under the common law, it was robbery “to seize another’s watch or purse, and use sufficient force to break a chain or guard by which it is attached to his person, or to run against another, or rudely push him about, for the purpose of diverting his attention and robbing him.” W. Clark & W. Marshall, Law of Crimes 554 (H. Lazell ed., 2d ed. 1905) (Clark & Marshall) (footnotes omitted). Similarly, it was robbery to pull a diamond pin out of a woman’s hair when doing so tore away hair attached to the pin. See 2 W. Russell, Crimes and Indictable Misdemeanors 68 (2d ed. 1828). But the crime was larceny, not robbery, if the thief did not have to overcome such resistance. In fact, common-law authorities frequently used the terms “violence” and “force” interchangeably. See ibid. (concluding that “if any injury be done to the person, or there be any struggle by the party to keep possession of the property before it be taken from him, there will be a sufficient actual ‘violence’ ” to establish robbery); Clark & Marshall 553 (“Sufficient force must be used to overcome resistance. . . . If there is any injury to the person of the owner, or if he resists the attempt to rob him, and his resistance is overcome, there is sufficient violence to make the taking robbery, however slight the resistance” (emphasis added)). The common law also did not distinguish between gradations of “violence.” If an act physically overcame a victim’s resistance, “however slight” that resistance might be, it necessarily constituted violence. Ibid.; 4 W. Blackstone, Commentaries on the Laws of England 242 (1769) (distinguishing “taking . . . by force” from “privately stealing,” and stating that the use of this “violence” differentiates robbery from other larcenies); see also 3 id., at 120 (explaining, in the battery context, that “the law cannot draw the line between different degrees of violence, and therefore totally prohibits the first and lowest stage of it”). The overlap between “force” and “violence” at common law is reflected in modern legal and colloquial usage of these terms. “Force” means “[p]ower, violence, or pressure directed against a person or thing,” Black’s Law Dictionary 656 (7th ed. 1999), or “unlawful violence threatened or committed against persons or property,” Random House Dictionary of the English Language 748 (2d ed. 1987). Likewise, “violence” implies force, including an “unjust or unwarranted use of force.” Black’s Law Dictionary, at 1564; accord, Random House Dictionary, at 2124 (“rough or injurious physical force, action, or treatment,” or “an unjust or unwarranted exertion of force or power, as against rights or laws”). Against this background, Congress, in the original ACCA, defined robbery as requiring the use of “force or violence”—a clear reference to the common law of robbery. See Samantar v. Yousuf, 560 U. S. 305, 320, n. 13 (2010) (“Congress ‘is understood to legislate against a background of common-law . . . principles’ ”). And the level of “force” or “violence” needed at common law was by this time well established: “Sufficient force must be used to overcome resistance . . . however slight the resistance.” Clark & Marshall 553. In 1986, Congress amended the relevant provisions of ACCA to their current form. The amendment was titled Expansion of Predicate Offenses for Armed Career Criminal Penalties. See Career Criminals Amendment Act of 1986, §1402, 100Stat. 3207–39. This amendment replaced the two enumerated crimes of “robbery or burglary” with the current elements clause, a new enumerated-offenses list, and a (now-defunct) residual clause. See Johnson v. United States, 576 U. S. ___ (2015). In the new statute, robbery was no longer enumerated as a predicate offense. But the newly created elements clause extended ACCA to cover any offense that has as an element “the use, attempted use, or threatened use of physical force.” 18 U. S. C. §924(e)(2)(B)(i) (2012 ed.) (emphasis added). “ ‘[I]f a word is obviously transplanted from another legal source, whether the common law or other legislation, it brings the old soil with it.’ ” Hall v. Hall, 584 U. S. ___, ___ (2018) (slip op., at 13) (quoting Frankfurter, Some Reflections on the Reading of Statutes, 47 Colum. L. Rev. 527, 537 (1947)). That principle supports our interpretation of the term “force” here. By retaining the term “force” in the 1986 version of ACCA and otherwise “[e]xpan[ding]” the predicate offenses under ACCA, Congress made clear that the “force” required for common-law robbery would be sufficient to justify an enhanced sentence under the new elements clause. We can think of no reason to read “force” in the revised statute to require anything more than the degree of “force” required in the 1984 statute. And it would be anomalous to read “force” as excluding the quintessential ACCA-predicate crime of robbery, despite the amendment’s retention of the term “force” and its stated intent to expand the number of qualifying offenses. The symmetry between the 1984 definition of robbery (requiring the use of “force or violence”) and the 1986 elements clause (requiring the use of “physical force”) is striking. By replacing robbery as an enumerated offense with a clause that has “force” as its touchstone, Congress made clear that “force” retained the same common-law definition that undergirded the original definition of robbery adopted a mere two years earlier. That conclusion is reinforced by the fact that the original 1984 statute defined “robbery” using terms with well-established common-law meanings. Our understanding of “physical force” is further buttressed by the then widely accepted definitions of robbery in the States. In 1986, a significant majority of the States defined nonaggravated robbery as requiring force that overcomes a victim’s resistance. The Government counts 43 States that measured force by this degree, 5 States that required “force” to cause bodily injury, and 2 States and the District of Columbia that permitted force to encompass something less, such as purse snatching. App. B to Brief for United States. Stokeling counters that, at most, 31 States defined force as overcoming victim resistance. Reply Brief 21. We need not declare a winner in this numbers game because, either way, it is clear that many States’ robbery statutes would not qualify as ACCA predicates under Stokeling’s reading. His reading would disqualify more than just basic-robbery statutes. Departing from the common-law understanding of “force” would also exclude other crimes that have as an element the force required to commit basic robbery. For instance, Florida requires the same element of “force” for both armed robbery and basic robbery. See Fla. Stat. §812.13(2)(a) (distinguishing armed robbery from robbery by requiring the additional element of “carr[ying] a firearm or other deadly weapon” during the robbery). Thus, as Stokeling’s counsel admitted at oral argument, “armed robbery in Florida” would not qualify under ACCA if his view were adopted. Tr. of Oral Arg. 3–4; see United States v. Lee, 886 F. 3d 1161, 1163, n. 1 (CA11 2018) (treating “Florida strong-arm robbery [i.e., basic robbery], armed robbery, and attempted robbery . . . the same for purposes of analyzing the ACCA’s elements clause”). Where, as here, the applicability of a federal criminal statute requires a state conviction, we have repeatedly declined to construe the statute in a way that would render it inapplicable in many States. See, e.g., United States v. Castleman, 572 U. S. 157, 167 (2014) (reading “physical force” to include common-law force, in part because a different reading would render 18 U. S. C. §922(g)(9) “ineffectual in at least 10 States”); Voisine v. United States, 579 U. S. ___, ___ (2016) (slip op., at 9) (declining to interpret §912(a)(33)(A) in a way that would “risk rendering §922(g)(9) broadly inoperative” in 34 States and the District of Columbia). That approach is appropriate here as well. B Our understanding of “physical force” comports with Johnson v. United States, 559 U. S. 133 (2010). There, the Court held that “ ‘actua[l] and intentiona[l] touching’ ”—the level of force necessary to commit common-law misdemeanor battery—did not require the “degree of force” necessary to qualify as a “violent felony” under ACCA’s elements clause. Id., at 138, 140. To reach this conclusion, the Court parsed the meaning of the phrase “physical force.” First, it explained that the modifier “physical” “plainly refers to force exerted by and through concrete bodies—distinguishing physical force, from, for example, intellectual force or emotional force.” Id., at 138. The Court then considered “whether the term ‘force’ in [the elements clause] has the specialized meaning that it bore in the common-law definition of battery.” Id., at 139. After reviewing the context of the statute, the Court rejected the Government’s suggestion that “force” encompassed even the “slightest offensive touching.” Ibid. Instead, it held that “physical force” means “violent force—that is, force capable of causing physical pain or injury to another person.” Id., at 140. Applying that standard to a Florida battery law criminalizing “any intentional physical contact,” the Court concluded that the law did not require the use of “physical force” within the meaning of ACCA. Ibid. Stokeling argues that Johnson rejected as insufficient the degree of “force” required to commit robbery under Florida law because it is not “substantial force.” We dis-agree. The nominal contact that Johnson addressed in- volved physical force that is different in kind from the violent force necessary to overcome resistance by a victim. The force necessary for misdemeanor battery does not require resistance or even physical aversion on the part of the victim; the “unwanted” nature of the physical contact itself suffices to render it unlawful. See State v. Hearns, 961 So. 2d 211, 216 (Fla. 2007). By contrast, the force necessary to overcome a victim’s physical resistance is inherently “violent” in the sense contemplated by Johnson, and “suggest[s] a degree of power that would not be satisfied by the merest touching.” 559 U. S., at 139. This is true because robbery that must overpower a victim’s will—even a feeble or weak-willed victim—necessarily involves a physical confrontation and struggle. The altercation need not cause pain or injury or even be prolonged; it is the physical contest between the criminal and the victim that is itself “capable of causing physical pain or injury.” Id., at 140. Indeed, Johnson itself relied on a definition of “physical force” that specifically encompassed robbery: “ ‘[f]orce consisting in a physical act, esp. a violent act directed against a robbery victim.’ ” Id., at 139 (quoting Black’s Law Dictionary 717 (9th ed. 2009); emphasis added). Robbery thus has always been within the “ ‘category of violent, active crimes’ ” that Congress included in ACCA. 559 U. S., at 140. To get around Johnson, Stokeling cherry picks adjectives from parenthetical definitions in the opinion, insisting that the level of force must be “severe,” “extreme,” “furious,” or “vehement.” These adjectives cannot bear the weight Stokeling would place on them. They merely supported Johnson’s actual holding: that common-law battery does not require “force capable of causing physical pain or injury.” Ibid. Johnson did not purport to establish a force threshold so high as to exclude even robbery from ACCA’s scope. Moreover, Stokeling ignores that the Court also defined “violence” as “ ‘unjust or improper force.’ ” Ibid. (emphasis added). As explained above, the common law similarly linked the terms “violence” and “force.” Overcoming a victim’s resistance was per se violence against the victim, even if it ultimately caused minimal pain or injury. See Russell, Crimes and Indictable Misdemeanors, at 68. C In the wake of Johnson, the Court has repeated its holding that “physical force” means “ ‘force capable of causing physical pain or injury.’ ” Sessions v. Dimaya, 584 U. S. ___, ___ (2018) (slip op., at 19–20) (quoting Johnson, supra, at 140); see also Castleman, supra, at 173–174 (Scalia, J., concurring in part and concurring in judgment). Finding this definition difficult to square with his position, Stokeling urges us to adopt a new, heightened reading of physical force: force that is “reasonably expected to cause pain or injury.” For the reasons already explained, that definition is inconsistent with the degree of force necessary to commit robbery at common law. Moreover, the Court declined to adopt that standard in Johnson, even after considering similar language employed in a nearby statutory provision, 18 U. S. C. §922(g)(8)(C)(ii). 559 U. S., at 143. The Court instead settled on “force capable of causing physical pain or injury.” Id., at 140 (emphasis added). “Capable” means “susceptible” or “having attributes . . . required for performance or accomplishment” or “having traits conducive to or features permitting.” Webster’s Ninth New Collegiate Dictionary 203 (1983); see also Oxford American Dictionary and Thesaurus 180 (2d ed. 2009) (“having the ability or quality necessary to do”). Johnson thus does not require any particular degree of likelihood or probability that the force used will cause physical pain or injury; only potentiality. Stokeling’s proposed standard would also prove exceedingly difficult to apply. Evaluating the statistical probability that harm will befall a victim is not an administrable standard under our categorical approach. Crimes can be committed in many different ways, and it would be difficult to assess whether a crime is categorically likely to harm the victim, especially when the statute at issue lacks fine-tuned gradations of “force.” We decline to impose yet another indeterminable line-drawing exercise on the lower courts. Stokeling next contends that Castleman held that minor uses of force do not constitute “violent force,” but he misreads that opinion. In Castleman, the Court noted that for purposes of a statute focused on domestic-violence misdemeanors, crimes involving relatively “minor uses of force” that might not “constitute ‘violence’ in the generic sense” could nevertheless qualify as predicate offenses. 572 U. S., at 165. The Court thus had no need to decide more generally whether, under Johnson, conduct that leads to relatively minor forms of injury—such as “a cut, abrasion, [or] bruise”—“necessitate[s]” the use of “violent force.” 572 U. S., at 170. Only Justice Scalia’s separate opinion addressed that question, and he concluded that force as small as “hitting, slapping, shoving, grabbing, pinching, biting, and hair pulling,” id., at 182 (alterations omitted), satisfied Johnson’s definition. He reasoned that “[n]one of those actions bears any real resemblance to mere offensive touching, and all of them are capable of causing physical pain or injury.” 572 U. S., at 182. This understanding of “physical force” is consistent with our holding today that force is “capable of causing physical injury” within the meaning of Johnson when it is sufficient to overcome a victim’s resistance. Such force satisfies ACCA’s elements clause. III We now apply these principles to Florida’s robbery statute to determine whether it “has as an element the use, attempted use, or threatened use of physical force against the person of another.” 18 U. S. C. §924(e)(2)(B)(i). We conclude that it does. As explained, Florida law defines robbery as “the taking of money or other property . . . from the person or custody of another, . . . when in the course of the taking there is the use of force, violence, assault, or putting in fear.” Fla. Stat. §812.13(1). The Florida Supreme Court has made clear that this statute requires “resistance by the victim that is overcome by the physical force of the offender.” Robinson v. State, 692 So. 2d 883, 886 (1997). Mere “snatching of property from another” will not suffice. Ibid. Several cases cited by the parties illustrate the application of the standard articulated in Robinson. For example, a defendant who grabs the victim’s fingers and peels them back to steal money commits robbery in Florida. Sanders v. State, 769 So. 2d 506, 507–508 (Fla. App. 2000). But a defendant who merely snatches money from the victim’s hand and runs away has not committed robbery. Goldsmith v. State, 573 So. 2d 445 (Fla. App. 1991). Similarly, a defendant who steals a gold chain does not use “ ‘force,’ within the meaning of the robbery statute,” simply because the victim “fe[els] his fingers on the back of her neck.” Walker v. State, 546 So. 2d 1165, 1166–1167 (Fla. App. 1989). It is worth noting that, in 1999, Florida enacted a separate “sudden snatching” statute that proscribes this latter category of conduct; under that statute, it is unnecessary to show either that the defendant “used any amount of force beyond that effort necessary to obtain possession of the money or other property” or that “[t]here was any resistance by the victim to the offender.” Fla. Stat. §812.131 (1999). Thus, the application of the categorical approach to the Florida robbery statute is straightforward. Because the term “physical force” in ACCA encompasses the degree of force necessary to commit common-law robbery, and because Florida robbery requires that same degree of “force,” Florida robbery qualifies as an ACCA-predicate offense under the elements clause. Cf. Descamps v. United States, 570 U. S. 254, 261 (2013) (“If the relevant statute has the same elemen[t],” “then the prior conviction can serve as an ACCA predicate”). IV In sum, “physical force,” or “force capable of causing physical pain or injury,” Johnson, 559 U. S., at 140, includes the amount of force necessary to overcome a victim’s resistance. Robbery under Florida law corresponds to that level of force and therefore qualifies as a “violent felony” under ACCA’s elements clause. For these reasons, we affirm the judgment of the Eleventh Circuit. It is so ordered. Notes 1 * The Government did not argue that Florida robbery should qualify under §924(e)(2)(B)(ii), presumably because robbery is not among the enumerated offenses and the Court held the “residual clause” unconstitutionally vague in Johnson v. United States, 576 U. S. ___ (2015).

587.US.2018_17-949

The Alaska National Interest Lands Conservation Act (ANILCA) set aside 104 million acres of federally owned land in Alaska for preservation purposes. With that land, ANILCA created ten new national parks, monuments, and preserves (areas known as “conservation system units”). 16 U. S. C. §3102(4). And in sketching those units’ boundary lines, Congress made an uncommon choice—to follow natural features rather than enclose only federally owned lands. It thus swept in a vast set of so-called inholdings—more than 18 million acres of state, Native, and private land. Had Congress done nothing more, those inholdings could have become subject to many National Park Service rules, as the Service has broad authority under its Organic Act to administer both lands and waters within parks across the country. 54 U. S. C. §100751. But Congress added Section 103(c), the provision principally in dispute in this case. Section 103(c)’s first sentence states that “[o]nly” the “public lands”—defined as most federally owned lands, waters, and associated interests—within any system unit’s boundaries are “deemed” a part of that unit. 16 U. S. C. §3103(c). The second sentence provides that no state, Native, or private lands “shall be subject to the regulations applicable solely to public lands within [system] units.” Ibid. And the third sentence permits the Service to “acquire such lands” from “the State, a Native Corporation, or other owner,” after which it may “administer[ ]” the land just as it does the other “public lands within such units.” Ibid. Petitioner John Sturgeon traveled for decades by hovercraft up a stretch of the Nation River that lies within the boundaries of the Yukon-Charley Preserve, a conservation system unit in Alaska. On one such trip, Park rangers informed him that the Service’s rules prohibit operating a hovercraft on navigable waters “located within [a park’s] boundaries.” 36 CFR §2.17(e). That regulation—issued under the Service’s Organic Act authority—applies to parks nationwide without any “regard to the ownership of submerged lands, tidelands, or lowlands.” §1.2(a)(3). Sturgeon complied with the order, but shortly thereafter sought an injunction that would allow him to resume using his hovercraft on his accustomed route. The District Court and the Ninth Circuit denied him relief, interpreting Section 103(c) to limit only the Service’s authority to impose Alaska-specific regulations on inholdings—not its authority to enforce nationwide regulations like the hovercraft rule. This Court granted review and rejected that ground for dismissal, but it remanded for consideration of two further questions: whether the Nation River “qualifies as ‘public land’ for purposes of ANILCA,” thus indisputably subjecting it to the Service’s regulatory authority; and, if not, whether the Service could nevertheless “regulate Sturgeon’s activities on the Nation River.” Sturgeon v. Frost, 577 U. S. ___, ___–___ (Sturgeon I). The Ninth Circuit never got past the first question, as it concluded that the Nation River was public land. Held: 1. The Nation River is not public land for purposes of ANILCA. “[P]ublic land” under ANILCA means (almost all) “lands, waters, and interests therein” the “title to which is in the United States.” 16 U. S. C. §3102(1)–(3). Because running waters cannot be owned, the United States does not have “title” to the Nation River in the ordinary sense. And under the Submerged Lands Act, it is the State of Alaska—not the United States—that holds “title to and ownership of the lands beneath [the River’s] navigable waters.” 43 U. S. C. §1311. The Service therefore argues that the United States has “title” to an “interest” in the Nation River under the reserved-water-rights doctrine, which provides that when the Federal Government reserves public land, it can retain rights to the specific “amount of water” needed to satisfy the purposes of that reservation. See Cappaert v. United States, 426 U.S. 128, 138–141. But even assuming that the Service held such a right, the Nation River itself would not thereby become “public land” in the way the Service contends. Under ANILCA, the “public land” would consist only of the Federal Government’s specific “interest” in the River—i.e., its reserved water right. And that right, the Service agrees, merely allows it to protect waters in the park from depletion or diversion. The right could not justify applying the hovercraft rule on the Nation River, as that rule targets nothing of the kind. Pp. 12–15. 2. Non-public lands within Alaska’s national parks are exempt from the Park Service’s ordinary regulatory authority. Section 103(c) arose out of concern from the State, Native Corporations, and private individuals that ANILCA’s broadly drawn boundaries might subject their properties to Park Service rules. Section 103(c)’s first sentence therefore sets out which land within those new parks qualify as parkland—“[o]nly” the “public lands” within any system unit’s boundaries are “deemed” a part of that unit. By negative implication, non-public lands are “deemed” outside the unit. In other words, non-federally owned lands inside system units (on a map) are declared outside them (for the law). The effect of that exclusion, as Section 103(c)’s second sentence affirms, is to exempt non-public lands, including waters, from Park Service regulations. That is, the Service’s rules will apply “solely” to public lands within the units. 16 U. S. C. §3103(c). And for that reason, the third sentence provides a kind of escape hatch—it allows the Service to acquire inholdings when it believes regulation of those lands is needed. The Service’s alternative interpretation of Section 103(c) is unpersuasive. The provision’s second sentence, it says, means that if a Park Service regulation on its face applies “solely” to public lands, then the regulation cannot apply to non-public lands. But if instead the regulation covers public and non-public lands alike, then the second sentence has nothing to say: The regulation can indeed cover both. On that view, Section 103(c)’s second sentence is a mere truism, not any kind of limitation. It does nothing to exempt inholdings from any regulation that might otherwise apply. And because that is so, the Government’s reading also strips the first and third sentences of their core functions. The first sentence’s “deeming” has no point, since there is no reason to pretend that inholdings are not part of a park if they can still be regulated as parklands. And the third sentence’s acquisition option has far less utility if the Service has its full regulatory authority over lands the Federal Government does not own. This sort of statute-gutting cannot be squared with ANILCA’s text and context. Pp. 16–26. 3. Navigable waters within Alaska’s national parks—no less than other non-public lands—are exempt from the Park Service’s normal regulatory authority. The Service argues that, if nothing else, ANILCA must at least allow it to regulate navigable waters. The Act, however, does not readily allow the decoupling of navigable waters from other non-federally owned areas in Alaskan national parks. ANILCA defines “land” to mean “lands, waters, and interests therein,” §3102(1)–(3); so when it refers to “lands” in Section 103(c) (and throughout the Act) it means waters as well. Nothing in the few aquatic provisions to which the Service points conflicts with reading Section 103(c)’s regulatory exemption to cover navigable waters. The Government largely relies on the Act’s statements of purpose, but this Court’s construction leaves the Service with multiple tools to “protect” and “preserve” rivers in Alaska’s national parks, as those provisions anticipate. See, e.g., §§3181(j), 3191(b)(7). While such authority might fall short of the Service’s usual power, it accords with ANILCA’s “repeated[ ] recogni[tion]” that Alaska is “the exception, not the rule.” Sturgeon I, 577 U. S., at ___. Pp. 26–29. 872 F.3d 927, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court. Sotomayor, J., filed a concurring opinion, in which Ginsburg, J., joined.

This Court first encountered John Sturgeon’s lawsuit three Terms ago. See Sturgeon v. Frost, 577 U. S. ___ (2016) (Sturgeon I ). As we explained then, Sturgeon hunted moose along the Nation River in Alaska for some 40 years. See id., at ___ (slip op., at 1). He traveled by hovercraft, an amphibious vehicle able to glide over land and water alike. To reach his favorite hunting ground, he would pilot the craft over a stretch of the Nation River that flows through the Yukon-Charley Rivers National Preserve, a unit of the federal park system managed by the National Park Service. On one such trip, park rangers informed Sturgeon that a Park Service regulation prohibits the use of hovercrafts on rivers within any federal preserve or park. Sturgeon complied with their order to remove his hovercraft from the Yukon-Charley, thus “heading home without a moose.” Id., at ___ (slip op., at 6). But soon afterward, Sturgeon sued the Park Service, seeking an injunction that would allow him to resume using his hovercraft on his accustomed route. The lower courts denied him relief. This Court, though, thought there was more to be said. See id., at ___–___ (slip op., at 15–16). As we put the matter then, Sturgeon’s case raises the issue how much “Alaska is different” from the rest of the country—how much it is “the exception, not the rule.” Id., at ___–___ (slip op., at 13–14). The rule, just as the rangers told Sturgeon, is that the Park Service may regulate boating and other activities on waters within national parks—and that it has banned the use of hovercrafts there. See 54 U. S. C. §100751(b); 36 CFR §2.17(e) (2018). But Sturgeon claims that Congress created an Alaska-specific exception to that broad authority when it enacted the Alaska National Interest Lands Conservation Act (ANILCA), 94Stat. 2371, 16 U. S. C. §3101 et seq. In Alaska, Sturgeon argues, the Park Service has no power to regulate lands or waters that the Federal Government does not own; rather, the Service may regulate only what ANILCA calls “public land” (essentially, federally owned land) in national parks. And, Sturgeon continues, the Federal Government does not own the Nation River—so the Service cannot ban hovercrafts there. When we last faced that argument, we disagreed with the reason the lower courts gave to reject it. But we remanded the case for consideration of two remaining questions. First, does “the Nation River qualif[y] as ‘public land’ for purposes of ANILCA”? 577 U. S., at ___ (slip op., at 15). Second, “even if the [Nation] is not ‘public land,’ ” does the Park Service have authority to “regulate Sturgeon’s activities” on the part of the river in the Yukon-Charley? Id., at ___ (slip op., at 16). Today, we take up those questions, and answer both “no.” That means Sturgeon can again rev up his hovercraft in search of moose. I A We begin, as Sturgeon I did, with a slice of Alaskan history. The United States purchased Alaska from Russia in 1867. It thereby acquired “[i]n a single stroke” 365 million acres of land—an area more than twice the size of Texas. Id., at ___ (slip op., at 2). You might think that would be enough to go around. But in the years since, the Federal Government and Alaskans (including Alaska Natives) have alternately contested and resolved and contested and . . . so forth who should own and manage that bounty. We offer here a few highlights because they are the backdrop against which Congress enacted ANILCA. As we do so, you might catch a glimpse of some former-day John Sturgeons—who (for better or worse) sought greater independence from federal control and, in the process, helped to shape the current law. For 90 years after buying Alaska, the Federal Government owned all its land. At first, those living in Alaska—a few settlers and some 30,000 Natives—were hardly aware of that fact. See E. Gruening, The State of Alaska 355 (1968). American citizens mocked the Alaska purchase as Secretary of State “Seward’s Folly” and President Johnson’s “Polar Bear Garden.” They paid no attention to the new area, leading to an “era of total neglect.” Id., at 31. But as Sturgeon I recounted, the turn of the century brought “newfound recognition of Alaska’s economic potential.” 577 U. S., at ___ (slip op., at 2). Opportunities to mine, trap, and fish attracted tens of thousands more settlers and sparked an emerging export economy. And partly because of that surge in commercial activity, the country’s foremost conservationists—President Theodore Roosevelt and Gifford Pinchot, chief of the fledgling Forest Service—took unprecedented action to protect Alaska’s natural resources. In particular, Roosevelt (and then President Taft) prevented settlers from logging or coal mining on substantial acreage. See W. Borneman, Alaska: Saga of a Bold Land 240–241 (2003). Alaskans responded by burning Pinchot in effigy and, more creatively, organizing the “Cordova Coal Party”—a mass dumping of imported Canadian coal (instead of English tea) into the Pacific Ocean (instead of Boston Harbor). See ibid. The terms of future conflict were thus set: resource conservation vs. economic development, federal management vs. local control. By the 1950s, Alaskans hankered for both statehood and land—and Congress decided to give them both. In pressing for statehood, Alaska’s delegate to the House of Representatives lamented that Alaskans were no better than “tenants upon the estate of the national landlord”; and Alaska’s Governor (then a Presidential appointee) called on the country to “[e]nd American [c]olonialism.” W. Everhart, The National Park Service 126–127 (1983) (Everhart). Ever more aware of Alaska’s economic and strategic importance, Congress agreed the time for statehood had come. The 1958 Alaska Statehood Act, 72Stat. 339, made Alaska the country’s 49th State. And because the new State would need property—to propel private industry and create a tax base—the Statehood Act made a land grant too. Over the next 35 years, Alaska could select for itself 103 million acres of “vacant, unappropri- ated, and unreserved” federal land—an area totaling the size of California. §§6(a)–(b), 72Stat. 340, as amended; see Everhart 127. And more: By incorporating the Submerged Lands Act of 1953, the Statehood Act gave Alaska “title to and ownership of the lands beneath navigable waters,” such as the Nation River. 43 U. S. C. §1311; see §6(m), 72Stat. 343. And a State’s title to the lands beneath navigable waters brings with it regulatory authority over “navigation, fishing, and other public uses” of those waters. United States v. Alaska, 521 U.S. 1, 5 (1997). All told, the State thus emerged a formidable property holder. But the State’s bonanza provoked land claims from Alaska Natives. Their ancestors had lived in the area for thousands of years, and they asserted aboriginal title to much of the property the State was now taking (and more besides). See Everhart 127. When their demands threatened to impede the trans-Alaska pipeline, Congress stepped in. The Alaska Native Claims Settlement Act of 1971 (ANCSA) extinguished the Natives’ aboriginal claims. See 85Stat. 688, as amended, 43 U. S. C. §1601 et seq. But it granted the Natives much in return. Under the law, corporations organized by groups of Alaska Natives could select for themselves 40 million acres of federal land—equivalent, when combined, to all of Pennsylvania. See §§1605, 1610–1615. So the Natives became large landowners too. Yet one more land dispute loomed. In addition to settling the Natives’ claims, ANCSA directed the Secretary of the Interior (Secretary) to designate, subject to congressional approval, 80 million more acres of federal land for inclusion in the national park, forest, or wildlife systems. See §1616(d)(2). The Secretary dutifully made his selections, but Congress failed to ratify them within the five-year period ANCSA had set. Rather than let the designations lapse, President Carter invoked another federal law (the 1906 Antiquities Act) to proclaim most of the lands (totaling 56 million acres) national monuments, under the National Park Service’s aegis. See 577 U. S., at ___ (slip op., at 4). Many Alaskans balked. “[R]egard[ing] national parks as just one more example of federal interference,” protesters demonstrated throughout the State and several thousand joined in the so-called Great Denali-McKinley Trespass. Everhart 129; see 577 U. S., at ___ (slip op., at 4). “The goal of the trespass,” as Sturgeon I explained, “was to break over 25 Park Service rules in a two-day period.” Ibid. One especially eager participant played a modern-day Paul Revere, riding on horseback through the crowd to deliver the message: “The Feds are coming! The Feds are coming!” Ibid. (internal quotation marks omitted). And so they were—but not in quite the way President Carter had contemplated. Responding to the uproar his proclamation had set off, Congress enacted a third major piece of legislation allocating land in Alaska. We thus reach ANILCA, the statute principally in dispute in this case, in which Congress set aside extensive land for national parks and preserves—but on terms different from those governing such areas in the rest of the country. B Starting with the statement of purpose in its first section, ANILCA sought to “balance” two goals, often thought conflicting. 16 U. S. C. §3101(d). The Act was designed to “provide[] sufficient protection for the national interest in the scenic, natural, cultural and environmental values on the public lands in Alaska.” Ibid. “[A]nd at the same time,” the Act was framed to “provide[] adequate opportunity for satisfaction of the economic and social needs of the State of Alaska and its people.” Ibid. So if, as you continue reading, you see some tension within the statute, you are not mistaken: It arises from Congress’s twofold ambitions. ANILCA set aside 104 million acres of federally owned land in Alaska for preservation purposes. See 577 U. S., at ___ (slip op., at 5). In doing so, the Act rescinded President Carter’s monument designations. But it brought into the national park, forest, or wildlife systems millions more acres than even ANCSA had contemplated. The park system’s share of the newly withdrawn land (to be administered, as usual, by the Park Service) was nearly 44 million acres—an amount that more than doubled the system’s prior (nationwide) size. See Everhart 132. With that land, ANILCA created ten new national parks, monuments, and preserves—including the Yukon-Charley Preserve—and expanded three old ones. See §§410hh, 410hh–1. In line with the Park Service’s usual terminol- ogy, ANILCA calls each such park or other area a “conservation system unit.” §3102(4) (“The term . . . means any unit in Alaska of the National Park System”); see 54 U. S. C. §100102(6) (similar). In sketching those units’ boundary lines, Congress made an uncommon choice—to follow “topographic or natural features,” rather than enclose only federally owned lands. §3103(b); see Brief for Respondents 24 (agreeing that “ANILCA [is] atypical in [this] respect”). In most parks outside Alaska, boundaries surround mainly federal property holdings. “[E]arly national parks were carved out of a larger public domain, in which virtually all land” was federally owned. Sax, Helpless Giants: The National Parks and the Regulation of Private Lands, 75 Mich. L. Rev. 239, 263 (1976); see Dept. of Interior, Nat. Park Serv., Statistical Abstract 87 (2017) (Table 9) (noting that only 2 of Yellowstone’s 2.2 million acres are in non-federal hands). And even in more recently established parks, Congress has used gerrymandered borders to exclude most non-federal land. See Sax, Buying Scenery, 1980 Duke L. J. 709, 712, and n. 12. But Congress had no real way to do that in Alaska. Its prior cessions of property to the State and Alaska Natives had created a “confusing patchwork of ownership” all but impossible to draw one’s way around. C. Naske & H. Slotnick, Alaska: A History 317 (3d ed. 2011). What’s more, an Alaskan Senator noted, the United States might want to reacquire state or Native holdings in the same “natural areas” as reserved federal land; that could occur most handily if Congress drew boundaries, “wherever possible, to encompass” those holdings and authorized the Secretary to buy whatever lay inside. 126 Cong. Rec. 21882 (1980) (remarks of Sen. Stevens). The upshot was a vast set of so-called inholdings—more than 18 million acres of state, Native, and private land—that wound up inside Alaskan system units. See 577 U. S., at ___–___ (slip op., at 5–6). Had Congress done nothing more, those inholdings could have become subject to many Park Service rules—the same kind of “restrictive federal regulations” Alaskans had protested in the years leading up to ANILCA (and further back too). Id., at ___ (slip op., at 4). That is because the Secretary, acting through the Director of the Park Service, has broad authority under the National Park Service Organic Act (Organic Act), 39Stat. 535, to administer both lands and waters within all system units in the country. See 54 U. S. C. §§100751, 100501, 100102. The Secretary “shall prescribe such regulations as [he] considers necessary or proper for the use and management of System units.” §100751(a). And he may, more specifically, issue regulations concerning “boating and other activities on or relating to water located within System units.” §100751(b). Those statutory grants of power make no distinctions based on the ownership of either lands or waters (or lands beneath waters).[1] And although the Park Service has sometimes chosen not to regulate non-federally owned lands and waters, it has also imposed major restrictions on their use. Rules about mining and solid-waste disposal, for example, apply to all lands within system units “whether federally or nonfederally owned.” 36 CFR §6.2; see §9.2. And (of particular note here) the Park Service freely regulates activities on all navigable (and some other) waters “within [a park’s] boundaries”—once more, “without regard to . . . ownership.” §1.2(a)(3). So Alaska and its Natives had reason to worry about how the Park Service would regulate their lands and waters within the new parks. Congress thus acted, as even the Park Service agrees, to give the State and Natives “assurance that their [lands] wouldn’t be treated just like” federally owned property. Tr. of Oral Arg. 50. (It is only—though this is quite a large “only”—the nature and extent of that assurance that is in dispute.) The key provision here is Section 103(c), which contains three sentences that may require some re-reading. We quote it first in one block; then provide some definitions; then go over it again a bit more slowly. But still, you should expect to return to this text as you proceed through this opinion. Section 103(c) provides in full: “Only those lands within the boundaries of any conservation system unit which are public lands (as such term is defined in this Act) shall be deemed to be included as a portion of such unit. No lands which, before, on, or after [the date of ANILCA’s passage], are conveyed to the State, to any Native Corporation, or to any private party shall be subject to the regulations applicable solely to public lands within such units. If the State, a Native Corporation, or other owner desires to convey any such lands, the Secretary may acquire such lands in accordance with applicable law (including this Act), and any such lands shall become part of the unit, and be administered accordingly.” §3103(c). Now for the promised definitions. The term “land,” as found in all three sentences, actually—and crucially for this case—“means lands, waters, and interests therein.” §3102(1). The term “public lands,” in the first two sentences, then means “lands” (including waters and interests therein) “the title to which is in the United States”—except for lands selected for future transfer to the State or Native Corporations (under the Statehood Act or ANCSA). §3102(2), (3); see supra, at 4–5. “Public lands” are therefore most but not quite all lands (and again, waters and interests) that the Federal Government owns. Finally, to recap. As explained in Sturgeon I, “Section 103(c) draws a distinction between ‘public’ and ‘non-public’ lands within the boundaries of conservation system units in Alaska.” 577 U. S., at __ (slip op., at 14). Section 103(c)’s first sentence makes clear that only public lands (again, defined as most federally owned lands, waters, and associated interests) would be considered part of a system unit (again, just meaning a national park, preserve, or similar area). By contrast, state, Native, or private lands would not be understood as part of such a unit, even though they in fact fall within its geographic boundaries. Section 103(c)’s second sentence then expressly exempts all those non-public lands (the inholdings) from certain regulations—though exactly which ones, as will soon become clear, is a matter of dispute. And last, Section 103(c)’s third sentence enables the Secretary to buy any inholdings. If he does, the lands (because now public) become part of the park, and may be administered in the usual way—e.g., without the provision’s regulatory exemption. C We can now return to John Sturgeon, on his way to a hunting ground alternatively dubbed “Moose Meadows” or “Sturgeon Fork.” As recounted above, Sturgeon used to travel by hovercraft up a stretch of the Nation River that lies within the boundaries of the Yukon-Charley Preserve. See supra, at 1. Until one day, three park rangers approached Sturgeon while he was repairing his steering cable and told him he was violating a Park Service rule. According to the specified regulation, “[t]he operation or use of hovercraft is prohibited” on navigable (and some other) waters “located within [a park’s] boundaries,” without any “regard to . . . ownership.” 36 CFR §§2.17(e), 1.2(a)(3); see supra, at 2. That regulation, issued under the Secretary’s Organic Act authority, applies on its face to parks across the country. See supra, at 8 (describing Organic Act). And Sturgeon did not doubt that the Nation River is a navigable water. But Sturgeon protested that in Alaska (even though nowhere else) the rule could not be enforced on a waterway—like, he said, the Nation River—that is not owned by the Federal Government. And when his objection got nowhere with the rangers (or with the Secretary, to whom he later petitioned), Sturgeon stopped using his hovercraft—but also brought this lawsuit, based on ANILCA’s Section 103(c). In Sturgeon I, we rejected one ground for dismissing Sturgeon’s case, but remanded for consideration of two further questions. The District Court and Court of Appeals for the Ninth Circuit had held that even assuming the Nation River is non-public land, the Park Service could enforce its hovercraft ban there. See 2013 WL 5888230 (Oct. 30, 2013); 768 F.3d 1066 (2014). Those two courts interpreted Section 103(c) to limit only the Service’s authority to impose Alaska-specific regulations on such lands—not its authority to apply nationwide regulations like the hovercraft rule. But we viewed that construction as “implausible.” 577 U. S., at ___ (slip op., at 15). ANILCA, we reasoned, “repeatedly recognizes that Alaska is different.” Id., at ___ (slip op., at 13); see id., at ___ (slip op., at 14) (The Act “reflect[s] the simple truth that Alaska is often the exception, not the rule”). Yet the lower courts’ reading would “prevent the Park Service from recognizing Alaska’s unique conditions”—thus producing a “topsy-turvy” result. Ibid. Still, we thought two hurdles remained before Sturgeon could take his hovercraft out of storage. We asked the Court of Appeals to decide whether the Nation River “qualifies as ‘public land’ for purposes of ANILCA,” thus indisputably subjecting it to the Service’s regulatory authority. Id., at ___ (slip op., at 15). And if the answer was “no,” we asked the Ninth Circuit to address whether the Service, on some different theory from the one just dispatched, could still “regulate Sturgeon’s activities on the Nation River.” Id., at ___ (slip op., at 16). The Ninth Circuit never got past the first question because it concluded that the Nation River is “public land[.]” See 872 F.3d 927, 936 (2017). The court explained that it was bound by three circuit decisions construing that term, when used in ANILCA’s provisions about subsistence fishing, as including all navigable waters. Id., at 933–934. Accordingly, the court again rejected Sturgeon’s challenge. Id., at 936. And we again granted certiorari. 585 U. S. ___ (2018). II We first address whether, as the Ninth Circuit found, the Nation River is “public land” under ANILCA. As defined, once again, that term means (almost all) “lands, waters, and interests therein” the “title to which is in the United States.” 16 U. S. C. §3102(1)–(3). If the Nation River comes within that definition, even Sturgeon agrees that the Park Service may enforce its hovercraft rule in the stretch traversing the Yukon-Charley. That is because the Organic Act authorizes the Park Service to regulate boating and similar activities in parks and other system units—and under ANILCA’s Section 103(c) those units include all “public land” within their boundaries. 54 U. S. C. §100751(a)–(b); 16 U. S. C. §3103(c); see supra, at 8–10. But the United States does not have “title” (as the just-quoted definition demands) to the Nation River in the ordinary sense. As the Park Service acknowledges, running waters cannot be owned—whether by a government or by a private party. See FPC v. Niagara Mohawk Power Corp., 347 U.S. 239, 247, n. 10 (1954); Brief for Respondents 33. In contrast, the lands beneath those waters—typically called submerged lands—can be owned, and the water regulated on that basis. But that does not help the Park Service because, as noted earlier, the Submerged Lands Act gives each State “title to and ownership of the lands beneath [its] navigable waters.” 43 U. S. C. §1311; see supra, at 4. That means Alaska, not the United States, has title to the lands beneath the Nation River. So the Park Service argues instead that the United States has “title” to an “interest” in the Nation River, under what is called the reserved-water-rights doctrine. See Brief for Respondents 32–37. The canonical statement of that doctrine goes as follows: “[W]hen the Federal Government withdraws its land from the public domain and reserves it for a federal purpose, the Government, by implication, reserves appurtenant water then unappropriated to the extent needed to accomplish the purpose of the reservation.” Cappaert v. United States, 426 U.S. 128, 138 (1976). For example, this Court decided that in reserving land for an Indian tribe, the Government impliedly reserved sufficient water from a nearby river to enable the tribe to farm the area. See Winters v. United States, 207 U.S. 564, 576 (1908). And similarly, we held that in creating a national monument to preserve a species of fish inhabiting an underground pool, the United States acquired an enforceable interest in preventing others from depleting the pool below the level needed for the fish to survive. See Cappaert, 426 U. S., at 147. According to the Park Service, the United States has an analogous interest in the Nation River and other navigable waters in Alaska’s national parks. “Because th[e] purposes [of those parks] require that the waters within [them] be safeguarded against depletion and diversion,” the Service contends, “Congress’s reservations of park lands also reserved interests in appurtenant navigable waters.” Brief for Respondents 35. That argument first raises the question whether it is even possible to hold “title,” as ANILCA uses the term, to reserved water rights. 16 U. S. C. §3102(2). Those rights, as all parties agree, are “usufructuary” in nature, meaning that they are rights for the Government to use—whether by withdrawing or maintaining—certain waters it does not own. See Niagara Mohawk Power Corp., 347 U. S., at 246; Brief for Petitioner 36; Brief for Respondents 36. The Park Service has found a couple of old cases suggesting that a person can hold “title” to such usufructuary interests. See ibid.; Crum v. Mt. Shasta Power Corp., 220 Cal. 295, 307, 30 P.2d 30, 36 (1934); Radcliff’s Ex’rs v. Mayor of Brooklyn, 4 N.Y. 195, 196 (1850). But the more common understanding, recently noted in another ANILCA case, is that “reserved water rights are not the type of property interests to which title can be held”; rather, “the term ‘title’ applies” to “fee ownership of property” and (sometimes) to “possessory interests” in property like those granted by a lease. See Totemoff v. State, 905 P.2d 954, 965 (Alaska 1995) (collecting cases); Brief for State of Idaho et al. as Amici Curiae 21–22 (same). And we see no evidence that the Congress enacting ANILCA meant to use the term in any less customary and more capacious sense. But even assuming so, the Nation River itself would not thereby become “public land” in the way the Park Service argues. Under ANILCA’s definition, the “public land” at issue would consist only of the Federal Government’s specific “interest” in the River—that is, its reserved water right. §3102(1), (3). And that reserved right, by its nature, is limited. It does not give the Government plenary authority over the waterway to which it attaches. Rather, the interest merely enables the Government to take or maintain the specific “amount of water”—and “no more”—required to “fulfill the purpose of [its land] reservation.” Cappaert, 426 U. S., at 141. So, for example, in the cases described above, the Government could control only the volume of water necessary for the tribe to farm or the fish to survive. See Winters, 207 U. S., at 576–577; Cappaert, 426 U. S., at 141. And likewise here, the Government could protect “only th[e] amount of water” in the Nation River needed to “accomplish the purpose of the [Yukon-Charley’s] reservation.” Id., at 138, 141. And whatever that volume, the Government’s (purported) reserved right could not justify applying the hovercraft rule on the Nation River. That right, to use the Park Service’s own phrase, would support a regulation preventing the “depletion or diversion” of waters in the River (up to the amount required to achieve the Yukon-Charley’s purposes). Brief for Respondents 34–35. But the hovercraft rule does nothing of that kind. A hovercraft moves above the water, on a thin cushion of air produced by downward-directed fans; it does not “deplet[e]” or “diver[t]” any water. Nor has the Park Service explained the hovercraft rule as an effort to protect the Nation River from pollution or other similar harm. To the contrary, that rule is directed against the “sight or sound” of “motorized equipment” in remote locations—concerns not related to safeguarding the water. 48 Fed. Reg. 30258 (1983). So the Park Service’s “public lands” argument runs aground: Even if the United States holds title to a reserved water right in the Nation River, that right (as opposed to title in the River itself) cannot prevent Sturgeon from wafting along the River’s surface toward his preferred hunting ground.[2] III We thus move on to the second question we posed in Sturgeon I, concerning the Park Service’s power to regulate even non-public lands and waters within Alaska’s system units (or, in our unofficial terminology, national parks). The Service principally relies on that sort of ownership-indifferent authority in defending its decision to expel Sturgeon’s hovercraft from the Nation River. See Brief for Respondents 16–18, 25–32. And we can see why. If Sturgeon lived in any other State, his suit would not have a prayer of success. As noted earlier, the Park Service has used its Organic Act authority to ban hovercrafts on navigable waters “located within [a national park’s] boundaries” without any “regard to . . . ownership.” 36 CFR §§2.17(e), 1.2(a)(3); see supra, at 10–11. And no one disputes that Sturgeon was driving his hovercraft on a stretch of the Nation River (a navigable water) inside the borders of the Yukon-Charley (a national park). So case closed. Except that Sturgeon lives in Alaska. And as we have said before, “Alaska is often the exception, not the rule.” Sturgeon I, 577 U. S., at ___ (slip op., at 14). Here, Section 103(c) of ANILCA makes it so. As explained below, that section provides that even when non-public lands—again, including waters—are geographically within a national park’s boundaries, they may not be regulated as part of the park. And that means the Park Service’s hovercraft regulation cannot apply there.[3] To understand why, first recall how Section 103(c) grew out of ANILCA’s unusual method for drawing park boundaries. See supra, at 7–8. Those lines followed the area’s “natural features,” rather than (as customary) the Federal Government’s property holdings. 16 U. S. C. §3103(b). The borders thus took in immense tracts owned by the State, Native Corporations, and private individuals. And as you might imagine, none of those parties was eager to have its lands newly regulated as national parks. To the contrary, all of them wanted to preserve the regulatory status quo—to prevent ANILCA’s maps from subjecting their properties to the Park Service’s rules. Hence arose Section 103(c). Cf. Tr. of Oral Arg. 50 (Solicitor General acknowledging that Section 103(c) responds to the State’s and Native Corporations’ “concern[s]” about the effects of “includ[ing their lands] within the outer boundaries” of the new parks). Now might be a good time to review that provision, block quoted above. See supra, at 9. In broad brush strokes, Sturgeon I described it as follows: “Section 103(c) draws a distinction between ‘public’ and ‘non-public’ lands,” including waters, “within the boundaries of [Alaska’s] conservation system units.” 577 U. S., at ___ (slip op., at 14). Section 103(c)’s first sentence sets out the essential distinction, relating to what qualifies as parkland. It provides, once again, that “[o]nly” the “public lands” (essentially, the federally owned lands) within any system unit’s boundaries would be “deemed” a part of that unit. §3103(c). The non-public lands (everything else) were, by negative implication, “deemed” not a part of the unit—even though within the unit’s geographic boundaries. The key word here is “deemed.” That term is used in legal materials “[t]o treat (something) as if . . . it were really something else.” Black’s Law Dictionary 504 (10th ed. 2014). Legislators (and other drafters) find the word “useful” when “it is necessary to establish a legal fiction,” either by “‘deeming’ something to be what it is not” or by “‘deeming’ something not to be what it is.” Ibid. (quoting G.C. Thornton, Legislative Drafting 99 (4th ed. 1996)). The fiction in Section 103(c) involves considering certain lands actually within the new national parks as instead without them. As a matter of geography, both public and non-public lands fall inside those parks’ boundaries. But as a matter of law, only public lands would be viewed as doing so. All non-public lands (again, including waters) would be “deemed,” abracadabra-style, outside Alaska’s system units.[4] The effect of that exclusion, as Section 103(c)’s second sentence affirms, is to exempt non-public lands, including waters, from the Park Service’s ordinary regulatory authority. Recall that the Organic Act pegs that authority to system units. See supra, at 8. The Service may issue rules thought “necessary or proper” for “System units.” 54 U. S. C. §100751(a). And more pertinently here, the Service may prescribe rules about activities on “water located within System units.” §100751(b). Absent Section 103(c), those grants of power enable the Service to administer even non-federally owned waters or lands inside national parks. See supra, at 8. But add Section 103(c), and the equation changes. Now, according to that section’s first sentence, non-federally owned waters and lands inside system units (on a map) are declared outside them (for the law). So those areas are no longer subject to the Service’s power over “System units” and the “water located within” them. §100751(a), (b). Instead, only the federal property in system units is subject to the Service’s authority.[5] And that is just what Section 103(c)’s second sentence pronounces, for waters and lands alike. Again, that sentence says that no state, Native, or private lands “shall be subject to the regulations applicable solely to public lands within [system] units.” 16 U. S. C. §3103(c). The sentence thus expressly states the consequence of the statute’s prior “deeming.” The Service’s rules will apply exclusively to public lands (meaning federally owned lands and waters) within system units. The rules cannot apply to any non-federal properties, even if a map would show they are within such a unit’s boundaries. Geographic inholdings thus become regulatory outholdings, impervious to the Service’s ordinary authority.[6] And for that reason, Section 103(c)’s third sentence provides a kind of escape hatch—for times when the Park Service believes regulation of the inholdings is needed. In that event, “the Secretary may acquire such lands” from “the State, a Native Corporation, or other owner.” §3103(c). (As noted earlier, facilitating those acquisitions was one reason Congress put non-federal lands inside park boundaries in the first instance. See supra, at 7.) When the Secretary makes such a purchase, the newly federal land “become[s] part of the [system] unit.” §3101(c). And the Park Service may then “administer[ ]” the land just as it does (in the second sentence’s phrase) the other “public lands within such units.” Ibid. In thus providing a way out of the Section’s first two sentences, the third underlines what they are doing: insulating the state, Native, or private lands that ANILCA enclosed in national parks from new and unexpected regulation. In sum, those lands may be regulated only as they could have been before ANILCA’s enactment, unless and until bought by the Federal Government. The Park Service interprets Section 103(c) differently, relying wholly on its second sentence and mostly on the single word “solely” there. True enough, the Service acknowledges, that anxiety about how it would regulate inholdings was “really what drove [Section] 103(c).” Tr. of Oral Arg. 46; see supra, at 9, 17. But still, the Service argues, the Section’s second sentence exempts those non-public lands from only “one particular class of Park Service regulations”—to wit, rules “ ‘applicable solely to public lands.’ ” Brief for Respondents 30 (quoting and adding emphasis to §3103(c)). In other words, if a Park Service regulation on its face applies only (“solely”) to public lands, then the regulation shall not apply to a park’s non-public lands. But if instead the regulation covers public and non-public lands alike, then the second sentence has nothing to say: The regulation can indeed cover both. See ibid. The Park Service labels that sentence a “tailored limitation” on its authority over inholdings. Ibid. And it concludes that the sentence has no bearing on the hovercraft rule, which expressly applies “without regard to . . . ownership.” 36 CFR §1.2(a)(3). But on the Park Service’s view, Section 103(c)’s second sentence is a mere truism, not any kind of limitation (however “tailored”). Once again: It tells Alaskans, so the Park Service says, that rules applying only to public lands . . . will apply only to public lands. And that rules applying to both public and non-public lands . . . will apply to both. (Or, to say the same thing, but with approximate statutory definitions plugged in: It tells Alaskans that rules applying only to the Federal Government’s lands . . . will apply only to the Federal Government’s lands. And that rules applying to federal, state, Native, and private lands alike . . . will apply to them all.) In short, under the Park Service’s reading, Section 103(c)’s second sentence does nothing but state the obvious. Its supposed exemption does not in fact exempt anyone from anything to which they would otherwise be subject. Remove the sentence from ANILCA and everything would be precisely the same. For it curtails none of the Service’s ordinary regulatory authority over inholdings.[7] And more: The Park Service’s reading of Section 103(c)’s second sentence also strips the first and third sentences of their core functions. Under the Service’s approach, the first sentence’s “deeming” has no point. There is no reason to pretend that inholdings are not part of a park if they can still be regulated as parklands. Nor is there a need to create a special legal fiction if the end result is to treat Alaskan inholdings no differently from those in the rest of the country. And similarly, the third sentence’s acquisition option has far less utility if the Service has its full regulatory authority over lands the Federal Government does not own. Why cough up money to “administer[ ]” property as “part of the [system] unit” unless doing so makes a real difference, by removing a regulatory exemption otherwise in effect? The Service’s reading effectively turns the whole of Section 103(c) into an inkblot. And still more (if implicit in all the above): That construction would undermine ANILCA’s grand bargain. Recall that ANILCA announced its Janus-faced nature in its statement of purpose, reflecting the century-long struggle over federal regulation of Alaska’s resources. See supra, at 3–6. In that opening section, ANILCA spoke about safeguarding “natural, scenic, historic[,] recreational, and wildlife values.” 16 U. S. C. §3101(a). Yet it in- sisted as well on “provid[ing] for” Alaska’s (and its citizens’) “economic and social needs.” §3101(d). In keeping with the statute’s conservation goal, Congress reserved huge tracts of land for national parks. But to protect Alaskans’ economic well-being, it mitigated the consequences to non-federal owners whose land wound up in those new system units. See supra, at 17–20. Once again, even the Park Service acknowledges that Section 103(c) was supposed to provide an “assurance” that those owners would not be subject to all the regulatory constraints placed on neighboring federal properties. See Tr. of Oral Arg. 50; see id., at 46–47; supra, at 9, 17, 20. But then the Service (head-spinningly) posits that it need only draft its regulations to cover both federal and non-federal lands in order to apply those rules to ANILCA’s inholdings. On that view, limitations on the Service’s authority are purely a matter of administrative grace, dependent on how narrowly (or broadly) the Service chooses to write its regulations. And ANILCA’s carefully drawn balance is thrown off-kilter, as Alaskan, Native, and private inholdings are exposed to the full extent of the Service’s regulatory authority. The word “solely” in Section 103(c)’s second sentence does not support that kind of statute-gutting. We do not gainsay that the Park Service has identified a grammatically possible way of viewing that word’s function: as pinpointing a narrow class of the Service’s regulations (those “solely applicable to public lands”).[8] But that reading, for all the reasons just stated, is “ultimately inconsistent” with the “text and context of the statute.” Sturgeon I, 577 U. S., at ___ (slip op., at 12). And a different understanding of “solely” instead aligns with that text and context. That word encapsulates Congress’s view that the Park Service’s regulations should apply “solely” to public lands (and not to state, Native, or private ones). See supra, at 19, and n. 5. And the word serves to distinguish between the Park Service’s rules and other regulations, both federal and state. Consider if Congress had exempted non-public lands in a system unit from regulations “applicable to public lands” there (without the “solely”). That language would apparently exempt those lands not just from park regulations but from a raft of others—e.g., pollution regulations of the Environmental Protection Agency, water safety regulations of the Coast Guard, even employment regulations of Alaska itself. For those rules, too, apply to public lands inside national parks. By adding “solely,” Congress made clear that the exemption granted was not from such generally applicable regulations. Instead, it was from rules applying only in national parks—i.e., the newly looming Park Service rules. Congress thus ensured that inholdings would emerge from ANILCA not worse off—but also not better off—than before.[9] The legislative history (for those who consider it) confirms, with unusual clarity, all we have said so far. The Senate Report notes that state, Native, and private lands in the new Alaskan parks would be subject to “[f ]ederal laws and regulations of general applicability,” such as “the Clean Air Act, the Water Pollution Control Act, [and] U. S. Army Corps of Engineers wetlands regulations.” S. Rep. No. 96–413, p. 303 (1980). But that would not be so of regulations applying only to parks. The Senate Report states: “Those private lands, and those public lands owned by the State of Alaska or a subordinate political entity, are not to be construed as subject to the management regulations which may be adopted to manage and administer any national conservation system unit which is adjacent to, or surrounds, the private or non-Federal public lands.” Ibid. The sponsor of Section 103(c) in the House of Representatives described that provision’s effect in similar terms. The section was designed, he observed, to ensure that ANILCA’s new boundary lines would “not in any way change the status” of the state, Native, and private lands placed within them. 125 Cong. Rec. 11158 (1979) (statement of Rep. Seiberling). Those lands, he continued, “are not parts of th[e system] unit and are not subject to regulations which are applied” by virtue of being “part of the unit.” Ibid. In short, whatever the new map might suggest, they are not subject to regulation as parkland. We thus arrive again at the conclusion that the Park Service may not prevent John Sturgeon from driving his hovercraft on the Nation River. We held in an earlier part of this opinion that the Nation is not public land. See supra, at 12–15. And here we hold that it cannot be regulated as if it were. Park Service regulations—like the hovercraft rule—do not apply to non-public lands in Alaska even when those lands lie within national parks. Section 103(c) “deem[s]” those lands outside the parks and in so doing deprives the Service of regulatory authority. IV Yet the Park Service makes one last plea—for some kind of special rule relating to Alaskan navigable waters. Even suppose, the argument runs, that those waters do not count as “public lands.” And even assume that Section 103(c) strips the Service of power to regulate most non-public lands. Still, the Service avers—invoking “the overall statutory scheme”—that ANILCA must at least allow it to regulate navigable waters. Brief for Respondents 40; see id., at 40–45; Tr. of Oral Arg. 42 (ANILCA’s regulatory restrictions were “not about navigable waters”); id., at 63–64 (similar). Here, the Service points to ANILCA’s general statement of purpose, which lists (among many other things) the “protect[ion] and preserv[ation]” of “rivers.” 16 U. S. C. §3101(b). Similarly, the Service notes that the statements of purpose associated with particular system units refer to “protect[ing]” named rivers there. E.g., §410hh–1(1). And the Service highlights several statutory sections that in some way speak to its ability to regulate motorboating and fishing within the new units. See §§3121, 3170, 3201, 3203(b), 3204.[10] According to the Service, all of those provisions show that “ANILCA preserves [its] authority to regulate conduct on navigable waters” in national parks. Brief for Respondents 42. But ANILCA does not readily allow the decoupling of navigable waters from other non-federally owned areas in Alaskan national parks for regulatory (or, indeed, any other) purposes. Section 103(c), as we have described, speaks of “lands (as such term is defined in th[e] Act).” 16 U. S. C. §3103(c); see supra, at 9. The Act, in turn, defines “land” to mean “lands, waters, and interests therein.” §3102(1)–(3); see supra, at 9. So according to an express definition, when ANILCA refers to “lands,” it means waters (including navigable waters) as well. And that kind of definition is “virtually conclusive.” A. Scalia & B. Garner, Reading Law: The Interpretation of Legal Texts 228 (2012); see ibid. (“It is very rare that a defined meaning can be replaced” or altered). Save for some exceptional reason, we must read ANILCA as treating identically solid ground and flowing water. So if the Park Service were right that it could regulate the Nation River under its ordinary authorities, then it also could regulate the private fields and farms in the surrounding park. And more to the point, once Section 103(c) is understood to preclude the regulation of those landed properties, then the same result follows—“virtually conclusive[ ly]”—for the river. And nothing in the few aquatic provisions to which the Park Service points can flip that strong presumption, for none conflicts with reading Section 103(c)’s regulatory exemption to cover non-federal waters. The most substantive of those provisions, as just noted, contemplate some role for the Service in regulating motorboating and fishing. But contra the Park Service, those sections have effect under our interpretation because both activities can occur on federally owned (and thus fully regulable) non-navigable waters. The other provisions the Service emphasizes are statements of purpose, which by their nature “cannot override [a statute’s] operative language.” Id., at 220. And anyway, our construction leaves the Park Service with multiple tools to “protect” rivers in Alaskan national parks, as those statements anticipate. §3101(b); §410hh–1(1). The Park Service may at a minimum regulate the public lands flanking rivers. It may, additionally, enter into “cooperative agreements” with the State (which holds the rivers’ submerged lands) to preserve the rivers themselves. §3181(j). It may similarly propose that state or other federal agencies with appropriate jurisdiction undertake needed regulatory action on those rivers. See §3191(b)(7); see also Kobuk Valley: Land Protection Plan, at 118, 121 (recommending that the Alaska Department of Natural Resources classify navigable parts of the Kobuk River for preservation efforts). And if all else fails, the Park Service may invoke Section 103(c)’s third sentence to buy from Alaska the submerged lands of navigable waters—and then administer them as public lands. See §§3103(c), 3192; see also Kobuk Valley: Land Protection Plan, at 133 (proposing that if Alaska does not adequately protect the Kobuk River, the Park Service should “seek to acquire title to th[o]se state lands through exchange”). Those authorities, though falling short of the Service’s usual power to administer navigable waters in system units, accord with ANILCA’s “repeated[ ] recogni[tion] that Alaska is different.” Sturgeon I, 577 U. S., at ___ (slip op., at 13). ANILCA’s broadly drawn parks include stretches of some of the State’s most important rivers, such as the Yukon and Kuskokwim. See Brief for State of Alaska as Amicus Curiae 12. And rivers function as the roads of Alaska, to an extent unknown anyplace else in the country. Over three-quarters of Alaska’s 300 communities live in regions unconnected to the State’s road system. See id., at 11. Residents of those areas include many of Alaska’s poorest citizens, who rely on rivers for access to necessities like food and fuel. See id., at 11–12. Who knows?—maybe John Sturgeon could have found a comparable hunting ground that did not involve traveling by hovercraft through a national park. But some Alaskans have no such options. The State’s extreme climate and rugged terrain make them dependent on rivers to reach a market, a hospital, or a home. So ANILCA recognized that when it came to navigable waters—just as to non-federal lands—in the new parks, Alaska should be “the exception, not the rule.” Sturgeon I, 577 U. S., at ___ (slip op., at 14). Which is to say, exempt from the Park Service’s normal regula- tory authority. V ANILCA, like much legislation, was a settlement. The statute set aside more than a hundred million acres of Alaska for conservation. In so doing, it enabled the Park Service to protect—if need be, through expansive regulation—“the national interest in the scenic, natural, cultural and environmental values on the public lands in Alaska.” 16 U. S. C. §3101(d). But public lands (and waters) was where it drew the line—or, at any rate, the legal one. ANILCA changed nothing for all the state, Native, and private lands (and waters) swept within the new parks’ boundaries. Those lands, of course, remain subject to all the regulatory powers they were before, exercised by the EPA, Coast Guard, and the like. But they did not become subject to new regulation by the happenstance of ending up within a national park. In those areas, Section 103(c) makes clear, Park Service administration does not replace local control. For that reason, park rangers cannot enforce the Service’s hovercraft rule on the Nation River. And John Sturgeon can once again drive his hovercraft up that river to Moose Meadows. We accordingly reverse the judgment below and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 None of the parties here have questioned the constitutional validity of the above statutory grants as applied to inholdings, and we therefore do not address the issue. Cf. Kleppe v. New Mexico, 426 U.S. 529, 536–541 (1976); Kansas v. Colorado, 206 U.S. 46, 88–89 (1907). 2 As noted earlier, the Ninth Circuit has held in three cases—the so-called Katie John trilogy—that the term “public lands,” when used in ANILCA’s subsistence-fishing provisions, encompasses navigable waters like the Nation River. See Alaska v. Babbitt, 72 F.3d 698 (1995); John v. United States, 247 F.3d 1032 (2001) (en banc); John v. United States, 720 F.3d 1214 (2013); supra, at 12. Those provisions are not at issue in this case, and we therefore do not disturb the Ninth Circuit’s holdings that the Park Service may regulate subsistence fishing on navigable waters. See generally Brief for State of Alaska as Amicus Curiae 29–35 (arguing that this case does not implicate those decisions); Brief for Ahtna, Inc., as Amicus Curiae 30–36 (same). 3 Because we see, for the reasons given below, no ambiguity as to Section 103(c)’s meaning, we cannot give deference to the Park Service’s contrary construction. See Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842 (1984) (“If the intent of Congress is clear, that is the end of the matter”). 4 Consistent with that approach, Congress left out non-public lands in calculating the acreage of every new or expanded system unit. Sections 201 and 202 of ANILCA, in describing those units, state the acreage of only their public lands. See, e.g., §410hh(1) (providing that Aniakchak National Preserve would “contain[ ] approximately [367,000] acres of public lands”); §410hh–1(3) (providing that Denali National Park would grow “by the addition of an area containing approximately [2,426,000] acres of public land”). 5 At times, the Park Service has argued here that the Organic Act gives it authority to regulate waters outside system units, so long as doing so protects waters or lands inside them. See Brief for Respondents 28–32. If so, the argument goes, that authority would similarly permit the Service to regulate the non-federally owned waters that Section 103(c) has deemed outside Alaskan system units, if and when needed to conserve those units’ federal waters or lands. But at other points in this litigation, the Service has all but disclaimed such out-of-the-park regulatory authority. See No. 14–1209, Tr. of Oral Arg. 58 (Jan. 20, 2016) (“The Park Service [has] consistently understood its authority to be regulating [within] the park’s boundaries. It’s never sought to enact a regulation outside of the park’s boundaries”). We take no position on the question because it has no bearing on the hovercraft rule at issue here. That rule, by its express terms, applies only inside system units. See supra, at 10–11. It therefore does not raise any question relating to the existence or scope of the Service’s authority over water outside system units. 6 Another provision of ANILCA reflects that result. Right after Sections 201 and 202 describe each new or expanded system unit by reference to how many acres of public land it contains, see n. 4, supra, Section 203 authorizes the Park Service to administer, under the Organic Act, the areas listed in “the foregoing sections.” §410hh–2. In other words, Section 203 of ANILCA ties the Service’s regulatory authority to the statute’s immediately preceding statements of public-land acreage. 7 And just to pile on: Even taken as a truism, the Park Service’s view of the second sentence misfires, because of the technical difference between “public lands” and federally owned lands in ANILCA. Recall that “public lands” is defined in the statute to mean most but not all federally owned lands: The term excludes those federal lands selected for future transfer to the State or Native Corporations. See §3102(3); supra, at 9–10. (That is why when we reframed the Park Service’s argument just above, we noted that we were using “approximate” statutory definitions.) But the Park Service’s existing regulations apply, at a minimum, to all federally owned lands within a park’s borders. See 36 CFR §1.2(a). That means there are no regulations “applicable solely to public lands” as defined in ANILCA. §3103(c). So when the Park Service argues that the second sentence exempts non-public lands from that single “class of [its] regulations,” Brief for Respondents 18, 30, it is not even exempting those lands from obviously inapplicable regulations (as we assume in the text); instead, it is exempting them from a null set of rules. 8 It is unfortunate for the Park Service’s argument that the narrow class of regulations thus identified does not in fact exist. See n. 7, supra. But we put that point aside for the remainder of this paragraph. 9 The Park Service points to one provision of ANILCA that (it says) contemplates application of its rules to inholdings; but as suggested in the text that provision really envisions other agencies’ regulations. Section 1301(b)(7) requires the Service to create for each system unit a land management plan that includes (among other things) a description of “privately owned areas” within the unit, the activities carried out there, and the “methods (such as cooperative agreements and issuance or enforcement of regulations)” for limiting those activities if appropriate. 16 U. S. C. §3191(b)(7). Nothing in that section “directs the Park Service” itself to issue or enforce regulations, as the Service now argues. See Brief for Respondents 30–31. Instead, the Service satisfies all its obligations under the provision by reporting on the panoply of federal and state statutes and regulations that apply to any non-public land (whether or not in a park). And indeed, the Service’s management plans have taken exactly that form. See, e.g., Dept. of Interior, Nat. Park Serv., Kobuk Valley National Park: Land Protection Plan 123–124 (1986) (noting that “[w]hile [Park Service] regulations do not generally apply to private lands in the park (Section 103, ANILCA),” the regulations “that do apply” include those issued under “the Alaska Anadromous Fish Act, the Endangered Species Act, the Clean Water and Clean Air acts, and the Protection of Wetlands, to name a few”); Dept. of Interior, Nat. Park Serv., Noatak National Preserve: Land Protection Plan 138–139, 142 (1986) (similar). 10 The Park Service also points to a separate title of ANILCA, which raises issues outside the scope of this case. Title VI designates 26 named rivers in Alaska as “wild and scenic rivers,” to be “administered by the Secretary” under the (nationwide) Wild and Scenic Rivers Act, 94Stat. 2412–2413. According to the Service, those special designations (and associated management instructions) enable it to “administer the [specified] rivers pursuant to its general statutory authorities”—notwithstanding anything in Section 103(c). Brief for Respondents 42–43. But the Nation River, all agree, is not a “wild and scenic river.” We may therefore leave for another day the interplay between Section 103(c) and Title VI.

587.US.2018_18-489

Petitioner Bradley Taggart formerly owned an interest in an Oregon company. That company and two of its other owners, who are among the respondents here, filed suit in Oregon state court, claiming that Taggart had breached the company’s operating agreement. Before trial, Taggart filed for bankruptcy under Chapter 7 of the Bankruptcy Code. At the conclusion of that proceeding, the Federal Bankruptcy Court issued a discharge order that released Taggart from liability for most prebankruptcy debts. After the discharge order issued, the Oregon state court entered judgment against Taggart in the prebankruptcy suit and awarded attorney’s fees to respondents. Taggart returned to the Federal Bankruptcy Court, seeking civil contempt sanctions against respondents for collecting attorney’s fees in violation of the discharge order. The Bankruptcy Court ultimately held respondents in civil contempt. The Bankruptcy Appellate Panel vacated the sanctions, and the Ninth Circuit affirmed the panel’s decision. Applying a subjective standard, the Ninth Circuit concluded that a “creditor’s good faith belief” that the discharge order “does not apply to the creditor’s claim precludes a finding of contempt, even if the creditor’s belief if unreasonable.” 888 F.3d 438, 444. Held: A court may hold a creditor in civil contempt for violating a discharge order if there is no fair ground of doubt as to whether the order barred the creditor’s conduct. Pp. 4–11. (a) This conclusion rests on a longstanding interpretive principle: When a statutory term is “ ‘obviously transplanted from another legal source,’ ” it “ ‘brings the old soil with it.’ ” Hall v. Hall, 584 U. S. ___, ___. Here, the bankruptcy statutes specifying that a discharge order “operates as an injunction,” 11 U. S. C. §524(a)(2), and that a court may issue any “order” or “judgment” that is “necessary or appropriate” to “carry out” other bankruptcy provisions, §105(a), bring with them the “old soil” that has long governed how courts enforce injunctions. In cases outside the bankruptcy context, this Court has said that civil contempt “should not be resorted to where there is [a] fair ground of doubt as to the wrongfulness of the defendant’s conduct.” California Artificial Stone Paving Co. v. Molitor, 113 U.S. 609, 618. This standard is generally an objective one. A party’s subjective belief that she was complying with an order ordinarily will not insulate her from civil contempt if that belief was objectively unreasonable. Subjective intent, however, is not always irrelevant. Civil contempt sanctions may be warranted when a party acts in bad faith, and a party’s good faith may help to determine an appropriate sanction. These traditional civil contempt principles apply straightforwardly to the bankruptcy discharge context. Under the fair ground of doubt standard, civil contempt may be appropriate when the creditor violates a discharge order based on an objectively unreasonable understanding of the discharge order or the statutes that govern its scope. Pp. 5–7. (b) The standard applied by the Ninth Circuit is inconsistent with traditional civil contempt principles, under which parties cannot be insulated from a finding of civil contempt based on their subjective good faith. Taggart, meanwhile, argues for a standard that would operate much like a strict-liability standard. But his proposal often may lead creditors to seek advance determinations as to whether debts have been discharged, creating the risk of additional federal litigation, additional costs, and additional delays. His proposal, which follows the standard some courts have used to remedy violations of automatic stays, also ignores key differences in text and purpose between the statutes governing automatic stays and discharge orders. Pp. 7–11. 888 F.3d 438, vacated and remanded. Breyer, J., delivered the opinion for a unanimous Court.

At the conclusion of a bankruptcy proceeding, a bankruptcy court typically enters an order releasing the debtor from liability for most prebankruptcy debts. This order, known as a discharge order, bars creditors from attempting to collect any debt covered by the order. See 11 U. S. C. §524(a)(2). The question presented here concerns the criteria for determining when a court may hold a creditor in civil contempt for attempting to collect a debt that a discharge order has immunized from collection. The Bankruptcy Court, in holding the creditors here in civil contempt, applied a standard that it described as akin to “strict liability” based on the standard’s expansive scope. In re Taggart, 522 B.R. 627, 632 (Bkrtcy. Ct. Ore. 2014). It held that civil contempt sanctions are permis- sible, irrespective of the creditor’s beliefs, so long as the creditor was “ ‘aware of the discharge’ ” order and “ ‘in- tended the actions which violate[d]’ ” it. Ibid. (quoting In re Hardy, 97 F.3d 1384, 1390 (CA11 1996)). The Court of Appeals for the Ninth Circuit, however, disagreed with that standard. Applying a subjective standard instead, it concluded that a court cannot hold a creditor in civil contempt if the creditor has a “good faith belief” that the discharge order “does not apply to the creditor’s claim.” In re Taggart, 888 F.3d 438, 444 (2018). That is so, the Court of Appeals held, “even if the creditor’s belief is unreasonable.” Ibid. We conclude that neither a standard akin to strict liability nor a purely subjective standard is appropriate. Rather, in our view, a court may hold a creditor in civil con- tempt for violating a discharge order if there is no fair ground of doubt as to whether the order barred the creditor’s conduct. In other words, civil contempt may be appropriate if there is no objectively reasonable basis for concluding that the creditor’s conduct might be lawful. I Bradley Taggart, the petitioner, formerly owned an interest in an Oregon company, Sherwood Park Business Center. That company, along with two of its other owners, brought a lawsuit in Oregon state court, claiming that Taggart had breached the Business Center’s operating agreement. (We use the name “Sherwood” to refer to the company, its two owners, and—in some instances—their former attorney, who is now represented by the executor of his estate. The company, the two owners, and the executor are the respondents in this case.) Before trial, Taggart filed for bankruptcy under Chapter 7 of the Bankruptcy Code, which permits insolvent debtors to discharge their debts by liquidating assets to pay creditors. See 11 U. S. C. §§704(a)(1), 726. Ultimately, the Federal Bankruptcy Court wound up the proceeding and issued an order granting him a discharge. Taggart’s discharge order, like many such orders, goes no further than the statute: It simply says that the debtor “shall be granted a discharge under §727.” App. 60; see United States Courts, Order of Discharge: Official Form 318 (Dec. 2015), http:/ /www.uscourts.gov / sites / default / files /form _ b318_0.pdf (as last visited May 31, 2019). Section 727, the statute cited in the discharge order, states that a discharge relieves the debtor “from all debts that arose before the date of the order for relief,” “[e]xcept as provided in section 523.” §727(b). Section 523 then lists in detail the debts that are exempt from discharge. §§523(a)(1)–(19). The words of the discharge order, though simple, have an important effect: A discharge order “operates as an injunction” that bars creditors from collecting any debt that has been discharged. §524(a)(2). After the issuance of Taggart’s federal bankruptcy discharge order, the Oregon state court proceeded to enter judgment against Taggart in the prebankruptcy suit involving Sherwood. Sherwood then filed a petition in state court seeking attorney’s fees that were incurred after Taggart filed his bankruptcy petition. All parties agreed that, under the Ninth Circuit’s decision in In re Ybarra, 424 F.3d 1018 (2005), a discharge order would normally cover and thereby discharge postpetition attorney’s fees stemming from prepetition litigation (such as the Oregon litigation) unless the discharged debtor “ ‘returned to the fray’ ” after filing for bankruptcy. Id., at 1027. Sherwood argued that Taggart had “returned to the fray” postpetition and therefore was liable for the postpetition attorney’s fees that Sherwood sought to collect. The state trial court agreed and held Taggart liable for roughly $45,000 of Sherwood’s postpetition attorney’s fees. At this point, Taggart returned to the Federal Bankruptcy Court. He argued that he had not returned to the state-court “fray” under Ybarra, and that the discharge order therefore barred Sherwood from collecting postpetition attorney’s fees. Taggart added that the court should hold Sherwood in civil contempt because Sherwood had violated the discharge order. The Bankruptcy Court did not agree. It concluded that Taggart had returned to the fray. Finding no violation of the discharge order, it refused to hold Sherwood in civil contempt. Taggart appealed, and the Federal District Court held that Taggart had not returned to the fray. Hence, it concluded that Sherwood violated the discharge order by trying to collect attorney’s fees. The District Court remanded the case to the Bankruptcy Court. The Bankruptcy Court, noting the District Court’s decision, then held Sherwood in civil contempt. In doing so, it applied a standard it likened to “strict liability.” 522 B. R., at 632. The Bankruptcy Court held that civil contempt sanctions were appropriate because Sherwood had been “ ‘aware of the discharge’ ” order and “ ‘intended the actions which violate[d]’ ” it. Ibid. (quoting In re Hardy, 97 F. 3d, at 1390). The court awarded Taggart approximately $105,000 in attorney’s fees and costs, $5,000 in damages for emotional distress, and $2,000 in punitive damages. Sherwood appealed. The Bankruptcy Appellate Panel vacated these sanctions, and the Ninth Circuit affirmed the panel’s decision. The Ninth Circuit applied a very different standard than the Bankruptcy Court. It concluded that a “creditor’s good faith belief” that the discharge order “does not apply to the creditor’s claim precludes a finding of contempt, even if the creditor’s belief is unreasonable.” 888 F. 3d, at 444. Because Sherwood had a “good faith belief” that the discharge order “did not apply” to Sherwood’s claims, the Court of Appeals held that civil contempt sanctions were improper. Id., at 445. Taggart filed a petition for certiorari, asking us to decide whether “a creditor’s good-faith belief that the discharge injunction does not apply precludes a finding of civil contempt.” Pet. for Cert. I. We granted certiorari. II The question before us concerns the legal standard for holding a creditor in civil contempt when the creditor attempts to collect a debt in violation of a bankruptcy discharge order. Two Bankruptcy Code provisions aid our efforts to find an answer. The first, section 524, says that a discharge order “operates as an injunction against the commencement or continuation of an action, the employment of process, or an act, to collect, recover or offset” a discharged debt. 11 U. S. C. §524(a)(2). The second, section 105, authorizes a court to “issue any order, process, or judgment that is necessary or appropriate to carry out the provisions of this title.” §105(a). In what circumstances do these provisions permit a court to hold a creditor in civil contempt for violating a discharge order? In our view, these provisions authorize a court to impose civil contempt sanctions when there is no objectively reasonable basis for concluding that the creditor’s conduct might be lawful under the discharge order. A Our conclusion rests on a longstanding interpretive principle: When a statutory term is “ ‘obviously transplanted from another legal source,’ ” it “ ‘brings the old soil with it.’ ” Hall v. Hall, 584 U. S. ___, ___ (2018) (slip op., at 13) (quoting Frankfurter, Some Reflections on the Reading of Statutes, 47 Colum. L. Rev. 527, 537 (1947)); see Field v. Mans, 516 U.S. 59, 69–70 (1995) (applying that principle to the Bankruptcy Code). Here, the statutes specifying that a discharge order “operates as an injunction,” §524(a)(2), and that a court may issue any “order” or “judgment” that is “necessary or appropriate” to “carry out” other bankruptcy provisions, §105(a), bring with them the “old soil” that has long governed how courts enforce injunctions. That “old soil” includes the “potent weapon” of civil contempt. Longshoremen v. Philadelphia Marine Trade Assn., 389 U.S. 64, 76 (1967). Under traditional principles of equity practice, courts have long imposed civil contempt sanctions to “coerce the defendant into compliance” with an injunction or “compensate the complainant for losses” stemming from the defendant’s noncompliance with an injunction. United States v. Mine Workers, 330 U.S. 258, 303–304 (1947); see D. Dobbs & C. Roberts, Law of Remedies §2.8, p. 132 (3d ed. 2018); J. High, Law of Injunctions §1449, p. 940 (2d ed. 1880). The bankruptcy statutes, however, do not grant courts unlimited authority to hold creditors in civil contempt. Instead, as part of the “old soil” they bring with them, the bankruptcy statutes incorporate the traditional standards in equity practice for determining when a party may be held in civil contempt for violating an injunction. In cases outside the bankruptcy context, we have said that civil contempt “should not be resorted to where there is [a] fair ground of doubt as to the wrongfulness of the defendant’s conduct.” California Artificial Stone Paving Co. v. Molitor, 113 U.S. 609, 618 (1885) (emphasis added). This standard reflects the fact that civil contempt is a “severe remedy,” ibid., and that principles of “basic fairness requir[e] that those enjoined receive explicit notice” of “what conduct is outlawed” before being held in civil contempt, Schmidt v. Lessard, 414 U.S. 473, 476 (1974) (per curiam). See Longshoremen, supra, at 76 (noting that civil contempt usually is not appropriate unless “those who must obey” an order “will know what the court intends to require and what it means to forbid”); 11A C. Wright, A. Miller, & M. Kane, Federal Practice and Procedure §2960, pp. 430–431 (2013) (suggesting that civil contempt may be improper if a party’s attempt at compliance was “reasonable”). This standard is generally an objective one. We have explained before that a party’s subjective belief that she was complying with an order ordinarily will not insulate her from civil contempt if that belief was objectively unreasonable. As we said in McComb v. Jacksonville Paper Co., 336 U.S. 187 (1949), “[t]he absence of wilfulness does not relieve from civil contempt.” Id., at 191. We have not held, however, that subjective intent is always irrelevant. Our cases suggest, for example, that civil contempt sanctions may be warranted when a party acts in bad faith. See Chambers v. NASCO, Inc., 501 U.S. 32, 50 (1991). Thus, in McComb, we explained that a party’s “record of continuing and persistent violations” and “persistent contumacy” justified placing “the burden of any uncertainty in the decree . . . on [the] shoulders” of the party who violated the court order. 336 U. S., at 192–193. On the flip side of the coin, a party’s good faith, even where it does not bar civil contempt, may help to determine an appropriate sanction. Cf. Young v. United States ex rel. Vuitton et Fils S. A., 481 U.S. 787, 801 (1987) (“[O]nly the least possible power adequate to the end proposed should be used in contempt cases” (quotation altered)). These traditional civil contempt principles apply straightforwardly to the bankruptcy discharge context. The typical discharge order entered by a bankruptcy court is not detailed. See supra, at 2–3. Congress, however, has carefully delineated which debts are exempt from discharge. See §§523(a)(1)–(19). Under the fair ground of doubt standard, civil contempt therefore may be appropriate when the creditor violates a discharge order based on an objectively unreasonable understanding of the discharge order or the statutes that govern its scope. B The Solicitor General, amicus here, agrees with the fair ground of doubt standard we adopt. Brief for United States as Amicus Curiae 13–15. And the respondents stated at oral argument that it would be appropriate for courts to apply that standard in this context. Tr. of Oral Arg. 43. The Ninth Circuit and petitioner Taggart, however, each believe that a different standard should apply. As for the Ninth Circuit, the parties and the Solicitor General agree that it adopted the wrong standard. So do we. The Ninth Circuit concluded that a “creditor’s good faith belief” that the discharge order “does not apply to the creditor’s claim precludes a finding of contempt, even if the creditor’s belief is unreasonable.” 888 F. 3d, at 444. But this standard is inconsistent with traditional civil contempt principles, under which parties cannot be insulated from a finding of civil contempt based on their subjective good faith. It also relies too heavily on difficult-to-prove states of mind. And it may too often lead creditors who stand on shaky legal ground to collect discharged debts, forcing debtors back into litigation (with its accompanying costs) to protect the discharge that it was the very purpose of the bankruptcy proceeding to provide. Taggart, meanwhile, argues for a standard like the one applied by the Bankruptcy Court. This standard would permit a finding of civil contempt if the creditor was aware of the discharge order and intended the actions that violated the order. Brief for Petitioner 19; cf. 522 B. R., at 632 (applying a similar standard). Because most creditors are aware of discharge orders and intend the actions they take to collect a debt, this standard would operate much like a strict-liability standard. It would authorize civil contempt sanctions for a violation of a discharge order regardless of the creditor’s subjective beliefs about the scope of the discharge order, and regardless of whether there was a reasonable basis for concluding that the creditor’s conduct did not violate the order. Taggart argues that such a standard would help the debtor obtain the “fresh start” that bankruptcy promises. He adds that a standard resembling strict liability would be fair to creditors because creditors who are unsure whether a debt has been discharged can head to federal bankruptcy court and obtain an advance determination on that question before trying to collect the debt. See Fed. Rule Bkrtcy. Proc. 4007(a). We doubt, however, that advance determinations would provide a workable solution to a creditor’s potential dilemma. A standard resembling strict liability may lead risk-averse creditors to seek an advance determination in bankruptcy court even where there is only slight doubt as to whether a debt has been discharged. And because discharge orders are written in general terms and operate against a complex statutory backdrop, there will often be at least some doubt as to the scope of such orders. Taggart’s proposal thus may lead to frequent use of the advance determination procedure. Congress, however, expected that this procedure would be needed in only a small class of cases. See 11 U. S. C. §523(c)(1) (noting only three categories of debts for which creditors must obtain advance determinations). The widespread use of this procedure also would alter who decides whether a debt has been discharged, moving litigation out of state courts, which have concurrent jurisdiction over such questions, and into federal courts. See 28 U. S. C. §1334(b); Advisory Committee’s 2010 Note on subd. (c)(1) of Fed. Rule Civ. Proc. 8, 28 U. S. C. App., p. 776 (noting that “whether a claim was excepted from discharge” is “in most instances” not determined in bankruptcy court). Taggart’s proposal would thereby risk additional federal litigation, additional costs, and additional delays. That result would interfere with “a chief purpose of the bankruptcy laws”: “ ‘to secure a prompt and effectual’ ” resolution of bankruptcy cases “ ‘within a limited period.’ ” Katchen v. Landy, 382 U.S. 323, 328 (1966) (quoting Ex parte Christy, 3 How. 292, 312 (1844)). These negative consequences, especially the costs associated with the added need to appear in federal proceedings, could work to the disadvantage of debtors as well as creditors. Taggart also notes that lower courts often have used a standard akin to strict liability to remedy violations of auto- matic stays. See Brief for Petitioner 21. An automatic stay is entered at the outset of a bankruptcy proceeding. The statutory provision that addresses the remedies for violations of automatic stays says that “an individual injured by any willful violation” of an automatic stay “shall recover actual damages, including costs and attorneys’ fees, and, in appropriate circumstances, may recover punitive damages.” 11 U. S. C. §362(k)(1). This language, however, differs from the more general language in section 105(a). Supra, at 5. The purposes of automatic stays and discharge orders also differ: A stay aims to prevent damaging disruptions to the administration of a bankruptcy case in the short run, whereas a discharge is entered at the end of the case and seeks to bind creditors over a much longer period. These differences in language and purpose sufficiently undermine Taggart’s proposal to warrant its rejection. (We note that the automatic stay provision uses the word “willful,” a word the law typically does not associate with strict liability but “ ‘whose construction is often dependent on the context in which it appears.’ ” Safeco Ins. Co. of America v. Burr, 551 U.S. 47, 57 (2007) (quoting Bryan v. United States, 524 U.S. 184, 191 (1998)). We need not, and do not, decide whether the word “willful” supports a standard akin to strict liability.) III We conclude that the Court of Appeals erred in applying a subjective standard for civil contempt. Based on the traditional principles that govern civil contempt, the proper standard is an objective one. A court may hold a creditor in civil contempt for violating a discharge order where there is not a “fair ground of doubt” as to whether the creditor’s conduct might be lawful under the discharge order. In our view, that standard strikes the “careful balance between the interests of creditors and debtors” that the Bankruptcy Code often seeks to achieve. Clark v. Rameker, 573 U.S. 122, 129 (2014). Because the Court of Appeals did not apply the proper standard, we vacate the judgment below and remand the case for further proceedings consistent with this opinion. It is so ordered.

588.US.2018_18-96

Tennessee law imposes durational-residency requirements on persons and companies wishing to operate retail liquor stores, requiring applicants for an initial license to have resided in the State for the prior two years; requiring an applicant for renewal of a license to reside in the State for 10 consecutive years; and providing that a corporation cannot obtain a license unless all of its stockholders are residents. Following the state attorney general’s opinion that the residency requirements discriminated against out-of-state economic interests in violation of the Commerce Clause, the Tennessee Alcoholic Beverage Commission (TABC) declined to enforce the requirements. Two businesses that did not meet the residency requirements (both respondents here) applied for licenses to own and operate liquor stores in Tennessee. Petitioner Tennessee Wine and Spirits Retailers Association (Association)—a trade association of in-state liquor stores—threatened to sue the TABC if it granted the licenses, so the TABC’s executive director (also a respondent) filed a declaratory judgment action in state court to settle the question of the residency requirements’ constitutionality. The case was removed to Federal District Court, which found the requirements unconstitutional. The State declined to appeal, but the Association took the case to the Sixth Circuit. It affirmed, concluding that the provisions violated the Commerce Clause. The Association petitioned for certiorari only with respect to the Sixth Circuit’s decision to invalidate the 2-year residency requirement applicable to initial liquor store license applicants. Held: Tennessee’s 2-year durational-residency requirement applicable to retail liquor store license applicants violates the Commerce Clause and is not saved by the Twenty-first Amendment. Pp. 6–37. (a) The Commerce Clause by its own force restricts state protectionism. Removing state trade barriers was a principal reason for the adoption of the Constitution, and at this point no provision other than the Commerce Clause could easily do that job. The Court has long emphasized the connection between the trade barriers that prompted the call for a new Constitution and its dormant Commerce Clause jurisprudence. See Guy v. Baltimore, 100 U.S. 434, 440; Granholm v. Heald, 544 U.S. 460, 472. Pp. 6–10. (b) Under the dormant Commerce Clause cases, a state law that discriminates against out-of-state goods or nonresident economic actors can be sustained only on a showing that it is narrowly tailored to “advanc[e] a legitimate local purpose.” Department of Revenue of Ky. v. Davis, 553 U.S. 328, 338. Tennessee’s 2-year residency requirement plainly favors Tennesseans over nonresidents. P. 10. (c) Because the 2-year residency requirement applies to the sale of alcohol, however, it must be evaluated in light of §2 of the Twenty-first Amendment. Pp. 10–20. (1) Section 2’s broad text—the “transportation or importation into any State, Territory, or possession of the United States for delivery or use therein of intoxicating liquors, in violation of the laws thereof, is hereby prohibited”—could be read to prohibit the transportation or importation of alcoholic beverages in violation of any state law. But the Court has declined to adopt that reading, instead interpreting §2 as one part of a unified constitutional scheme and in light of the provision’s history. History teaches that §2’s thrust is to “constitutionaliz[e]” the basic structure of federal-state alcohol regulatory authority that prevailed prior to the Eighteenth Amendment’s adoption. Craig v. Boren, 429 U.S. 190, 206. Pp. 10–12. (2) This Court invalidated many state liquor regulations before the Eighteenth Amendment’s ratification, and by the late 19th century it had concluded that the Commerce Clause both prevented States from discriminating “against citizens and products of other States,” Walling v. Michigan, 116 U.S. 446, 460, and “prevented States from passing facially neutral laws that placed an impermissible burden on interstate commerce,” Granholm, 544 U. S., at 477. State bans on the production and sale of alcohol within state borders were rendered ineffective by the “original-package doctrine,” which made “goods shipped in interstate commerce . . . immune from state regulation while in their original package.” Ibid. Congress responded by passing the Wilson Act, which provided that all alcoholic beverages “transported into any State or Territory” were subject “upon arrival” to the same restrictions imposed by the State “in the exercise of its police powers” over alcohol produced in the State, i.e., bona fide health and safety measures. This Court, however, narrowly construed the term “arrival” in the Wilson Act as arrival to the consignee rather than arrival within the State’s borders, which allowed consumers to continue to receive direct shipments of alcohol from out of State. Congress passed the Webb-Kenyon Act to close that loophole. But, as this Court’s decision in Granholm determined, the Webb-Kenyon Act was not intended to override the rule barring States from discriminating against out-of-state citizens and products, nor the traditional limits on state police power. Thereafter, the Eighteenth Amendment was ratified, prohibiting the manufacture, sale, transportation, and importation of alcoholic beverages across the country. Pp. 12–20. (d) Section 2 of the Twenty-first Amendment grants the States latitude with respect to the regulation of alcohol, but it does not allow the States to violate the “nondiscrimination principle” that was a central feature of the regulatory regime that the provision was meant to constitutionalize. Granholm, supra, at 487. Pp. 20–32. (1) The Twenty-first Amendment ended nationwide Prohibition, but §2 gave each State the option of banning alcohol if its citizens so chose. Its text “closely follow[ed]” the Webb-Kenyon Act’s operative language, suggesting that it was meant to have a similar meaning. Craig v. Boren, 429 U. S., at 205–206. The provision was meant to “constitutionaliz[e]” the basic understanding of the extent of the States’ power to regulate alcohol that prevailed before Prohibition. Id., at 206. And during that period, the Commerce Clause did not permit the States to impose protectionist measures clothed as police-power regulations. Pp. 20–22. (2) At first, the Court did not take account of this history. But it has since recognized that §2 cannot be interpreted to override all previously adopted constitutional provisions, scrutinizing state alcohol laws for compliance with, e.g., the Free Speech Clause, 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484; the Establishment Clause, Larkin v. Grendel’s Den, Inc., 459 U.S. 116; the Equal Protection Clause, Craig v. Boren, supra; the Due Process Clause, Wisconsin v. Constantineau, 400 U.S. 433; and the Import-Export Clause, Department of Revenue v. James B. Beam Distilling Co., 377 U.S. 341. Section 2 also does not entirely supersede Congress’s power to regulate commerce, see, e.g., Hostetter v. Idlewild Bon Voyage Liquor Corp., 377 U.S. 324, 333–334, nor is its aim to permit States to restrict the importation of alcohol for purely protectionist purposes, see, e.g., Granholm, supra, at 486–487. Pp. 22–23. (3) Protectionism is not a legitimate §2 interest shielding state alcohol laws that burden interstate commerce. Bacchus Imports, Ltd. v. Dias, 468 U.S. 263, 276. The Court has applied that principle to invalidate state alcohol laws aimed at giving a competitive advantage to in-state businesses. See, e.g., id., at 274. Pp. 24–26. (4) The Association and the dissent’s overly broad understanding of §2 is unpersuasive. They claim that, while §2 does not give the States the power to discriminate against out-of-state alcohol products and producers, a different rule applies to state laws regulating in-state alcohol distribution. There is no sound basis for this distinction. The Association and the dissent also claim that discriminatory distribution laws, including in-state residency requirements, long predate Prohibition and were adopted by many States following the Twenty-first Amendment’s ratification. State laws adopted soon after ratification, however, may have been based on an overly expansive interpretation of §2 that can no longer be defended, and many state laws adopted before Prohibition were never tested in this Court. Nor have States historically enjoyed absolute authority to police alcohol within their borders. Section 2 allows each State leeway to enact measures to address the public health and safety effects of alcohol use and other legitimate interests, but it does not license the States to adopt protectionist measures with no demonstrable connection to those interests. Pp. 26–32. (d) Applying the appropriate §2 analysis here, Tennessee’s 2-year residency requirement cannot be sustained. The provision expressly discriminates against nonresidents and has at best a highly attenuated relationship to public health or safety. The Association claims that the requirement ensures that retailers are subject to process in state courts, but does not explain why that objective could not easily be achieved by, e.g., requiring a nonresident to designate an agent to receive process. Similarly unpersuasive is its claim that the requirement allows the State to ensure that only law-abiding and responsible applicants receive licenses. The State can thoroughly investigate applicants without requiring them to reside in the State for two years, and in any event the requirement poorly serves that goal since the TABC would have no reason to investigate a nonresident who moves to the State with the intention of applying for a license once the 2-year period ends. Nor is the residency requirement needed to enable the State to maintain oversight over liquor store operators; they can be monitored through any number of nondiscriminatory means, including on-site inspections, audits, and the like. There is also no evidence to support the claim that the requirement would promote responsible alcohol consumption because retailers who know the communities they serve will be more likely to engage in responsible sales practices. The residency requirement is poorly designed for such a purpose, and the State could better serve the goal without discriminating against nonresidents by, e.g., limiting both the number of retail licenses and the amount of alcohol that may be sold to an individual, mandating more extensive training for managers and employees, or monitoring retailer practices and taking action against those who violate the law. Pp. 32–36. 883 F.3d 608, affirmed. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Sotomayor, Kagan, and Kavanaugh, JJ., joined. Gorsuch, J., filed a dissenting opinion, in which Thomas, J., joined.

The State of Tennessee imposes demanding durational-residency requirements on all individuals and businesses seeking to obtain or renew a license to operate a liquor store. One provision precludes the renewal of a license unless the applicant has resided in the State for 10 consecutive years. Another provides that a corporation cannot obtain a license unless all of its stockholders are residents. The Court of Appeals for the Sixth Circuit struck down these provisions as blatant violations of the Commerce Clause, and neither petitioner—an association of Tennessee liquor retailers—nor the State itself defends them in this Court. The Sixth Circuit also invalidated a provision requiring applicants for an initial license to have resided in the State for the prior two years, and petitioner does challenge that decision. But while this requirement is less extreme than the others that the Sixth Circuit found to be unconstitutional, we now hold that it also violates the Commerce Clause and is not shielded by §2 of the Twenty-first Amendment. Section 2 was adopted as part of the scheme that ended prohibition on the national level. It gives each State leeway in choosing the alcohol-related public health and safety measures that its citizens find desirable. But §2 is not a license to impose all manner of protectionist restrictions on commerce in alcoholic beverages. Because Tennessee’s 2-year residency requirement for retail license applicants blatantly favors the State’s residents and has little relationship to public health and safety, it is unconstitutional. I A Tennessee, like many other States, requires alcoholic beverages distributed in the State to pass through a specified three-tiered system.[1] Acting through the Tennessee Alcoholic Beverage Commission (TABC), the State issues different types of licenses to producers, wholesalers, and retailers of alcoholic beverages. See Tenn. Code Ann. §57–3–201 (2018). Producers may sell only to licensed wholesalers; wholesalers may sell only to licensed retailers or other wholesalers; and only licensed retailers may sell to consumers. §57–3–404. No person may lawfully participate in the sale of alcohol without the appropriate license. See, e.g., §57–3–406. Included in the Tennessee scheme are onerous durational-residency requirements for all persons and compa- nies wishing to operate “retail package stores” that sell alcoholic beverages for off-premises consumption (hereinafter liquor stores). See §57–3–204(a). To obtain an initial retail license, an individual must demonstrate that he or she has “been a bona fide resident” of the State for the previous two years. §57–3–204(b)(2)(A). And to renew such a license—which Tennessee law requires after only one year of operation—an individual must show continuous residency in the State for a period of 10 consecutive years. Ibid. The rule for corporations is also extraordinarily restrictive. A corporation cannot get a retail license unless all of its officers, directors, and owners of capital stock satisfy the durational-residency requirements applicable to individuals. §57–3–204(b)(3). In practice, this means that no corporation whose stock is publicly traded may operate a liquor store in the State. In 2012, the Tennessee attorney general was asked whether the State’s durational-residency requirements violate the Commerce Clause, and his answer was that the requirements constituted “trade restraints and barriers that impermissibly discriminate against interstate commerce.” App. to Brief in Opposition 11a; see also id., at 12a (citing Jelovsek v. Bredesen, 545 F.3d 431, 435 (CA6 2008)). In light of that opinion, the TABC stopped enforcing the requirements against new applicants. See App. 51, ¶9; id., at 76, ¶10. The Tennessee General Assembly responded by amending the relevant laws to include a statement of legislative intent. Citing the alcohol content of the beverages sold in liquor stores, the Assembly found that protection of “the health, safety and welfare” of Tennesseans called for “a higher degree of oversight, control and accountability for individuals involved in the ownership, management and control” of such outlets. §57–3–204(b)(4). After the amendments became law, the attorney gen- eral was again asked about the constitutionality of the durational-residency requirements, but his answer was the same as before. See App. to Brief in Opposition 13a. Consequently, the TABC continued its practice of nonenforcement. B In 2016, respondents Tennessee Fine Wines and Spirits, LLC dba Total Wine Spirits Beer & More (Total Wine) and Affluere Investments, Inc. dba Kimbrough Fine Wine & Spirits (Affluere) applied for licenses to own and operate liquor stores in Tennessee. At the time, neither Total Wine nor Affluere satisfied the durational-residency requirements. Total Wine was formed as a Tennessee limited liability company but is owned by residents of Maryland, Brief for Respondent Total Wine 10; App. 51, ¶4–5, and Affluere was owned and controlled by two individuals who, by the time their application was considered, had only recently moved to the State, see App. 11–12, 20, 22. TABC staff recommended approval of the applications, but petitioner Tennessee Wine and Spirits Retailers Association (the Association)—a trade association of in-state liquor stores—threatened to sue the TABC if it granted them. Id., at 15, ¶17. The TABC’s executive director (a respondent here) filed a declaratory judgment action in state court to settle the question of the residency requirements’ constitutionality. Id., at 17. The case was removed to the United States District Court for the Middle District of Tennessee, and that court, relying on our decision in Granholm v. Heald, 544 U.S. 460 (2005), concluded that the requirements are unconstitutional. Byrd v. Tennessee Wine and Spirits Retailers Assn., 259 F. Supp. 3d 785, 797 (2017). The State de- clined to appeal, and Total Wine and Affluere were issued licenses. The Association, however, took the case to the Court of Appeals for the Sixth Circuit, where a divided panel affirmed. See Byrd v. Tennessee Wine and Spirits Retailers Assn., 883 F.3d 608 (2018). All three judges acknowledged that the Tennessee residency requirements facially discriminate against out-of-state economic interests. See id., at 624; id., at 634 (Sutton, J., concurring in part and dissenting in part). And all three also agreed that neither the 10-year residency requirement for license renewals nor the 100-percent-resident shareholder requirement is constitutional under this Court’s Twenty-first Amendment and dormant Commerce Clause precedents. See id., at 625–626; id., at 635 (opinion of Sutton, J.). The panel divided, however, over the constitutionality of the 2-year residency requirement for individuals seeking initial retail licenses, as well as the provision applying those requirements to officers and directors of corporate applicants. Applying standard dormant Commerce Clause scrutiny, the majority struck down the challenged restrictions, reasoning that they facially discriminate against interstate commerce and that the interests they are claimed to further can be adequately served through reasonable, nondiscriminatory alternatives. Id., at 623–626. The dissent disagreed, reading §2 of the Twenty-first Amendment to grant States “ ‘virtually’ limitless” authority to regulate the in-state distribution of alcohol, the only exception being for laws that “serve no purpose besides ‘economic protectionism.’ ” Id., at 633 (quoting Bacchus Imports, Ltd. v. Dias, 468 U.S. 263, 276 (1984)). Applying that highly deferential standard, the dissent would have upheld the 2-year residency requirement, as well as the provision applying that requirement to all officers and directors of corporate applicants. The dissent argued that these provisions help to promote the State’s interests in “responsible consumption” of alcohol and “orderly liquor markets.” 883 F. 3d, at 633. The Association filed a petition for a writ of certiorari challenging the decision on the 2-year residency requirement for initial licenses. Tennessee declined to seek certiorari but filed a letter with the Court expressing agreement with the Association’s position.[2] We granted certiorari, 585 U. S. ___ (2018), in light of the disagreement among the Courts of Appeals about how to reconcile our modern Twenty-first Amendment and dormant Commerce Clause precedents. See 883 F. 3d, at 616 (collecting cases). II A The Court of Appeals held that Tennessee’s 2-year residency requirement violates the Commerce Clause, which provides that “[t]he Congress shall have Power . . . [t]o regulate Commerce with foreign Nations, and among the several States, and with the Indian Tribes.” Art. I, §8, cl. 3. “Although the Clause is framed as a positive grant of power to Congress,” Comptroller of Treasury of Md. v. Wynne, 575 U. S. ___, ___ (2015) (slip op., at 5), we have long held that this Clause also prohibits state laws that unduly restrict interstate commerce. See, e.g., ibid.; Philadelphia v. New Jersey, 437 U.S. 617, 623–624 (1978); Cooley v. Board of Wardens of Port of Philadelphia ex rel. Soc. for Relief of Distressed Pilots, 12 How. 299, 318–319 (1852); Willson v. Black Bird Creek Marsh Co., 2 Pet. 245, 252 (1829). “This ‘negative’ aspect of the Commerce Clause” prevents the States from adopting protectionist measures and thus preserves a national market for goods and services. New Energy Co. of Ind. v. Limbach, 486 U.S. 269, 273 (1988). This interpretation, generally known as “the dormant Commerce Clause,” has a long and complicated history. Its roots go back as far as Gibbons v. Ogden, 9 Wheat. 1 (1824), where Chief Justice Marshall found that a version of the dormant Commerce Clause argument had “great force.” Id., at 209. His successor disagreed, see License Cases, 5 How. 504, 578–579 (1847) (Taney, C. J.), but by the latter half of the 19th century the dormant Commerce Clause was firmly established, see, e.g., Case of the State Freight Tax, 15 Wall. 232, 279–280 (1873), and it played an important role in the economic history of our Nation. See Cushman, Formalism and Realism in Commerce Clause Jurisprudence, 67 U. Chi. L. Rev. 1089, 1107 (2000). In recent years, some Members of the Court have authored vigorous and thoughtful critiques of this interpretation. See, e.g., Camps Newfound/Owatonna, Inc. v. Town of Harrison, 520 U.S. 564, 609–620 (1997) (Thomas, J., dissenting); Tyler Pipe Industries, Inc. v. Washington State Dept. of Revenue, 483 U.S. 232, 259–265 (1987) (Scalia, J., concurring in part and dissenting in part); cf. post, at 2–3 (Gorsuch, J., dissenting) (deeming doctrine “peculiar”). But the proposition that the Commerce Clause by its own force restricts state protectionism is deeply rooted in our case law. And without the dormant Commerce Clause, we would be left with a constitutional scheme that those who framed and ratified the Constitution would surely find surprising. That is so because removing state trade barriers was a principal reason for the adoption of the Constitution. Under the Articles of Confederation, States notoriously obstructed the interstate shipment of goods. “Interference with the arteries of commerce was cutting off the very life-blood of the nation.” M. Farrand, The Framing of the Constitution of the United States 7 (1913). The Annapolis Convention of 1786 was convened to address this critical problem, and it culminated in a call for the Philadelphia Convention that framed the Constitution in the summer of 1787.[3] At that Convention, discussion of the power to regulate interstate commerce was almost uniformly linked to the removal of state trade barriers, see Abel, The Commerce Clause in the Constitutional Convention and in Contemporary Comment, 25 Minn. L. Rev. 432, 470–471 (1941), and when the Constitution was sent to the state conventions, fostering free trade among the States was prominently cited as a reason for ratification. In The Federalist No. 7, Hamilton argued that state protectionism could lead to conflict among the States, see The Federalist No. 7, pp. 62–63 (C. Rossiter ed. 1961), and in No. 11, he touted the benefits of a free national market, id., at 88–89. In The Federalist No. 42, Madison sounded a similar theme. Id., at 267–268. In light of this background, it would be strange if the Constitution contained no provision curbing state protectionism, and at this point in the Court’s history, no provision other than the Commerce Clause could easily do the job. The only other provisions that the Framers might have thought would fill that role, at least in part, are the Import-Export Clause, Art. I, §10, cl. 2, which generally prohibits a State from “lay[ing] any Imposts or Duties on Imports or Exports,” and the Privileges and Immunities Clause, Art. IV, §2, which provides that “[t]he Citizens of each State shall be entitled to all Privileges and Immunities of Citizens in the several States.” But the Import-Export Clause was long ago held to refer only to international trade. See Woodruff v. Parham, 8 Wall. 123, 136–137 (1869). And the Privileges and Immunities Clause has been interpreted not to protect corporations, Western & Southern Life Ins. Co. v. State Bd. of Equalization of Cal., 451 U.S. 648, 656 (1981) (citing Hemphill v. Orloff, 277 U.S. 537, 548–550 (1928)), and may not guard against certain discrimination scrutinized under the dormant Commerce Clause, see Denning, Why the Privileges and Immunities Clause of Article IV Cannot Replace the Dormant Commerce Clause Doctrine, 88 Minn. L. Rev. 384, 393–397 (2003). So if we accept the Court’s established interpretation of those provisions, that leaves the Commerce Clause as the primary safeguard against state protectionism.[4] It is not surprising, then, that our cases have long emphasized the connection between the trade barriers that prompted the call for a new Constitution and our dormant Commerce Clause jurisprudence. In Guy v. Baltimore, 100 U.S. 434, 440 (1880), for example, the Court wrote that state protectionist measures, “if maintained by this court, would ultimately bring our commerce to that ‘oppressed and degraded state,’ existing at the adoption of the present Constitution, when the helpless, inadequate Confederation was abandoned and the national government instituted.” More recently, we observed that our dormant Commerce Clause cases reflect a “ ‘central concern of the Framers that was an immediate reason for calling the Constitutional Convention: the conviction that in order to succeed, the new Union would have to avoid the tendencies toward economic Balkanization that had plagued relations among the Colonies and later among the States under the Articles of Confederation.’ ” Granholm, 544 U. S., at 472 (quoting Hughes v. Oklahoma, 441 U.S. 322, 325–326 (1979)). In light of this history and our established case law, we reiterate that the Commerce Clause by its own force restricts state protectionism. B Under our dormant Commerce Clause cases, if a state law discriminates against out-of-state goods or nonresident economic actors, the law can be sustained only on a showing that it is narrowly tailored to “ ‘advanc[e] a legitimate local purpose.’ ” Department of Revenue of Ky. v. Davis, 553 U.S. 328, 338 (2008). See also, e.g., Oregon Waste Systems, Inc. v. Department of Environmental Quality of Ore., 511 U.S. 93, 100–101 (1994); Maine v. Taylor, 477 U.S. 131, 138 (1986). Tennessee’s 2-year durational-residency requirement plainly favors Tennesseans over nonresidents, and neither the Association nor the dissent below defends that requirement under the standard that would be triggered if the requirement applied to a person wishing to operate a retail store that sells a commodity other than alcohol. See 883 F. 3d, at 626. Instead, their arguments are based on §2 of the Twenty-first Amendment, to which we will now turn. III A Section 2 of the Twenty-first Amendment provides as follows: “The transportation or importation into any State, Territory, or possession of the United States for delivery or use therein of intoxicating liquors, in violation of the laws thereof, is hereby prohibited.” Although the interpretation of any provision of the Constitution must begin with a consideration of the literal meaning of that particular provision, reading §2 to prohibit the transportation or importation of alcoholic beverages in violation of any state law[5] would lead to absurd results that the provision cannot have been meant to produce. Under the established rule that a later adopted provision takes precedence over an earlier, conflicting provision of equal stature, see, e.g., United States v. Tynen, 11 Wall. 88, 92 (1871); Posadas v. National City Bank, 296 U.S. 497, 503 (1936); A. Scalia & B. Garner, Reading Law 327–328 (2012); 1A N. Singer & J. Singer, Sutherland on Statutory Construction §23:9 (7th ed. 2009), such a reading of §2 would mean that the provision would trump any irreconcilable provision of the original Constitution, the Bill of Rights, the Fourteenth Amendment, and every other constitutional provision predating ratification of the Twenty-first Amendment in 1933. This would mean, among other things, that a state law prohibiting the importation of alcohol for sale to persons of a particular race, religion, or sex would be immunized from challenge under the Equal Protection Clause. Similarly, if a state law prohibited the importation of alcohol for sale by proprietors who had expressed an unpopular point of view on an important public issue, the First Amendment would provide no protection. If a State imposed a duty on the importation of foreign wine or spirits, the Import-Export Clause would have to give way. If a state law retroactively made it a crime to have bought or sold imported alcohol under specified conditions, the Ex Post Facto Clause would provide no barrier to conviction. The list goes on. Despite the ostensibly broad text of §2, no one now contends that the provision must be interpreted in this way. Instead, we have held that §2 must be viewed as one part of a unified constitutional scheme. See California Retail Liquor Dealers Assn. v. Midcal Aluminum, Inc., 445 U.S. 97, 109 (1980); Hostetter v. Idlewild Bon Voyage Liquor Corp., 377 U.S. 324, 331–332 (1964); cf. Scalia & Garner, supra, at 167–169, 180–182. In attempting to understand how §2 and other constitutional provisions work together, we have looked to history for guidance, and history has taught us that the thrust of §2 is to “constitutionaliz[e]” the basic structure of federal-state alcohol regulatory authority that prevailed prior to the adoption of the Eighteenth Amendment. Craig v. Boren, 429 U.S. 190, 206 (1976). We therefore examine that history. B Throughout the 19th century, social problems attributed to alcohol use prompted waves of state regulation, and these measures were often challenged as violations of various provisions of the Federal Constitution. One wave of state regulation occurred during the first half of the century. The country’s early years were a time of notoriously hard drinking, see D. Okrent, Last Call: The Rise and Fall of Prohibition 7 (2010),[6] and the problems that this engendered prompted States to enact a variety of regulations, including licensing requirements, age restrictions, and Sunday-closing laws. See Byse, Alcoholic Beverage Control Before Repeal, 7 Law & Contemp. Prob. 544, 546–551 (1940). Three States’ alcohol licensing laws came before this Court in 1847 in the License Cases, 5 How. 504. The principal claim in those cases was similar to the one now before us; licensing laws enacted in three States were challenged under the Commerce Clause. The Court unanimously rejected those claims, but six Justices authored opinions; no opinion commanded a majority; and the general status of dormant Commerce Clause claims was left uncertain. See 5 C. Swisher, The Taney Period, 1836–64, History of the Supreme Court of the United States 373–374 (1974). Following the Civil War, the Court considered a steady stream of alcohol-regulation cases. The postwar period saw a great proliferation of saloons,[7] and myriad social problems were attributed to this development. In response, many States passed laws restricting the sale of alcohol. By 1891, six States had banned alcohol production and sale completely. R. Hamm, Shaping the Eighteenth Amendment 25 (1995) (Hamm). During this period, state laws regulating the alcohol trade were unsuccessfully challenged in this Court on a variety of constitutional grounds. See, e.g., Mugler v. Kansas, 123 U.S. 623 (1887) (Privileges or Immunities and Due Process Clauses of Fourteenth Amendment); Beer Co. v. Massachusetts, 97 U.S. 25 (1878) (Contracts Clause); Bartemeyer v. Iowa, 18 Wall. 129 (1874) (Privileges or Immunities and Due Process Clauses of Fourteenth Amendment). In those decisions, the Court staunchly affirmed the “right of the States,” in exercising their “police power,” to “protect the health, morals, and safety of their people,” but the Court also cautioned that this objective could be pursued only “by regulations that do not interfere with the execution of the powers of the general government, or violate rights secured by the Constitution of the United States.” Mugler, 123 U. S., at 659. For that reason, the Court continued, “mere pretences” could not sustain a law regulating alcohol; rather, if “a statute purporting to have been enacted to protect the public health, the public morals, or the public safety, has no real or substantial relation to those objects, or is a palpable invasion of rights secured by the fundamental law, it is the duty of the courts to so adjudge, and thereby give effect to the Constitution.” Id., at 661. Dormant Commerce Clause challenges also reached the Court. States that banned the production and sale of alcohol within their borders found that these laws did not stop residents from consuming alcohol shipped in from other States. To curb that traffic, States passed laws regulating or prohibiting the importation of alcohol, and these enactments were quickly challenged. By the late 19th century, the Court was firmly of the view that the Commerce Clause by its own force restricts state regulation of interstate commerce. See Bowman v. Chicago & Northwestern R. Co., 125 U.S. 465 (1888); Leisy v. Hardin, 135 U.S. 100 (1890). Dormant Commerce Clause cases from that era “advanced two distinct principles,” an understanding of which is critical to gauging the States’ pre-Prohibition power to regulate alcohol. Granholm, 544 U. S., at 476. First, the Court held that the Commerce Clause prevented States from discriminating “against the citizens and products of other States,” Walling v. Michigan, 116 U.S. 446, 460 (1886). See also Scott v. Donald, 165 U.S. 58 (1897); Tiernan v. Rinker, 102 U.S. 123 (1880). Applying that rule, the Walling Court struck down a discriminatory state fee that applied only to those in the business of selling imported alcohol. 116 U. S., at 454, 458. Similarly, in Scott, the Court invalidated a law that gave an “unjust preference [to] the products of the enacting State as against similar products of the other States.” 165 U. S., at 101. The Court did not question the States’ use of the police power to regulate the alcohol trade but stressed that such regulation must have a “bona fide” relation to protecting “ ‘the public health, the public morals or the public safety,’ ” id., at 91 (quoting Mugler, supra, at 661), and could not encroach upon Congress’s “power to regulate commerce among the several States,” Walling, supra, at 458. Second, the Court “held that the Commerce Clause prevented States from passing facially neutral laws that placed an impermissible burden on interstate commerce.” Granholm, 544 U. S., at 477. At the time of these decisions, the “original-package doctrine” defined the outer limits of Congress’s authority to regulate interstate commerce. Ibid. See Brown v. Maryland, 12 Wheat. 419 (1827). Under that doctrine, “goods shipped in interstate commerce were immune from state regulation while in their original package,” because at that point they had not yet been comingled with the mass of domestic property subject to state jurisdiction. Granholm, 544 U. S., at 477; see id., at 477–478 (citing Vance v. W. A. Vandercook Co., 170 U.S. 438, 444–445 (1898)). Applying this doctrine to state alcohol laws, the Court struck down an Iowa statute that required importers to obtain special certificates, Bowman, supra, as well as another Iowa law that, with limited exceptions, banned the importation of liquor, Leisy, supra. These decisions left dry States “in a bind.” Granholm, supra, at 478. See Rogers, Interstate Commerce in Intoxicating Liquors Before the Webb-Kenyon Act, 4 Va. L. Rev. 174 (1916), 288 (1917) (noting “practical nullification of state laws” by original-package decisions). States could ban the production and sale of alcohol within their borders, but those bans “were ineffective because out-of-state liquor was immune from any state regulation as long as it remained in its original package.” Granholm, supra, at 478. In effect, the Court’s interpretation of the dormant Commerce Clause conferred favored status on out-of-state alcohol, and that hamstrung the dry States’ efforts to enforce local prohibition laws. Representatives of those States and temperance advocates thus turned to Congress, which passed two laws to solve the problem. The first of these was the Wilson Act, enacted in 1890. Ch. 728, 26Stat. 313, 27 U. S. C. §121. Named for Senator James F. Wilson of Iowa, whose home State’s laws had fallen in Bowman and Leisy, the Wilson Act aimed to obviate the problem presented by the “original-package” rule. Dormant Commerce Clause restrictions apply only when Congress has not exercised its Commerce Clause power to regulate the matter at issue, cf. Bowman, supra, at 485; Leisy, supra, at 123–124, and the strategy of those who favored the Wilson Act was for Congress to eliminate the problem that had surfaced in Bowman and Leisy by regulating the interstate shipment of alcohol, see Hamm 77–80; Rogers, supra, at 194–195. During the late 19th century and early 20th century, Congress enacted laws that entirely prohibited the transportation of certain goods and persons across state lines, and some but not all of these measures were held to be valid exercises of the commerce power. See Lottery Case, 188 U.S. 321 (1903) (upholding law prohibiting interstate shipment of lottery tickets); Hoke v. United States, 227 U.S. 308 (1913) (sustaining Mann Act prohibition on bringing women across state lines for prostitution); Hammer v. Dagenhart, 247 U.S. 251 (1918) (striking down provision banning interstate shipment of goods produced by child labor). Unlike these laws, the Wilson Act did not attempt to ban all interstate shipment of alcohol. Its goal was more modest: to leave it up to each State to decide whether to admit alcohol. Its critical provision specified that all alcoholic beverages “transported into any State or Terri- tory” were subject “upon arrival” to the same restrictions imposed by the State “in the exercise of its police powers” over alcohol produced in the State.[8] Thus, the Wilson Act mandated equal treatment for alcohol produced within and outside a State, not favorable treatment for local products. See Granholm, supra, at 479 (discussing Scott, 165 U. S., at 100–101). And the only state laws that it attempted to shield were those enacted by a State “in the exercise of its police powers,” which, as we have seen, applied only to bona fide health and safety measures. See, e.g., id., at 91 (citing Mugler, 123 U. S., at 661). Despite Congress’s clear aim, the Wilson Act failed to relieve the dry States’ predicament. In Rhodes v. Iowa, 170 U.S. 412 (1898), and Vance v. W. A. Vandercook Co., supra, the Court read the Act’s reference to the “arrival” of alcohol in a State to mean delivery to the consignee, not arrival within the State’s borders. Granholm, 544 U. S., at 480. The upshot was that residents of dry States could continue to order and receive imported alcohol. Ibid. See also Hamm 178. In 1913, Congress tried to patch this hole by passing the Webb-Kenyon Act, ch. 90, 37Stat. 699, 27 U. S. C. §122. The aim of the Webb-Kenyon Act was to give each State a measure of regulatory authority over the importation of alcohol, but this created a drafting problem. There were those who thought that a federal law giving the States this authority would amount to an unconstitutional delegation of Congress’s legislative power over interstate commerce.[9] So the Act was framed not as a measure conferring power on the States but as one prohibiting conduct that violated state law. The Act provided that the shipment of alcohol into a State for use in any manner, “either in the original package or otherwise,” “in violation of any law of such State,” was prohibited.[10] This formulation is significant for present purposes because it would provide a model for §2 of the Twenty-first Amendment. The Webb-Kenyon Act attempted to fix the hole in the Wilson Act and thus to “eliminate the regulatory advantage . . . afforded imported liquor,” Granholm, supra, at 482; see also Clark Distilling Co. v. Western Maryland R. Co., 242 U.S. 311, 324 (1917), but its wording, unlike the Wilson Act’s, did not explicitly mandate equal treatment for imported and domestically produced alcohol. And it referred to “any law of such State,” 37Stat. 700 (emphasis added), whereas the Wilson Act referred to “the laws of such State or Territory enacted in the exercise of its police powers.” 26Stat. 313 (emphasis added). But despite these differences, Granholm held, over a strenuous dissent, 544 U. S., at 505–514 (opinion of Thomas, J.), that the Webb-Kenyon Act did not purport to authorize States to enact protectionist measures. There is good reason for this holding. As we have noted, the Court’s pre-Webb-Kenyon Act decisions upholding state liquor laws against challenges based on constitutional provisions other than the Commerce Clause had cau- tioned that protectionist laws disguised as exercises of the police power would not escape scrutiny. See supra, at 14–15.[11] The Webb-Kenyon Act, by regulating commerce, could obviate dormant Commerce Clause problems, but it could not override the limitations imposed by these other constitutional provisions and the traditional understanding regarding the bounds of the States’ inherent police powers. Therefore the Wilson Act’s reference to laws “enacted in the exercise of [a State’s] police powers,” 26Stat. 313, merely restated what this Court had already found to be a constitutional necessity, and consequently, there was no need to include such language in the Webb-Kenyon Act. Even without limiting language like that in the Wilson Act, the shelter given by the Webb-Kenyon Act applied only where “the States treated in-state and out-of-state liquor on the same terms.” Granholm, supra, at 481.[12] Following passage of the Webb-Kenyon Act, temperance advocates began the final push for nationwide Prohibition, and with the ratification of the Eighteenth Amendment in 1919, their goal was achieved. The manufacture, sale, transportation, and importation of alcoholic beverages anywhere in the country were prohibited. IV A By 1933, support for Prohibition had substantially diminished but not vanished completely. Thirty-eight state conventions eventually ratified the Twenty-first Amendment, but 10 States either rejected or took no action on the Amendment. Section 1 of the Twenty-first Amendment repealed the Eighteenth Amendment and thus ended nationwide Prohibition, but §2, the provision at issue here, gave each State the option of banning alcohol if its citizens so chose. As we have previously noted, the text of §2 “closely follow[ed]” the operative language of the Webb-Kenyon Act, and this naturally suggests that §2 was meant to have a similar meaning. Craig, 429 U. S., at 205–206. The decision to follow that unusual formulation is especially revealing since the drafters of §2, unlike those who framed the Webb-Kenyon Act, had no need to worry that a more straightforward wording might trigger a constitutional challenge. Accordingly, we have inferred that §2 was meant to “constitutionaliz[e]” the basic understanding of the extent of the States’ power to regulate alcohol that prevailed before Prohibition. Id., at 206. See also Granholm, supra, at 484. And as recognized during that period, the Commerce Clause did not permit the States to impose protectionist measures clothed as police-power regulations. See supra, at 14–15. See also, e.g., Railroad Co. v. Husen, 95 U.S. 465, 472 (1878) (a State “may not, under the cover of exerting its police powers, substantially prohibit or burden either foreign or inter-state commerce”). This understanding is supported by the debates on the Amendment in Congress[13] and the state ratifying conventions. The records of the state conventions provide no evidence that §2 was understood to give States the power to enact protectionist laws,[14] “a privilege [the States] had not enjoyed at any earlier time.” Granholm, supra, at 485. B Although our later cases have recognized that §2 cannot be given an interpretation that overrides all previously adopted constitutional provisions, the Court’s earliest cases interpreting §2 seemed to feint in that direction. In 1936, the Court found that §2’s text was “clear” and saw no need to consider whether history supported a more modest interpretation, State Bd. of Equalization of Cal. v. Young’s Market Co., 299 U.S. 59, 63–64 (1936)—an approach even the dissent rejects, see infra, at 24, n. 16; post, at 2.[15] The Court read §2 as granting each State plenary “power to forbid all importations which do not comply with the conditions which it prescribes,” Young’s Market, supra, at 62; see also Ziffrin, Inc. v. Reeves, 308 U.S. 132, 138–139 (1939), including laws that discriminated against out-of-state products. See, e.g., Young’s Market, supra, at 62; Mahoney v. Joseph Triner Corp., 304 U.S. 401, 403 (1938); Indianapolis Brewing Co. v. Liquor Control Comm’n, 305 U.S. 391, 394 (1939). The Court went so far as to assume that the Fourteenth Amendment imposed no barrier to state legislation in the field of alcohol regulation. See Young’s Market, supra, at 64 (“A classification recognized by the Twenty-first Amendment cannot be deemed forbidden by the Fourteenth”). With subsequent cases, however, the Court saw that §2 cannot be read that way, and it therefore scrutinized state alcohol laws for compliance with many constitutional provisions. See, e.g., 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484 (1996) (Free Speech Clause); Larkin v. Grendel’s Den, Inc., 459 U.S. 116 (1982) (Establishment Clause); Craig v. Boren, supra (Equal Protection Clause); Wisconsin v. Constantineau, 400 U.S. 433 (1971) (Due Process Clause); Department of Revenue v. James B. Beam Distilling Co., 377 U.S. 341 (1964) (Import-Export Clause). The Court also held that §2 does not entirely supersede Congress’s power to regulate commerce. Instead, after evaluating competing federal and state interests, the Court has ruled against state alcohol laws that conflicted with federal regulation of the export of alcohol, Hostetter, 377 U. S., at 333–334, federal antitrust law, Midcal Aluminum, 445 U. S., at 110–111, 113–114; 324 Liquor Corp. v. Duffy, 479 U.S. 335, 346–347, 350–351 (1987), and federal regulation of the airwaves, Capital Cities Cable, Inc. v. Crisp, 467 U.S. 691, 713, 716 (1984). As for the dormant Commerce Clause, the developments leading to the adoption of the Twenty-first Amendment have convinced us that the aim of §2 was not to give States a free hand to restrict the importation of alcohol for purely protectionist purposes. See Granholm, supra, at 486–487; Bacchus, 468 U. S., at 276. C Although some Justices have argued that §2 shields all state alcohol regulation—including discriminatory laws—from any application of dormant Commerce Clause doctrine,[16] the Court’s modern §2 precedents have repeatedly rejected that view. We have examined whether state alcohol laws that burden interstate commerce serve a State’s legitimate §2 interests. And protectionism, we have stressed, is not such an interest. Ibid. Applying that principle, we have invalidated state alcohol laws aimed at giving a competitive advantage to in-state businesses. The Court’s decision in Bacchus “provides a particularly telling example.” Granholm, supra, at 487. There, the Court was confronted with a tax exemption that favored certain in-state alcohol producers. In defending the law, the State argued that even if the discriminatory exemption violated “ordinary Commerce Clause principles, it [was] saved by the Twenty-first Amendment.” Bacchus, 468 U. S., at 274. We rejected that argument and held instead that the relevant question was “whether the principles underlying the Twenty-first Amendment are sufficiently implicated by the [discriminatory] exemption . . . to outweigh the Commerce Clause principles that would otherwise be offended.” Id., at 275. Ultimately, we held that §2 did not save the disputed tax because it clearly aimed “ ‘to promote a local industry’ ” rather than “to promote temperance or to carry out any other purpose of the Twenty-first Amendment.” Id., at 276. The same went for the state law in Healy v. Beer Institute, 491 U.S. 324 (1989), which required out-of-state shippers of beer to affirm that their wholesale price for products sold in Connecticut was no higher than the prices they charged to wholesalers in bordering States. Connecticut argued that the “Twenty-first Amendment sanction[ed]” this law “regardless of its effect on interstate commerce,” id., at 341, but we held that the law violated the Commerce Clause, noting that it “discriminate[d] against brewers and shippers of beer engaged in interstate commerce” without justification “by a valid factor unrelated to economic protectionism,” id., at 340–341.[17] Most recently, in Granholm, we struck down a set of discriminatory direct-shipment laws that favored in-state wineries over out-of-state competitors. After surveying the history of §2, we affirmed that “the Twenty-first Amendment does not immunize all laws from Commerce Clause challenge.” 544 U. S., at 488. We therefore examined whether the challenged laws were reasonably necessary to protect the States’ asserted interests in policing underage drinking and facilitating tax collection. Id., at 489–493. Concluding that the answer to that question was no, we invalidated the laws as inconsistent with the dormant Commerce Clause’s nondiscrimination principle. Id., at 492–493. To summarize, the Court has acknowledged that §2 grants States latitude with respect to the regulation of alcohol, but the Court has repeatedly declined to read §2 as allowing the States to violate the “nondiscrimination principle” that was a central feature of the regulatory regime that the provision was meant to constitutionalize. Id., at 487. D The Association resists this reading. Although it concedes (as it must under Granholm) that §2 does not give the States the power to discriminate against out-of-state alcohol products and producers, the Association presses the argument, echoed by the dissent, that a different rule applies to state laws that regulate in-state alcohol distribution. There is no sound basis for this distinction.[18] 1 The Association’s argument encounters a problem at the outset. The argument concedes that §2 does not shield state laws that discriminate against interstate commerce with respect to the very activity that the provision explicitly addresses—the importation of alcohol. But at the same time, the Association claims that §2 protects something that §2’s text, if read literally, does not cover—laws restricting the licensing of domestic retail alcohol stores. That reading is implausible. Surely if §2 granted States the power to discriminate in the field of alcohol regulation, that power would be at its apex when it comes to regulating the activity to which the provision expressly refers. The Association and the dissent point out that Granholm repeatedly spoke of discrimination against out-of-state products and producers, but there is an obvious explanation: The state laws at issue in Granholm discriminated against out-of-state producers. See 883 F. 3d, at 621. And Granholm never said that its reading of history or its Commerce Clause analysis was limited to discrimination against products or producers. On the contrary, the Court stated that the Clause prohibits state discrimination against all “ ‘out-of-state economic interests,’ ” Granholm, 544 U. S., at 472 (emphasis added), and noted that the direct-shipment laws in question “contradict[ed]” dormant Commerce Clause principles because they “deprive[d] citizens of their right to have access to the markets of other States on equal terms.” Id., at 473 (emphasis added). Granholm also described its analysis as consistent with the rule set forth in Bacchus, Brown-Forman Distillers Corp. v. New York State Liquor Authority, 476 U.S. 573 (1986), and Healy that “ ‘[w]hen a state statute directly regulates or discriminates against interstate commerce, or when its effect is to favor in-state economic interests over out-of-state interests, we have generally struck down the statute without further inquiry.’ ” Granholm, supra, at 487 (quoting Brown-Forman, supra, at 579; emphasis added). The Association counters that even if the Granholm Court did not explicitly limit its holding to products and producers, the Court implicitly did so when it rejected the argument that its analysis would call into question the constitutionality of state laws setting up three-tiered alcohol distribution systems. See Granholm, supra, at 488–489. This argument, which the dissent also advances, see post, at 12–13, reads far too much into Granholm’s discussion of the three-tiered model. Although Granholm spoke approvingly of that basic model, it did not suggest that §2 sanctions every discriminatory feature that a State may incorporate into its three-tiered scheme. At issue in the present case is not the basic three-tiered model of separating producers, wholesalers, and retailers, but the durational-residency requirement that Tennessee has chosen to impose on new applicants for liquor store licenses. Such a requirement is not an essential feature of a three-tiered scheme. Many such schemes do not impose durational-residency requirements—or indeed any residency requirements—on individual or corporate liquor store owners. See, e.g., Brief for State of Illinois et al. as Amici Curiae 24–25, 27 (identifying States that have either “dispos[ed] with the durational aspect of the [residency] requirement” or “d[o] not regulate the residency of the applicant corporation or partnership”). Other three-tiered schemes differ in other ways. See, e.g., id., at 24–28 (noting variations); FTC, Possible Anticompetitive Barriers to E-Commerce: Wine 7–9 (July 2003), https:// www .ftc . gov / sites /default/files/documents/reports/possible-anticompetitive-barriers-e-commerce-wine/winereport2_0. pdf (as last visited June 24, 2019) (same). Because we agree with the dissent that, under §2, States “remai[n] free to pursue” their legitimate interests in regulating the health and safety risks posed by the alcohol trade, post, at 12, each variation must be judged based on its own features. 2 In support of the argument that the Tennessee scheme is constitutional, the Association and its amici claim that discriminatory distribution laws, including in-state presence and residency requirements, long predate Prohibition and were adopted by many States following ratification of the Twenty-first Amendment.[19] Indeed, the Association notes that the 2-year durational-residency requirement now before us dates back to 1939 and is consistent with durational-residency regimes adopted by several other States around the same time.[20] According to the Association, that history confirms that §2 was intended to broadly exempt all in-state distribution laws from dormant Commerce Clause scrutiny. The dissent relies heavily on this same argument. This argument fails for several reasons. Insofar as it relies on state laws enacted shortly after the ratification of the Twenty-first Amendment and this Court’s early decisions interpreting it, the Association and the dissent’s argument does not take into account the overly expansive interpretation of §2 that took hold for a time in the immediate aftermath of its adoption. See supra, at 22–23. Thus, some state laws adopted soon after the ratification of the Twenty-first Amendment may have been based on an understanding of §2 that can no longer be defended. It is telling that an argument similar to the one now made by the Association would have dictated a contrary result in Granholm, since state laws disfavoring imported products were passed during this same period. See, e.g., Young’s Market Co., 299 U. S., at 62 (discriminatory license fee on imported beer); Mahoney, 304 U. S., at 403 (prohibition on import of certain liquors); Indianapolis Brewing Co., 305 U. S., at 394 (same). But our later cases have rejected this interpretation of §2. See Granholm, supra, at 487. Insofar as the Association’s argument is based on state laws adopted prior to Prohibition, it infers too much from the existence of laws that were never tested in this Court. Had they been tested here, there is no reason to conclude that they would have been sustained. During that time, the Court repeatedly invalidated, on dormant Commerce Clause grounds, a variety of state and local efforts to license those engaged in interstate business,[21] and as noted, pre-Prohibition decisions of this Court and the lower courts held that state alcohol laws that discriminated against interstate commerce were unconstitutional, see supra, at 15. Contrary to the Association’s contention, not all of these decisions involved discrimination against alcohol produced out of State or alcohol importers. The tax in Walling, for example, applied to those engaged in the business of selling imported alcohol within the State. 116 U.S. 446. And in concluding that the law violated the Commerce Clause, the Court affirmed that, without the dormant Commerce Clause, there would “be no security against conflicting regulations of different states, each discriminating in favor of its own products and citizens, and against the products and citizens of other states.” Id., at 456–457 (emphasis added). So too, the dispensary law in Scott was challenged on the ground that it discriminated “against products of other States and against citizens of other States.” 165 U. S., at 62 (emphasis added); see also id., at 94. Nor have States historically enjoyed absolute authority to police alcohol within their borders. As discussed earlier, far from granting the States plenary authority to adopt domestic regulations, the Court’s police-power precedents required an examination of the actual purpose and effect of a challenged law. See, e.g., Mugler, 123 U. S., at 661 (“It does not at all follow that every statute enacted ostensibly for the promotion” of “the public health, the public morals, or the public safety” is “to be accepted as a legitimate exertion of the police powers of the State”); see also Husen, 95 U. S., at 472; Welton v. Missouri, 91 U.S. 275, 278 (1876). Cf. H. Black, Intoxicating Liquors §30, p. 40 (1892) (stating that certain 19th-century licensing and residency requirements were valid because their “purpose and effect” was to prevent “the unlawful selling of liquors, and not to discriminate against citizens of other states” (emphasis added)). For these reasons, we reject the Association’s overly broad understanding of §2. That provision allows each State leeway to enact the measures that its citizens believe are appropriate to address the public health and safety effects of alcohol use and to serve other legitimate interests, but it does not license the States to adopt protectionist measures with no demonstrable connection to those interests. V Having concluded that §2 does not confer limitless authority to regulate the alcohol trade, we now apply the §2 analysis dictated by the provision’s history and our precedents. If we viewed Tennessee’s durational-residency requirements as a package, it would be hard to avoid the conclusion that their overall purpose and effect is protectionist. Indeed, two of those requirements—the 10-year residency requirement for license renewal and the provision that shuts out all publicly traded corporations—are so plainly based on unalloyed protectionism that neither the Association nor the State is willing to come to their defense. The provision that the Association and the State seek to preserve—the 2-year residency requirement for initial license applicants—forms part of that scheme. But we assume that it can be severed from its companion provisions, see 883 F. 3d, at 626–628, and we therefore analyze that provision on its own. Since the 2-year residency requirement discriminates on its face against nonresidents, it could not be sustained if it applied across the board to all those seeking to operate any retail business in the State. Cf. C & A Carbone, Inc. v. Clarkstown, 511 U.S. 383, 391–392 (1994); Lewis v. BT Investment Managers, Inc., 447 U.S. 27, 39 (1980). But because of §2, we engage in a different inquiry. Recognizing that §2 was adopted to give each State the authority to address alcohol-related public health and safety issues in accordance with the preferences of its citizens, we ask whether the challenged requirement can be justified as a public health or safety measure or on some other legitimate nonprotectionist ground. Section 2 gives the States regulatory authority that they would not otherwise enjoy, but as we pointed out in Granholm, “mere speculation” or “unsupported assertions” are insufficient to sustain a law that would otherwise violate the Commerce Clause. 544 U. S., at 490, 492. Where the predominant effect of a law is protectionism, not the protection of public health or safety, it is not shielded by §2. The provision at issue here expressly discriminates against nonresidents and has at best a highly attenuated relationship to public health or safety. During the course of this litigation, the Association relied almost entirely on the argument that Tennessee’s residency requirements are simply “not subject to Commerce Clause challenge,” 259 F. Supp. 3d, at 796, and the State itself mounted no independent defense. As a result, the record is devoid of any “concrete evidence” showing that the 2-year residency requirement actually promotes public health or safety; nor is there evidence that nondiscriminatory alternatives would be insufficient to further those interests. Granholm, supra, at 490; see 883 F. 3d, at 625–626. In this Court, the Association has attempted to defend the 2-year residency requirement on public health and safety grounds, but this argument is implausible on its face. The Association claims that the requirement ensures that retailers are “amenable to the direct process of state courts,” Brief for Petitioner 48 (internal quotation marks omitted), but the Association does not explain why this objective could not easily be achieved by ready alternatives, such as requiring a nonresident to designate an agent to receive process or to consent to suit in the Tennessee courts. See Cooper v. McBeath, 11 F.3d 547, 554 (CA5 1994). Similarly unpersuasive is the Association’s claim that the 2-year requirement gives the State a better opportu- nity to determine an applicant’s fitness to sell alcohol and guards against “undesirable nonresidents” moving into the State for the purpose of operating a liquor store. Brief for Petitioner 10 (internal quotation marks omitted). The State can thoroughly investigate applicants without requiring them to reside in the State for two years before obtaining a license. Tennessee law already calls for criminal background checks on all applicants, see Tenn. Code Ann. §57–3–208, and more searching checks could be demanded if necessary. As the Fifth Circuit observed in a similar case, “[i]f [the State] desires to scrutinize its applicants thoroughly, as is its right, it can devise nondiscriminatory means short of saddling applicants with the ‘burden’ of residing” in the State. Cooper, 11 F. 3d, at 554. The 2-year residency requirement, in any event, poorly serves the goal of enabling the State to ensure that only law-abiding and responsible applicants receive licenses. As the Tennessee attorney general explained, if a nonresident moves to the State with the intention of applying for a license once the 2-year period ends, the TABC will not necessarily have any inkling of the future applicant’s intentions until that individual applies for a license, and consequently, the TABC will have no reason to begin an investigation until the 2-year period has ended. App. to Brief in Opposition 17a. And all that the 2-year requirement demands is residency. A prospective applicant is not obligated during that time “to be educated about liquor sales, submit to inspections, or report to the State.” Ibid. The 2-year residency requirement is not needed to en- able the State to maintain oversight over liquor store operators. In Granholm, it was argued that the prohibition on the shipment of wine from out-of-state sources was justified because the State could not adequately monitor the activities of nonresident entities. Citing “improvements in technology,” we found that argument insufficient. 544 U. S., at 492. See also Cooper, supra, at 554 (“In this age of split-second communications by means of computer networks . . . there is no shortage of less burdensome, yet still suitable, options”). In this case, the argument is even less persuasive since the stores at issue are physically located within the State. For that reason, the State can monitor the stores’ operations through on-site inspections, audits, and the like. See §57–3–104. Should the State conclude that a retailer has “fail[ed] to comply with state law,” it may revoke its operating license. Granholm, 544 U. S., at 490. This “provides strong incentives not to sell alcohol” in a way that threatens public health or safety. Ibid. In addition to citing the State’s interest in regulatory control, the Association argues that the 2-year residency requirement would promote responsible alcohol consumption. According to the Association, the requirement makes it more likely that retailers will be familiar with the communities served by their stores, and this, it is suggested, will lead to responsible sales practices. Brief for Petitioner 48–49. The idea, it seems, is that a responsible neighborhood proprietor will counsel or cut off sales to patrons who are known to be abusing alcohol, who manifest the effects of alcohol abuse, or who perhaps appear to be purchas- ing too much alcohol. No evidence has been offered that durational-residency requirements actually foster such sales practices, and in any event, the requirement now before us is very poorly designed to do so. For one thing, it applies to those who hold a license, not to those who actually make sales. For another, it requires residence in the State, not in the community that a store serves. The Association cannot explain why a proprietor who lives in Bristol, Virginia, will be less knowledgeable about the needs of his neighbors right across the border in Bristol, Tennessee, than someone who lives 500 miles away in Memphis. And the rationale is further undermined by other features of Tennessee law, particularly the lack of durational-residency requirements for owners of bars and other establishments that sell alcohol for on-premises consumption. §57–4–201. Not only is the 2-year residency requirement ill suited to promote responsible sales and consumption practices (an interest that we recognize as legitimate, contrary to the dissent’s suggestion, post, at 9, 12, 14), but there are obvious alternatives that better serve that goal without discriminating against nonresidents. State law empowers the relevant authorities to limit both the number of retail licenses and the amount of alcohol that may be sold to an individual. Cf. §57–3–208(c) (permitting local governments to “limit . . . the number of licenses issued within their jurisdictions”); §57–3–204(d)(7)(C) (imposing volume limits on certain sales of alcohol to patrons); Rules of TABC, ch. 0100–01, §0100–01–.03(15) (2018) (same). The State could also mandate more extensive training for managers and employees and could even demand that they demonstrate an adequate connection with and knowledge of the local community. Cf., e.g., Tenn. Code Ann. §57–3–221 (requiring managers of liquor stores to obtain permits, satisfy background checks, and undergo “alcohol awareness” training). And the State of course remains free to monitor the practices of retailers and to take action against those who violate the law. Given all this, the Association has fallen far short of showing that the 2-year durational-residency requirement for license applicants is valid. Like the other discriminatory residency requirements that the Association is unwilling to defend, the predominant effect of the 2-year residency requirement is simply to protect the Association’s members from out-of-state competition. We therefore hold that this provision violates the Commerce Clause and is not saved by the Twenty-first Amendment.[22] * * * The judgment of the Court of Appeals for the Sixth Circuit is affirmed. It is so ordered. Notes 1 For purposes of the provisions at issue here, Tennessee law defines “alcoholic beverage[s]” to include “spirits, liquor, wine, high alcohol content beer,” and “any liquid product containing distilled alcohol capable of being consumed by a human being, manufactured or made with distilled alcohol, regardless of alcohol content,” Tenn. Code Ann. §57–3–101(a)(1)(A) (2018). This definition excludes “beer,” which is defined and regulated by separate statutory provisions, see §57–5–101(b). 2 See Letter from H. Slatery III, Tenn. Atty. Gen., to S. Harris, Clerk of Court (Nov. 13, 2018). 3 See, e.g., R. Beeman, Plain, Honest Men: The Making of the American Constitution 18–20 (2009); D. Stewart, The Summer of 1787: The Men Who Invented the Constitution 9–10 (2007); M. Farrand, The Framing of the Constitution of the United States 7–10 (1913). 4 Before Woodruff, there was authority suggesting that the Import-Export Clause applied to trade between States. See Brown v. Maryland, 12 Wheat. 419, 449 (1827) (Marshall, C. J.); Almy v. California, 24 How. 169 (1861). And more recently Woodruff has been questioned. See Camps Newfound/Owatonna, Inc. v. Town of Harrison, 520 U.S. 564, 624–636 (1997) (Thomas, J., dissenting). But one way or the other, it would grossly distort the Constitution to hold that it provides no protection against a broad swath of state protectionist measures. Even at the time of the adoption of the Constitution, it would have been asking a lot to require that Congress pass a law striking down every protectionist measure that a State or unit of local government chose to enact. Cf. Friedman & Deacon, A Course Unbroken: The Constitutional Legitimacy of the Dormant Commerce Clause, 97 Va. L. Rev. 1877, 1898–1903 (2011); 3 The Records of the Federal Convention of 1787, p. 549 (M. Farrand ed. 1911) (the Virginia Plan’s proposal of a congressional negative was “justly abandoned, as, apart from other objections, it was not practicable among so many States, increasing in number, and enacting, each of them, so many laws”). 5 As we will explain, §2 followed the wording of the 1913 Webb-Kenyon Act, ch. 90, 37Stat. 699, see Craig v. Boren, 429 U.S. 190, 205–206 (1976), and, given this Court’s case law at the time, it went without saying that the only state laws that Congress could protect from constitutional challenge were those that represented the valid exercise of the police power, which was not understood to authorize purely protectionist measures with no bona fide relation to public health or safety. See infra, at 14–15, 18–19. 6 Between 1780 and 1830, Americans consumed “more alcohol, on an individual basis, than at any other time in the history of the nation,” with per capita consumption double that of the modern era. R. Mendelson, From Demon to Darling: A Legal History of Wine in America 11 (2009). 7 By 1872, about 100,000 had sprung up across the country, and by the end of the century, that number had climbed to almost 300,000. Id., at 31. This increase has been linked to the introduction of the English “tied-house” system. Under the tied-house system, an alcohol producer, usually a brewer, would set up saloonkeepers, providing them with premises and equipment, and the saloonkeepers, in exchange, agreed to sell only that producer’s products and to meet set sales requirements. Ibid.; T. Pegram, Battling Demon Rum: The Struggle for a Dry America, 1800–1933, p. 95 (1998). To meet those requirements, saloonkeepers often encouraged irresponsible drinking. Id., at 97. The three-tiered distribution model was adopted by States at least in large part to preclude this system. See Arnold’s Wines, Inc. v. Boyle, 571 F.3d 185, 187 (CA2 2009). 8 The provision read as follows: “That all fermented, distilled, or other intoxicating liquors or liquids transported into any State or Territory or remaining therein for use, consumption, sale or storage therein, shall upon arrival in such State or Territory be subject to the operation and effect of the laws of such State or Territory enacted in the exercise of its police powers, to the same extent and in the same manner as though such liquids or liquors had been produced in such State or Territory, and shall not be exempt therefrom by reason of being introduced therein in original packages or otherwise.” Ch. 728, 26Stat. 313, 27 U. S. C. §121. 9 That was the position expressed in an opinion issued by Attorney General Wickersham, 30 Op. Atty. Gen. 88 (1913), and President Taft’s veto, which Congress overrode, was based on exactly this ground. 49 Cong. Rec. 4291 (1913) (Veto Message of the President). 10 The Act provided: “That the shipment or transportation . . . of any spirituous, vinous, malted, fermented, or other intoxicating liquor of any kind, from one State . . . into any other State . . . which said spirituous, vinous, malted, fermented, or other intoxicating liquor is intended, by any person interested therein, to be received, possessed, sold, or in any manner used, either in the original package or otherwise, in violation of any law of such State . . . is hereby prohibited.” 37Stat. 699–700. 11 This principle was also invoked in dormant Commerce Clause cases involving other products. See, e.g., Minnesota v. Barber, 136 U.S. 313, 319, 323 (1890); Railroad Co. v. Husen, 95 U.S. 465, 472 (1878). 12 Lower court decisions issued between the enactment of the Webb-Kenyon Act and the ratification of the Eighteenth Amendment interpreted the Act this way. See Evansville Brewing Assn. v. Excise Comm’n of Jefferson Cty., Ala., 225 F. 204 (ND Ala. 1915); Southern Express Co. v. Whittle, 194 Ala. 406, 69 So. 652 (1915); Brennen v. Southern Express Co., 106 S. C. 102, 90 S.E. 402 (1916); Charleston & W. C. R. Co. v. Gosnell, 106 S. C. 84, 90 S.E. 264 (1916) (Hydrick, J., concurring); Monumental Brewing Co. v. Whitlock, 111 S. C. 198, 97 S.E. 56 (1918). See also Pacific Fruit & Produce Co. v. Martin, 16 F. Supp. 34, 39–40 (WD Wash. 1936); Friedman, Constitutional Law: State Regulation of Importation of Intoxicating Liquor Under Twenty-first Amendment, 21 Cornell L. Q. 504, 509 (1936). 13 See, e.g., 76 Cong. Rec. 4172 (1933) (statement of Sen. Borah) (§2 of Twenty-first Amendment would “incorporat[e] [Webb-Kenyon] permanently in the Constitution of the United States”); id., at 4168 (statement of Sen. Fess) (“[T]he second section of the joint resolution . . . is designed to permit the Federal authority to assist the States that want to be dry to remain dry”); id., at 4518 (statement of Rep. Robinson) (“Section 2 attempts to protect dry states”). 14 See Nielson, No More “Cherry-Picking”: The Real History of the 21st Amendment’s §2, 28 Harv. J. L. & Pub. Pol’y 281, 286, n. 21 (2004). See generally E. Brown, Ratification of the Twenty-first Amendment to the Constitution of the United States; State Convention Records and Laws (1938). 15 The dissent characterizes the Court as a “committee of nine” that has “stray[ed] from the text” of the Twenty-first Amendment and “impose[d] [its] own free-trade rules” on the States. Post, at 8, 15 (opinion of J. Gorsuch). This is empty rhetoric. The dissent itself strays from a blinkered reading of the Amendment. The dissent interprets §2 of the Amendment to mean more than it literally says, arguing that §2 covers the residency requirements at issue even though they are not tied in any way to what the Amendment actually addresses, namely, “the transportation or importation” of alcohol across state lines. See post, at 2, n. 1. And the dissent agrees that §2 cannot be read as broadly as one might think if its language were read in isolation and not as part of an integrated constitutional scheme. See post, at 2. The dissent asserts that §2 does not abrogate all previously adopted constitutional provisions, just the dormant Commerce Clause. But the dissent does not say whether it thinks §2 allows the States to adopt alcohol regulations that serve no conceivable purpose other than protectionism. Even the dissent below did not go that far. See n. 18, infra. If §2 gives the States carte blanche to engage in protectionism, we suppose that Tennessee could restrict licenses to persons who can show that their lineal ancestors have lived in the State since 1796 when the State entered the Union. Does the dissent really think that this is what §2 was meant to permit? 16 See, e.g., Granholm v. Heald, 544 U.S. 460, 497–498 (2005) (Thomas, J., dissenting); Healy v. Beer Institute, 491 U.S. 324, 349 (1989) (Rehnquist, C. J., dissenting); 324 Liquor Corp. v. Duffy, 479 U.S. 335, 352–353 (1987) (O’Connor, J., dissenting); Bacchus Imports, Ltd. v. Dias, 468 U.S. 263, 281–282 (1984) (Stevens, J., dissenting). The dissent rehashes this debate, see post, at 5–8, 14, asserting that the Webb-Kenyon Act, and thus §2, were “understood” to repudiate not only the original-package cases, but also the antidiscrimination rule articulated in cases including Scott v. Donald, 165 U.S. 58 (1897). But this Court’s modern §2 decisions—not simply the lower court decisions at which the dissent takes aim, see post, at 6, n. 3—establish that those enactments, though no doubt aimed at granting States additional “discretion to calibrate alcohol regulations to local preferences,” post, at 2, did not exempt States from “the nondiscrimination principle of the Commerce Clause.” Granholm, supra, at 487. 17 Justice Scalia, for his part, thought the “statute’s invalidity [was] fully established by its facial discrimination against interstate commerce”—discrimination that in his view “eliminate[d] the immunity afforded by the Twenty-first Amendment.” Healy, supra, at 344 (opinion concurring in part and concurring in judgment) (citing Bacchus, supra, at 275–276). 18 The Association’s argument is more extreme than that of the dissent below, which recognized that in-state distribution laws that “serve no purpose besides ‘economic protectionism’ ” remain subject to dormant Commerce Clause scrutiny. Byrd v. Tennessee Wine and Spirits Retailers Assn., 883 F.3d 608, 633 (CA6 2018) (Sutton, J., concurring in part and dissenting in part) (quoting Bacchus, supra,at 276). 19 See Granholm, 544 U. S., at 518, and n. 6 (Thomas, J., dissenting) (licensing schemes adopted by three-tier States following ratification of Twenty-first Amendment discriminated “by requiring in-state residency or physical presence as a condition of obtaining licenses”) (collecting statutes); Brief for Petitioner 33–34 (collecting residency-requirement statutes). See also Brief for State of Illinois et al. as Amici Curiae 7–8 (referencing 19th-century state statutes that required “retailers to reside in-state or to maintain an in-state presence”). 20 See 1939 Tenn. Pub. Acts, ch. 49, §§5–8; Brief for Petitioner 34 (collecting durational-residency requirement statutes); Brief for State of Illinois et al. as Amici Curiae 24 (same). 21 Real Silk Hosiery Mills v. Portland, 268 U.S. 325, 335–336 (1925) (license tax on solicitors of orders to be filled by an out-of-state manufacturer); Shafer v. Farmers Grain Co. of Embden, 268 U.S. 189, 197–201 (1925) (license requirement for the purchase of grain shipped immediately out of the State); Stewart v. Michigan, 232 U.S. 665, 669–670 (1914) (state law requiring a license for catalog sales); Crenshaw v. Arkansas, 227 U.S. 389, 399–401 (1913) (state law requiring a foreign corporation actively soliciting sales in State to obtain a license); Dozier v. Alabama, 218 U.S. 124, 127–128 (1910) (licensing requirement on the solicitors of photography enlargement services and frames manufactured out of State); International Textbook Co. v. Pigg, 217 U.S. 91, 107–111 (1910) (state law requiring an out-of-state educational publishing company to pay a license fee for exchanging materials with customers); Rearick v. Pennsylvania, 203 U.S. 507, 510–511 (1906) (ordinance requiring license to solicit orders for out-of-state goods); Norfolk & Western R. Co. v. Sims, 191 U.S. 441, 449–451 (1903) (state licensing requirement on express company acting as agent for importer of a sewing machine); Brennan v. Titusville, 153 U.S. 289, 306–308 (1894) (licensing tax on persons engaged in trade on behalf of firms doing business outside the State); Corson v. Maryland, 120 U.S. 502, 505–506 (1887) (state licensing requirement as applied to agent of out-of-state firm soliciting sales); Welton v. Missouri, 91 U.S. 275, 278, 282–283 (1876) (state law requiring payment of license tax by sellers of out-of-state goods). 22 Our analysis and conclusion apply as well to the provision requiring all officers and directors of corporate applicants to satisfy the 2-year residency requirement. See 883 F. 3d, at 623.

587.US.2018_17-1201

The Tennessee Valley Authority (TVA), a Government-owned corporation, provides electric power to millions of Americans. In creating the TVA, Congress decided that the corporation could “sue and be sued in its corporate name,” 16 U. S. C. §831c(b), thus waiving at least some of the sovereign immunity from suit that it would have enjoyed as a Federal Government entity. Congress subsequently waived immunity from tort suits involving agencies across the Government in the Federal Tort Claims Act (FTCA), but it carved out an exception for claims based on a federal employee’s performance of a “discretionary function.” 28 U. S. C. §2680(a). Congress specifically excluded from the FTCA’s provisions—including the discretionary function exception—“[a]ny claim arising from the activities of the [TVA].” §2680(l). In this case, TVA employees were raising a downed power line that was partially submerged in the Tennessee River when petitioner Gary Thacker drove his boat into the area at high speed. Thacker’s boat collided with the power line, seriously injuring him and killing his passenger. He sued for negligence. The TVA moved to dismiss, claiming sovereign immunity, and the District Court granted the motion. Affirming, the Eleventh Circuit used the same test it applies when evaluating whether the Government is immune from suit under the discretionary function exception to the FTCA, and it held that Thacker’s suit was foreclosed because the challenged actions were “a matter of choice.” Held: 1. The waiver of immunity in the TVA’s sue-and-be-sued clause is not subject to a discretionary function exception of the kind in the FTCA. By the terms of the Tennessee Valley Authority Act of 1933, the TVA’s sue-and-be-sued clause contains no exception for suits based on discretionary functions. Nor does the FTCA’s discretionary function exception apply to the TVA. See 28 U. S. C. §2680(l). But this Court recognized in Federal Housing Administration v. Burr, 309 U.S. 242, that a sue-and-be-sued clause might be subject to an “implied restriction,” id., at 245. In particular, a court should recognize such a restriction if the type of suit at issue is “not consistent with the statutory or constitutional scheme” or the restriction is “necessary to avoid grave interference with the performance of a governmental function.” Ibid. The Government tries to use the framework of Burr to argue that this Court should imply an FTCA-like limit on the TVA’s sue-and-be-sued clause for all suits challenging discretionary functions because those suits would conflict with separation-of-powers principles and interfere with important governmental functions. At the outset, Congress made a considered decision not to apply the FTCA to the TVA, and the Government is effectively asking this Court to negate that legislative choice. In any event, the Government errs in arguing that waiving the TVA’s immunity from suits based on discretionary functions would offend the separation of powers. And the Government overreaches when it says that all suits based on the TVA’s discretionary conduct would interfere with governmental functions. The discretionary acts of hybrid entities like the TVA may be commercial in nature, and a suit challenging a commercial act will not interfere with governmental functions. Ibid. Pp. 4–10. 2. The courts below, which wrongly relied on the discretionary function exception, should have the first chance to address the issues this Court finds relevant in deciding whether this suit may go forward. To determine if the TVA has immunity, the court on remand must first decide whether the conduct alleged to be negligent is governmental or commercial in nature. If it is commercial, the TVA cannot invoke sovereign immunity. If it is governmental, the court might decide that an implied limitation on the clause bars the suit, but only if it finds that prohibiting the “type[ ] of suit [at issue] is necessary to avoid grave interference” with that function’s performance. Burr, 309 U. S., at 245. Pp. 10–11. 868 F.3d 979, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court.

Federal law provides that the Tennessee Valley Authority (TVA), a Government-owned corporation supplying electric power to millions of Americans, “[m]ay sue and be sued in its corporate name.” Tennessee Valley Authority Act of 1933 (TVA Act), 48Stat. 60, 16 U. S. C. §831c(b). That provision serves to waive sovereign immunity from suit. Today, we consider how far the waiver goes. We reject the view, adopted below and pressed by the Government, that the TVA remains immune from all tort suits arising from its performance of so-called discretionary functions. The TVA’s sue-and-be-sued clause is broad and contains no such limit. Under the clause—and consistent with our precedents construing similar ones—the TVA is subject to suits challenging any of its commercial activities. The law thus places the TVA in the same position as a private corporation supplying electricity. But the TVA might have immunity from suits contesting one of its governmental activities, of a kind not typically carried out by private parties. We remand this case for consideration of whether that limited immunity could apply here. I Congress created the TVA—a “wholly owned public corporation of the United States”—in the throes of the Great Depression to promote the Tennessee Valley’s economic development. TVA v. Hill, 437 U.S. 153, 157 (1978). In its early decades, the TVA focused on reforesting the countryside, improving farmers’ fertilization practices, and building dams on the Tennessee River. See Brief for Respondent 3. The corporation also soon began constructing new power plants for the region. And over the years, as it completed other projects, the TVA devoted more and more of its efforts to producing and selling electric power. Today, the TVA operates around 60 power plants and provides electricity to more than nine million people in seven States. See id., at 3–4. The rates it charges (along with the bonds it issues) bring in over $10 billion in annual revenues, making federal appropriations unnecessary. See ibid.; GAO, FY 2018 Financial Report of the United States Government 53 (GAO–19–294R, 2019). As even that short description may suggest, the TVA is something of a hybrid, combining traditionally governmental functions with typically commercial ones. On the one hand, the TVA possesses powers and responsibilities reserved to sovereign actors. It may, for example, “exercise the right of eminent domain” and “condemn all property” necessary to carry out its goals. 16 U. S. C. §§831c(h), (i). Similarly, it may appoint employees as “law enforcement agents” with powers to investigate crimes and make arrests. §831c–3(a); see §831c–3(b)(2). But on the other hand, much of what the TVA does could be done—no, is done routinely—by non-governmental parties. Just as the TVA produces and sells electricity in its region, privately owned power companies (e.g., Con Edison, Dominion Energy) do so in theirs. As to those commonplace commercial functions, the emphasis in the oft-used label “public corporation” rests heavily on the latter word. Hill, 437 U. S., at 157. In establishing this mixed entity, Congress decided (as it had for similar government businesses) that the TVA could “sue and be sued in its corporate name.” §831c(b); see, e.g., Reconstruction Finance Corporation Act, §4, 47Stat. 6; Federal Home Loan Bank Act, §12, 47Stat. 735. Without such a clause, the TVA (as an entity of the Fed- eral Government) would have enjoyed sovereign immunity from suit. See Loeffler v. Frank, 486 U.S. 549, 554 (1988). By instead providing that the TVA could “be sued,” Congress waived at least some of the corporation’s immunity. (Just how much is the question here.) Slightly more than a decade after creating the TVA, Congress enacted the Federal Tort Claims Act of 1946 (FTCA), 28 U. S. C. §§1346(b), 2671 et seq., to waive immunity from tort suits involving agencies across the Government. See §1346(b)(1) (waiving immunity from damages claims based on “the negligent or wrongful act or omission of any employee of the Government”). That statute carved out an exception for claims based on a federal employee’s performance of a “discretionary function.” §2680(a). But Congress specifically excluded from all the FTCA’s provisions—including the discretionary function exception—“[a]ny claim arising from the activities of the [TVA].” §2680(l). This case involves such a claim. See App. 22–33 (Complaint). One summer day, TVA employees embarked on work to replace a power line over the Tennessee River. When a cable they were using failed, the power line fell into the water. The TVA informed the Coast Guard, which announced that it was closing part of the river; and the TVA itself positioned two patrol boats near the downed line. But several hours later, just as the TVA workers began to raise the line, petitioner Gary Thacker drove his boat into the area at high speed. The boat and line col- lided, seriously injuring Thacker and killing a passenger. Thacker sued for negligence, alleging that the TVA had failed to “exercise reasonable care” in “assembl[ing] and install[ing] power lines” and in “warning boaters” like him “of the hazards it created.” Id., at 31. The TVA moved to dismiss the suit, claiming sovereign immunity. The District Court granted the motion. It reasoned that the TVA, no less than other government agencies, is entitled to immunity from any suit based on an employee’s exercise of discretionary functions. See 188 F. Supp. 3d 1243, 1245 (ND Ala. 2016). And it thought that the TVA’s actions surrounding the boating accident were discretionary because “they involve[d] some judgment and choice.” Ibid. The Court of Appeals for the Eleventh Circuit affirmed on the same ground. According to the circuit court, the TVA has immunity for discretionary functions even when they are part of the “TVA’s commercial, power-generating activities.” 868 F.3d 979, 981 (2017). In deciding whether a suit implicates those functions, the court explained that it “use[s] the same test that applies when the government invokes the discretionary-function exception to the [FTCA].” Id., at 982. And that test, the court agreed, foreclosed Thacker’s suit because the challenged actions were “a matter of choice.” Ibid. (internal quotation marks omitted). We granted certiorari to decide whether the waiver of sovereign immunity in TVA’s sue-and-be-sued clause is subject to a discretionary function exception, of the kind in the FTCA. 585 U. S. ___ (2018). We hold it is not. II Nothing in the statute establishing the TVA (again, the TVA Act for short) expressly recognizes immunity for discretionary functions. As noted above, that law provides simply that the TVA “[m]ay sue and be sued.” 16 U. S. C. §831c(b); see supra, at 3. Such a sue-and-be-sued clause serves to waive sovereign immunity otherwise belonging to an agency of the Federal Government. See Loeffler, 486 U. S., at 554. By the TVA Act’s terms, that waiver is subject to “[e]xcept[ions] as “specifically provided in” the statute itself. §831c. But the TVA Act contains no exceptions relevant to tort claims, let alone one turning on whether the challenged conduct is discretionary. Nor does the FTCA’s exception for discretionary functions apply to the TVA. As described earlier, see supra, at 3, the FTCA retained the Federal Government’s immunity from tort suits challenging discretionary conduct, even while allowing other tort claims to go forward. See 28 U. S. C. §§1346(b), 2680(a); United States v. Gaubert, 499 U.S. 315, 322–325 (1991) (describing the discretionary function exception’s scope). But Congress made clear that the FTCA does “not apply to[] [a]ny claim arising from the activities of the [TVA].” §2680(l). That means the FTCA’s discretionary function provision has no relevance to this case. Even the Government concedes as much. It acknowledges that the FTCA’s discretionary function exception “does not govern [Thacker’s] suit.” Brief for Respondent 15. Rather, it says, the TVA Act’s sue-and-be-sued clause does so. See id., at 6. And that is the very clause we have just described as containing no express exception for discretionary functions. But that is not quite the end of the story because in Federal Housing Administration v. Burr, 309 U.S. 242 (1940), this Court recognized that a sue-and-be-sued clause might contain “implied exceptions.” Id., at 245. The Court in that case permitted a suit to proceed against a government entity (providing mortgage insurance) whose organic statute had a sue-and-be-sued clause much like the TVA Act’s. And the Court made clear that in green-lighting the suit, it was doing what courts normally should. Sue-and-be-sued clauses, the Court explained, “should be liberally construed.” Ibid.; see FDIC v. Meyer, 510 U.S. 471, 475 (1994) (similarly calling such clauses “broad”). Those words “in their usual and ordinary sense,” the Court noted, “embrace all civil process incident to the commencement or continuance of legal proceedings.” Burr, 309 U. S., at 245–246. And Congress generally “intend[s] the full consequences of what it sa[ys]”—even if “inconvenient, costly, and inefficient.” Id., at 249 (quotation modified). But not quite always, the Court continued. And when not—when Congress meant to use the words “sue and be sued” in a more “narrow sense”—a court should recognize “an implied restriction.” Id., at 245. In particular, Burr stated, a court should take that route if one of the following circumstances is “clearly shown”: either the “type[] of suit [at issue is] not consistent with the statutory or constitutional scheme” or the restriction is “necessary to avoid grave interference with the performance of a governmental function.” Ibid. Although the courts below never considered Burr, the Government tries to use its framework to defend their decisions. See Brief for Respondent 17–40. According to the Government, we should establish a limit on the TVA’s sue-and-be-sued clause—like the one in the FTCA—for all suits challenging discretionary functions. That is for two reasons, tracking Burr’s statement of when to recognize an “implied exception” to a sue-and-be-sued clause. 309 U. S., at 245. First, the Government argues that allowing those suits would conflict with the “constitutional scheme”—more precisely, with “separation-of-powers principles”—by subjecting the TVA’s discretionary conduct to “judicial second-guessing.” Brief for Respondent 19, 21 (internal quotation marks omitted). Second, the Government maintains that permitting those suits would necessarily “interfere[ ] with important governmental functions.” Id., at 36; see id., at 39–40; Tr. of Oral Arg. 39–41. We disagree. At the outset, we balk at using Burr to provide a government entity excluded from the FTCA with a replica of that statute’s discretionary function exception. Congress made a considered decision not to apply the FTCA to the TVA (even as Congress applied that legislation to some other public corporations, see 28 U. S. C. §2679(a)). See supra, at 3, 5. The Government effectively asks us to negate that legislative choice. Or otherwise put, it asks us to let the FTCA in through the back door, when Congress has locked the front one. We have once before rejected such a maneuver. In FDIC v. Meyer, a plaintiff brought a constitutional tort claim against a government agency with another broad sue-and-be-sued clause. The agency claimed immunity, stressing that the claim would have fallen outside the FTCA’s immunity waiver (which extends only to conventional torts). We dismissed the argument. “In essence,” we observed, the “FDIC asks us to engraft” a part of the FTCA “onto [the agency’s] sue-and-be-sued clause.” 510 U. S., at 480. But that would mean doing what Congress had not. See id., at 483. And so too here, if we were to bestow the FTCA’s discretionary function exception on the TVA through the conduit of Burr. Indeed, the Government’s proposal would make the TVA’s tort liability largely coextensive with that of all the agencies the FTCA governs. See Tr. of Oral Arg. 33–34. Far from acting to achieve such parity, Congress did everything possible to avoid it. In any event, the Government is wrong to think that waiving the TVA’s immunity from suits based on discretionary functions would offend the separation of powers. As this Court explained in Burr, the scope of immunity that federal corporations enjoy is up to Congress. That body “has full power to endow [such an entity] with the government’s immunity from suit.” 309 U. S., at 244. And equally, it has full power to “waive [that] immunity” and “subject[ the entity] to the judicial process” to whatever extent it wishes. Ibid. When Congress takes the latter route—even when it goes so far as to waive the corporation’s immunity for discretionary functions—its action raises no separation of powers problems. The right governmental actor (Congress) is making a decision within its bailiwick (to waive immunity) that authorizes an appropriate body (a court) to render a legal judgment. Indeed, the Government itself conceded at oral argument that Congress, when creating a public corporation, may constitutionally waive its “immunity [for] discretionary functions.” Tr. of Oral Arg. 37. But once that is acknowledged, the Government’s argument from “separation-of-powers principles” collapses. Brief for Respondent 19. Those principles can offer no reason to limit a statutory waiver that even without any emendation complies with the constitutional scheme. Finally, the Government overreaches when it says that all suits based on the TVA’s discretionary conduct will “grave[ly] interfere[]” with “governmental function[s].” Burr, 309 U. S., at 245. That is so, at the least, because the discretionary acts of hybrid entities like the TVA may be not governmental but commercial in nature. And a suit challenging a commercial act will not “grave[ly]”—or, indeed, at all—interfere with the “governmental functions” Burr cared about protecting. The Government contests that point, arguing that this Court has not meant to distinguish between the governmental and the commercial in construing sue-and-be-sued clauses. See Brief for Respondent 39–40. But both Burr and later decisions do so explicitly. Burr took as its “premise” that an agency “launched [with such a clause] into the commercial world” and “authorize[d] to engage” in “business transactions with the public” should have the same “amenab[ility] to judicial process [as] a private enterprise under like circumstances.” 309 U. S., at 245. Meyer also made clear that such an agency “could not escape the liability a private enterprise would face in similar circumstances.” 510 U. S., at 482; see ibid. (“[T]he liability of a private enterprise [is] a floor below which the agency’s liability [may] not fall”). And twice the Court held that the liability of the Postal Service (another sue-and-be-sued agency) should be “similar[ ] to [that of] other self-sustaining commercial ventures.” Franchise Tax Bd. of Cal. v. Postal Service, 467 U.S. 512, 525 (1984); see Loeffler, 486 U. S., at 556. The point of those decisions, contra the Government, is that (barring special constitutional or statutory issues not present here) suits based on a public corporation’s commercial activity may proceed as they would against a private company; only suits challenging the entity’s governmental activity may run into an implied limit on its sue-and-be-sued clause. Burr and its progeny thus require a far more refined analysis than the Government offers here. The reasons those decisions give to recognize a restriction on a sue-and-be-sued clause do not justify the wholesale incorporation of the discretionary function exception. As explained above, the “constitutional scheme” has nothing to say about lawsuits challenging a public corporation’s discretionary activity—except to leave their fate to Congress. Burr, 309 U. S., at 245; see supra, at 8. For its part, Congress has not said in enacting sue-and-be-sued clauses that it wants to prohibit all such suits—quite the contrary. And no concern for “governmental functions” can immunize discretionary activities that are commercial in kind. Burr, 309 U. S., at 245; see supra, at 8–9. When the TVA or similar body operates in the marketplace as private companies do, it is as liable as they are for choices and judgments. The possibility of immunity arises only when a suit challenges governmental activities—the kinds of functions private parties typically do not perform. And even then, an entity with a sue-and-be-sued clause may receive immunity only if it is “clearly shown” that prohibiting the “type[ ] of suit [at issue] is necessary to avoid grave interference” with a governmental function’s performance. Burr, 309 U. S., at 245. That is a high bar. But it is no higher than appropriate given Congress’s enactment of so broad an immunity waiver—which demands, as we have held, a “liberal construction.” Ibid. (quotation modified). III All that remains is to decide this case in accord with what we have said so far. But as we often note at this point, “we are a court of review, not of first view.” Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7 (2005). In wrongly relying on the discretionary function exception, the courts below never addressed the issues we have found relevant in deciding whether this suit may go forward. Those courts should have the first chance to do so, as guided by the principles set out above and a few last remarks about applying them here. As described earlier, the TVA sometimes resembles a government actor, sometimes a commercial one. See supra, at 2–3. Consider a few diverse examples. When the TVA exercises the power of eminent domain, taking landowners’ property for public purposes, no one would confuse it for a private company. So too when the TVA exercises its law enforcement powers to arrest individuals. But in other operations—and over the years, a growing number—the TVA acts like any other company producing and supplying electric power. It is an accident of history, not a difference in function, that explains why most Tennesseans get their electricity from a public enterprise and most Virginians get theirs from a private one. Whatever their ownership structures, the two companies do basically the same things to deliver power to customers. So to determine if the TVA has immunity here, the court on remand must first decide whether the conduct alleged to be negligent is governmental or commercial in nature. For the reasons given above, if the conduct is commercial—the kind of thing any power company might do—the TVA cannot invoke sovereign immunity. In that event, the TVA’s sue-and-be-sued clause renders it liable to the same extent as a private party. Only if the conduct at issue is governmental might the court decide that an implied limit on the clause bars the suit. But even assuming governmental activity, the court must find that prohibiting the “type[] of suit [at issue] is necessary to avoid grave interference” with that function’s performance. Burr, 309 U. S., at 245. Unless it is, Congress’s express statement that the TVA may “be sued” continues to demand that this suit go forward. We accordingly reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered.

588.US.2018_17-1717

In 1918, residents of Prince George’s County, Maryland, formed a committee for the purpose of erecting a memorial for the county’s soldiers who fell in World War I. The committee decided that the memorial should be a cross, which was not surprising since the plain Latin cross had become a central symbol of the war. The image of row after row of plain white crosses marking the overseas graves of soldiers was emblazoned on the minds of Americans at home. The memorial would stand at the terminus of another World War I memorial—the National Defense Highway connecting Washington to Annapolis. When the committee ran out of funds, the local American Legion took over the project, completing the memorial in 1925. The 32-foot tall Latin cross displays the American Legion’s emblem at its center and sits on a large pedestal bearing, inter alia, a bronze plaque that lists the names of the 49 county soldiers who had fallen in the war. At the dedication ceremony, a Catholic priest offered an invocation and a Baptist pastor offered a benediction. The Bladensburg Cross (Cross) has since been the site of patriotic events honoring veterans on, e.g., Veterans Day, Memorial Day, and Independence Day. Monuments honoring the veterans of other conflicts have been added in a park near the Cross. As the area around the Cross developed, the monument came to be at the center of a busy intersection. In 1961, the Maryland-National Capital Park and Planning Commission (Commission) acquired the Cross and the land where it sits, but the American Legion reserved the right to continue using the site for ceremonies. The Commission has used public funds to maintain the monument ever since. In 2014, the American Humanist Association (AHA) and others filed suit in District Court, alleging that the Cross’s presence on public land and the Commission’s maintenance of the memorial violate the First Amendment’s Establishment Clause. The American Legion intervened to defend the Cross. The District Court granted summary judgment for the Commission and the American Legion, concluding that the Cross satisfies both the test announced in Lemon v. Kurtzman, 403 U.S. 602, and the analysis applied by Justice Breyer in upholding a Ten Commandments monument in Van Orden v. Perry, 545 U.S. 677. The Fourth Circuit reversed. Held: The judgment is reversed and remanded. 874 F.3d 195, reversed and remanded. Justice Alito delivered the opinion of the Court with respect to Parts I, II–B, II–C, III, and IV, concluding that the Bladensburg Cross does not violate the Establishment Clause. Pp. 16–24, 28–31. (a) At least four considerations show that retaining established, religiously expressive monuments, symbols, and practices is quite different from erecting or adopting new ones. First, these cases often concern monuments, symbols, or practices that were first established long ago, and thus, identifying their original purpose or purposes may be especially difficult. See Salazar v. Buono, 559 U.S. 700. Second, as time goes by, the purposes associated with an established monument, symbol, or practice often multiply, as in the Ten Commandments monuments addressed in Van Orden and McCreary County v. American Civil Liberties Union of Ky., 545 U.S. 844. Even if the monument’s original purpose was infused with religion, the passage of time may obscure that sentiment and the monument may be retained for the sake of its historical significance or its place in a common cultural heritage. Third, the message of a monument, symbol, or practice may evolve, Pleasant Grove City v. Summum, 555 U.S. 460, 477, as is the case with a city name like Bethlehem, Pennsylvania; Arizona’s motto “Ditat Deus” (“God enriches”), adopted in 1864; or Maryland’s flag, which has included two crosses since 1904. Familiarity itself can become a reason for preservation. Fourth, when time’s passage imbues a religiously expressive monument, symbol, or practice with this kind of familiarity and historical significance, removing it may no longer appear neutral, especially to the local community. The passage of time thus gives rise to a strong presumption of constitutionality. Pp. 16–21. (b) The cross is a symbol closely linked to World War I. The United States adopted it as part of its military honors, establishing the Distinguished Service Cross and the Navy Cross in 1918 and 1919, respectively. And the fallen soldiers’ final resting places abroad were marked by white crosses or Stars of David, a solemn image that became inextricably linked with and symbolic of the ultimate price paid by 116,000 soldiers. This relationship between the cross and the war may not have been the sole or dominant motivation for the design of the many war memorials that sprang up across the Nation, but that is all but impossible to determine today. The passage of time means that testimony from the decisionmakers may not be available. And regardless of the original purposes for erecting the monument, a community may wish to preserve it for very different reasons, such as the historic preservation and traffic-safety concerns noted here. The area surrounding a monument like the Bladensburg Cross may also have been altered in ways that change its meaning and provide new reasons for its preservation. Even the AHA recognizes that the monument’s surroundings are important, as it concedes that the presence of a cross monument in a cemetery is unobjectionable. But a memorial’s placement in a cemetery is not necessary to create the connection to those it honors. Memorials took the place of gravestones for those parents and other relatives who lacked the means to travel to Europe to visit the graves of their war dead and for those soldiers whose bodies were never recovered. Similarly, memorials and monuments honoring important historical figures e.g., Dr. Martin Luther King, Jr., often include a symbol of the faith that was important to the persons whose lives are commemorated. Finally, as World War I monuments have endured through the years and become a familiar part of the physical and cultural landscape, requiring their removal or alteration would not be viewed by many as a neutral act. Few would say that California is attempting to convey a religious message by retaining the many city names, like Los Angeles and San Diego, given by the original Spanish settlers. But it would be something else entirely if the State undertook to change those names. Much the same is true about monuments to soldiers who sacrificed their lives for this country more than a century ago. Pp. 21–24. (c) Applying these principles here, the Bladensburg Cross does not violate the Establishment Clause. The image of the simple wooden cross that originally marked the graves of American soldiers killed in World War I became a symbol of their sacrifice, and the design of the Bladensburg Cross must be understood in light of that background. That the cross originated as a Christian symbol and retains that meaning in many contexts does not change the fact that the symbol took on an added secular meaning when used in World War I memorials. The Cross has also acquired historical importance with the passage of time, reminding the townspeople of the deeds and sacrifices of their predecessors as it stands among memorials to veterans of later wars. It has thus become part of the community. It would not serve that role had its design deliberately disrespected area soldiers, but there is no evidence that the names of any area Jewish soldiers were either intentionally left off the memorial’s list or included against the wishes of their families. The AHA tries to connect the Cross and the American Legion with anti-Semitism and the Ku Klux Klan, but the monument, which was dedicated during a period of heightened racial and religious animosity, includes the names of both Black and White soldiers; and both Catholic and Baptist clergy participated in the dedication. It is also natural and appropriate for a monument commemorating the death of particular individuals to invoke the symbols that signify what death meant for those who are memorialized. Excluding those symbols could make the memorial seem incomplete. This explains why Holocaust memorials invariably feature a Star of David or other symbols of Judaism and why the memorial at issue features the same symbol that marks the graves of so many soldiers near the battlefields where they fell. Pp. 28–30. (d) The fact that the cross is undoubtedly a Christian symbol should not blind one to everything else that the Bladensburg Cross has come to represent: a symbolic resting place for ancestors who never returned home, a place for the community to gather and honor all veterans and their sacrifices for this Nation, and a historical landmark. For many, destroying or defacing the Cross would not be neutral and would not further the ideals of respect and tolerance embodied in the First Amendment. P. 31. Justice Alito, joined by The Chief Justice, Justice Breyer, and Justice Kavanaugh, concluded in Parts II–A and II–D: (a) Lemon ambitiously attempted to fashion a test for all Establishment Clause cases. The test called on courts to examine the purposes and effects of a challenged government action, as well as any entanglement with religion that it might entail. The expectation of a ready framework has not been met, and the Court has many times either expressly declined to apply the test or simply ignored it. See, e.g., Zobrest v. Catalina Foothills Sch. Dist., 509 U.S. 1; Town of Greece v. Galloway, 572 U.S. 565. Pp. 12–16. (b) The Lemon Court ambitiously attempted to find a grand unified theory of the Establishment Clause, but the Court has since taken a more modest approach that focuses on the particular issue at hand and looks to history for guidance. The cases involving prayer before legislative sessions are illustrative. In Marsh v. Chambers, 463 U.S. 783, the Court upheld a State Legislature’s practice of beginning each session with a prayer by an official chaplain, finding it highly persuasive that Congress for over 200 years had opened its sessions with a prayer and that many state legislatures had followed suit. And the Court in Town of Greece reasoned that the historical practice of having, since the First Congress, chaplains in Congress showed “that the Framers considered legislative prayer a benign acknowledgment of religion’s role in society.” 572 U. S., at 576. Where monuments, symbols, and practices with a longstanding history follow in the tradition of the First Congress in respecting and tolerating different views, endeavoring to achieve inclusivity and nondiscrimination, and recognizing the important role religion plays in the lives of many Americans, they are likewise constitutional. Pp. 24–28. Justice Thomas, agreeing that the Bladensburg Cross is constitutional, concluded: (a) The text and history of the Clause—which reads “Congress shall make no law respecting an establishment of religion”—suggest that it should not be incorporated against the States. When the Court incorporated the Clause in Everson v. Board of Ed. of Ewing, 330 U.S. 1, 15, it apparently did not consider that an incorporated Establishment Clause would prohibit exactly what the text of the Clause seeks to protect: state establishments of religion. The appropriate question is whether any longstanding right of citizenship restrains the States in the establishment context. Further confounding the incorporation question is the fact that the First Amendment by its terms applies only to “law[s]” enacted by “Congress.” Pp. 1–3. (b) Even if the Clause applied to state and local governments in some fashion, “[t]he mere presence of the monument along [respondents’] path involves no [actual legal] coercion,” the sine qua non of an establishment of religion. Van Orden v. Perry, 545 U.S. 677, 694 (opinion of Thomas, J.). The plaintiff claiming an unconstitutional establishment of religion must demonstrate that he was actually coerced by government conduct that shares the characteristics of an establishment as understood at the founding. Respondents have not demonstrated that maintaining a religious display on public property shares any of the historical characteristics of an establishment of religion. Town of Greece v. Galloway, 572 U.S. 565, 608 (same). The Bladensburg Cross is constitutional even though the cross has religious significance. Religious displays or speech need not be limited to those considered nonsectarian. Insisting otherwise is inconsistent with this Nation’s history and traditions, id., at 578–580 (majority opinion), and would force the courts “to act as supervisors and censors of religious speech,” id., at 581. Pp. 3–5. (c) The plurality rightly rejects the relevance of the test set forth in Lemon v. Kurtzman, 403 U.S. 602, 612–613, to claims like this one, which involve religiously expressive monuments, symbols, displays, and similar practices, but Justice Thomas would take the logical next step and overrule the Lemon test in all contexts. The test has no basis in the original meaning of the Constitution; it has “been manipulated to fit whatever result the Court aimed to achieve,” McCreary County v. American Civil Liberties Union of Ky., 545 U.S. 844, 900 (Scalia, J., dissenting); and it continues to cause enormous confusion in the States and the lower courts. Pp. 6–7. Justice Gorsuch, joined by Justice Thomas, concludes that a suit like this one should be dismissed for lack of standing. Pp. 1–11. (a) The American Humanist Association claims that its members come into regular, unwelcome contact with the Bladensburg Cross when they drive through the area, but this “offended observer” theory of standing has no basis in law. To establish standing to sue consistent with the Constitution, a plaintiff must show: (1) injury-in-fact, (2) causation, and (3) redressability. And the injury-in-fact must be “concrete and particularized.” Lujan v. Defenders of Wildlife, 504 U.S. 555, 560. This Court has already rejected the notion that offense alone qualifies as a “concrete and particularized” injury sufficient to confer standing, Diamond v. Charles, 476 U.S. 54, 62, and it has done so in the context of the Establishment Clause itself, see Valley Forge Christian College v. Americans United for the Separation of Church and State, 454 U.S. 464. Offended observer standing is deeply inconsistent, too, with many other longstanding principles and precedents, including the rule that “ ‘generalized grievances’ about the conduct of Government” are insufficient to confer standing to sue, Schlesinger v. Reservists Comm. to Stop the War, 418 U.S. 208, 217, and “the rule that a party ‘generally must assert his own legal rights and interests,’ ” not those “ ‘of third parties,’ ” Kowalski v. Tesmer, 543 U.S. 125, 129. Pp. 1–6. (b) Lower courts invented offended observer standing for Establishment Clause cases in response to Lemon v. Kurtzman, 403 U.S. 602, reasoning that if the Establishment Clause forbids anything that a reasonable observer would view as an endorsement of religion, then such an observer must be able to sue. Lemon, however, was a misadventure, and the Court today relies on a more modest, historically sensitive approach, interpreting the Establishment Clause with reference to historical practices and understandings. The monument here is clearly constitutional in light of the nation’s traditions. Al- though the plurality does not say it in as many words, the message of today’s decision for the lower courts must be this: whether a monument, symbol, or practice is old or new, apply Town of Greece v. Galloway, 572 U.S. 565, not Lemon, because what matters when it comes to assessing a monument, symbol, or practice is not its age but its compliance with ageless principles. Pp. 6–9. (c) With Lemon now shelved, little excuse will remain for the anomaly of offended observer standing, and the gaping hole it tore in standing doctrine in the courts of appeals should now begin to close. Abandoning offended observer standing will mean only a return to the usual demands of Article III, requiring a real controversy with real impact on real persons to make a federal case out of it. Pp. 9–11. Alito, J., announced the judgment of the Court and delivered the opinion of the Court with respect to Parts I, II–B, II–C, III, and IV, in which Roberts, C. J., and Breyer, Kagan, and Kavanaugh, JJ., joined, and an opinion with respect to Parts II–A and II–D, in which Roberts, C. J., and Breyer and Kavanaugh, JJ., joined. Breyer, J., filed a concurring opinion, in which Kagan, J., joined. Kavanaugh, J., filed a concurring opinion. Kagan, J., filed an opinion concurring in part. Thomas, J., filed an opinion concurring in the judgment. Gorsuch, J., filed an opinion concurring in the judgment, in which Thomas, J., joined. Ginsburg, J., filed a dissenting opinion, in which Sotomayor, J., joined. Notes 1 Together with No. 18–18, Maryland-National Capital Park and Planning Commission v. American Humanist Assn. et al., also on certiorari to the same court.

of the Court with respect to Parts I, II–B, II–C, III, and IV, and an opinion with respect to Parts II–A and II–D, in which The Chief Justice, Justice Breyer, and Justice Kavanaugh join. Since 1925, the Bladensburg Peace Cross (Cross) has stood as a tribute to 49 area soldiers who gave their lives in the First World War. Eighty-nine years after the dedication of the Cross, respondents filed this lawsuit, claiming that they are offended by the sight of the memorial on public land and that its presence there and the expenditure of public funds to maintain it violate the Establishment Clause of the First Amendment. To remedy this violation, they asked a federal court to order the relocation or demolition of the Cross or at least the removal of its arms. The Court of Appeals for the Fourth Circuit agreed that the memorial is unconstitutional and remanded for a determination of the proper remedy. We now reverse. Although the cross has long been a preeminent Christian symbol, its use in the Bladensburg memorial has a special significance. After the First World War, the picture of row after row of plain white crosses marking the overseas graves of soldiers who had lost their lives in that horrible conflict was emblazoned on the minds of Americans at home, and the adoption of the cross as the Bladensburg memorial must be viewed in that historical context. For nearly a century, the Bladensburg Cross has expressed the community’s grief at the loss of the young men who perished, its thanks for their sacrifice, and its dedication to the ideals for which they fought. It has become a prominent community landmark, and its removal or radical alteration at this date would be seen by many not as a neutral act but as the manifestation of “a hostility toward religion that has no place in our Establishment Clause traditions.” Van Orden v. Perry, 545 U.S. 677, 704 (2005) (Breyer, J., concurring in judgment). And con- trary to respondents’ intimations, there is no evidence of discriminatory intent in the selection of the design of the memorial or the decision of a Maryland commission to maintain it. The Religion Clauses of the Constitution aim to foster a society in which people of all beliefs can live together harmoniously, and the presence of the Bladensburg Cross on the land where it has stood for so many years is fully consistent with that aim. I A The cross came into widespread use as a symbol of Christianity by the fourth century,[1] and it retains that meaning today. But there are many contexts in which the symbol has also taken on a secular meaning. Indeed, there are instances in which its message is now almost entirely secular. A cross appears as part of many registered trademarks held by businesses and secular organizations, including Blue Cross Blue Shield, the Bayer Group, and some Johnson & Johnson products.[2] Many of these marks relate to health care, and it is likely that the association of the cross with healing had a religious origin. But the current use of these marks is indisputably secular. The familiar symbol of the Red Cross—a red cross on a white background—shows how the meaning of a symbol that was originally religious can be transformed. The International Committee of the Red Cross (ICRC) selected that symbol in 1863 because it was thought to call to mind the flag of Switzerland, a country widely known for its neutrality.[3] The Swiss flag consists of a white cross on a red background. In an effort to invoke the message associated with that flag, the ICRC copied its design with the colors inverted. Thus, the ICRC selected this symbol for an essentially secular reason, and the current secular message of the symbol is shown by its use today in nations with only tiny Christian populations.[4] But the cross was originally chosen for the Swiss flag for religious reasons.[5] So an image that began as an expression of faith was transformed. The image used in the Bladensburg memorial—a plain Latin cross[6]—also took on new meaning after World War I. “During and immediately after the war, the army marked soldiers’ graves with temporary wooden crosses or Stars of David”—a departure from the prior practice of marking graves in American military cemeteries with uniform rectangular slabs. G. Piehler, Remembering War the American Way 101 (1995); App. 1146. The vast majority of these grave markers consisted of crosses,[7] and thus when Americans saw photographs of these cemeteries, what struck them were rows and rows of plain white crosses. As a result, the image of a simple white cross “developed into a ‘central symbol’ ” of the conflict. Ibid. Contemporary literature, poetry, and art reflected this powerful imagery. See Brief for Veterans of Foreign Wars of the United States et al. as Amici Curiae 10–16. Perhaps most famously, John McCrae’s poem, In Flanders Fields, began with these memorable lines: “In Flanders fields the poppies blow Between the crosses, row on row.” In Flanders Fields and Other Poems 3 (G. P. Putnam’s Sons ed. 1919). The poem was enormously popular. See P. Fussell, The Great War and Modern Memory 248–249 (1975). A 1921 New York Times article quoted a description of McCrae’s composition as “ ‘the poem of the army’ ” and “ ‘of all those who understand the meaning of the great conflict.’ ”[8] The image of “the crosses, row on row,” stuck in people’s minds, and even today for those who view World War I cemeteries in Europe, the image is arresting.[9] After the 1918 armistice, the War Department announced plans to replace the wooden crosses and Stars of David with uniform marble slabs like those previously used in American military cemeteries. App. 1146. But the public outcry against that proposal was swift and fierce. Many organizations, including the American War Mothers, a nonsectarian group founded in 1917, urged the Department to retain the design of the temporary markers. Id., at 1146–1147. When the American Battle Monuments Commission took over the project of designing the headstones, it responded to this public sentiment by opting to replace the wooden crosses and Stars of David with marble versions of those symbols. Id., at 1144. A Member of Congress likewise introduced a resolution noting that “these wooden symbols have, during and since the World War, been regarded as emblematic of the great sacrifices which that war entailed, have been so treated by poets and artists and have become peculiarly and inseparably associated in the thought of surviving relatives and comrades and of the Nation with these World War graves.” H. Res. 15, 68th Cong., 1 (1924), App. 1163–1164. This national debate and its outcome confirmed the cross’s widespread resonance as a symbol of sacrifice in the war. B Recognition of the cross’s symbolism extended to local communities across the country. In late 1918, residents of Prince George’s County, Maryland, formed a committee for the purpose of erecting a memorial for the county’s fallen soldiers. App. 988–989, 1014. Among the committee’s members were the mothers of 10 deceased soldiers. Id., at 989. The committee decided that the memorial should be a cross and hired sculptor and architect John Joseph Earley to design it. Although we do not know precisely why the committee chose the cross, it is unsurprising that the committee—and many others commemorating World War I[10]—adopted a symbol so widely associated with that wrenching event. After selecting the design, the committee turned to the task of financing the project. The committee held fundraising events in the community and invited donations, no matter the size, with a form that read: “We, the citizens of Maryland, trusting in God, the Supreme Ruler of the Universe, Pledge Faith in our Brothers who gave their all in the World War to make [the] World Safe for Democracy. Their Mortal Bodies have turned to dust, but their spirit Lives to guide us through Life in the way of Godliness, Justice and Liberty. “With our Motto, ‘One God, One Country, and One Flag’ We contribute to this Memorial Cross Commemorating the Memory of those who have not Died in Vain.” Id., at. 1251. Many of those who responded were local residents who gave small amounts: Donations of 25 cents to 1 dollar were the most common. Id., at 1014. Local businesses and political leaders assisted in this effort. Id., at 1014, 1243. In writing to thank United States Senator John Walter Smith for his donation, committee treasurer Mrs. Martin Redman explained that “[t]he chief reason I feel as deeply in this matter [is that], my son, [Wm.] F. Redman, lost his life in France and because of that I feel that our memorial cross is, in a way, his grave stone.” Id., at 1244. The Cross was to stand at the terminus of another World War I memorial—the National Defense Highway, which connects Washington to Annapolis. The community gathered for a joint groundbreaking ceremony for both memorials on September 28, 1919; the mother of the first Prince George’s County resident killed in France broke ground for the Cross. Id., at 910. By 1922, however, the committee had run out of funds, and progress on the Cross had stalled. The local post of the American Legion took over the project, and the monument was finished in 1925. The completed monument is a 32-foot tall Latin cross that sits on a large pedestal. The American Legion’s emblem is displayed at its center, and the words “Valor,” “Endurance,” “Courage,” and “Devotion” are inscribed at its base, one on each of the four faces. The pedestal also features a 9- by 2.5-foot bronze plaque explaining that the monument is “Dedicated to the heroes of Prince George’s County, Maryland who lost their lives in the Great War for the liberty of the world.” Id., at 915 (capitalization omitted). The plaque lists the names of 49 local men, both Black and White, who died in the war. It identifies the dates of American involvement, and quotes President Woodrow Wilson’s request for a declaration of war: “The right is more precious than peace. We shall fight for the things we have always carried nearest our hearts. To such a task we dedicate our lives.” Ibid. At the dedication ceremony, a local Catholic priest offered an invocation. Id., at 217–218. United States Representative Stephen W. Gambrill delivered the keynote address, honoring the “ ‘men of Prince George’s County’ ” who “ ‘fought for the sacred right of all to live in peace and security.’ ” Id., at 1372. He encouraged the commu- nity to look to the “ ‘token of this cross, symbolic of Calvary,’ ” to “ ‘keep fresh the memory of our boys who died for a righteous cause.’ ” Ibid. The ceremony closed with a benediction offered by a Baptist pastor. Since its dedication, the Cross has served as the site of patriotic events honoring veterans, including gatherings on Veterans Day, Memorial Day, and Independence Day. Like the dedication itself, these events have typically included an invocation, a keynote speaker, and a benediction. Id., at 182, 319–323. Over the years, memorials honoring the veterans of other conflicts have been added to the surrounding area, which is now known as Veterans Memorial Park. These include a World War II Honor Scroll; a Pearl Harbor memorial; a Korea-Vietnam veterans memorial; a September 11 garden; a War of 1812 memorial; and two recently added 38-foot-tall markers depicting British and American soldiers in the Battle of Bladensburg. Id., at 891–903, 1530. Because the Cross is located on a traffic island with limited space, the closest of these other monuments is about 200 feet away in a park across the road. Id., at 36, 44. As the area around the Cross developed, the monument came to be at the center of a busy intersection. In 1961, the Maryland-National Capital Park and Planning Commission (Commission) acquired the Cross and the land on which it sits in order to preserve the monument and address traffic-safety concerns.[11] Id., at 420–421, 1384–1387. The American Legion reserved the right to continue using the memorial to host a variety of ceremonies, including events in memory of departed veterans. Id., at 1387. Over the next five decades, the Commission spent approximately $117,000 to maintain and preserve the monument. In 2008, it budgeted an additional $100,000 for renovations and repairs to the Cross.[12] C In 2012, nearly 90 years after the Cross was dedicated and more than 50 years after the Commission acquired it, the American Humanist Association (AHA) lodged a complaint with the Commission. The complaint alleged that the Cross’s presence on public land and the Commission’s maintenance of the memorial violate the Establishment Clause of the First Amendment. Id., at 1443–1451. The AHA, along with three residents of Washington, D. C., and Maryland, also sued the Commission in the District Court for the District of Maryland, making the same claim. The AHA sought declaratory and injunctive relief requiring “removal or demolition of the Cross, or removal of the arms from the Cross to form a non-religious slab or obelisk.” 874 F.3d 195, 202, n. 7 (CA4 2017) (internal quotation marks omitted). The American Legion intervened to defend the Cross. The District Court granted summary judgment for the Commission and the American Legion. The Cross, the District Court held, satisfies both the three-pronged test announced in Lemon v. Kurtzman, 403 U.S. 602 (1971), and the analysis applied by Justice Breyer in upholding the Ten Commandments monument at issue in Van Orden v. Perry, 545 U.S. 677. Under the Lemon test, a court must ask whether a challenged government action (1) has a secular purpose; (2) has a “principal or primary effect” that “neither advances nor inhibits religion”; and (3) does not foster “an excessive government entanglement with religion,” 403 U. S., at 612–613 (internal quotation marks omitted). Applying that test, the District Court determined that the Commission had secular purposes for acquiring and maintaining the Cross—namely, to commemorate World War I and to ensure traffic safety. The court also found that a reasonable observer aware of the Cross’s history, setting, and secular elements “would not view the Monument as having the effect of impermissibly endorsing religion.” 147 F. Supp. 3d 373, 387 (Md. 2015). Nor, according to the court, did the Commission’s maintenance of the memorial create the kind of “continued and repeated government involvement with religion” that would constitute an excessive entanglement. Ibid. (internal quotation marks and emphasis omitted). Finally, in light of the factors that informed its analysis of Lemon’s “effects” prong, the court concluded that the Cross is constitutional under Justice Breyer’s approach in Van Orden. 147 F. Supp. 3d, at 388–390. A divided panel of the Court of Appeals for the Fourth Circuit reversed. The majority relied primarily on the Lemon test but also took cognizance of Justice Breyer’s Van Orden concurrence. While recognizing that the Commission acted for a secular purpose, the court held that the Bladensburg Cross failed Lemon’s “effects” prong because a reasonable observer would view the Commission’s ownership and maintenance of the monument as an endorsement of Christianity. The court emphasized the cross’s “inherent religious meaning” as the “ ‘preeminent symbol of Christianity.’ ” 874 F. 3d, at 206–207. Although conceding that the monument had several “secular elements,” the court asserted that they were “overshadow[ed]” by the Cross’s size and Christian connection—especially because the Cross’s location and condition would make it difficult for “passers-by” to “read” or otherwise “examine” the plaque and American Legion emblem. Id., at 209–210. The court rejected as “too simplistic” an argument dppefending the Cross’s constitutionality on the basis of its 90-year history, suggesting that “[p]erhaps the longer a violation persists, the greater the affront to those offended.” Id., at 208. In the alternative, the court concluded, the Commission had become excessively entangled with religion by keeping a display that “aggrandizes the Latin cross” and by spending more than de minimis public funds to maintain it. Id., at 211–212. Chief Judge Gregory dissented in relevant part, contending that the majority misapplied the “effects” test by failing to give adequate consideration to the Cross’s “physical setting, history, and usage.” Id., at 218 (opinion concurring in part and dissenting in part). He also disputed the majority’s excessive-entanglement analysis, noting that the Commission’s maintenance of the Cross was not the kind of “comprehensive, discriminating, and continuing state surveillance” of religion that Lemon was con-cerned to rule out. 874 F. 3d, at 221 (internal quotation marks omitted). The Fourth Circuit denied rehearing en banc over dissents by Chief Judge Gregory, Judge Wilkinson, and Judge Niemeyer. 891 F.3d 117 (2018). The Commission and the American Legion each petitioned for certiorari. We granted the petitions and consolidated them for argument. 586 U. S. ___ (2016). II A The Establishment Clause of the First Amendment provides that “Congress shall make no law respecting an establishment of religion.” While the concept of a formally established church is straightforward, pinning down the meaning of a “law respecting an establishment of religion” has proved to be a vexing problem. Prior to the Court’s decision in Everson v. Board of Ed. of Ewing, 330 U.S. 1 (1947), the Establishment Clause was applied only to the Federal Government, and few cases involving this provision came before the Court. After Everson recognized the incorporation of the Clause, however, the Court faced a steady stream of difficult and controversial Establishment Clause issues, ranging from Bible reading and prayer in the public schools, Engel v. Vitale, 370 U.S. 421 (1962); School Dist. of Abington Township v. Schempp, 374 U.S. 203 (1963), to Sunday closing laws, McGowan v. Maryland, 366 U.S. 420 (1961), to state subsidies for church-related schools or the parents of students attending those schools, Board of Ed. of Central School Dist. No. 1 v. Allen, 392 U.S. 236 (1968); Everson, supra. After grappling with such cases for more than 20 years, Lemon ambitiously attempted to distill from the Court’s existing case law a test that would bring order and predictability to Establishment Clause decisionmaking. That test, as noted, called on courts to examine the purposes and effects of a challenged government action, as well as any entanglement with religion that it might entail. Lemon, 403 U. S., at 612–613. The Court later elaborated that the “effect[s]” of a challenged action should be assessed by asking whether a “reasonable observer” would conclude that the action constituted an “endorsement” of religion. County of Allegheny v. American Civil Liberties Union, Greater Pittsburgh Chapter, 492 U.S. 573, 592 (1989); id., at 630 (O’Connor, J., concurring in part and concurring in judgment). If the Lemon Court thought that its test would provide a framework for all future Establishment Clause decisions, its expectation has not been met. In many cases, this Court has either expressly declined to apply the test or has simply ignored it. See Zobrest v. Catalina Foothills School Dist., 509 U.S. 1 (1993); Board of Ed. of Kiryas Joel Village School Dist. v. Grumet, 512 U.S. 687 (1994); Rosenberger v. Rector and Visitors of Univ. of Va., 515 U.S. 819 (1995); Capitol Square Review and Advisory Bd. v. Pinette, 515 U.S. 753 (1995); Good News Club v. Milford Central School, 533 U.S. 98 (2001); Zelman v. Simmons-Harris, 536 U.S. 639 (2002); Cutter v. Wilkinson, 544 U.S. 709 (2005); Van Orden, 545 U.S. 677; Hosanna-Tabor Evangelical Lutheran Church and School v. EEOC, 565 U.S. 171 (2012); Town of Greece v. Galloway, 572 U.S. 565 (2014); Trump v. Hawaii, 585 U. S. ___ (2018). This pattern is a testament to the Lemon test’s shortcomings. As Establishment Clause cases involving a great array of laws and practices came to the Court, it became more and more apparent that the Lemon test could not resolve them. It could not “explain the Establishment Clause’s tolerance, for example, of the prayers that open legislative meetings, . . . certain references to, and invocations of, the Deity in the public words of public officials; the public references to God on coins, decrees, and buildings; or the attention paid to the religious objectives of certain holidays, including Thanksgiving.” Van Orden, supra, at 699 (opinion of Breyer, J.). The test has been harshly criticized by Members of this Court,[13] lamented by lower court judges,[14] and questioned by a diverse roster of scholars.[15] For at least four reasons, the Lemon test presents particularly daunting problems in cases, including the one now before us, that involve the use, for ceremonial, celebratory, or commemorative purposes, of words or symbols with religious associations.[16] Together, these considera- tions counsel against efforts to evaluate such cases under Lemon and toward application of a presumption of constitutionality for longstanding monuments, symbols, and practices. B First, these cases often concern monuments, symbols, or practices that were first established long ago, and in such cases, identifying their original purpose or purposes may be especially difficult. In Salazar v. Buono, 559 U.S. 700 (2010), for example, we dealt with a cross that a small group of World War I veterans had put up at a remote spot in the Mojave Desert more than seven decades earlier. The record contained virtually no direct evidence regarding the specific motivations of these men. We knew that they had selected a plain white cross, and there was some evidence that the man who looked after the monument for many years—“a miner who had served as a medic and had thus presumably witnessed the carnage of the war firsthand”—was said not to have been “particularly religious.” Id., at 724 (Alito, J., concurring in part and concurring in judgment). Without better evidence about the purpose of the monument, different Justices drew different inferences. The plurality thought that this particular cross was meant “to commemorate American servicemen who had died in World War I” and was not intended “to promote a Christian message.” Id., at 715. The dissent, by contrast, “presume[d]” that the cross’s purpose “was a Christian one, at least in part, for the simple reason that those who erected the cross chose to commemorate American veterans in an explicitly Christian manner.” Id., at 752 (opinion of Stevens, J.). The truth is that 70 years after the fact, there was no way to be certain about the motivations of the men who were responsible for the creation of the monument. And this is often the case with old monuments, symbols, and practices. Yet it would be inappropriate for courts to compel their removal or termination based on supposition. Second, as time goes by, the purposes associated with an established monument, symbol, or practice often multiply. Take the example of Ten Commandments monuments, the subject we addressed in Van Orden, 545 U.S. 677, and McCreary County v. American Civil Liberties Union of Ky., 545 U.S. 844 (2005). For believing Jews and Christians, the Ten Commandments are the word of God handed down to Moses on Mount Sinai, but the image of the Ten Commandments has also been used to convey other meanings. They have historical significance as one of the foundations of our legal system, and for largely that reason, they are depicted in the marble frieze in our courtroom and in other prominent public buildings in our Nation’s capital. See Van Orden, supra, at 688–690. In Van Orden and McCreary, no Member of the Court thought that these depictions are unconstitutional. 545 U. S., at 688–690; id., at 701 (opinion of Breyer, J.); id., at 740 (Souter, J., dissenting). Just as depictions of the Ten Commandments in these public buildings were intended to serve secular purposes, the litigation in Van Orden and McCreary showed that secular motivations played a part in the proliferation of Ten Commandments monuments in the 1950s. In 1946, Minnesota Judge E. J. Ruegemer proposed that the Ten Commandments be widely disseminated as a way of combating juvenile delinquency.[17] With this prompting, the Fraternal Order of the Eagles began distributing paper copies of the Ten Commandments to churches, school groups, courts, and government offices. The Eagles, “while interested in the religious aspect of the Ten Commandments, sought to highlight the Commandments’ role in shaping civic morality.” Van Orden, supra, at 701 (opinion of Breyer, J.). At the same time, Cecil B. DeMille was filming The Ten Commandments.[18] He learned of Judge Ruegemer’s campaign, and the two collaborated, deciding that the Commandments should be carved on stone tablets and that DeMille would make arrangements with the Eagles to help pay for them, thus simultaneously promoting his film and public awareness of the Decalogue. Not only did DeMille and Judge Ruegemer have different purposes, but the motivations of those who accepted the monuments and those responsible for maintaining them may also have differed. As we noted in Pleasant Grove City v. Summum, 555 U.S. 460, 476 (2009), “the thoughts or sentiments expressed by a government entity that accepts and displays [a monument] may be quite different from those of either its creator or its donor.” The existence of multiple purposes is not exclusive to longstanding monuments, symbols, or practices, but this phenomenon is more likely to occur in such cases. Even if the original purpose of a monument was infused with religion, the passage of time may obscure that sentiment. As our society becomes more and more religiously diverse, a community may preserve such monuments, symbols, and practices for the sake of their historical significance or their place in a common cultural heritage. Cf. Schempp, 374 U. S., at 264–265 (Brennan, J., concurring) (“[The] government may originally have decreed a Sunday day of rest for the impermissible purpose of supporting religion but abandoned that purpose and retained the laws for the permissible purpose of furthering overwhelmingly secular ends”). Third, just as the purpose for maintaining a monument, symbol, or practice may evolve, “[t]he ‘message’ conveyed . . . may change over time.” Summum, 555 U. S., at 477. Consider, for example, the message of the Statue of Lib- erty, which began as a monument to the solidarity and friendship between France and the United States and only decades later came to be seen “as a beacon welcoming immigrants to a land of freedom.” Ibid. With sufficient time, religiously expressive monuments, symbols, and practices can become embedded features of a community’s landscape and identity. The community may come to value them without necessarily embracing their religious roots. The recent tragic fire at Notre Dame in Paris provides a striking example. Although the French Republic rigorously enforces a secular public square,[19] the cathedral remains a symbol of national importance to the religious and nonreligious alike. Notre Dame is fundamentally a place of worship and retains great religious importance, but its meaning has broadened. For many, it is inextricably linked with the very idea of Paris and France.[20] Speaking to the nation shortly after the fire, President Macron said that Notre Dame “ ‘is our history, our literature, our imagination. The place where we survived epidemics, wars, liberation. It has been the epicenter of our lives.’ ”[21] In the same way, consider the many cities and towns across the United States that bear religious names. Religion undoubtedly motivated those who named Bethlehem, Pennsylvania; Las Cruces, New Mexico; Providence, Rhode Island; Corpus Christi, Texas; Nephi, Utah, and the countless other places in our country with names that are rooted in religion. Yet few would argue that this history requires that these names be erased from the map. Or take a motto like Arizona’s, “Ditat Deus” (“God enriches”), which was adopted in 1864,[22] or a flag like Maryland’s, which has included two crosses since 1904.[23] Familiarity itself can become a reason for preservation. Fourth, when time’s passage imbues a religiously expressive monument, symbol, or practice with this kind of familiarity and historical significance, removing it may no longer appear neutral, especially to the local community for which it has taken on particular meaning. A government that roams the land, tearing down monuments with religious symbolism and scrubbing away any reference to the divine will strike many as aggressively hostile to religion. Militantly secular regimes have carried out such projects in the past,[24] and for those with a knowledge of history, the image of monuments being taken down will be evocative, disturbing, and divisive. Cf. Van Orden, 545 U. S., at 704 (opinion of Breyer, J.) (“[D]isputes concerning the removal of longstanding depictions of the Ten Commandments from public buildings across the Nation . . . could thereby create the very kind of religiously based divisiveness that the Establishment Clause seeks to avoid”). These four considerations show that retaining established, religiously expressive monuments, symbols, and practices is quite different from erecting or adopting new ones. The passage of time gives rise to a strong presumption of constitutionality. C The role of the cross in World War I memorials is il- lustrative of each of the four preceding considerations. Immediately following the war, “[c]ommunities across America built memorials to commemorate those who had served the nation in the struggle to make the world safe for democracy.” G. Piehler, The American Memory of War, App. 1124. Although not all of these communities included a cross in their memorials, the cross had become a symbol closely linked to the war. “[T]he First World War witnessed a dramatic change in . . . the symbols used to commemorate th[e] service” of the fallen soldiers. Id., at 1123. In the wake of the war, the United States adopted the cross as part of its military honors, establishing the Distinguished Service Cross and the Navy Cross in 1918 and 1919, respectively. See id., at 147–148. And as already noted, the fallen soldiers’ final resting places abroad were marked by white crosses or Stars of David. The solemn image of endless rows of white crosses became inextricably linked with and symbolic of the ultimate price paid by 116,000 soldiers. And this relationship between the cross and the war undoubtedly influenced the design of the many war memorials that sprang up across the Nation. This is not to say that the cross’s association with the war was the sole or dominant motivation for the inclusion of the symbol in every World War I memorial that features it. But today, it is all but impossible to tell whether that was so. The passage of time means that testimony from those actually involved in the decisionmaking process is generally unavailable, and attempting to uncover their motivations invites rampant speculation. And no matter what the original purposes for the erection of a monument, a community may wish to preserve it for very different reasons, such as the historic preservation and traffic-safety concerns the Commission has pressed here. In addition, the passage of time may have altered the area surrounding a monument in ways that change its meaning and provide new reasons for its preservation. Such changes are relevant here, since the Bladensburg Cross now sits at a busy traffic intersection, and numerous additional monuments are located nearby. Even the AHA recognizes that there are instances in which a war memorial in the form of a cross is unobjectionable. The AHA is not offended by the sight of the Argonne Cross or the Canadian Cross of Sacrifice, both Latin crosses commemorating World War I that rest on public grounds in Arlington National Cemetery. The difference, according to the AHA, is that their location in a cemetery gives them a closer association with individual gravestones and interred soldiers. See Brief for Respondents 96; Tr. of Oral Arg. 52. But a memorial’s placement in a cemetery is not necessary to create such a connection. The parents and other relatives of many of the war dead lacked the means to travel to Europe to visit their graves, and the bodies of approximately 4,400 American soldiers were either never found or never identified.[25] Thus, for many grieving relatives and friends, memorials took the place of gravestones. Recall that the mother of one of the young men memorialized by the Bladensburg Cross thought of the memorial as, “in a way, his grave stone.” App. 1244. Whether in a cemetery or a city park, a World War I cross remains a memorial to the fallen. Similar reasoning applies to other memorials and monuments honoring important figures in our Nation’s his- tory. When faith was important to the person whose life is commemorated, it is natural to include a symbolic reference to faith in the design of the memorial. For example, many memorials for Dr. Martin Luther King, Jr., make reference to his faith. Take the Martin Luther King, Jr. Civil Rights Memorial Park in Seattle, which contains a sculpture in three segments representing “both the Christian Trinity and the union of the family.”[26] In Atlanta, the Ebenezer Baptist Church sits on the grounds of the Martin Luther King, Jr. National Historical Park. National Statuary Hall in the Capitol honors a variety of religious figures: for example, Mother Joseph Pariseau kneeling in prayer; Po’Pay, a Pueblo religious leader with symbols of the Pueblo religion; Brigham Young, president of the Church of Jesus Christ of Latter-day Saints; and Father Eusebio Kino with a crucifix around his neck and his hand raised in blessing.[27] These monuments honor men and women who have played an important role in the history of our country, and where religious symbols are included in the monuments, their presence acknowledges the centrality of faith to those whose lives are commemorated. Finally, as World War I monuments have endured through the years and become a familiar part of the physical and cultural landscape, requiring their removal would not be viewed by many as a neutral act. And an alteration like the one entertained by the Fourth Circuit—amputating the arms of the Cross, see 874 F. 3d, at 202, n. 7—would be seen by many as profoundly disrespectful. One member of the majority below viewed this objection as inconsistent with the claim that the Bladensburg Cross serves secular purposes, see 891 F. 3d, at 121 (Wynn, J., concurring in denial of en banc), but this argument misunderstands the complexity of monuments. A monument may express many purposes and convey many different messages, both secular and religious. Cf. Van Orden, 545 U. S., at 690 (plurality opinion) (describing simultaneous religious and secular meaning of the Ten Commandments display). Thus, a campaign to obliterate items with religious associations may evidence hostility to religion even if those religious associations are no longer in the forefront. For example, few would say that the State of California is attempting to convey a religious message by retaining the names given to many of the State’s cities by their original Spanish settlers—San Diego, Los Angeles, Santa Barbara, San Jose, San Francisco, etc. But it would be something else entirely if the State undertook to change all those names. Much the same is true about monuments to soldiers who sacrificed their lives for this country more than a century ago. D While the Lemon Court ambitiously attempted to find a grand unified theory of the Establishment Clause, in later cases, we have taken a more modest approach that focuses on the particular issue at hand and looks to history for guidance. Our cases involving prayer before a legislative session are an example. In Marsh v. Chambers, 463 U.S. 783 (1983), the Court upheld the Nebraska Legislature’s practice of beginning each session with a prayer by an official chaplain, and in so holding, the Court conspicuously ignored Lemon and did not respond to Justice Brennan’s argument in dissent that the legislature’s practice could not satisfy the Lemon test. Id., at 797–801. Instead, the Court found it highly persuasive that Congress for more than 200 years had opened its sessions with a prayer and that many state legislatures had followed suit. Id., at 787–788. We took a similar approach more recently in Town of Greece, 572 U. S., at 577. We reached these results even though it was clear, as stressed by the Marsh dissent, that prayer is by definition religious. See Marsh, supra, at 797–798 (opinion of Brennan, J.). As the Court put it in Town of Greece: “Marsh must not be understood as permitting a practice that would amount to a constitutional violation if not for its historical foundation.” 572 U. S., at 576. “The case teaches instead that the Establishment Clause must be interpreted ‘by reference to historical practices and understandings’ ” and that the decision of the First Congress to “provid[e] for the appointment of chaplains only days after approving language for the First Amendment demonstrates that the Framers considered legislative prayer a benign acknowledgment of religion’s role in society.” Ibid. The prevalence of this philosophy at the time of the founding is reflected in other prominent actions taken by the First Congress. It requested—and President Washington proclaimed—a national day of prayer, see 1 J. Richardson, Messages and Papers of the Presidents, 1789–1897, p. 64 (1897) (President Washington’s Thanksgiving Proclamation), and it reenacted the Northwest Territory Ordinance, which provided that “[r]eligion, morality, and knowledge, being necessary to good government and the happiness of mankind, schools and the means of education shall forever be encouraged,” 1Stat. 52, n. (a). President Washington echoed this sentiment in his Farewell Address, calling religion and morality “indispensable supports” to “political prosperity.” Farewell Address (1796), in 35 The Writings of George Washington 229 (J. Fitzpatrick ed. 1940). See also P. Hamburger, Separation of Church and State 66 (2002). The First Congress looked to these “supports” when it chose to begin its sessions with a prayer. This practice was designed to solemnize congressional meetings, unifying those in attendance as they pursued a common goal of good governance. To achieve that purpose, legislative prayer needed to be inclusive rather than divisive, and that required a determined effort even in a society that was much more religiously homogeneous than ours today. Although the United States at the time was overwhelmingly Christian and Protestant,[28] there was considerable friction between Protestant denominations. See M. Noll, America’s God: From Jonathan Edwards to Abraham Lincoln 228 (2002). Thus, when an Episcopal clergyman was nominated as chaplain, some Congregationalist Members of Congress objected due to the “ ‘diversity of religious sentiments represented in Congress.’ ” D. Davis, Religion and the Continental Congress 74 (2000). Nevertheless, Samuel Adams, a staunch Congregationalist, spoke in favor of the motion: “ ‘I am no bigot. I can hear a prayer from a man of piety and virtue, who is at the same time a friend of his country.’ ” Ibid. Others agreed and the chaplain was appointed. Over time, the members of the clergy invited to offer prayers at the opening of a session grew more and more diverse. For example, an 1856 study of Senate and House Chaplains since 1789 tallied 22 Methodists, 20 Presbyterians, 19 Episcopalians, 13 Baptists, 4 Congregationalists, 2 Roman Catholics, and 3 that were characterized as “miscellaneous.”[29] Four years later, Rabbi Morris Raphall became the first rabbi to open Congress.[30] Since then, Congress has welcomed guest chaplains from a variety of faiths, including Islam, Hinduism, Buddhism, and Native American religions.[31] In Town of Greece, which concerned prayer before a town council meeting, there was disagreement about the inclusiveness of the town’s practice. Compare 572 U. S., at 585 (opinion of the Court) (“The town made reasonable efforts to identify all of the congregations located within its borders and represented that it would welcome a prayer by any minister or layman who wished to give one”), with id., at 616 (Kagan, J., dissenting) (“Greece’s Board did nothing to recognize religious diversity”). But there was no disagreement that the Establishment Clause permits a nondiscriminatory practice of prayer at the beginning of a town council session. See ibid. (“I believe that pluralism and inclusion [in legislative prayer] in a town hall can satisfy the constitutional requirement of neutrality”). Of course, the specific practice challenged in Town of Greece lacked the very direct connection, via the First Congress, to the thinking of those who were responsible for framing the First Amendment. But what mattered was that the town’s practice “fi[t] within the tradition long followed in Congress and the state legislatures.” Id., at 577 (opinion of the Court). The practice begun by the First Congress stands out as an example of respect and tolerance for differing views, an honest endeavor to achieve inclusivity and nondiscrimination, and a recognition of the important role that religion plays in the lives of many Americans. Where categories of monuments, symbols, and practices with a longstand- ing history follow in that tradition, they are likewise constitutional. III Applying these principles, we conclude that the Bladensburg Cross does not violate the Establishment Clause. As we have explained, the Bladensburg Cross carries special significance in commemorating World War I. Due in large part to the image of the simple wooden crosses that originally marked the graves of American soldiers killed in the war, the cross became a symbol of their sacrifice, and the design of the Bladensburg Cross must be understood in light of that background. That the cross originated as a Christian symbol and retains that meaning in many contexts does not change the fact that the symbol took on an added secular meaning when used in World War I memorials. Not only did the Bladensburg Cross begin with this meaning, but with the passage of time, it has acquired historical importance. It reminds the people of Bladensburg and surrounding areas of the deeds of their predecessors and of the sacrifices they made in a war fought in the name of democracy. As long as it is retained in its original place and form, it speaks as well of the community that erected the monument nearly a century ago and has maintained it ever since. The memorial represents what the relatives, friends, and neighbors of the fallen soldiers felt at the time and how they chose to express their sentiments. And the monument has acquired additional layers of historical meaning in subsequent years. The Cross now stands among memorials to veterans of later wars. It has become part of the community. The monument would not serve that role if its design had deliberately disrespected area soldiers who perished in World War I. More than 3,500 Jewish soldiers gave their lives for the United States in that conflict,[32] and some have wondered whether the names of any Jewish soldiers from the area were deliberately left off the list on the memorial or whether the names of any Jewish soldiers were included on the Cross against the wishes of their families. There is no evidence that either thing was done, and we do know that one of the local American Legion leaders responsible for the Cross’s construction was a Jewish veteran. See App. 65, 205, 990. The AHA’s brief strains to connect the Bladensburg Cross and even the American Legion with anti-Semitism and the Ku Klux Klan, see Brief for Respondents 5–7, but the AHA’s disparaging intimations have no evidentiary support. And when the events surrounding the erection of the Cross are viewed in historical context, a very different picture may perhaps be discerned. The monument was dedicated on July 12, 1925, during a period when the country was experiencing heightened racial and religious animosity. Membership in the Ku Klux Klan, which preached hatred of Blacks, Catholics, and Jews, was at its height.[33] On August 8, 1925, just two weeks after the dedication of the Bladensburg Cross and less than 10 miles away, some 30,000 robed Klansmen marched down Pennsylvania Avenue in the Nation’s Capital. But the Bladensburg Cross memorial included the names of both Black and White soldiers who had given their lives in the war; and despite the fact that Catholics and Baptists at that time were not exactly in the habit of participating together in ecumenical services, the ceremony dedicating the Cross began with an invocation by a Catholic priest and ended with a benediction by a Baptist pastor. App. 1559–1569, 1373. We can never know for certain what was in the minds of those responsible for the memorial, but in light of what we know about this ceremony, we can perhaps make out a picture of a community that, at least for the moment, was united by grief and patriotism and rose above the divisions of the day. Finally, it is surely relevant that the monument commemorates the death of particular individuals. It is natural and appropriate for those seeking to honor the deceased to invoke the symbols that signify what death meant for those who are memorialized. In some circumstances, the exclusion of any such recognition would make a memorial incomplete. This well explains why Holocaust memorials invariably include Stars of David or other symbols of Judaism.[34] It explains why a new memorial to Native American veterans in Washington, D. C., will portray a steel circle to represent “ ‘the hole in the sky where the creator lives.’ ”[35] And this is why the memorial for soldiers from the Bladensburg community features the cross—the same symbol that marks the graves of so many of their comrades near the battlefields where they fell. IV The cross is undoubtedly a Christian symbol, but that fact should not blind us to everything else that the Bladensburg Cross has come to represent. For some, that monument is a symbolic resting place for ancestors who never returned home. For others, it is a place for the community to gather and honor all veterans and their sacrifices for our Nation. For others still, it is a historical landmark. For many of these people, destroying or defacing the Cross that has stood undisturbed for nearly a century would not be neutral and would not further the ideals of respect and tolerance embodied in the First Amendment. For all these reasons, the Cross does not offend the Constitution. * * * We reverse the judgment of the Court of Appeals for the Fourth Circuit and remand the cases for further proceedings. It is so ordered. Notes 1 B. Longenecker, The Cross Before Constantine: The Early Life of a Christian Symbol 2 (2015). 2 See Blue Cross, Blue Shield, https://www.bcbs.com; The Bayer Group, The Bayer Cross—Logo and Landmark, https://www.bayer.com/en/logo-history.aspx; Band-Aid Brand Adhesive Bandages, Johnson & Johnson All Purpose First Aid Kit, https://www.band-aid.com/products/first-aid-kits/all-purpose (all Internet materials as last visited June 18, 2019). 3 International Committee of the Red Cross, The History of the Emblems, https://www.icrc.org/en/doc/resources/documents/misc/emblem-history.htm. 4 For example, the Indian and Japanese affiliates of the ICRC and Red Crescent Societies use the symbol of the cross. See Indian Red Cross Society, https://www.indianredcross.org/ircs/index.php; Japanese Red Cross Society, http://www.jrc.or.jp/english /. 5 See “Flag of Switzerland,” Britannica Academic, https://academic.eb.com/levels/collegiate/article/flag-of-Switzerland/93966. 6 The Latin form of the cross “has a longer upright than crossbar. The intersection of the two is usually such that the upper and the two horizontal arms are all of about equal length, but the lower arm is conspicuously longer.” G. Ferguson, Signs & Symbols in Christian Art 294 (1954). See also Webster’s Third New International Dictionary 1276 (1981) (“latin cross, n.”: “a figure of a cross having a long upright shaft and a shorter crossbar traversing it above the middle”). 7 Of the roughly 116,000 casualties the United States suffered in World War I, some 3,500 were Jewish soldiers. J. Fredman & L. Falk, Jews in American Wars 100 (5th ed. 1954). In the congressional hearings involving the appropriate grave markers for those buried abroad, one Representative stated that approximately 1,600 of these Jewish soldiers were buried in overseas graves marked by Stars of David. See Hearings before the Committee on Military Affairs, 68th Cong., 1st Sess., 3 (1924). That would constitute about 5.2% of the 30,973 graves in American World War I cemeteries abroad. See American Battle Monuments Commission (ABMC), World War I Burials and Memorializations, https://www.abmc.gov/node/1273. 8 “In Flanders Fields,” N. Y. Times, Dec. 18, 1921, p. 96. 9 See ABMC, Cemeteries and Memorials, https://www.abmc.gov/cemeteries-memorials. 10 Other World War I memorials that incorporate the cross include the Argonne Cross and the Canadian Cross of Sacrifice in Arlington National Cemetery; the Wayside Cross in Towson, Maryland; the Wayside Cross in New Canaan, Connecticut; the Troop K Georgia Cavalry War Memorial Front in Augusta, Georgia; the Chestnut Hill and Mt. Airy World War Memorial in Philadelphia, Pennsylvania; and the Great War for Democracy Memorial in Waterbury, Connecticut. 11 There is some ambiguity as to whether the American Legion ever owned the land on which the Cross rests. When the Legion took over the Cross, the town of Bladensburg passed a resolution “assign[ing] and grant[ing] to the said Snyder-Farmer Post #3, American Legion, that parcel of ground upon which the cross now stands and that part necessary to complete . . . the park around said cross, to the perpetual care of the Snyder-Farmer Post #3 as long as it is in existence, and should the said Post go out of existence the plot to revert to the Town of Bladensburg, together with the cross and its surroundings.” App. 65. In 1935, a statute authorized the State Roads Commission of Maryland to “investigate the ownership and possessory rights” of the tract surrounding the Cross and to “acquire the same by purchase or condemnation.” Id., at 421. It appears that in 1957, a court determined that it was necessary for the State to condemn the property. Id., at 1377–1379. The State Roads Commission thereafter conveyed the property to the Commission in 1960. Id., at 1380, 1382. To resolve any ambiguities, in 1961, the local American Legion post “transfer[ed] and assign[ed] to [the Commission] all its right, title and interest in and to the Peace Cross, also originally known as the Memorial Cross, and the tract upon which it is located.” Id., at 1387. At least by 1961, then, both the land and the Cross were publicly owned. 12 Of the budgeted $100,000, the Commission had spent only $5,000 as of 2015. The Commission put off additional spending and repairs in light of this lawsuit. Id., at 823. 13 See, e.g., Utah Highway Patrol Assn. v. American Atheists, Inc., 565 U.S. 994, 995 (2011) (Thomas, J., dissenting from denial of certiorari); County of Allegheny v. American Civil Liberties Union, Greater Pittsburgh Chapter, 492 U.S. 573, 655–656 (1989) (Kennedy, J., concurring in judgment in part and dissenting in part); Lamb’s Chapel v. Center Moriches Union Free School Dist., 508 U.S. 384, 398–399 (1993) (Scalia, J., concurring in judgment); Wallace v. Jaffree, 472 U.S. 38, 112 (1985) (Rehnquist, J., dissenting). 14 See, e.g., Green v. Haskell Cty. Bd. of Comm’rs, 574 F.3d 1235, n. 1 (CA10 2009) (Kelly, J., dissenting from denial of rehearing en banc) (discussing the “judicial morass resulting from the Supreme Court’s opinions”); Cooper v. United States Postal Service, 577 F.3d 479, 494 (CA2 2009) (“Lemon is difficult to apply and not a particularly useful test”); Roark v. South Iron R–1 School Dist., 573 F.3d 556, 563 (CA8 2009) (“[T]he Lemon test has had a ‘checkered career’ ”); Skoros v. New York, 437 F.3d 1, 15 (CA2 2006) (government officials “confront a ‘jurisprudence of minutiae’ that leaves them to rely on ‘little more than intuition and a tape measure’ to ensure the constitutionality of public holiday displays” (quoting County of Allegheny, supra, at 674–675 (opinion of Kennedy, J.)); Felix v. Bloomfield, 841 F.3d 848, 864 (CA10 2016) (court “cannot speculate what precise actions a government must take” to comply with the Establishment Clause); Separation of Church and State Comm. v. Eugene, 93 F.3d 617, 627 (CA9 1996) (O’Scannlain, J., concurring in result) (The standards announced by this Court “are not always clear, consistent or coherent”). 15 See McConnell, Religious Freedom at a Crossroads, 59 U. Chi. L. Rev. 115, 118–120 (1992) (describing doctrinal “chaos” Lemon created, allowing the Court to “reach almost any result in almost any case”); Laycock, Towards a General Theory of the Religion Clauses: The Case of Church Labor Relations and the Right to Church Autonomy, 81 Colum. L. Rev. 1373, 1380–1388 (1981) (criticizing the “unstructured expansiveness of the entanglement notion” and the potential that certain constructions of the effects prong may result in “the establishment clause threaten[ing] to swallow the free exercise clause”); Smith, Symbols, Perceptions, and Doctrinal Illusions: Establishment Neutral-ity and the “No Endorsement” Test, 86 Mich. L. Rev. 266, 269 (1987) (criticizing both the Lemon test and the endorsement gloss); Tushnet, Reflections on the Role of Purpose in the Jurisprudence of the Religion Clauses, 27 Wm. & Mary L. Rev. 997, 1004 (1986) (describing cases involving “ ‘deeply ingrained practices’ ” as “not readily susceptible to analysis under the ordinary Lemon approach”); Choper, The Endorsement Test: Its Status and Desirability, 18 J. L. & Politics 499 (2002) (criticizing both Lemon and the endorsement gloss); Paulsen, Religion, Equality, and the Constitution: An Equal Protection Approach to Establishment Clause Adjudication, 61 Notre Dame L. Rev. 311, 315 (1986) (criticizing the Court’s reading of the Establishment Clause as “producing a schizophrenic pattern of decisions”); Marshall, “We Know It When We See It”: The Supreme Court and Establishment, 59 S. Cal. L. Rev. 495, 526 (1986) (explaining that the purpose prong of Lemon, “[t]aken to its logical conclusion . . . suggests that laws which respect free exercise rights . . . are unconstitutional”). 16 While we do not attempt to provide an authoritative taxonomy of the dozens of Establishment Clause cases that the Court has decided since Everson v. Board of Ed. of Ewing, 330 U.S. 1 (1947), most can be divided into six rough categories: (1) religious references or imagery in public monuments, symbols, mottos, displays, and ceremonies, e.g., Lynch v. Donnelly, 465 U.S. 668 (1984); Van Orden v. Perry, 545 U.S. 677 (2005); (2) religious accommodations and exemptions from gener-ally applicable laws, e.g., Cutter v. Wilkinson, 544 U.S. 709 (2005); Corporation of Presiding Bishop of Church of Jesus Christ of Latter-day Saints v. Amos, 483 U.S. 327 (1987); (3) subsidies and tax exemptions, e.g., Walz v. Tax Comm’n of City of New York, 397 U.S. 664 (1970); Zelman v. Simmons-Harris, 536 U.S. 639 (2002); (4) religious expression in public schools, e.g., School Dist. of Abington Township v. Schempp, 374 U.S. 203 (1963); Lee v. Weisman, 505 U.S. 577 (1992); (5) regulation of private religious speech, e.g., Capitol Square Review and Advisory Bd. v. Pinette, 515 U.S. 753 (1995); and (6) state interference with internal church affairs, e.g., Hosanna-Tabor Evangelical Lutheran Church and School v. EEOC, 565 U.S. 171 (2012). A final, miscellaneous category, including cases involving such issues as Sunday closing laws, see McGowan, v. Maryland, 366 U.S. 420 (1961), and church involvement in governmental decisionmaking, see Larkin v. Grendel’s Den, Inc., 459 U.S. 116 (1982); Board of Ed. of Kiryas Joel Village School Dist. v. Grumet, 512 U.S. 687 (1994), might be added. We deal here with an issue that falls into the first category. 17 See Bravin, When Moses’ Laws Run Afoul of the U. S.’s, Get Me Cecil B. deMille—Ten Commandment Memorial Has Novel Defense in Suit, Wall Street Journal, Apr. 18, 2001, p. A1. 18 See D. Davis, The Oxford Handbook of Church and State in the United States 284 (2010). 19 See French Constitution, Art. 1 (proclaiming that France is a “secular . . . Republic”). 20 See Erlanger, What the Notre-Dame Fire Reveals About the Soul of France, N. Y. Times, Apr. 16, 2019. 21 Hinnant, Petrequin, & Ganley, Fire Ravages Soaring Notre Dame Cathedral, Paris Left Aghast, AP News, Apr. 16, 2019. 22 See B. Shearer & B. Shearer, State Names, Seals, Flags, and Symbols: A Historical Guide 17–18 (3d ed. 2002). See also id., at 18 (Connecticut motto: “Qui Tanstulit Sustinet” (“He Who Transplanted Still Sustains”), dating back to the colonial era and adapted from the Book of Psalms 79:3); ibid. (Florida motto: “In God We Trust,” adopted in 1868); id., at 20 (Maryland motto: “Scuto Bonae Volantatis Tuae Coronasti Nos” (“With Favor Wilt Thou Compass Us as with a Shield”), which appeared on the seal adopted in 1876 and comes from Psalms 5:12); id., at 21–22 (Ohio motto: “With God, All Things Are Possible,” adopted in 1959 and taken from Matthew 19:26); id., at 22 (South Dakota motto: “Under God the People Rule,” adopted in 1885); id., at 23 (American Samoa motto: “Samoa—Muamua le Atua” (“Samoa—Let God Be First”), adopted in 1975). 23 The current flag was known and used since at least October 1880, and was officially adopted by the General Assembly in 1904. See History of the Maryland Flag, https://sos.maryland.gov/Pages/Services/Flag-History.aspx. 24 For example, the French Revolution sought to “dechristianize” the nation and thus removed “plate[s], statues and other fittings from places of worship,” destroyed “crosses, bells, shrines and other, ‘external signs of worship,’ ” and altered “personal and place names which had any ecclesiastical connotations to more suitably Revolutionary ones.” Tallett, Dechristianizing France: The Year II and the Revolutionary Experience, in Religion, Society and Politics in France Since 1789, pp. 1–2 (F. Tallett & N. Atkin eds. 1991). 25 See App. 141, 936; M. Sledge, Soldier Dead 67 (2005). 26 Local Memorials Honoring Dr. King, https://www.kingcounty.gov/elected/executive/equity-social-justice/mlk/local-memorials.aspx. 27 The National Statuary Hall Collection, https://www.aoc.gov/the-national-statuary-hall-collection. 28 W. Hutchison, Religious Pluralism in America 20–21 (2003). 29 A. Stokes, 3 Church and State in the United States 130 (1950). 30 Korn, Rabbis, Prayers, and Legislatures, 23 Hebrew Union College Annual, No. 2, pp. 95, 96 (1950). 31 See Lund, The Congressional Chaplaincies, 17 Wm. & Mary Bill of Rights J. 1171, 1204–1205 (2009). See also 160 Cong. Rec. 3853 (2014) (prayer by the Dalai Lama). 32 J. Fredman & L. Falk, Jews in American Wars 100–101 (5th ed. 1954). 33 Fryer & Levitt, Hatred and Profits: Under the Hood of the Ku Klux Klan, 127 Q. J. Econ. 1883 (2012). 34 For example, the South Carolina Holocaust Memorial depicts a large Star of David “ ‘in sacred memory of the six million,’ ”see https://www.onecolumbiasc.com/public-art/south-carolina-holocaust-memorial/, and the Philadelphia Monument to Six Million Jewish Martyrs depicts a burning bush, Torah scrolls, and a blazing men-orah, see https://www.associationforpublicart.org/artwork/monument-to-six-million-jewish-martyrs/. 35 Hedgpeth, “A Very Deep Kind of Patriotism”: Memorial to Honor Native American Veterans Is Coming to the Mall, Washington Post, Mar. 31, 2019.

588.US.2018_18-266

Respondent Christopher Batterton was working on a vessel owned by petitioner Dutra Group when a hatch blew open and injured his hand. Batterton sued Dutra, asserting a variety of claims, including unseaworthiness, and seeking general and punitive damages. Dutra moved to dismiss the claim for punitive damages, arguing that they are not available on claims for unseaworthiness. The District Court denied Dutra’s motion, and the Ninth Circuit affirmed. Held: A plaintiff may not recover punitive damages on a claim of unseaworthiness. Pp. 10–19. (a) This case is governed by Miles v. Apex Marine Corp., 498 U.S. 19, and Atlantic Sounding Co. v. Townsend, 557 U.S. 404. Miles establishes that the Court “should look primarily to . . . legislative enactments for policy guidance” when exercising its inherent common-law authority over maritime and admiralty cases, while recognizing that such statutory remedies may be supplemented to “achieve the uniform vindication” of the policies served by the relevant statutes. 498 U. S., at 27. And in Atlantic Sounding, the Court allowed recovery of punitive damages but justified that departure from the statutory remedial scheme based on the established history of awarding punitive damages for certain maritime torts, including maintenance and cure. 557 U. S., at 413–414. P. 10. (b) The overwhelming historical evidence suggests that punitive damages are not available for unseaworthiness claims. Neither The Rolf, 293 F. 269, nor The Noddleburn, 28 F. 855—on which Batterton relies—contains a relevant discussion of exemplary or punitive damages. And two other cases to which Batterton points—The City of Carlisle, 39 F. 807, and The Troop, 118 F. 769—both involve maintenance and cure, not unseaworthiness, claims. The lack of punitive damages in traditional maritime law cases is practically dispositive. Pp. 11–13. (c) This Court cannot sanction a novel remedy here unless it is required to maintain uniformity with Congress’s clearly expressed policies, particularly those in the Merchant Marine Act of 1920 (Jones Act)—which codified the rights of injured mariners by incorporating the rights provided to railway workers under the Federal Employers’ Liability Act (FELA). Early decisions held that FELA damages were strictly compensatory. See, e.g., American R. Co. of P. R. v. Didricksen, 227 U.S. 145, 149. And the Federal Courts of Appeals have unanimously held that punitive damages are not available under FELA. This Court’s early discussions of the Jones Act followed the same practices, see, e.g., Pacific S. S. Co. v. Peterson, 278 U.S. 130, 135, and lower courts have uniformly held that punitive damages are not available under the Jones Act. Adopting Batterton’s rule would be contrary to Miles’s command that federal courts should seek to promote a “uniform rule applicable to all actions” for the same injury, whether under the Jones Act or the general maritime law. 498 U. S., at 33. Pp. 13–15. (d) Batterton argues that punitive damages are justified on policy grounds or as a regulatory measure. But unseaworthiness in its current strict-liability form is this Court’s own invention and came after passage of the Jones Act, and a claim of unseaworthiness serves as a duplicate and substitute for a Jones Act claim. It would, therefore, exceed the Court’s objectives of pursuing policies found in congressional enactments and promoting uniformity between maritime statutory law and maritime common law to introduce novel remedies contradictory to those provided by Congress in similar areas. Allowing punitive damages on unseaworthiness claims would also create bizarre disparities in the law. First, due to Miles’s holding, which limited recovery to compensatory damages in wrongful-death actions, a mariner could make a claim for punitive damages if he was injured onboard a ship, but his estate would lose the right to seek punitive damages if he died from his injuries. Second, because unseaworthiness claims run against the owner of the vessel, the owner could be liable for punitive damages while the ship’s master or operator—who could be more culpable—would not be liable for such damages under the Jones Act. Finally, allowing punitive damages would place American shippers at a significant competitive disadvantage and discourage foreign-owned vessels from employing American seamen. The maritime doctrine mentioned by Batterton, which encourages special solicitude for the welfare of seamen, has its roots in the paternalistic approach taken toward mariners by 19th century courts and has never been a commandment that maritime law must favor seamen whenever possible. Pp. 15–18. 880 F.3d 1089, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Ginsburg, J., filed a dissenting opinion, in which Breyer and Sotomayor, JJ., joined.

By granting federal courts jurisdiction over maritime and admiralty cases, the Constitution implicitly directs federal courts sitting in admiralty to proceed “in the manner of a common law court.” Exxon Shipping Co. v. Baker, 554 U.S. 471, 489–490 (2008). Thus, where Congress has not prescribed specific rules, federal courts must develop the “amalgam of traditional common-law rules, modifications of those rules, and newly created rules” that forms the general maritime law. East River S. S. Corp. v. Transamerica Delaval Inc., 476 U.S. 858, 864–865 (1986). But maritime law is no longer solely the province of the Federal Judiciary. “Congress and the States have legislated extensively in these areas.” Miles v. Apex Marine Corp., 498 U.S. 19, 27 (1990). When exercising its inherent common-law authority, “an admiralty court should look primarily to these legislative enactments for policy guidance.” Ibid. We may depart from the policies found in the statutory scheme in discrete instances based on long-established history, see, e.g., Atlantic Sounding Co. v. Townsend, 557 U.S. 404, 424–425 (2009), but we do so cautiously in light of Congress’s persistent pursuit of “uniformity in the exercise of admiralty jurisdiction.” Miles, supra, at 26 (quoting Moragne v. States Marine Lines, Inc., 398 U.S. 375, 401 (1970)). This case asks whether a mariner may recover punitive damages on a claim that he was injured as a result of the unseaworthy condition of the vessel. We have twice confronted similar questions in the past several decades, and our holdings in both cases were based on the particular claims involved. In Miles, which concerned a wrongful-death claim under the general maritime law, we held that recovery was limited to pecuniary damages, which did not include loss of society. 498 U. S., at 23. And in Atlantic Sounding, after examining centuries of relevant case law, we held that punitive damages are not categorically barred as part of the award on the traditional maritime claim of maintenance and cure. 557 U. S., at 407. Here, because there is no historical basis for allowing punitive damages in unseaworthiness actions, and in order to promote uniformity with the way courts have applied parallel statutory causes of action, we hold that punitive damages remain unavailable in unseaworthiness actions. I In order to determine the remedies for unseaworthiness, we must consider both the heritage of the cause of action in the common law and its place in the modern statutory framework. A The seaman’s right to recover damages for personal injury on a claim of unseaworthiness originates in the admiralty court decisions of the 19th century. At the time, “seamen led miserable lives.” D. Robertson, S. Friedell, & M. Sturley, Admiralty and Maritime Law in the United States 163 (2d ed. 2008). Maritime law was largely judge-made, and seamen were viewed as “emphatically the wards of the admiralty.” Harden v. Gordon, 11 F. Cas. 480, 485 (No. 6,047) (CC Me. 1823). In that era, the primary responsibility for protecting seamen lay in the courts, which saw mariners as “peculiarly entitled to”—and particularly in need of—judicial protection “against the effects of the superior skill and shrewdness of masters and owners of ships.” Brown v. Lull, 4 F. Cas. 407, 409 (No. 2,018) (CC Mass. 1836) (Story, J.).[1] Courts of admiralty saw it as their duty not to be “confined to the mere dry and positive rules of the common law” but to “act upon the enlarged and liberal jurisprudence of courts of equity; and, in short, so far as their powers extend[ed], they act[ed] as courts of equity.” Ibid. This Court interpreted the Constitution’s grant of admiralty jurisdiction to the Federal Judiciary as “the power to . . . dispose of [a case] as justice may require.” The Resolute, 168 U.S. 437, 439 (1897). Courts used this power to protect seamen from injury primarily through two causes of action. The first, maintenance and cure, has its roots in the medieval and renaissance law codes that form the ancient foundation of maritime common law.[2] The duty of maintenance and cure requires a ship’s master “to provide food, lodging, and medical services to a seaman injured while serving the ship.” Lewis v. Lewis & Clark Marine, Inc., 531 U.S. 438, 441 (2001). This duty, “which arises from the contract of employment, does not rest upon negligence or culpability on the part of the owner or master, nor is it restricted to those cases where the seaman’s employment is the cause of the injury or illness.” Calmar S. S. Corp. v. Taylor, 303 U.S. 525, 527 (1938) (citations omitted). The second claim, unseaworthiness, is a much more recent development and grew out of causes of action unrelated to personal injury. In its earliest forms, an unseaworthiness claim gave sailors under contract to sail on a ship the right to collect their wages even if they had refused to board an unsafe vessel after discovering its condition. See, e.g., Dixon v. The Cyrus, 7 F. Cas. 755, 757 (No. 3,930) (Pa. 1789); Rice v. The Polly & Kitty, 20 F. Cas. 666, 667 (No. 11,754) (Pa. 1789). Similarly, unseaworthiness was a defense to criminal charges against seamen who refused to obey a ship master’s orders. See, e.g., United States v. Nye, 27 F. Cas. 210, 211 (No. 15,906) (CC Mass. 1855); United States v. Ashton, 24 F. Cas. 873, 874–875 (No. 14,470) (CC Mass. 1834). A claim of unseaworthiness could also be asserted by a shipper to recover damages or by an insurer to deny coverage when the poor condition of the ship resulted in damage to or loss of the cargo. See The Caledonia, 157 U.S. 124, 132–136 (1895) (cataloging cases). Only in the latter years of the 19th century did unseaworthiness begin a long and gradual evolution toward remedying personal injury. Courts began to extend the cases about refusals to serve to allow recovery for mariners who were injured because of the unseaworthy condition of the vessel on which they had served.[3] These early cases were sparse, and they generally allowed recovery only when a vessel’s owner failed to exercise due diligence to ensure that the ship left port in a seaworthy condition. See, e.g., The Robert C. McQuillen, 91 F. 685, 686–687 (Conn. 1899); The Lizzie Frank, 31 F. 477, 480 (SD Ala. 1887); The Tammerlane, 47 F. 822, 824 (ND Cal. 1891). Unseaworthiness remained a suspect basis for personal injury claims until 1903, when, in dicta, this Court concluded that “the vessel and her owner are . . . liable to an indemnity for injuries received by seamen in consequence of the unseaworthiness of the ship.” The Osceola, 189 U.S. 158, 175 (1903). Although this was the first recognition of unseaworthiness as a personal injury claim in this Court, we took pains to note that the claim was strictly cabined. Ibid. Some of the limitations on recovery were imported from the common law. The fellow-servant doctrine, in particular, prohibited recovery when an employee suffered an injury due to the negligent act of another employee without negligence on the part of the employer. Ibid.; see, e.g., The Sachem, 42 F. 66 (EDNY 1890) (denying recovery based on fellow-servant doctrine). Because a claimant had to show that he was injured by some aspect of the ship’s condition that rendered the vessel unseaworthy, a claim could not prevail based on “the negligence of the master, or any member of the crew.” [4] The Osceola, supra, at 175; see also The City of Alexandria, 17 F. 390 (SDNY 1883) (no recovery based on negligence that does not render vessel unseaworthy). Instead, a seaman had to show that the owner of the vessel had failed to exercise due diligence in ensuring the ship was in seaworthy condition. See generally Dixon v. United States, 219 F.2d 10, 12–14 (CA2 1955) (Harlan, J.) (cataloging evolution of the claim). B In the early 20th century, then, under “the general maritime law . . . a vessel and her owner . . . were liable to an indemnity for injuries received by a seaman in consequence of the unseaworthiness of the ship and her appliances; but a seaman was not allowed to recover an indemnity for injuries sustained through the negligence of the master or any member of the crew.” Pacific S. S. Co. v. Peterson, 278 U.S. 130, 134 (1928); see also Plamals v. S. S. “Pinar Del Rio,” 277 U.S. 151, 155 (1928) (vessel was not unseaworthy when mate negligently selected defective rope but sound rope was available on board). Because of these severe limitations on recovery, “the seaman’s right to recover damages for injuries caused by unseaworthiness of the ship was an obscure and relatively little used rem- edy.” G. Gilmore & C. Black, The Law of Admiralty §6–38, p. 383 (2d ed. 1975) (Gilmore & Black). Tremendous shifts in mariners’ rights took place between 1920 and 1950. First, during and after the First World War, Congress enacted a series of laws regulating maritime liability culminating in the Merchant Marine Act of 1920, §33, 41Stat. 1007 (Jones Act), which codified the rights of injured mariners and created new statutory claims that were freed from many of the common-law limitations on recovery. The Jones Act provides injured seamen with a cause of action and a right to a jury. 46 U. S. C. §30104. Rather than create a new structure of substantive rights, the Jones Act incorporated the rights provided to railway workers under the Federal Employers’ Liability Act (FELA), 45 U. S. C. §51 et seq. 46 U. S. C. §30104. In the 30 years after the Jones Act’s passage, “the Act was the vehicle for almost all seamen’s personal injury and death actions.” Gilmore & Black §6–20, at 327. But the Jones Act was overtaken in the 1950s by the second fundamental change in personal injury maritime claims—and it was this Court, not Congress, that played the leading role. In a pair of decisions in the late 1940s, the Court transformed the old claim of unseaworthiness, which had demanded only due diligence by the vessel owner, into a strict-liability claim. In Mahnich v. Southern S. S. Co., 321 U.S. 96 (1944), the Court stated that “the exercise of due diligence does not relieve the owner of his obligation” to provide a seaworthy ship and, in the same ruling, held that the fellow-servant doctrine did not provide a defense. Id., at 100, 101. Mahnich’s interpretation of the early cases may have been suspect, see Tetreault 397–398 (Mahnich rests on “startling misstatement” of relevant precedents), but its assertion triggered a sea-change in maritime personal injury. Less than two years later, we affirmed that the duty of seaworthiness was “essentially a species of liability without fault . . . neither limited by conceptions of negligence nor contractual in character. It is a form of absolute duty owing to all within the range of its humanitarian policy.” Seas Shipping Co. v. Sieracki, 328 U.S. 85, 94–95 (1946) (citations omitted). From Mahnich forward, “the decisions of this Court have undeviatingly reflected an understanding that the owner’s duty to furnish a seaworthy ship is absolute and completely independent of his duty under the Jones Act to exercise reasonable care.” Mitchell v. Trawler Racer, Inc., 362 U.S. 539, 549 (1960). As a result of Mahnich and Sieracki, between the 1950s and 1970s “the unseaworthiness count [was] the essential basis for recovery with the Jones Act count preserved merely as a jury-getting device.”[5] Gilmore & Black §6–20, at 327–328. The shifts in plaintiff preferences between Jones Act and unseaworthiness claims were possible because of the significant overlap between the two causes of action. See id., §6–38, at 383. One leading treatise goes so far as to describe the two claims as “alternative ‘grounds’ of recovery for a single cause of action.” 2 R. Force & M. Norris, The Law of Seamen §30:90, p. 30–369 (5th ed. 2003). The two claims are so similar that, immediately after the Jones Act’s passage, we held that plaintiffs could not submit both to a jury. Plamals, supra, at 156–157 (“Seamen may invoke, at their election, the relief accorded by the old rules against the ship, or that provided by the new against the employer. But they may not have the benefit of both”). We no longer require such election. See McAllister v. Magnolia Petroleum Co., 357 U.S. 221, 222, n. 2 (1958). But a plaintiff still cannot duplicate his recovery by collecting full damages on both claims because, “whether or not the seaman’s injuries were occasioned by the unseaworthiness of the vessel or by the negligence of the master or members of the crew, . . . there is but a single wrongful invasion of his primary right of bodily safety and but a single legal wrong.” Peterson, 278 U. S., at 138; see also 2 Force, supra, §§26:73, 30:90. II Christopher Batterton worked as a deckhand and crew member on vessels owned and operated by the Dutra Group. According to Batterton’s complaint, while working on a scow near Newport Beach, California, Batterton was injured when his hand was caught between a bulkhead and a hatch that blew open as a result of unventilated air accumulating and pressurizing within the compartment. Batterton sued Dutra and asserted a variety of claims, including negligence, unseaworthiness, maintenance and cure, and unearned wages. He sought to recover general and punitive damages. Dutra moved to strike Batterton’s claim for punitive damages, arguing that they are not available on claims for unseaworthiness. The District Court denied Dutra’s motion, 2014 WL 12538172 (CD Cal., Dec. 15, 2014), but agreed to certify an interlocutory appeal on the question, 2015 WL 13752889 (CD Cal., Feb. 6, 2015). The Court of Appeals affirmed. 880 F.3d 1089 (CA9 2018). Applying Circuit precedent, see Evich v. Morris, 819 F.2d 256, 258–259 (CA9 1987), the Court of Appeals held that punitive damages are available for unseaworthiness claims. 880 F. 3d, at 1096. This holding reaffirmed a division of authority between the Circuits. Compare McBride v. Estis Well Serv., L. L. C., 768 F.3d 382, 391 (CA5 2014) (en banc) (punitive damages are not recover- able), and Horsley v. Mobil Oil Corp., 15 F.3d 200, 203 (CA1 1994) (same), with Self v. Great Lakes Dredge & Dock Co., 832 F.2d 1540, 1550 (CA11 1987) (“Punitive damages should be available in cases where the shipowner willfully violated the duty to maintain a safe and seaworthy ship . . .”). We granted certiorari to resolve this division. 586 U. S. ___ (2018). III Our resolution of this question is governed by our decisions in Miles and Atlantic Sounding. Miles establishes that we “should look primarily to . . . legislative enactments for policy guidance,” while recognizing that we “may supplement these statutory remedies where doing so would achieve the uniform vindication” of the policies served by the relevant statutes. 498 U. S., at 27. In Atlantic Sounding, we allowed recovery of punitive damages, but we justified our departure from the statutory remedial scheme based on the established history of awarding punitive damages for certain maritime torts, including maintenance and cure. 557 U. S., at 411–414 (discussing cases of piracy and maintenance and cure awarding damages with punitive components). We were explicit that our decision represented a gloss on Miles rather than a departure from it. Atlantic Sounding, supra, at 420 (“The reasoning of Miles remains sound”). And we recognized the importance of viewing each claim in its proper historical context. “ ‘[R]emedies for negligence, unseaworthiness, and maintenance and cure have different origins and may on occasion call for application of slightly different principles and procedures.’ ” 557 U. S., at 423. In accordance with these decisions, we consider here whether punitive damages have traditionally been awarded for claims of unseaworthiness and whether conformity with parallel statutory schemes would require such damages. Finally, we consider whether we are compelled on policy grounds to allow punitive damages for unseaworthiness claims. A For claims of unseaworthiness, the overwhelming historical evidence suggests that punitive damages are not available. Batterton principally relies on two cases to establish that punitive damages were traditionally avail- able for breach of the duty of seaworthiness. Upon close inspection, neither supports this argument. The Rolph, 293 F. 269, 271 (ND Cal. 1923), involved a mate who brutally beat members of the crew, rendering one seaman blind and leaving another with impaired hearing. The central question in the case was not the form of damages, but rather whether the viciousness of the mate rendered the vessel unseaworthy. The Rolph, 299 F. 52, 54 (CA9 1924). The court concluded that the master, by staffing the vessel with such an unsuitable officer, had rendered it unseaworthy. Id., at 55. To the extent the court described the basis for the damages awarded, it explained that the judgment was supported by testimony as to “the expectation of life and earnings of these men.” 293 F., at 272. And the Court of Appeals discussed only the seamen’s entitlement “to recover an indemnity” for their injuries. 299 F., at 56. These are discussions of compensatory damages—nowhere does the court speak in terms of an exemplary or punitive award.[6] The Noddleburn, 28 F. 855, 857–858 (Ore. 1886), involved an injury to a British seaman serving on a British vessel and was decided under English law. The plaintiff in the case was injured when he fell to the deck after being ordered aloft and stepping on an inadequately secured line. Id., at 855. After the injury, the master neglected the man’s wounds, thinking the injury a mere sprain. Id., at 856. The leg failed to heal and the man had to insist on being discharged to a hospital, where he learned that he would be permanently disabled. Ibid. As damages, the court awarded him accrued wages, as well as $1,000 to compensate for the loss in future earnings from his dis- ability and $500 for his pain and suffering. Id., at 860. But these are purely compensatory awards—the only discussion of exemplary damages comes at the very close of the opinion, and it is clear that they were considered because of the master’s failure to provide maintenance and cure. Ibid. (discussing additional award “in consideration of the neglect and indifference with which the libelant was treated by the master after his injury” (emphasis added)). Finally, Batterton points to two other cases, The City of Carlisle, 39 F. 807 (Ore. 1889), and The Troop, 118 F. 769 (Wash. 1902). But these cases, like The Noddleburn, both involve maintenance and cure claims that rest on the willful failure of the master and mate to provide proper care for wounded sailors after they were injured. 39 F., at 812 (“master failed and neglected to procure or provide any medical aid or advice . . . and was contriving and intending to get rid of him as easily as possible”); 118 F., at 771 (assessing damages based on provision of Laws of Oleron requiring maintenance). Batterton characterizes these as unseaworthiness actions on the theory that the seamen could have pursued that claim. But, because courts award damages for the claims a plaintiff actually pleads rather than those he could have brought, these cases are irrelevant. The lack of punitive damages in traditional maritime law cases is practically dispositive. By the time the claim of unseaworthiness evolved to remedy personal injury, punitive damages were a well-established part of the common law. Exxon Shipping, 554 U. S., at 491. American courts had awarded punitive (or exemplary) damages from the Republic’s earliest days. See, e.g., Genay v. Norris, 1 S. C. L. 6, 7 (1784); Coryell v. Colbaugh, 1 N. J. L. 77, 78 (1791). And yet, beyond the decisions discussed above, Batterton presents no decisions from the formative years of the personal injury unseaworthiness claim in which exemplary damages were awarded. From this we conclude that, unlike maintenance and cure, unseawor- thiness did not traditionally allow recovery of punitive damages. B In light of this overwhelming historical evidence, we cannot sanction a novel remedy here unless it is required to maintain uniformity with Congress’s clearly expressed policies. Therefore, we must consider the remedies typically recognized for Jones Act claims. The Jones Act adopts the remedial provisions of FELA, and by the time of the Jones Act’s passage, this Court and others had repeatedly interpreted the scope of damages available to FELA plaintiffs. These early decisions held that “[t]he damages recoverable [under FELA] are limited . . . strictly to the financial loss . . . sustained.”[7] American R. Co. of P. R. v. Didricksen, 227 U.S. 145, 149 (1913); see also Gulf, C. & S. F. R. Co. v. McGinnis, 228 U.S. 173, 175 (1913) (FELA is construed “only to compensate . . . for the actual pecuniary loss resulting” from the worker’s injury or death); Michigan Central R. Co. v. Vreeland, 227 U.S. 59, 68 (1913) (FELA imposes “a liability for the pecuniary damage resulting to [the worker] and for that only”). In one particularly illuminating case, in deciding whether a complaint alleged a claim under FELA or state law, the Court observed that if the complaint “were read as manifestly demanding exemplary damages, that would point to the state law.” Seaboard Air Line R. Co. v. Koennecke, 239 U.S. 352, 354 (1915). And in the years since, Federal Courts of Appeals have unanimously held that punitive damages are not available under FELA. Miller v. American President Lines, Ltd., 989 F.2d 1450, 1457 (CA6 1993); Wildman v. Burlington No. R. Co., 825 F.2d 1392, 1395 (CA9 1987); Kozar v. Chesapeake & Ohio R. Co., 449 F.2d 1238, 1243 (CA6 1971). Our early discussions of the Jones Act followed the same practices. We described the Act shortly after its passage as creating “an action for compensatory damages, on the ground of negligence.”[8] Peterson, 278 U. S., at 135. And we have more recently observed that the Jones Act “limits recovery to pecuniary loss.” Miles, 498 U. S., at 32. Looking to FELA and these decisions, the Federal Courts of Appeals have uniformly held that punitive damages are not available under the Jones Act. McBride, 768 F. 3d, at 388 (“[N]o cases have awarded punitive damages under the Jones Act”); Guevara v. Maritime Overseas Corp., 59 F.3d 1496, 1507, n. 9 (CA5 1995) (en banc); Horsley, 15 F. 3d, at 203; Miller, supra, at 1457 (“Punitive damages are not . . . recoverable under the Jones Act”); Kopczynski v. The Jacqueline, 742 F.2d 555, 560 (CA9 1984). Batterton argues that these cases are either inapposite or wrong, but because of the absence of historical evidence to support punitive damages—evidence that was central to our decision in Atlantic Sounding—we need not reopen this question of statutory interpretation. It is enough for us to note the general consensus that exists in the lower courts and to observe that the position of those courts conforms with the discussion and holding in Miles. Adopting the rule urged by Batterton would be contrary to Miles’s command that federal courts should seek to promote a “uniform rule applicable to all actions” for the same injury, whether under the Jones Act or the general maritime law. 498 U. S., at 33. C To the extent Batterton argues that punitive damages are justified on policy grounds or as a regulatory measure, we are unpersuaded. In contemporary maritime law, our overriding objective is to pursue the policy expressed in congressional enactments, and because unseaworthiness in its current strict-liability form is our own invention and came after passage of the Jones Act, it would exceed our current role to introduce novel remedies contradictory to those Congress has provided in similar areas. See id., at 36 (declining to create remedy “that goes well beyond the limits of Congress’ ordered system of recovery”). We are particularly loath to impose more expansive liabilities on a claim governed by strict liability than Congress has imposed for comparable claims based in negligence. Ibid. And with the increased role that legislation has taken over the past century of maritime law, we think it wise to leave to the political branches the development of novel claims and remedies. We are also wary to depart from the practice under the Jones Act because a claim of unseaworthiness—more than a claim for maintenance and cure—serves as a duplicate and substitute for a Jones Act claim. The duty of maintenance and cure requires the master to provide medical care and wages to an injured mariner in the period after the injury has occurred. Calmar S. S. Corp., 303 U. S., at 527–528. By contrast, both the Jones Act and unseaworthiness claims compensate for the injury itself and for the losses resulting from the injury. Peterson, supra, at 138. In such circumstances, we are particularly mindful of the rule that requires us to promote uniformity between maritime statutory law and maritime common law.[9] See Miles, supra, at 27. See also Mobil Oil Corp. v. Higginbotham, 436 U.S. 618, 625 (1978) (declining to recognize loss-of-society damages under general maritime law because that would “rewrit[e the] rules that Congress has affirmatively and specifically enacted”). Unlike a claim of maintenance and cure, which addresses a situation where the vessel owner and master have “just about every economic incentive to dump an injured seaman in a port and abandon him to his fate,” in the unseaworthiness context the interests of the owner and mariner are more closely aligned. McBride, supra, at 394, n. 12 (Clement, J., concurring). That is because there are significant economic incentives prompting owners to ensure that their vessels are seaworthy. Most obviously, an owner who puts an unseaworthy ship to sea stands to lose the ship and the cargo that it carries. And if a vessel’s unseaworthiness threatens the crew or cargo, the owner risks losing the protection of his insurer (who may not cover losses incurred by the owner’s negligence) and the work of the crew (who may refuse to serve on an unseaworthy vessel). In some instances, the vessel owner may even face criminal penalties. See, e.g., 46 U. S. C. §10908. Allowing punitive damages on unseaworthiness claims would also create bizarre disparities in the law. First, due to our holding in Miles, which limited recovery to compensatory damages in wrongful-death actions, a mariner could make a claim for punitive damages if he was injured onboard a ship, but his estate would lose the right to seek punitive damages if he died from his injuries. Second, because unseaworthiness claims run against the owner of the vessel, the ship’s owner could be liable for punitive damages while the master or operator of the ship—who has more control over onboard conditions and is best positioned to minimize potential risks—would not be liable for such damages under the Jones Act. See Sieracki, 328 U. S., at 100 (The duty of seaworthiness is “peculiarly and exclusively the obligation of the owner. It is one he cannot delegate”). Finally, because “[n]oncompensatory damages are not part of the civil-code tradition and thus unavailable in such countries,” Exxon Shipping, 554 U. S., at 497, allowing punitive damages would place American shippers at a significant competitive disadvantage and would discourage foreign-owned vessels from employing American seamen. See Gotanda, Punitive Damages: A Comparative Analysis, 42 Colum. J. Transnat’l L. 391, 396, n. 24 (2004) (listing civil-law nations that restrict private plaintiffs to compensatory damages). This would frustrate another “fundamental interest” served by federal maritime jurisdiction: “the protection of maritime commerce.” Norfolk Southern R. Co. v. James N. Kirby, Pty Ltd., 543 U.S. 14, 25 (2004) (internal quotation marks omitted; emphasis deleted). Against this, Batterton points to the maritime doctrine that encourages special solicitude for the welfare of seamen. But that doctrine has its roots in the paternalistic approach taken toward mariners by 19th century courts. See, e.g., Harden, 11 F. Cas., at 485; Brown, 4 F. Cas., at 409. The doctrine has never been a commandment that maritime law must favor seamen whenever possible. Indeed, the doctrine’s apex coincided with many of the harsh common-law limitations on recovery that were not set aside until the passage of the Jones Act. And, while sailors today face hardships not encountered by those who work on land, neither are they as isolated nor as dependent on the master as their predecessors from the age of sail. In light of these changes and of the roles now played by the Judiciary and the political branches in protecting sailors, the special solicitude to sailors has only a small role to play in contemporary maritime law. It is not sufficient to overcome the weight of authority indicating that punitive damages are unavailable. IV Punitive damages are not a traditional remedy for unseaworthiness. The rule of Miles—promoting uniformity in maritime law and deference to the policies expressed in the statutes governing maritime law—prevents us from recognizing a new entitlement to punitive damages where none previously existed. We hold that a plaintiff may not recover punitive damages on a claim of unseaworthiness. We reverse the judgment of the United States Court of Appeals for the Ninth Circuit and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 Riding circuit, Justice Story described mariners in markedly paternalistic terms: “Seamen are a class of persons remarkable for their rashness, thoughtlessness and improvidence. They are generally necessitous, ignorant of the nature and extent of their own rights and privileges, and for the most part incapable of duly appreciating their value. They combine, in a singular manner, the apparent anomalies of gallantry, extravagance, profusion in expenditure, indifference to the future, credulity, which is easily won, and confidence, which is readily surprised.” Brown, 4 F. Cas., at 409. 2 A right resembling maintenance and cure appears in the Laws of Oleron, promulgated by Eleanor of Aquitaine around 1160, in the 13th-century Laws of Wisbuy, in the Laws of the Hanse Towns, published in 1597, and in the Marine Ordinances of Louis XIV, published in 1681. See 30 F. Cas. 1169 (collecting sources). The relevant passages are the Laws of Oleron, Arts. VI and VII, 30 F. Cas., at 1174–1175; the Laws of Wisbuy, Arts. XVIII, XIX, and XXXIII, 30 F. Cas., at 1191–1192; the Laws of the Hanse Towns, Arts. XXXIX and XLV, 30 F. Cas., at 1200; the Marine Ordinances of Louis XIV, Tit. IV, Arts. XI and XII, 30 F. Cas., at 1209. 3 Most of these cases allowed recovery for personal injury in “erroneous reliance” on certain passages in Dixon v. The Cyrus, 7 F. Cas. 755 (No. 3,930) (Pa. 1789). Tetreault, Seamen, Seaworthiness, and the Rights of Harbor Workers, 39 Cornell L. Q. 381, 390 (1954) (Tetreault). These cases misread The Cyrus as resting on an implied warranty of seaworthiness. Tetreault 390. But The Cyrus is more fairly read to turn on a theory of true implied condition. While a warranty would provide a basis for damages if the breach caused an injury, an implied condition would only allow the mariner to escape performance without surrendering the benefit of the contract. In other words, “[t]he manifest unseaworthiness of the vessel at the commencement of the voyage would excuse non-performance by the mariners but did not constitute a basis for damages.” Tetreault 390. 4 To be sure, in some instances the concept of “unseaworthiness” expanded to embrace conditions that resulted from the negligence of fellow servants, see, e.g., Carlisle Packing Co. v. Sandanger, 259 U.S. 255, 259 (1922) (vessel was rendered unseaworthy when it left port with gasoline in a container labeled “coal oil”); see also G. Robinson, Handbook of Admiralty Law in the United States §37, p. 305–307 (1st ed. 1939) (collecting cases). But it was only after the passage of the Jones Act that negligence by a fellow mariner provided a reliable basis for recovery. See Part I–B, infra. 5 The decline of Jones Act claims was arrested, although not reversed, by our holding that some negligent actions on a vessel may create Jones Act liability without rendering the vessel unseaworthy. See Usner v. Luckenbach Overseas Corp., 400 U.S. 494 (1971); see also 1B Benedict on Admiralty §23, p. 3–35 (7th rev. ed. 2018). 6 Even if this case did involve a sub silentio punitive award, we share the Fifth Circuit’s reluctance to “rely on one dust-covered case to establish that punitive damages were generally available in unseaworthiness cases.” McBride v. Estis Well Serv., L. L. C., 768 F.3d 382, 397 (2014) (Clement, J., concurring). Absent a clear historical pattern, Miles v. Apex Marine Corp., 498 U.S. 19 (1990), commands us to seek conformity with the policy preferences the political branches have expressed in legislation. 7 Treatises from the same period lend further support to the view that “in all actions under [FELA], an award of exemplary damages is not permitted.” 2 M. Roberts, Federal Liabilities of Carriers §621, p. 1093 (1918); 1 id., §417, at 708; 5 J. Berryman, Sutherland on Damages §1333, p. 5102 (4th ed. 1916) (FELA “provid[es] compensation for pecuniary loss or damage only”). 8 We also note that Congress declined to allow punitive damages when it enacted the Death on the High Seas Act. 46 U. S. C. §30303 (allowing “fair compensation for the pecuniary loss sustained” for a death on the high seas). 9 The dissent, post at 9, and n. 7 (opinion of Ginsburg, J.), suggests that because of the existing differences between a Jones Act claim and an unseaworthiness claim, recognizing punitive damages would not be a cause of disparity. But, as the dissent acknowledges, much of the expanded reach of the modern unseaworthiness doctrine can be attributed to innovations made by this Court following the enactment of the Jones Act. See post at 8, and n. 6; supra, at 7–8. Although Batterton and the dissent would continue this evolution by recognizing damages previously unavailable, Miles dictates that such innovation is the prerogative of the political branches, our past expansion of the unseaworthiness doctrine notwithstanding. Of course, Miles recognized that the general maritime law need not be static. For example, our decision in Moragne v. States Marine Lines, Inc., 398 U.S. 375 (1970), smoothed a disjunction created by the imperfect alignment of statutory claims with past decisions limiting maritime claims for wrongful death. But when there is no disjunction—as here, where traditional remedies align with modern statutory remedies—we are unwilling to endorse doctrinal changes absent legislative changes.

586.US.2018_17-1091

Tyson Timbs pleaded guilty in Indiana state court to dealing in a controlled substance and conspiracy to commit theft. At the time of Timbs’s arrest, the police seized a Land Rover SUV Timbs had purchased for $42,000 with money he received from an insurance policy when his father died. The State sought civil forfeiture of Timbs’s vehicle, charging that the SUV had been used to transport heroin. Observing that Timbs had recently purchased the vehicle for more than four times the maximum $10,000 monetary fine assessable against him for his drug conviction, the trial court denied the State’s request. The vehicle’s forfeiture, the court determined, would be grossly disproportionate to the gravity of Timbs’s offense, and therefore unconstitutional under the Eighth Amendment’s Excessive Fines Clause. The Court of Appeals of Indiana affirmed, but the Indiana Supreme Court reversed, holding that the Excessive Fines Clause constrains only federal action and is inapplicable to state impositions. Held: The Eighth Amendment’s Excessive Fines Clause is an incorporated protection applicable to the States under the Fourteenth Amendment’s Due Process Clause. Pp. 2–9. (a) The Fourteenth Amendment’s Due Process Clause incorporates and renders applicable to the States Bill of Rights protections “fundamental to our scheme of ordered liberty,” or “deeply rooted in this Nation’s history and tradition.” McDonald v. Chicago, 561 U.S. 742, 767 (alterations omitted). If a Bill of Rights protection is incorporated, there is no daylight between the federal and state conduct it prohibits or requires. Pp. 2–3. (b) The prohibition embodied in the Excessive Fines Clause carries forward protections found in sources from Magna Carta to the English Bill of Rights to state constitutions from the colonial era to the present day. Protection against excessive fines has been a constant shield throughout Anglo-American history for good reason: Such fines undermine other liberties. They can be used, e.g., to retaliate against or chill the speech of political enemies. They can also be employed, not in service of penal purposes, but as a source of revenue. The historical and logical case for concluding that the Fourteenth Amendment incorporates the Excessive Fines Clause is indeed overwhelming. Pp. 3–7. (c) Indiana argues that the Clause does not apply to its use of civil in rem forfeitures, but this Court held in Austin v. United States, 509 U.S. 602, that such forfeitures fall within the Clause’s protection when they are at least partially punitive. Indiana cannot prevail unless the Court overrules Austin or holds that, in light of Austin, the Excessive Fines Clause is not incorporated because its application to civil in rem forfeitures is neither fundamental nor deeply rooted. The first argument, overturning Austin, is not properly before this Court. The Indiana Supreme Court held only that the Excessive Fines Clause did not apply to the States. The court did not address the Clause’s application to civil in rem forfeitures, nor did the State ask it to do so. Timbs thus sought this Court’s review only of the question whether the Excessive Fines Clause is incorporated by the Fourteenth Amendment. Indiana attempted to reformulate the question to ask whether the Clause restricted States’ use of civil in rem forfeitures and argued on the merits that Austin was wrongly decided. Respondents’ “right, . . . to restate the questions presented,” however, “does not give them the power to expand [those] questions,” Bray v. Alexandria Women’s Health Clinic, 506 U.S. 263, 279, n. 10 (emphasis deleted), particularly where the proposed reformulation would lead the Court to address a question neither pressed nor passed upon below, cf. Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7. The second argument, that the Excessive Fines Clause cannot be incorporated if it applies to civil in rem forfeitures, misapprehends the nature of the incorporation inquiry. In considering whether the Fourteenth Amendment incorporates a Bill of Rights protection, this Court asks whether the right guaranteed—not each and every particular application of that right—is fundamental or deeply rooted. To suggest otherwise is inconsistent with the approach taken in cases concerning novel applications of rights already deemed incorporated. See, e.g., Packingham v. North Carolina, 582 U. S. ___, ___. The Excessive Fines Clause is thus incorporated regardless of whether application of the Clause to civil in rem forfeitures is itself fundamental or deeply rooted. Pp. 7–9. 84 N. E. 3d 1179, vacated and remanded. Ginsburg, J., delivered the opinion of the Court, in which Roberts, C. J., and Breyer, Alito, Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Gorsuch, J., filed a concurring opinion. Thomas, J., filed an opinion concurring in the judgment.

Tyson Timbs pleaded guilty in Indiana state court to dealing in a controlled substance and conspiracy to commit theft. The trial court sentenced him to one year of home detention and five years of probation, which included a court-supervised addiction-treatment program. The sentence also required Timbs to pay fees and costs totaling $1,203. At the time of Timbs’s arrest, the police seized his vehicle, a Land Rover SUV Timbs had purchased for about $42,000. Timbs paid for the vehicle with money he received from an insurance policy when his father died. The State engaged a private law firm to bring a civil suit for forfeiture of Timbs’s Land Rover, charging that the vehicle had been used to transport heroin. After Timbs’s guilty plea in the criminal case, the trial court held a hearing on the forfeiture demand. Although finding that Timbs’s vehicle had been used to facilitate violation of a criminal statute, the court denied the requested forfeiture, observing that Timbs had recently purchased the vehicle for $42,000, more than four times the maximum $10,000 monetary fine assessable against him for his drug conviction. Forfeiture of the Land Rover, the court determined, would be grossly disproportionate to the gravity of Timbs’s offense, hence unconstitutional under the Eighth Amendment’s Excessive Fines Clause. The Court of Appeals of Indiana affirmed that determination, but the Indiana Supreme Court reversed. 84 N. E. 3d 1179 (2017). The Indiana Supreme Court did not decide whether the forfeit-ure would be excessive. Instead, it held that the Exces- sive Fines Clause constrains only federal action and is inapplicable to state impositions. We granted certiorari. 585 U. S. __ (2018). The question presented: Is the Eighth Amendment’s Excessive Fines Clause an “incorporated” protection applicable to the States under the Fourteenth Amendment’s Due Process Clause? Like the Eighth Amendment’s proscriptions of “cruel and unusual punishment” and “[e]xcessive bail,” the protection against excessive fines guards against abuses of government’s punitive or criminal-law-enforcement authority. This safeguard, we hold, is “fundamental to our scheme of ordered liberty,” with “dee[p] root[s] in [our] history and tradition.” McDonald v. Chicago, 561 U.S. 742, 767 (2010) (internal quotation marks omitted; emphasis deleted). The Excessive Fines Clause is therefore incorporated by the Due Process Clause of the Fourteenth Amendment. I A When ratified in 1791, the Bill of Rights applied only to the Federal Government. Barron ex rel. Tiernan v. Mayor of Baltimore, 7 Pet. 243 (1833). “The constitutional Amendments adopted in the aftermath of the Civil War,” however, “fundamentally altered our country’s federal system.” McDonald, 561 U. S., at 754. With only “a handful” of exceptions, this Court has held that the Fourteenth Amendment’s Due Process Clause incorporates the protections contained in the Bill of Rights, rendering them applicable to the States. Id., at 764–765, and nn. 12–13. A Bill of Rights protection is incorporated, we have explained, if it is “fundamental to our scheme of ordered liberty,” or “deeply rooted in this Nation’s history and tradition.” Id., at 767 (internal quotation marks omitted; emphasis deleted). Incorporated Bill of Rights guarantees are “enforced against the States under the Fourteenth Amendment according to the same standards that protect those personal rights against federal encroachment.” Id., at 765 (internal quotation marks omitted). Thus, if a Bill of Rights protection is incorporated, there is no daylight between the federal and state conduct it prohibits or requires.[1] B Under the Eighth Amendment, “[e]xcessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.” Taken together, these Clauses place “parallel limitations” on “the power of those entrusted with the criminal-law function of government.” Browning-Ferris Industries of Vt., Inc. v. Kelco Disposal, Inc., 492 U.S. 257, 263 (1989) (quoting Ingraham v. Wright, 430 U.S. 651, 664 (1977)). Directly at issue here is the phrase “nor excessive fines imposed,” which “limits the government’s power to extract payments, whether in cash or in kind, ‘as punishment for some offense.’ ” United States v. Bajakajian, 524 U.S. 321, 327–328 (1998) (quoting Austin v. United States, 509 U.S. 602, 609–610 (1993)). The Fourteenth Amendment, we hold, incorporates this protection. The Excessive Fines Clause traces its venerable lineage back to at least 1215, when Magna Carta guaranteed that “[a] Free-man shall not be amerced for a small fault, but after the manner of the fault; and for a great fault after the greatness thereof, saving to him his contenement . . . .” §20, 9 Hen. III, ch. 14, in 1 Eng. Stat. at Large 5 (1225).[2] As relevant here, Magna Carta required that economic sanctions “be proportioned to the wrong” and “not be so large as to deprive [an offender] of his livelihood.” Browning-Ferris, 492 U. S., at 271. See also 4 W. Blackstone, Commentaries on the Laws of England 372 (1769) (“[N]o man shall have a larger amercement imposed upon him, than his circumstances or personal estate will bear . . . .”). But cf. Bajakajian, 524 U. S., at 340, n. 15 (taking no position on the question whether a person’s income and wealth are relevant considerations in judging the excessiveness of a fine). Despite Magna Carta, imposition of excessive fines persisted. The 17th century Stuart kings, in particular, were criticized for using large fines to raise revenue, harass their political foes, and indefinitely detain those un-able to pay. E.g., The Grand Remonstrance ¶¶17, 34 (1641), in The Constitutional Documents of the Puritan Revolution 1625–1660, pp. 210, 212 (S. Gardiner ed., 3d ed. rev. 1906); Browning-Ferris, 492 U. S., at 267. When James II was overthrown in the Glorious Revolution, the attendant English Bill of Rights reaffirmed Magna Carta’s guarantee by providing that “excessive Bail ought not to be required, nor excessive Fines imposed; nor cruel and unusual Punishments inflicted.” 1 Wm. & Mary, ch. 2, §10, in 3 Eng. Stat. at Large 441 (1689). Across the Atlantic, this familiar language was adopted almost verbatim, first in the Virginia Declaration of Rights, then in the Eighth Amendment, which states: “Excessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.” Adoption of the Excessive Fines Clause was in tune not only with English law; the Clause resonated as well with similar colonial-era provisions. See, e.g., Pa. Frame of Govt., Laws Agreed Upon in England, Art. XVIII (1682), in 5 Federal and State Constitutions 3061 (F. Thorpe ed. 1909) (“[A]ll fines shall be moderate, and saving men’s contenements, merchandize, or wainage.”). In 1787, the constitutions of eight States—accounting for 70% of the U. S. population—forbade excessive fines. Calabresi, Agudo, & Dore, State Bills of Rights in 1787 and 1791, 85 S. Cal. L. Rev. 1451, 1517 (2012). An even broader consensus obtained in 1868 upon ratification of the Fourteenth Amendment. By then, the constitutions of 35 of the 37 States—accounting for over 90% of the U. S. population—expressly prohibited excessive fines. Calabresi & Agudo, Individual Rights Under State Constitutions When the Fourteenth Amendment Was Ratified in 1868, 87 Texas L. Rev. 7, 82 (2008). Notwithstanding the States’ apparent agreement that the right guaranteed by the Excessive Fines Clause was fundamental, abuses continued. Following the Civil War, Southern States enacted Black Codes to subjugate newly freed slaves and maintain the prewar racial hierarchy. Among these laws’ provisions were draconian fines for violating broad proscriptions on “vagrancy” and other dubious offenses. See, e.g., Mississippi Vagrant Law, Laws of Miss. §2 (1865), in 1 W. Fleming, Documentary History of Reconstruction 283–285 (1950). When newly freed slaves were unable to pay imposed fines, States often demanded involuntary labor instead. E.g., id. §5; see Finkelman, John Bingham and the Background to the Fourteenth Amendment, 36 Akron L. Rev 671, 681–685 (2003) (describing Black Codes’ use of fines and other methods to “replicate, as much as possible, a system of involuntary servitude”). Congressional debates over the Civil Rights Act of 1866, the joint resolution that became the Fourteenth Amendment, and similar measures repeatedly mentioned the use of fines to coerce involuntary labor. See, e.g., Cong. Globe, 39th Cong., 1st Sess., 443 (1866); id., at 1123–1124. Today, acknowledgment of the right’s fundamental nature remains widespread. As Indiana itself reports, all 50 States have a constitutional provision prohibiting the imposition of excessive fines either directly or by requiring proportionality. Brief in Opposition 8–9. Indeed, Indiana explains that its own Supreme Court has held that the Indiana Constitution should be interpreted to impose the same restrictions as the Eighth Amendment. Id., at 9 (citing Norris v. State, 271 Ind. 568, 576, 394 N.E.2d 144, 150 (1979)). For good reason, the protection against excessive fines has been a constant shield throughout Anglo-American history: Exorbitant tolls undermine other constitutional liberties. Excessive fines can be used, for example, to retaliate against or chill the speech of political enemies, as the Stuarts’ critics learned several centuries ago. See Browning-Ferris, 492 U. S., at 267. Even absent a political motive, fines may be employed “in a measure out of accord with the penal goals of retribution and deterrence,” for “fines are a source of revenue,” while other forms of punishment “cost a State money.” Harmelin v. Michigan, 501 U.S. 957, 979, n. 9 (1991) (opinion of Scalia, J.) (“it makes sense to scrutinize governmental action more closely when the State stands to benefit”). This concern is scarcely hypothetical. See Brief for American Civil Liberties Union et al. as Amici Curiae 7 (“Perhaps because they are politically easier to impose than generally applicable taxes, state and local governments nationwide increasingly depend heavily on fines and fees as a source of general revenue.”). In short, the historical and logical case for concluding that the Fourteenth Amendment incorporates the Excessive Fines Clause is overwhelming. Protection against excessive punitive economic sanctions secured by the Clause is, to repeat, both “fundamental to our scheme of ordered liberty” and “deeply rooted in this Nation’s history and tradition.” McDonald, 561 U. S., at 767 (internal quotation marks omitted; emphasis deleted). II The State of Indiana does not meaningfully challenge the case for incorporating the Excessive Fines Clause as a general matter. Instead, the State argues that the Clause does not apply to its use of civil in rem forfeitures because, the State says, the Clause’s specific application to such forfeitures is neither fundamental nor deeply rooted. In Austin v. United States, 509 U.S. 602 (1993), however, this Court held that civil in rem forfeitures fall within the Clause’s protection when they are at least partially punitive. Austin arose in the federal context. But when a Bill of Rights protection is incorporated, the protection applies “identically to both the Federal Government and the States.” McDonald, 561 U. S., at 766, n. 14. Accordingly, to prevail, Indiana must persuade us either to overrule our decision in Austin or to hold that, in light of Austin, the Excessive Fines Clause is not incorporated because the Clause’s application to civil in rem forfeitures is neither fundamental nor deeply rooted. The first argument is not properly before us, and the second misapprehends the nature of our incorporation inquiry. A In the Indiana Supreme Court, the State argued that forfeiture of Timbs’s SUV would not be excessive. See Brief in Opposition 5. It never argued, however, that civil in rem forfeitures were categorically beyond the reach of the Excessive Fines Clause. The Indiana Supreme Court, for its part, held that the Clause did not apply to the States at all, and it nowhere addressed the Clause’s application to civil in rem forfeitures. See 84 N. E. 3d 1179. Accordingly, Timbs sought our review of the question “[w]hether the Eighth Amendment’s Excessive Fines Clause is incorporated against the States under the Fourteenth Amendment.” Pet. for Cert. i. In opposing review, Indiana attempted to reformulate the question to ask “[w]hether the Eighth Amendment’s Excessive Fines Clause restricts States’ use of civil asset forfeitures.” Brief in Opposition i. And on the merits, Indiana has argued not only that the Clause is not incorporated, but also that Austin was wrongly decided. Respondents’ “right, in their brief in opposition, to restate the questions presented,” however, “does not give them the power to expand [those] questions.” Bray v. Alexandria Women’s Health Clinic, 506 U.S. 263, 279, n. 10 (1993) (emphasis deleted). That is particularly the case where, as here, a respondent’s reformulation would lead us to address a question neither pressed nor passed upon below. Cf. Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7 (2005) (“[W]e are a court of review, not of first view . . . .”). We thus decline the State’s invitation to reconsider our unanimous judgment in Austin that civil in rem forfeitures are fines for purposes of the Eighth Amendment when they are at least partially punitive. B As a fallback, Indiana argues that the Excessive Fines Clause cannot be incorporated if it applies to civil in rem forfeitures. We disagree. In considering whether the Fourteenth Amendment incorporates a protection contained in the Bill of Rights, we ask whether the right guaranteed—not each and every particular application of that right—is fundamental or deeply rooted. Indiana’s suggestion to the contrary is inconsistent with the approach we have taken in cases concerning novel applications of rights already deemed incorporated. For example, in Packingham v. North Carolina, 582 U. S. ___ (2017), we held that a North Carolina statute prohibiting registered sex offenders from accessing certain commonplace social media websites violated the First Amendment right to freedom of speech. In reaching this conclusion, we noted that the First Amendment’s Free Speech Clause was “applicable to the States under the Due Process Clause of the Fourteenth Amendment.” Id., at ___ (slip op., at 1). We did not, however, inquire whether the Free Speech Clause’s application specifically to social media websites was fundamental or deeply rooted. See also, e.g., Riley v. California, 573 U.S. 373 (2014) (holding, without separately considering incorporation, that States’ warrantless search of digital information stored on cell phones ordinarily violates the Fourth Amendment). Similarly here, regardless of whether application of the Excessive Fines Clause to civil in rem forfeitures is itself fundamental or deeply rooted, our conclusion that the Clause is incorporated remains unchanged. * * * For the reasons stated, the judgment of the Indiana Supreme Court is vacated, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered. Notes 1 The sole exception is our holding that the Sixth Amendment requires jury unanimity in federal, but not state, criminal proceedings. Apodaca v. Oregon, 406 U.S. 404 (1972). As we have explained, that “exception to th[e] general rule . . . was the result of an unusual division among the Justices,” and it “does not undermine the well-established rule that incorporated Bill of Rights protections apply identically to the States and the Federal Government.” McDonald, 561 U. S., at 766, n. 14. 2 “Amercements were payments to the Crown, and were required of individuals who were ‘in the King’s mercy,’ because of some act offensive to the Crown.” Browning-Ferris, 492 U. S., at 269. “[T]hough fines and amercements had distinct historical antecedents, they served fundamentally similar purposes—and, by the seventeenth and eighteenth centuries, the terms were often used interchangeably.” Brief for Eighth Amendment Scholars as Amici Curiae 12.

588.US.2018_18-431

Respondents Maurice Davis and Andre Glover were charged with multiple counts of Hobbs Act robbery and one count of conspiracy to commit Hobbs Act robbery. They were also charged under 18 U. S. C. §924(c), which authorizes heightened criminal penalties for using, carrying, or possessing a firearm in connection with any federal “crime of violence or drug trafficking crime.” §924(c)(1)(A). “Crime of violence” is defined in two subparts: the elements clause, §924(c)(3)(A), and the residual clause, §924(c)(3)(B). The residual clause in turn defines a “crime of violence” as a felony “that by its nature, involves a substantial risk that physical force against the person or property of another may be used in the course of committing the offense.” Ibid. A jury convicted the men on most of the underlying charges and on two separate §924(c) charges for brandishing a firearm in connection with their crimes. The Fifth Circuit initially rejected their argument that §924(c)’s residual clause is unconstitutionally vague, but on remand in light of Sessions v. Dimaya, 584 U. S. ___, the court reversed course and held §924(c)(3)(B) unconstitutional. It then held that Mr. Davis’s and Mr. Glover’s convictions on the §924(c) count charging robbery as the predicate crime of violence could be sustained under the elements clause, but that the other count—which charged conspiracy as a predicate crime of violence—could not be upheld because it depended on the residual clause. Held: Section 924(c)(3)(B) is unconstitutionally vague. Pp. 4–25. (a) In our constitutional order, a vague law is no law at all. The vagueness doctrine rests on the twin constitutional pillars of due process and separation of powers. This Court has recently applied the doctrine in two cases involving statutes that bear more than a passing resemblance to §924(c)(3)(B)’s residual clause—Johnson v. United States, 576 U. S. ___, which addressed the residual clause of the Armed Career Criminal Act (ACCA), and Sessions v. Dimaya, which addressed the residual clause of 18 U. S. C. §16. The residual clause in each case required judges to use a “categorical approach” to determine whether an offense qualified as a violent felony or crime of violence. Judges had to disregard how the defendant actually committed the offense and instead imagine the degree of risk that would attend the idealized “ ‘ordinary case’ ” of the offense. Johnson, 576 U. S., at ___. The Court held in each case that the imposition of criminal punishments cannot be made to depend on a judge’s estimation of the degree of risk posed by a crime’s imagined “ordinary case.” The government and lower courts have long understood §924(c)(3)(B) to require the same categorical approach. Now, the government asks this Court to abandon the traditional categorical approach and hold that the statute commands a case-specific approach that would look at the defendant’s actual conduct in the predicate crime. The government’s case-specific approach would avoid the vagueness problems that doomed the statutes in Johnson and Dimaya and would not yield to the same practical and Sixth Amendment complications that a case-specific approach under the ACCA and §16 would, but this approach finds no support in §924(c)’s text, context, and history. Pp. 4–9. (b) This Court has already read the nearly identical language of §16(b) to mandate a categorical approach. See Leocal v. Ashcroft, 543 U.S. 1, 7. And what is true of §16(b) seems at least as true of §924(c)(3)(B). The government claims that the singular term “offense” carries the “generic” meaning in connection with the elements clause but a “specific act” meaning in connection with the residual clause, but nothing in §924(c)(3)(B) rebuts the presumption that the single term “offense” bears a consistent meaning. This reading is reinforced by the language of the residual clause itself, which speaks of an offense that, “by its nature,” involves a certain type of risk. Pp. 9–12. (c) The categorical reading is also reinforced by §924(c)(3)(B)’s role in the broader context of the federal criminal code. Dozens of federal statutes use the phrase “crime of violence” to refer to presently charged conduct. Some cross-reference §924(c)(3)’s definition, while others are governed by the virtually identical definition in §16. The choice appears completely random. To hold that §16(b) requires the categorical approach while §924(c)(3)(B) requires the case-specific approach would make a hash of the federal criminal code. Pp. 12–13. (d) Section 924(c)(3)(B)’s history provides still further evidence that it carries the same categorical-approach command as §16(b). When Congress enacted the definition of “crime of violence” in §16 in 1984, it also employed the term in numerous places in the Act, including §924(c). The two statutes, thus, were originally designed to be read together. And when Congress added a definition of “crime of violence” to §924(c) in 1986, it copied the definition from §16 without making any material changes to the language of the residual clause, which would have been a bizarre way of suggesting that the two clauses should bear drastically different meanings. Moreover, §924(c) originally prohibited the use of a firearm in connection with any federal felony, before Congress narrowed §924(c) in 1984 by limiting its predicate offenses to “crimes of violence.” The case-specific reading would go a long way toward nullifying that limitation and restoring the statute’s original breadth. Pp. 14–17. (e) Relying on the canon of constitutional avoidance, the government insists that if the case-specific approach does not represent the best reading of the statute, it is nevertheless the Court’s duty to adopt any “fairly possible” reading to save the statute from being unconstitutional. But it is doubtful the canon could play a proper role in this case even if the government’s reading were “possible.” This Court has sometimes adopted the narrower construction of a criminal statute to avoid having to hold it unconstitutional if it were construed more broadly, but it has not invoked the canon to expand the reach of a criminal statute in order to save it. To do so would risk offending the very same due process and separation of powers principles on which the vagueness doctrine itself rests and would sit uneasily with the rule of lenity’s teaching that ambiguities about a criminal statute’s breadth should be resolved in the defendant’s favor. Pp. 17–19. 903 F.3d 483, affirmed in part, vacated in part, and remanded. Gorsuch, J., delivered the opinion of the Court, in which Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined. Kavanaugh, J., filed a dissenting opinion, in which Thomas and Alito, JJ., joined, and in which Roberts, C. J., joined as to all but Part II–C.

In our constitutional order, a vague law is no law at all. Only the people’s elected representatives in Congress have the power to write new federal criminal laws. And when Congress exercises that power, it has to write statutes that give ordinary people fair warning about what the law demands of them. Vague laws transgress both of those constitutional requirements. They hand off the legislature’s responsibility for defining criminal behavior to unelected prosecutors and judges, and they leave people with no sure way to know what consequences will attach to their conduct. When Congress passes a vague law, the role of courts under our Constitution is not to fashion a new, clearer law to take its place, but to treat the law as a nullity and invite Congress to try again. Today we apply these principles to 18 U. S. C. §924(c). That statute threatens long prison sentences for anyone who uses a firearm in connection with certain other federal crimes. But which other federal crimes? The statute’s residual clause points to those felonies “that by [their] nature, involv[e] a substantial risk that physical force against the person or property of another may be used in the course of committing the offense.” §924(c)(3)(B). Even the government admits that this language, read in the way nearly everyone (including the government) has long understood it, provides no reliable way to determine which offenses qualify as crimes of violence and thus is unconstitutionally vague. So today the government attempts a new and alternative reading designed to save the residual clause. But this reading, it turns out, cannot be squared with the statute’s text, context, and history. Were we to adopt it, we would be effectively stepping outside our role as judges and writing a new law rather than applying the one Congress adopted. I After Maurice Davis and Andre Glover committed a string of gas station robberies in Texas, a federal prosecutor charged both men with multiple counts of robbery affecting interstate commerce in violation of the Hobbs Act, 18 U. S. C. §1951(a), and one count of conspiracy to commit Hobbs Act robbery. The prosecutor also charged Mr. Davis with being a felon in possession of a firearm. In the end, a jury acquitted Mr. Davis of one robbery charge and otherwise found the men guilty on all counts. And these convictions, none of which are challenged here, authorized the court to impose prison sentences of up to 70 years for Mr. Davis and up to 100 years for Mr. Glover. But that was not all. This appeal concerns additional charges the government pursued against the men under §924(c). That statute authorizes heightened criminal penalties for using or carrying a firearm “during and in relation to,” or possessing a firearm “in furtherance of,” any federal “crime of violence or drug trafficking crime.” §924(c)(1)(A). The statute proceeds to define the term “crime of violence” in two subparts—the first known as the elements clause, and the second the residual clause. According to §924(c)(3), a crime of violence is “an offense that is a felony” and “(A) has as an element the use, attempted use, or threatened use of physical force against the person or property of another, or “(B) that by its nature, involves a substantial risk that physical force against the person or property of an- other may be used in the course of committing the offense.” Violators of §924(c) face a mandatory minimum sentence of five years in prison, over and above any sentence they receive for the underlying crime of violence or drug trafficking crime. The minimum sentence rises to 7 years if the defendant brandishes the firearm and 10 years if he discharges it. Certain types of weapons also trigger enhanced penalties—for example, a defendant who uses a short-barreled shotgun faces a minimum sentence of 10 years. And repeat violations of §924(c) carry a minimum sentence of 25 years.[1] At trial, the government argued that Mr. Davis and Mr. Glover had each committed two separate §924(c) violations by brandishing a short-barreled shotgun in connection with their crimes. Here, too, the jury agreed. These convictions yielded a mandatory minimum sentence for each man of 35 years, which had to run consecutively to their other sentences. Adding the §924(c) mandatory minimums to its discretionary sentences for their other crimes, the district court ultimately sentenced Mr. Glover to more than 41 years in prison and Mr. Davis to more than 50 years. On appeal, both defendants argued that §924(c)’s residual clause is unconstitutionally vague. At first, the Fifth Circuit rejected the argument. United States v. Davis, 677 Fed. Appx. 933, 936 (2017) (per curiam). But after we vacated its judgment and remanded for further consideration in light of our decision in Sessions v. Dimaya, 584 U. S. ___ (2018), striking down a different, almost identically worded statute, the court reversed course and held §924(c)(3)(B) unconstitutional. 903 F.3d 483, 486 (2018) (per curiam). It then held that Mr. Davis’s and Mr. Glover’s convictions on one of the two §924(c) counts, the one that charged robbery as a predicate crime of violence, could be sustained under the elements clause. But it held that the other count, which charged conspiracy as a predicate crime of violence, depended on the residual clause; and so it vacated the men’s convictions and sentences on that count. Because the Fifth Circuit’s ruling deepened a dispute among the lower courts about the constitutionality of §924(c)’s residual clause, we granted certiorari to resolve the question. 586 U. S. ___ (2018).[2] II Our doctrine prohibiting the enforcement of vague laws rests on the twin constitutional pillars of due process and separation of powers. See Dimaya, 584 U. S., at ___–___ (plurality opinion) (slip op., at 4–5); id., at ___–___ (Gorsuch, J., concurring in part and concurring in judgment) (slip op., at 2–9). Vague laws contravene the “first essential of due process of law” that statutes must give people “of common intelligence” fair notice of what the law demands of them. Connally v. General Constr. Co., 269 U.S. 385, 391 (1926); see Collins v. Kentucky, 234 U.S. 634, 638 (1914). Vague laws also undermine the Constitution’s separation of powers and the democratic self-governance it aims to protect. Only the people’s elected representatives in the legislature are authorized to “make an act a crime.” United States v. Hudson, 7 Cranch 32, 34 (1812). Vague statutes threaten to hand responsibility for defining crimes to relatively unaccountable police, prosecutors, and judges, eroding the people’s ability to oversee the creation of the laws they are expected to abide. See Kolender v. Lawson, 461 U.S. 352, 357–358, and n. 7 (1983); United States v. L. Cohen Grocery Co., 255 U.S. 81, 89–91 (1921); United States v. Reese, 92 U.S. 214, 221 (1876). In recent years, this Court has applied these principles to two statutes that bear more than a passing resemblance to §924(c)(3)(B)’s residual clause. In Johnson v. United States, 576 U. S. ___ (2015), the Court addressed the residual clause of the Armed Career Criminal Act (ACCA), which defined a “violent felony” to include offenses that presented a “serious potential risk of physical injury to another.” §924(e)(2)(B)(ii). The ACCA’s residual clause required judges to use a form of what we’ve called the “categorical approach” to determine whether an offense qualified as a violent felony. Following the categorical approach, judges had to disregard how the defendant actually committed his crime. Instead, they were required to imagine the idealized “ ‘ordinary case’ ” of the defendant’s crime and then guess whether a “ ‘serious potential risk of physical injury to another’ ” would attend its commission. Id., at ___ (slip op., at 4). Johnson held this judicial inquiry produced “more unpredictability and arbitrariness” when it comes to specifying unlawful conduct than the Constitution allows. Id., at ___–___ (slip op., at 5–6). Next, in Sessions v. Dimaya, we considered the residual clause of 18 U. S. C. §16, which defines a “crime of violence” for purposes of many federal statutes. Like §924(c)(3), §16 contains an elements clause and a residual clause. The only difference is that §16’s elements clause, unlike §924(c)(3)’s elements clause, isn’t limited to felonies; but there’s no material difference in the language or scope of the statutes’ residual clauses.[3] As with the ACCA, our precedent under §16’s residual clause required courts to use the categorical approach to determine whether an offense qualified as a crime of violence. Dimaya, 584 U. S., at ___–___ (slip op., at 2–3); see Leocal v. Ashcroft, 543 U.S. 1, 7, 10 (2004). And, again as with the ACCA, we held that §16’s residual clause was unconstitutionally vague because it required courts “to picture the kind of conduct that the crime involves in the ordinary case, and to judge whether that abstraction presents some not-well-specified-yet-sufficiently-large degree of risk.” Dimaya, 584 U. S., at ___ (slip op., at 11) (internal quotation marks omitted). What do Johnson and Dimaya have to say about the statute before us? Those decisions teach that the imposition of criminal punishment can’t be made to depend on a judge’s estimation of the degree of risk posed by a crime’s imagined “ordinary case.” But does §924(c)(3)(B) require that sort of inquiry? The government and lower courts have long thought so. For years, almost everyone understood §924(c)(3)(B) to require exactly the same categorical approach that this Court found problematic in the residual clauses of the ACCA and §16.[4] Today, the government acknowledges that, if this understanding is correct, then §924(c)(3)(B) must be held unconstitutional too. But the government thinks it has now found a way around the problem. In the aftermath of our decisions holding the residual clauses of the ACCA and §16(b) unconstitutionally vague, the government “abandon[ed] its longstanding position” that §924(c)(3)(B) requires a categorical analysis and began urging lower courts to “adopt a new ‘case specific’ method” that would look to “the ‘defendant’s actual conduct’ in the predicate offense.” 903 F. 3d, at 485. Now, the government tries the same strat- egy in this Court, asking us to abandon the traditional categorical approach and hold that the statute actually commands the government’s new case-specific approach. So, while the consequences in this case may be of constitutional dimension, the real question before us turns out to be one of pure statutory interpretation. In approaching the parties’ dispute over the statute’s meaning, we begin by acknowledging that the government is right about at least two things. First, a case-specific approach would avoid the vagueness problems that doomed the statutes in Johnson and Dimaya. In those cases, we recognized that there would be no vagueness problem with asking a jury to decide whether a defendant’s “ ‘real-world conduct’ ” created a substantial risk of physical violence. Dimaya, 584 U. S., at ___–___ (slip op., at 10–11); see Johnson, 576 U. S., at ___, ___ (slip op., at 6, 12). Second, a case-specific approach wouldn’t yield the same practical and Sixth Amendment complications under §924(c) that it would have under the ACCA or §16. Those other statutes, in at least some of their applications, required a judge to determine whether a defendant’s prior conviction was for a “crime of violence” or “violent felony.” In that context, a case-specific approach would have entailed “reconstruct[ing], long after the original conviction, the conduct underlying that conviction.” Id., at ___ (slip op., at 13). And having a judge, not a jury, make findings about that underlying conduct would have “raise[d] serious Sixth Amendment concerns.” Descamps v. United States, 570 U.S. 254, 269–270 (2013). By contrast, a §924(c) prosecution focuses on the conduct with which the defendant is currently charged. The government already has to prove to a jury that the defendant committed all the acts necessary to punish him for the underlying crime of violence or drug trafficking crime. So it wouldn’t be that difficult to ask the jury to make an additional finding about whether the defendant’s conduct also created a substantial risk that force would be used. But all this just tells us that it might have been a good idea for Congress to have written a residual clause for §924(c) using a case-specific approach. It doesn’t tell us whether Congress actually wrote such a clause. To answer that question, we need to examine the statute’s text, context, and history. And when we do that, it becomes clear that the statute simply cannot support the government’s newly minted case-specific theory. III A Right out of the gate, the government faces a challenge. This Court, in a unanimous opinion, has already read the nearly identical language of 18 U. S. C. §16(b) to mandate a categorical approach. And, importantly, the Court did so without so much as mentioning the practical and constitutional concerns described above. Instead, the Court got there based entirely on the text. In Leocal, the Court wrote: “In determining whether petitioner’s conviction falls within the ambit of §16, the statute directs our focus to the ‘offense’ of conviction. See §16(a) (defining a crime of violence as ‘an offense that has as an element the use . . . of physical force against the person or property of another’ (emphasis added)); §16(b) (defining the term as ‘any other offense that is a felony and that, by its nature, involves a substantial risk that physical force against the person or property of an- other may be used in the course of committing the offense’ (emphasis added)). This language requires us to look to the elements and the nature of the offense of conviction, rather than to the particular facts relating to petitioner’s crime.” 543 U. S., at 7. Leocal went on to suggest that burglary would always be a crime of violence under §16(b) “because burglary, by its nature, involves a substantial risk that the burglar will use force against a victim in completing the crime,” regardless of how any particular burglar might act on a specific occasion. Id., at 10 (emphasis added); see also Dimaya, 584 U. S., at ___ (slip op., at 14) (plurality opinion) (reaffirming that “§16(b)’s text . . . demands a categorical approach”). And what was true of §16(b) seems to us at least as true of §924(c)(3)(B): It’s not even close; the statutory text commands the categorical approach. Consider the word “offense.” It’s true that “in ordinary speech,” this word can carry at least two possible meanings. It can refer to “a generic crime, say, the crime of fraud or theft in general,” or it can refer to “the specific acts in which an offender engaged on a specific occasion.” Nijhawan v. Holder, 557 U.S. 29, 33–34 (2009). But the word “offense” appears just once in §924(c)(3), in the statute’s prefatory language. And everyone agrees that, in connection with the elements clause, the term “offense” carries the first, “generic” meaning. Cf. id., at 36 (similar language of the ACCA’s elements clause “refers directly to generic crimes”). So reading this statute most naturally, we would expect “offense” to retain that same meaning in connection with the residual clause. After all, “[i]n all but the most unusual situations, a single use of a statutory phrase must have a fixed meaning.” Cochise Consultancy, Inc. v. United States ex rel. Hunt, 587 U. S. ___, ___ (2019) (slip op., at 5). To prevail, the government admits it must persuade us that the singular term “offense” bears a split personality in §924(c), carrying the “generic” meaning in connection with the elements clause but then taking on the “specific act” meaning in connection with the residual clause. And, the government suggests, this isn’t quite as implausible as it may sound; sometimes the term “offense” can carry both meanings simultaneously. To illustrate its point, the government posits a statute defining a “youthful gun crime” as “an offense that has as an element the use of a gun and is committed by someone under the age of 21.” Tr. of Oral Arg. 16. This statute, the government suggests, would leave us little choice but to understand the single word “offense” as encompassing both the generic crime and the manner of its commission on a specific occasion. To which we say: Fair enough. It’s possible for surrounding text to make clear that “offense” carries a double meaning. But absent evidence to the contrary, we presume the term is being used consistently. And nothing in §924(c)(3)(B) comes close to rebutting that presumption. Just the opposite. The language of the residual clause itself reinforces the conclusion that the term “offense” carries the same “generic” meaning throughout the statute. Section 924(c)(3)(B), just like §16(b), speaks of an offense that, “by its nature,” involves a certain type of risk. And that would be an exceedingly strange way of referring to the circumstances of a specific offender’s conduct. As both sides agree, the “nature” of a thing typically denotes its “ ‘normal and characteristic quality,’ ” Dimaya, 584 U. S., at ___ (slip op., at 14) (quoting Webster’s Third New International Dictionary 1507 (2002)), or its “ ‘basic or inherent features,’ ” United States v. Barrett, 903 F.3d 166, 182 (CA2 2018) (quoting Oxford Dictionary of English 1183 (A. Stevenson ed., 3d ed. 2010)). So in plain English, when we speak of the nature of an offense, we’re talking about “what an offense normally—or, as we have repeat- edly said, ‘ordinarily’—entails, not what happened to occur on one occasion.” Dimaya, 584 U. S., at ___ (slip op., at 14); see Leocal, 543 U. S., at 7 (contrasting the “nature of the offense” with “the particular facts [of] petitioner’s crime”).[5] Once again, the government asks us to overlook this obvious reading of the text in favor of a strained one. It suggests that the statute might be referring to the “nature” of the defendant’s conduct on a particular occasion. But while this reading may be linguistically feasible, we struggle to see why, if it had intended this meaning, Congress would have used the phrase “by its nature” at all. The government suggests that “by its nature” keeps the focus on the offender’s conduct and excludes evidence about his personality, such as whether he has violent tendencies. But even without the words “by its nature,” nothing in the statute remotely suggests that courts are allowed to consider character evidence—a type of evidence usually off-limits during the guilt phase of a criminal trial. Cf. Fed. Rule Evid. 404. B Things become clearer yet when we consider §924(c)(3)(B)’s role in the broader context of the federal criminal code. As we’ve explained, the language of §924(c)(3)(B) is almost identical to the language of §16(b), which this Court has read to mandate a categorical approach. And we normally presume that the same language in related statutes carries a consistent meaning. See, e.g., Sullivan v. Stroop, 496 U.S. 478, 484 (1990). This case perfectly illustrates why we do that. There are dozens of federal statutes that use the phrase “crime of violence” to refer to presently charged conduct rather than a past conviction. Some of those statutes cross-reference the definition of “crime of violence” in §924(c)(3), while others are governed by the virtually identical definition in §16. The choice appears completely random. Reading the similar language in §924(c)(3)(B) and §16(b) similarly yields sensibly congruent applications across all these other statutes. But if we accepted the government’s invitation to reinterpret §924(c)(3)(B) as alone endorsing a case-specific approach, we would produce a series of seemingly inexplicable results. Take just a few examples. If the government were right, Congress would have mandated the case-specific approach in a prosecution for providing explosives to facilitate a crime of violence, 18 U. S. C. §844(o), but the (now-invalidated) categorical approach in a prosecution for providing information about explosives to facilitate a crime of violence, §842(p)(2). It would have mandated the case-specific approach in a prosecution for using false identification documents in connection with a crime of violence, §1028(b)(3)(B), but the categorical approach in a prosecution for using confidential phone records in connection with a crime of violence, §1039(e)(1). It would have mandated the case-specific approach in a prosecution for giving someone a firearm to use in a crime of violence, §924(h), but the categorical approach in a prosecution for giving a minor a handgun to use in a crime of violence, §924(a)(6)(B)(ii). It would have mandated the case-specific approach in a prosecution for traveling to another State to acquire a firearm for use in a crime of violence, §924(g), but the categorical approach in a prosecution for traveling to another State to commit a crime of violence, §1952(a)(2). And it would have mandated the case-specific approach in a prosecution for carrying armor-piercing ammunition in connection with a crime of violence, §924(c)(5), but the categorical approach in a prosecution for carrying a firearm while “in possession of armor piercing ammunition capable of being fired in that firearm” in connection with a crime of violence, §929(a)(1). There would be no rhyme or reason to any of this. Nor does the government offer any plausible account why Congress would have wanted courts to take such dramatically different approaches to classifying offenses as crimes of violence in these various provisions. To hold, as the government urges, that §16(b) requires the categorical approach while §924(c)(3)(B) requires the case-specific approach would make a hash of the federal criminal code. C Section 924(c)(3)(B)’s history provides still further evidence that it carries the same categorical-approach command as §16(b). It’s no accident that the language of the two laws is almost exactly the same. The statutory term “crime of violence” traces its origins to the Comprehensive Crime Control Act of 1984. There, Congress enacted the definition of “crime of violence” in §16. §1001(a), 98Stat. 2136. It also “employed the term ‘crime of violence’ in numerous places in the Act,” Leocal, 543 U. S., at 6, including in §924(c). §1005(a), 98Stat. 2138. At that time, Congress didn’t provide a separate definition of “crime of violence” in §924(c) but relied on §16’s general definition. The two statutes, thus, were originally designed to be read together. Admittedly, things changed a bit over time. Eventually, Congress expanded §924(c)’s predicate offenses to include drug trafficking crimes as well as crimes of violence. §§104(a)(2)(B)–(C), 100Stat. 457. When it did so, Congress added a subsection-specific definition of “drug trafficking crime” in §924(c)(2)—and, perhaps thinking that both terms should be defined in the same place, it also added a subsection-specific definition of “crime of violence” in §924(c)(3). §104(a)(2)(F), id., at 457. But even then, Congress didn’t write a new definition of that term. Instead, it copied and pasted the definition from §16 without making any material changes to the language of the residual clause. The government suggests that, in doing so, Congress “intentionally separated” and “decoupled” the two definitions. Brief for United States 34, 37. But importing the residual clause from §16 into §924(c)(3) almost word for word would have been a bizarre way of suggesting that the two clauses should bear drastically different meanings. Usually when statutory language “ ‘is obviously transplanted from . . . other legislation,’ ” we have reason to think “ ‘it brings the old soil with it.’ ” Sekhar v. United States, 570 U.S. 729, 733 (2013). What’s more, when Congress copied §16(b)’s language into §924(c) in 1986, it proceeded on the premise that the language required a categorical approach. By then courts had, as the government puts it, “beg[u]n to settle” on the view that §16(b) demanded a categorical analysis. Brief for United States 36–37. Of particular significance, the Second Circuit, along with a number of district courts, had relied on the categorical approach to hold that selling drugs could never qualify as a crime of violence because “[w]hile the traffic in drugs is often accompanied by violence,” it can also be carried out through consensual sales and thus “does not by its nature involve substantial risk that physical violence will be used.” United States v. Diaz, 778 F.2d 86, 88 (1985). Congress moved quickly to abrogate those decisions. But, notably, it didn’t do so by directing a case-specific approach or changing the language courts had read to require the categorical approach. Instead, it accepted the categorical approach as given and simply declared that certain drug trafficking crimes automatically trigger §924 penalties, regardless of the risk of violence that attends them. §§104(a)(2)(B)–(C), 100Stat. 457. The government’s reply to this development misses the mark. The government argues that §16(b) had not acquired such a well-settled judicial construction by 1986 that the reenactment of its language in §924(c)(3)(B) should be presumed to have incorporated the same construction. We agree. See Jerman v. Carlisle, McNellie, Rini, Kramer & Ulrich, L. P. A., 559 U.S. 573, 590 (2010) (interpretations of three courts of appeals “may not have ‘settled’ the meaning” of a statute for purposes of the reenactment canon). But Congress in 1986 did more than just reenact language that a handful of courts had interpreted to require the categorical approach. It amended §924(c) specifically to abrogate the results of those decisions, without making any attempt to overturn the categorical reading on which they were based. And that would have been an odd way of proceeding if Congress had thought the categorical reading erroneous. There’s yet one further and distinct way in which §924(c)’s history undermines the government’s case-specific reading of the residual clause. As originally enacted in 1968, §924(c) prohibited the use of a firearm in connection with any federal felony. §102, 82Stat. 1224. The 1984 amendments narrowed §924(c) by limiting its predicate offenses to “crimes of violence.” But the case-specific reading would go a long way toward nullifying that limitation and restoring the statute’s original breadth. After all, how many felonies don’t involve a substantial risk of physical force when they’re committed using a firearm—let alone when the defendant brandishes or discharges the firearm? Recognizing this difficulty, the government assures us that a jury wouldn’t be allowed to find a felony to be a crime of violence solely because the defendant used a firearm, although it could consider the firearm as a “factor.” Tr. of Oral Arg. 8. But the government identifies no textual basis for this rule, and exactly how it would work in practice is anyone’s guess. The government says, for example, that “selling counterfeit handbags” while carrying a gun wouldn’t be a crime of violence under its approach. Id., at 9. But why not? Because the counterfeit-handbag trade is so inherently peaceful that there’s no substantial risk of a violent confrontation with dissatisfied customers, territorial competitors, or dogged police officers? And how are jurors supposed to determine that? The defendant presumably knew the risks of his trade, and he chose to arm himself. See United States v. Simms, 914 F.3d 229, 247–248 (CA4 2019) (en banc) (refusing to “condem[n] jurors to such an ill-defined inquiry”). Even granting the government its handbag example, we suspect its approach would result in the vast majority of federal felonies becoming potential predicates for §924(c) charges, contrary to the limitation Congress deliberately imposed when it restricted the statute’s application to crimes of violence. D With all this statutory evidence now arrayed against it, the government answers that it should prevail anyway because of the canon of constitutional avoidance. Maybe the case-specific approach doesn’t represent the best reading of the statute—but, the government insists, it is our duty to adopt any “ ‘fairly possible’ ” reading of a statute to save it from being held unconstitutional. Brief for United States 45.[6] We doubt, however, the canon could play a proper role in this case even if the government’s reading were “possible.” True, when presented with two “fair alternatives,” this Court has sometimes adopted the narrower construction of a criminal statute to avoid having to hold it unconstitutional if it were construed more broadly. United States v. Rumely, 345 U.S. 41, 45, 47 (1953); see, e.g., Skilling v. United States, 561 U.S. 358, 405–406, and n. 40 (2010); United States v. Lanier, 520 U.S. 259, 265–267, and n. 6 (1997). But no one before us has identified a case in which this Court has invoked the canon to expand the reach of a criminal statute in order to save it. Yet that is exactly what the government seeks here. Its case-specific reading would cause §924(c)(3)(B)’s penalties to apply to conduct they have not previously been understood to reach: categorically nonviolent felonies committed in violent ways. See Simms, 914 F. 3d, at 256–257 (Wynn, J., concurring).[7] Employing the avoidance canon to expand a criminal statute’s scope would risk offending the very same due process and separation-of-powers principles on which the vagueness doctrine itself rests. See supra, at 4–5. Everyone agrees that Mr. Davis and Mr. Glover did many things that Congress had declared to be crimes; and no matter how we rule today, they will face substantial prison sentences for those offenses. But does §924(c)(3)(B) require them to suffer additional punishment, on top of everything else? Even if you think it’s possible to read the statute to impose such additional punishment, it’s impossible to say that Congress surely intended that result, or that the law gave Mr. Davis and Mr. Glover fair warning that §924(c)’s mandatory penalties would apply to their conduct. Respect for due process and the separation of powers suggests a court may not, in order to save Congress the trouble of having to write a new law, construe a criminal statute to penalize conduct it does not clearly proscribe. Employing the canon as the government wishes would also sit uneasily with the rule of lenity’s teaching that ambiguities about the breadth of a criminal statute should be resolved in the defendant’s favor. That rule is “perhaps not much less old than” the task of statutory “construction itself.” United States v. Wiltberger, 5 Wheat. 76, 95 (1820) (Marshall, C. J.). And much like the vagueness doctrine, it is founded on “the tenderness of the law for the rights of individuals” to fair notice of the law “and on the plain principle that the power of punishment is vested in the legislative, not in the judicial department.” Ibid.; see Lanier, 520 U. S., at 265–266, and n. 5. Applying constitutional avoidance to narrow a criminal statute, as this Court has historically done, accords with the rule of lenity. By contrast, using the avoidance canon instead to adopt a more expansive reading of a criminal statute would place these traditionally sympathetic doctrines at war with one another.[8] IV What does the dissent have to say about all this? It starts by emphasizing that §924(c)(3)(B) has been used in “tens of thousands of federal prosecutions” since its enactment 33 years ago. Post, at 2 (opinion of Kavanaugh, J.). And the dissent finds it “surprising” and “extraordinary” that, after all those prosecutions over all that time, the statute could “suddenly” be deemed unconstitutional. Post, at 2–3. But the government concedes that §924(c)(3)(B) is unconstitutional if it means what everyone has understood it to mean in nearly all of those prosecutions over all those years. So the only way the statute can be saved is if we were “suddenly” to give it a new meaning different from the one it has borne for the last three decades. And if we could do that, it would indeed be “surprising” and “extraordinary.” The dissent defends giving this old law a new meaning by appealing to intuition. It suggests that a categorical reading of §924(c)(3)(B) is “unnatural” because “[i]f you were to ask John Q. Public whether a particular crime posed a substantial risk of violence, surely he would respond, ‘Well, tell me how it went down—what happened?’ ” Post, at 13 (some internal quotation marks omitted). Maybe so. But the language in the statute before us isn’t the language posited in the dissent’s push poll. Section 924(c)(3)(B) doesn’t ask about the risk that “a particular crime posed” but about the risk that an “offense . . . by its nature, involves.” And a categorical reading of this categorical language seemed anything but “unnatural” to the unanimous Court in Leocal or the plurality in Dimaya.[9] Nor did the government think the categorical reading of §924(c)(3)(B) “unnatural” when it embraced that reading for decades. The dissent asks us to overlook the government’s prior view, explaining that the government only defended a categorical reading of the statute “when it did not matter for constitutional vagueness purposes”—that is, before Johnson and Dimaya identified constitutional problems with the categorical approach. Post, at 34. But isn’t that exactly the point? Isn’t it at least a little revealing that, when the government had no motive to concoct an alternative reading, even it thought the best reading of §924(c)(3)(B) demanded a categorical analysis? If this line of attack won’t work, the dissent tries another by telling us that we have “not fully account[ed] for the long tradition of substantial-risk criminal statutes.” Post, at 34. The dissent proceeds to offer a lengthy bill of particulars, citing dozens of state and federal laws that do not use the categorical approach. Post, at 7–10, and nn. 4–17. But what does this prove? Most of the statutes the dissent cites impose penalties on whoever “creates,” or “engages in conduct that creates,” or acts under “circumstances that create” a substantial risk of harm; others employ similar language. Not a single one imposes penalties for committing certain acts during “an offense . . . that by its nature, involves” a substantial risk, or anything similar. Marching through the dissent’s own catalog thus only winds up confirming that legislatures know how to write risk-based statutes that require a case-specific analysis—and that §924(c)(3)(B) is not a statute like that. When the dissent finally turns to address the words Congress actually wrote in §924(c)(3)(B), its main argument seems to be that a categorical reading violates the canon against superfluity. On this account, reading “offense” generically in connection with the residual clause makes the residual clause “duplicate” the elements clause and leaves it with “virtually nothing” to do. Post, at 20. But that is a surprising assertion coming from the dissent, which devotes several pages to describing the “many” offenders who have been convicted under the residual clause using the categorical approach but who “might not” be prosecutable under the elements clause. Post, at 30–33. It is also wrong. As this Court has long understood, the residual clause, read categorically, “sweeps more broadly” than the elements clause—potentially reaching offenses, like burglary, that do not have violence as an element but that arguably create a substantial risk of violence. Leocal, 543 U. S., at 10. So even under the categorical reading, the residual clause is far from superfluous. Without its misplaced reliance on the superfluity canon, there is little left of the dissent’s textual analysis. The dissent asserts that the phrase “by its nature” must “focu[s] on the defendant’s actual conduct”—but only because this “follows” from the dissent’s earlier (and mistaken) superfluity argument. Post, at 21. Next, the dissent claims that “the word ‘involves’ ” and “the phrase ‘in the course of committing the offense’ ” both support a case-specific approach. Post, at 22. But these words do not favor either reading: It is just as natural to ask whether the offense of robbery ordinarily “involves” a substantial risk that violence will be used “in the course of committing the offense” as it is to ask whether a particular robbery “involved” a substantial risk that violence would be used “in the course of committing the offense.” If anything, the statute’s use of the present and not the past tense lends further support to the categorical reading.[10] The dissent thinks it significant, too, that the statute before us “does not use the term ‘conviction,’ ” post, at 23; but that word is hardly a prerequisite for the categorical approach, as Dimaya makes clear. Remarkably, the dissent has nothing at all to say about §924(c)(3)’s history or its relationship with other criminal statutes; it just ignores those arguments. And when it comes to the constitutional avoidance canon, the dissent does not even try to explain how using that canon to criminalize conduct that isn’t criminal under the fairest reading of a statute might be reconciled with traditional principles of fair notice and separation of powers. Instead, the dissent seems willing to consign “ ‘thousands’ ” of defendants to prison for “years—potentially decades,” not because it is certain or even likely that Congress ordained those penalties, but because it is merely “possible” Congress might have done so. Post, at 30, 33–34. In our republic, a speculative possibility that a man’s conduct violated the law should never be enough to justify taking his liberty. In the end, the dissent is forced to argue that holding §924(c)(3)(B) unconstitutional would invite “bad” social policy consequences. Post, at 34. In fact, the dissent’s legal analysis only comes sandwiched between a lengthy paean to laws that impose severe punishments for gun crimes and a rogue’s gallery of offenses that may now be punished somewhat less severely. See post, at 1–2, 30–34. The dissent acknowledges that “the consequences cannot change our understanding of the law.” Post, at 34. But what’s the point of all this talk of “bad” consequences if not to suggest that judges should be tempted into reading the law to satisfy their policy goals? Even taken on their own terms, too, the dissent’s policy concerns are considerably overblown. While the dissent worries that our ruling may elicit challenges to past §924(c) convictions, post, at 33, the dissent’s preferred approach—saving §924(c)(3)(B) by changing its meaning—would also call into question countless convictions premised on the categorical reading. And defendants whose §924(c) convictions are overturned by virtue of today’s ruling will not even necessarily receive lighter sentences: As this Court has noted, when a defendant’s §924(c) conviction is invalidated, courts of appeals “routinely” vacate the defendant’s entire sentence on all counts “so that the district court may increase the sentences for any remaining counts” if such an increase is warranted. Dean v. United States, 581 U. S. ___, ___ (2017) (slip op., at 5). Of course, too, Congress always remains free to adopt a case-specific approach to defining crimes of violence for purposes of §924(c)(3)(B) going forward. As Mr. Davis and Mr. Glover point out, one easy way of achieving that goal would be to amend the statute so it covers any felony that, “based on the facts underlying the offense, involved a substantial risk” that physical force against the person or property of another would be used in the course of committing the offense. Brief for Respondents 46 (quoting H. R. 7113, 115th Cong., 2d Sess. (2018); emphasis deleted); see also Tr. of Oral Arg. 19 (government’s counsel agreeing that this language would offer “clearer” support for the case-specific approach than the current version of the statute does). The dissent’s catalog of case-specific, risk-based criminal statutes supplies plenty of other models Congress could follow. Alternatively still, Congress might choose to retain the categorical approach but avoid vagueness in other ways, such as by defining crimes of violence to include certain enumerated offenses or offenses that carry certain minimum penalties. All these options and more are on the table. But these are options that belong to Congress to consider; no matter how tempting, this Court is not in the business of writing new statutes to right every social wrong it may perceive. * We agree with the court of appeals’ conclusion that §924(c)(3)(B) is unconstitutionally vague. At the same time, exactly what that holding means for Mr. Davis and Mr. Glover remains to be determined. After the Fifth Circuit vacated their convictions and sentences on one of the two §924(c) counts at issue, both men sought rehearing and argued that the court should have vacated their sentences on all counts. In response, the government con- ceded that, if §924(c)(3)(B) is held to be vague, then the de- fendants are entitled to a full resentencing, not just the more limited remedy the court had granted them. The Fifth Circuit has deferred ruling on the rehearing petitions pending our decision, so we remand the case to allow the court to address those petitions. The judgment below is affirmed in part and vacated in part, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 When this case was tried, a defendant convicted of two §924(c) violations in a single prosecution faced a 25-year minimum for the second violation. See Deal v. United States, 508 U.S. 129, 132 (1993); §1(a)(1), 112Stat. 3469. In 2018, Congress changed the law so that, going forward, only a second §924(c) violation committed “after a prior [§924(c)] conviction . . . has become final” will trigger the 25-year minimum. Pub. L. 115–391, §403(a), 132Stat. 5221. 2 Compare United States v. Simms, 914 F.3d 229, 236–246 (CA4 2019) (en banc), United States v. Salas, 889 F.3d 681, 685–686 (CA10 2018), and United States v. Eshetu, 898 F.3d 36, 37–38 (CADC 2018) (holding that §924(c)(3)(B) is vague), with United States v. Douglas, 907 F.3d 1, 11–16 (CA1 2018), Ovalles v. United States, 905 F.3d 1231, 1240–1252 (CA11 2018) (en banc), and United States v. Barrett, 903 F.3d 166, 178–184 (CA2 2018) (taking the opposite view). 3 Section 16 provides that the term “crime of violence” means “(a) an offense that has as an element the use, attempted use, or threatened use of physical force against the person or property of another, or (b) any other offense that is a felony and that, by its nature, involves a substantial risk that physical force against the person or property of another may be used in the course of committing the offense.” 4 See, e.g., United States v. Acosta, 470 F.3d 132, 134–135 (CA2 2006); United States v. Butler, 496 Fed. Appx. 158, 161 (CA3 2012); United States v. Fuertes, 805 F.3d 485, 498 (CA4 2015); United States v. Williams, 343 F.3d 423, 431 (CA5 2003); Evans v. Zych, 644 F.3d 447, 453 (CA6 2011); United States v. Jackson, 865 F.3d 946, 952 (CA7 2017), vacated and remanded, 584 U. S. ___ (2018); United States v. Moore, 38 F.3d 977, 979–980 (CA8 1994); United States v. Amparo, 68 F.3d 1222, 1225–1226 (CA9 1995); United States v. Munro, 394 F.3d 865, 870 (CA10 2005); United States v. McGuire, 706 F.3d 1333, 1336–1337 (CA11 2013); United States v. Kennedy, 133 F.3d 53, 56 (CADC 1998); see also Ovalles v. United States, 905 F.3d 1231, 1295 (CA11 2018) (en banc) (J. Pryor, J., dissenting) (“For years, and even after Johnson, the government consistently has urged that we apply a categorical approach to §924(c)”). 5 The government’s own regulations reflect this understanding of the ordinary meaning of “by its nature.” A Department of Justice regulation provides that an inmate is not eligible for early release if he was convicted of an offense “that, by its nature or conduct, presents a serious potential risk of physical force.” 28 CFR §550.55(b)(5)(iii) (2017) (emphasis added); see Bush v. Pitzer, 133 F.3d 455, 458 (CA7 1997) (denying early release because “[c]onspiracy does not by its ‘nature’ present a serious risk; but Bush’s ‘conduct’ did so”). 6 There are at least two different canons of construction that sometimes go by the name “constitutional avoidance.” The one the government invokes here is perhaps better termed the presumption of constitutionality. Of long lineage, it holds that courts should, if possible, interpret ambiguous statutes to avoid rendering them unconstitutional, see, e.g., Parsons v. Bedford, 3 Pet. 433, 448–449 (1830) (Story, J.), and it is distinct from the more modern (and more debated) constitutional doubt canon, which suggests courts should construe ambiguous statutes to avoid the need even to address serious questions about their constitutionality, see Rust v. Sullivan, 500 U.S. 173, 190–191 (1991). 7 The government claims to have found cases invoking the canon to expand a statute’s reach, but none actually stands for that proposition. Each simply remarks in passing that a construction the Court arrived at for other reasons had the additional benefit of avoiding vagueness concerns; none suggests that a narrower construction was available. See United States v. Grace, 461 U.S. 171, 176 (1983) (accepting government’s construction, which was “not contested by appellees”); United States v. Culbert, 435 U.S. 371, 379 (1978) (finding statute clear and refusing to “manufacture ambiguity where none exists”); United States v. Shreveport Grain & Elevator Co., 287 U.S. 77, 82–83 (1932) (finding statute unambiguous and construing it according to “the natural import of its terms”). And the dissent, despite compiling a page-long list of constitutional avoidance cases spanning “more than 200 years,” post, at 25–26, has been unable to find any better examples. See post, at 29–30 (opinion of Kavanaugh, J.). 8 Admittedly, abandoning the categorical approach in favor of the case-specific approach would also have the effect of excluding from the statute’s coverage defendants who commit categorically violent felonies in nonviolent ways, and in that respect would be more “lenient” for some defendants. Regardless, the constitutional principles underlying the rule of lenity counsel caution before invoking constitutional avoidance to construe the statute to punish conduct that it does not unambiguously proscribe. 9 To be sure, the dissent suggests that Leocal and Dimaya adopted a categorical reading simply to avoid practical and constitutional problems. Post, at 15–16, 23, and n. 23. But, as we have seen, this too is mistaken. Leocal did not even mention those problems, and Dimaya held that the text demanded a categorical approach. See supra, at 9. 10 The dissent claims that Taylor v. United States, 495 U.S. 575 (1990), and Nijhawan v. Holder, 557 U.S. 29 (2009), pointed to “the absence of the word ‘involved’ ” as one reason to adopt a categorical approach. Post, at 22. Not true. Taylor explained that the ACCA’s elements clause requires a categorical approach in part because it refers to a crime “that ‘has as an element’—not any crime that, in a particular case, involves—the use or threat of force.” 495 U. S., at 600. All the work in that sentence was being done by the phrase “in a particular case,” not by the word “involves.” And Nijhawan noted that the Court had construed the ACCA’s residual clause, which refers to crimes “that ‘involv[e] conduct that presents a serious potential risk of physical injury,” to require the categorical approach. 557 U. S., at 36.

588.US.2018_17-1672

Respondent Andre Haymond was convicted of possessing child pornography, a crime that carries a prison term of zero to 10 years. After serving a prison sentence of 38 months, and while on supervised release, Mr. Haymond was again found with what appeared to be child pornography. The government sought to revoke his supervised release and secure a new and additional prison sentence. A district judge, acting without a jury, found by a preponderance of the evidence that Mr. Haymond knowingly downloaded and possessed child pornography. Under 18 U. S. C. §3583(e)(3), the judge could have sentenced him to a prison term of between zero and two additional years. But because possession of child pornography is an enumerated offense under §3583(k), the judge instead imposed that provision’s 5-year mandatory minimum. On appeal, the Tenth Circuit observed that whereas a jury had convicted Mr. Haymond beyond a reasonable doubt of a crime carrying a prison term of zero to 10 years, this new prison term included a new and higher mandatory minimum resting on facts found only by a judge by a preponderance of the evidence. The Tenth Circuit therefore held that §3583(k) violated the right to trial by jury guaranteed by the Fifth and Sixth Amendments. Held: The judgment is vacated, and the case is remanded. 869 F.3d 1153, vacated and remanded. Justice Gorsuch, joined by Justice Ginsburg, Justice Sotomayor, and Justice Kagan, concluded that the application of §3583(k) in this case violated Mr. Haymond’s right to trial by jury. Pp. 5–22. (a) As at the time of the Fifth and Sixth Amendments’ adoption, a judge’s sentencing authority derives from, and is limited by, the jury’s factual findings of criminal conduct. A jury must find beyond a reasonable doubt every fact “ ‘which the law makes essential to [a] punishment’ ” that a judge might later seek to impose. Blakely v. Washington, 542 U.S. 296, 304. Historically, that rule’s application proved straightforward, but recent legislative innovations have raised difficult questions. In Apprendi v. New Jersey, 530 U.S. 466, for example, this Court held unconstitutional a sentencing scheme that allowed a judge to increase a defendant’s sentence beyond the statutory maximum based on the judge’s finding of new facts by a preponderance of the evidence. And in Alleyne v. United States, 570 U.S. 99, the Court held that Apprendi’s principle “applies with equal force to facts increasing the mandatory minimum.” 570 U. S., at 111–112. The lesson for this case is clear: Based solely on the facts reflected in the jury’s verdict, Mr. Haymond faced a lawful prison term of between zero and 10 years. But just like the facts the judge found at the defendant’s sentencing hearing in Alleyne, the facts the judge found here increased “the legally prescribed range of allowable sentences” in violation of the Fifth and Sixth Amendments. Id., at 115. Pp. 5–11. (b) The government’s various replies are unpersuasive. First, it stresses that Alleyne arose in a different procedural posture, but this Court has repeatedly rejected efforts to dodge the demands of the Fifth and Sixth Amendments by the simple expedient of relabeling a criminal prosecution. And this Court has already recognized that punishments for revocation of supervised release arise from and are “treat[ed] . . . as part of the penalty for the initial offense.” Johnson v. United States, 529 U.S. 694, 700. Because a defendant’s final sentence includes any revocation sentence he may receive, §3583(k)’s 5-year mandatory minimum mirrors the unconstitutional sentencing enhancement in Alleyne. Second, the government suggests that Mr. Haymond’s sentence for violating the terms of his supervised release was actually fully authorized by the jury’s verdict, because his supervised release was from the outset always subject to the possibility of judicial revocation and §3583(k)’s mandatory prison sentence. But what is true in Apprendi and Alleyne can be no less true here: A mandatory minimum 5-year sentence that comes into play only as a result of additional judicial factual findings by a preponderance of the evidence cannot stand. Finally, the government contends that §3583(k)’s supervised release revocation procedures are practically identical to historic parole and probation revocation procedures, which have usually been understood to comport with the Fifth and Sixth Amendments. That argument overlooks a critical difference between §3583(k) and traditional parole and probation practices. Where parole and probation violations traditionally exposed a defendant only to the remaining prison term authorized for his crime of conviction, §3583(k) exposes a defendant to an additional mandatory minimum prison term beyond that authorized by the jury’s verdict—all based on facts found by a judge by a mere preponderance of the evidence. Pp. 11–18. (c) The Tenth Circuit may address on remand the question whether its remedy—declaring the last two sentences of §3583(k) “unconstitutional and unenforceable”—sweeps too broadly, including any question concerning whether the government’s argument to that effect was adequately preserved. Pp. 22–23. Justice Breyer agreed that the particular provision at issue, 18 U. S. C. §3583(k), is unconstitutional. Three features of §3583(k), considered together, make it less like ordinary supervised-release revocation and more like punishment for a new offense, to which the jury right would typically attach. First, §3583(k) applies only when a defendant commits a discrete set of criminal offenses specified in the statute. Second, §3583(k) takes away the judge’s discretion to decide whether violation of the conditions of supervised release should result in imprisonment and for how long. Third, §3583(k) limits the judge’s discretion in a particular manner: by imposing a mandatory minimum term of imprisonment of “not less than 5 years” upon a judge’s finding that a defendant has committed a listed offense. But because the role of the judge in a typical supervised-release proceeding is consistent with traditional parole and because Congress clearly did not intend the supervised release system to differ from parole in this respect, Justice Breyer would not transplant the Apprendi line of cases to the supervised-release context. Pp. 1–3. Gorsuch, J., announced the judgment of the Court and delivered an opinion, in which Ginsburg, Sotomayor, and Kagan, JJ., joined. Breyer, J., filed an opinion concurring in the judgment. Alito, J., filed a dissenting opinion, in which Roberts, C. J., and Thomas and Kavanaugh, JJ., joined.

in which Justice Ginsburg, Justice Sotomayor, and Justice Kagan joined. Only a jury, acting on proof beyond a reasonable doubt, may take a person’s liberty. That promise stands as one of the Constitution’s most vital protections against arbitrary government. Yet in this case a congressional statute compelled a federal judge to send a man to prison for a minimum of five years without empaneling a jury of his peers or requiring the government to prove his guilt beyond a reasonable doubt. As applied here, we do not hesitate to hold that the statute violates the Fifth and Sixth Amendments. I After a jury found Andre Haymond guilty of possessing child pornography in violation of federal law, the question turned to sentencing. The law authorized the district judge to impose a prison term of between zero and 10 years, 18 U. S. C. §2252(b)(2), and a period of supervised release of between 5 years and life, §3583(k). Because Mr. Haymond had no criminal history and was working to help support his mother who had suffered a stroke, the judge concluded that Mr. Haymond was “not going to get much out of being in prison” and sentenced him to a prison term of 38 months, followed by 10 years of supervised release. After completing his prison sentence, however, Mr. Haymond encountered trouble on supervised release. He sat for multiple polygraph tests in which he denied possessing or viewing child pornography, and each time the test indicated no deception. But when the government conducted an unannounced search of his computers and cellphone, it turned up 59 images that appeared to be child pornography. Based on that discovery, the government sought to revoke Mr. Haymond’s supervised release and secure a new and additional prison sentence. A hearing followed before a district judge acting without a jury, and under a preponderance of the evidence rather than a reasonable doubt standard. In light of expert testimony regarding the manner in which cellphones can “cache” images without the user’s knowledge, the judge found insufficient evidence to show that Mr. Haymond knowingly possessed 46 of the images. At the same time, the judge found it more likely than not that Mr. Haymond knowingly downloaded and possessed the remaining 13 images. With that, the question turned once more to sentencing. Under 18 U. S. C. §3583(e)(3), enacted as part of the Sentencing Reform Act of 1984, a district judge who finds that a defendant has violated the conditions of his supervised release normally may (but is not required to) impose a new prison term up to the maximum period of supervised release authorized by statute for the defendant’s original crime of conviction, subject to certain limits.[1] Under that provision, the judge in this case would have been free to sentence Mr. Haymond to between zero and two additional years in prison. But there was a complication. Under §3583(k), added to the Act in 2003 and amended in 2006, if a judge finds by a preponderance of the evidence that a defendant on supervised release committed one of several enumerated offenses, including the possession of child pornography, the judge must impose an additional prison term of at least five years and up to life without regard to the length of the prison term authorized for the defendant’s initial crime of conviction.[2] Because Mr. Haymond had committed an offense cov- ered by §3583(k), the judge felt bound to impose an additional prison term of at least five years. He did so, though, with reservations. It’s one thing, Judge Terence Kern said, for a judge proceeding under a preponderance of the evidence standard to revoke a defendant’s supervised release and order him to serve additional time in prison within the range already authorized by the defendant’s original conviction; after all, the jury’s verdict, reached under the reasonable doubt standard, permitted that much punishment. But the judge found it “ ‘repugnant’ ” that a statute might impose a new and additional “mandatory five-year” punishment without those traditional protections. Were it not for §3583(k)’s mandatory minimum, the judge added, he “probably would have sentenced in the range of two years or less.” On appeal to the Tenth Circuit, Mr. Haymond challenged both the factual support for his new punishment and its constitutionality. On the facts, the court of appeals held that the district court’s findings against Mr. Haymond were clearly erroneous in certain respects. Even so, the court concluded, just enough evidence remained to sustain a finding that Mr. Haymond had knowingly possessed the 13 images at issue, in violation of §3583(k). That left the question of the statute’s constitutionality, and there the Tenth Circuit concluded that §3583(k) violated the Fifth and Sixth Amendments. The court explained that a jury had convicted Mr. Haymond beyond a reasonable doubt of a crime carrying a prison term of zero to 10 years. Yet now Mr. Haymond faced a new potential prison term of five years to life. Because this new prison term included a new and higher mandatory minimum resting only on facts found by a judge by a preponderance of the evidence, the court held, the statute violated Mr. Haymond’s right to trial by jury. By way of remedy, the court held the last two sentences of §3583(k), which mandate a 5-year minimum prison term, “unconstitutional and unenforceable.” 869 F.3d 1153, 1168 (2017). The court then vacated Mr. Haymond’s revocation sentence and remanded the case to the district court for resentencing without regard to those provisions. In effect, the court of appeals left the district court free to issue a new sentence under the preexisting statute governing most every other supervised release violation, §3583(e). Following the Tenth Circuit’s directions, the district court proceeded to resentence Mr. Haymond to time served, as he had already been detained by that point for approximately 28 months. We granted review to consider the Tenth Circuit’s constitutional holding. 586 U. S. ___ (2018). II Together with the right to vote, those who wrote our Constitution considered the right to trial by jury “the heart and lungs, the mainspring and the center wheel” of our liberties, without which “the body must die; the watch must run down; the government must become arbitrary.” Letter from Clarendon to W. Pym (Jan. 27, 1766), in 1 Papers of John Adams 169 (R. Taylor ed. 1977). Just as the right to vote sought to preserve the people’s authority over their government’s executive and legislative functions, the right to a jury trial sought to preserve the people’s authority over its judicial functions. J. Adams, Diary Entry (Feb. 12, 1771), in 2 Diary and Autobiography of John Adams 3 (L. Butterfield ed. 1961); see also 2 J. Story, Commentaries on the Constitution §1779, pp. 540–541 (4th ed. 1873). Toward that end, the Framers adopted the Sixth Amendment’s promise that “[i]n all criminal prosecutions the accused shall enjoy the right to a speedy and public trial, by an impartial jury.” In the Fifth Amendment, they added that no one may be deprived of liberty without “due process of law.” Together, these pillars of the Bill of Rights ensure that the government must prove to a jury every criminal charge beyond a reasonable doubt, an ancient rule that has “extend[ed] down centuries.” Apprendi v. New Jersey, 530 U.S. 466, 477 (2000). But when does a “criminal prosecution” arise implicating the right to trial by jury beyond a reasonable doubt? At the founding, a “prosecution” of an individual simply referred to “the manner of [his] formal accusation.” 4 W. Blackstone, Commentaries on the Laws of England 298 (1769) (Blackstone); see also N. Webster, An American Dictionary of the English Language (1st ed. 1828) (defining “prosecution” as “the process of exhibiting formal charges against an offender before a legal tribunal”). And the concept of a “crime” was a broad one linked to punishment, amounting to those “acts to which the law affixes . . . punishment,” or, stated differently, those “element[s] in the wrong upon which the punishment is based.” 1 J. Bishop, Criminal Procedure §§80, 84, pp. 51–53 (2d ed. 1872) (Bishop); see also J. Archbold, Pleading and Evidence in Criminal Cases *106 (5th Am. ed. 1846) (Archbold) (discussing a crime as including any fact that “annexes a higher degree of punishment”); Blakely v. Washington, 542 U.S. 296, 309 (2004); Apprendi, 530 U. S., at 481. Consistent with these understandings, juries in our constitutional order exercise supervisory authority over the judicial function by limiting the judge’s power to punish. A judge’s authority to issue a sentence derives from, and is limited by, the jury’s factual findings of criminal conduct. In the early Republic, if an indictment or “accusation . . . lack[ed] any particular fact which the laws ma[d]e essential to the punishment,” it was treated as “no accusation” at all. 1 Bishop §87, at 55; see also 2 M. Hale, Pleas of the Crown *170 (1736); Archbold *106. And the “truth of every accusation” that was brought against a person had to “be confirmed by the unanimous suffrage of twelve of his equals and neighbours.” 4 Blackstone 343. Because the Constitution’s guarantees cannot mean less today than they did the day they were adopted, it remains the case today that a jury must find beyond a reasonable doubt every fact “ ‘which the law makes essential to [a] punishment’ ” that a judge might later seek to impose. Blakely, 542 U. S., at 304 (quoting 1 Bishop §87, at 55). For much of our history, the application of this rule of jury supervision proved pretty straightforward. At common law, crimes tended to carry with them specific sanctions, and “once the facts of the offense were determined by the jury, the judge was meant simply to impose the prescribed sentence.” Alleyne v. United States, 570 U.S. 99, 108 (2013) (plurality opinion) (internal quotation marks and brackets omitted). Even when judges did enjoy discretion to adjust a sentence based on judge-found aggravating or mitigating facts, they could not “ ‘swell the penalty above what the law ha[d] provided for the acts charged’ ” and found by the jury. Apprendi, 530 U. S., at 519 (Thomas, J., concurring) (quoting 1 Bishop §85, at 54); see also 1 J. Bishop, Criminal Law §§933–934(1), p. 690 (9th ed. 1923) (“[T]he court determines in each case what within the limits of the law shall be the punishment” (emphasis added)). In time, of course, legislatures adopted new laws allowing judges or parole boards to suspend part (parole) or all (probation) of a defendant’s prescribed prison term and afford him a period of conditional liberty as an “act of grace,” subject to revocation. Escoe v. Zerbst, 295 U.S. 490, 492 (1935); see Anderson v. Corall, 263 U.S. 193, 196–197 (1923). But here, too, the prison sentence a judge or parole board could impose for a parole or probation violation normally could not exceed the remaining balance of the term of imprisonment already authorized by the jury’s verdict. So even these developments did not usually implicate the historic concerns of the Fifth and Sixth Amendments. See Blakely, 542 U. S., at 309; Ap- prendi, 530 U. S., at 498 (Scalia, J., concurring); 4 Atty. Gen.’s Survey of Release Proc. 22 (1939); 2 id., at 333. More recent legislative innovations have raised harder questions. In Apprendi, for example, a jury convicted the defendant of a gun crime that carried a maximum prison sentence of 10 years. But then a judge sought to impose a longer sentence pursuant to a statute that authorized him to do so if he found, by a preponderance of the evidence, that the defendant had committed the crime with racial bias. Apprendi held this scheme unconstitutional. “[A]ny fact that increases the penalty for a crime beyond the prescribed statutory maximum,” this Court explained, “must be submitted to a jury, and proved beyond a reasonable doubt” or admitted by the defendant. 530 U. S., at 490. Nor may a State evade this traditional restraint on the judicial power by simply calling the process of finding new facts and imposing a new punishment a judicial “sentencing enhancement.” Id., at 495. “[T]he relevant inquiry is one not of form, but of effect—does the required [judicial] finding expose the defendant to a greater punishment than that authorized by the jury’s guilty verdict?” Id., at 494. While “trial practices ca[n] change in the course of centuries and still remain true to the principles that emerged from the Framers’ ” design, id., at 483, in the years since Apprendi this Court has not hesitated to strike down other innovations that fail to respect the jury’s supervisory function. See, e.g., Ring v. Arizona, 536 U.S. 584 (2002) (imposition of death penalty based on judicial factfinding); Blakely, 542 U. S., at 303 (mandatory state sentencing guidelines); Cunningham v. California, 549 U.S. 270 (2007) (same); United States v. Booker, 543 U.S. 220 (2005) (mandatory federal sentencing guidelines); Southern Union Co. v. United States, 567 U.S. 343 (2012) (imposition of criminal fines based on judicial factfinding).[3] Still, these decisions left an important gap. In Apprendi, this Court recognized that “ ‘[i]t is unconstitutional for a legislature to remove from the jury the assessment of facts that increase the prescribed range of penalties.’ ” 530 U. S., at 490. But by definition, a range of punishments includes not only a maximum but a minimum. And logically it would seem to follow that any facts necessary to increase a person’s minimum punishment (the “floor”) should be found by the jury no less than facts necessary to increase his maximum punishment (the “ceiling”). Before Apprendi, however, this Court had held that facts elevating the minimum punishment need not be proven to a jury beyond a reasonable doubt. McMillan v. Pennsylvania, 477 U.S. 79 (1986); see also Harris v. United States, 536 U.S. 545 (2002) (adhering to McMillan). Eventually, the Court confronted this anomaly in Alleyne. There, a jury convicted the defendant of a crime that ordinarily carried a sentence of five years to life in prison. But a separate statutory “sentencing enhancement” increased the mandatory minimum to seven years if the defendant “brandished” the gun. At sentencing, a judge found by a preponderance of the evidence that the defendant had indeed brandished a gun and imposed the mandatory minimum 7-year prison term. This Court reversed. Finding no basis in the original understanding of the Fifth and Sixth Amendments for McMillan and Harris, the Court expressly overruled those decisions and held that “the principle applied in Apprendi applies with equal force to facts increasing the mandatory minimum” as it does to facts increasing the statutory maximum penalty. Alleyne, 570 U. S., at 112. Nor did it matter to Alleyne’s analysis that, even without the mandatory minimum, the trial judge would have been free to impose a 7-year sentence because it fell within the statutory sentencing range authorized by the jury’s findings. Both the “floor” and “ceiling” of a sentencing range “define the legally prescribed penalty.” Ibid. And under our Constitution, when “a finding of fact alters the legally prescribed punishment so as to aggravate it” that finding must be made by a jury of the defendant’s peers beyond a reasonable doubt. Id., at 114. Along the way, the Court observed that there can be little doubt that “[e]levating the low end of a sentencing range heightens the loss of liberty associated with the crime: The defendant’s expected punishment has increased as a result of the narrowed range and the prosecution is empowered, by invoking the mandatory minimum, to require the judge to impose a higher punishment than he might wish.” Id., at 113 (internal quotation marks omitted). By now, the lesson for our case is clear. Based on the facts reflected in the jury’s verdict, Mr. Haymond faced a lawful prison term of between zero and 10 years under §2252(b)(2). But then a judge—acting without a jury and based only on a preponderance of the evidence—found that Mr. Haymond had engaged in additional conduct in violation of the terms of his supervised release. Under §3583(k), that judicial factfinding triggered a new punishment in the form of a prison term of at least five years and up to life. So just like the facts the judge found at the defendant’s sentencing hearing in Alleyne, the facts the judge found here increased “the legally prescribed range of allowable sentences” in violation of the Fifth and Sixth Amendments. Id., at 115. In this case, that meant Mr. Haymond faced a minimum of five years in prison instead of as little as none. Nor did the absence of a jury’s finding beyond a reasonable doubt only infringe the rights of the accused; it also divested the “ ‘people at large’ ”—the men and women who make up a jury of a defendant’s peers—of their constitutional authority to set the metes and bounds of judicially administered criminal punishments. Blakely, 542 U. S., at 306 (quoting Letter XV by the Federal Farmer (Jan. 18, 1788), in 2 The Complete Anti-Federalist 315, 320 (H. Storing ed. 1981)).[4] III In reply, the government and the dissent offer many and sometimes competing arguments, but we find none persuasive. A The government begins by pointing out that Alleyne arose in a different procedural posture. There, the trial judge applied a “sentencing enhancement” based on his own factual findings at the defendant’s initial sentencing hearing; meanwhile, Mr. Haymond received his new punishment from a judge at a hearing to consider the revocation of his term of supervised release. This procedural distinction makes all the difference, we are told, because the Sixth Amendment’s jury trial promise applies only to “criminal prosecutions,” which end with the issuance of a sentence and do not extend to “postjudgment sentence-administration proceedings.” Brief for United States 24; see also post, at 13–17 (Alito, J., dissenting) (echoing this argument). But we have been down this road before. Our precedents, Apprendi, Blakely, and Alleyne included, have repeatedly rejected efforts to dodge the demands of the Fifth and Sixth Amendments by the simple expedient of relabeling a criminal prosecution a “sentencing enhancement.” Calling part of a criminal prosecution a “sentence modification” imposed at a “postjudgment sentence-administration proceeding” can fare no better. As this Court has repeatedly explained, any “increase in a defendant’s authorized punishment contingent on the finding of a fact” requires a jury and proof beyond a reasonable doubt “no matter” what the government chooses to call the exercise. Ring, 536 U. S., at 602. To be sure, and as the government and dissent emphasize, founding-era prosecutions traditionally ended at final judgment. But at that time, generally, “questions of guilt and punishment both were resolved in a single proceeding” subject to the Fifth and Sixth Amendment’s demands. Douglass, Confronting Death: Sixth Amendment Rights at Capital Sentencing, 105 Colum. L. Rev. 1967, 2011 (2005); see also supra, at 7. Over time, procedures changed as legislatures sometimes bifurcated criminal prosecutions into separate trial and penalty phases. But none of these developments licensed judges to sentence individuals to punishments beyond the legal limits fixed by the facts found in the jury’s verdict. See ibid. To the contrary, we recognized in Apprendi and Alleyne, a “criminal prosecution” continues and the defendant remains an “accused” with all the rights provided by the Sixth Amendment, until a final sentence is imposed. See Apprendi, 530 U. S., at 481–482. Today, we merely acknowledge that an accused’s final sentence includes any supervised release sentence he may receive. Nor in saying that do we say anything new: This Court has already recognized that supervised release punishments arise from and are “treat[ed] . . . as part of the penalty for the initial offense.” Johnson v. United States, 529 U.S. 694, 700 (2000). The defendant receives a term of supervised release thanks to his initial offense, and whether that release is later revoked or sustained, it constitutes a part of the final sentence for his crime. As at the initial sentencing hearing, that does not mean a jury must find every fact in a revocation hearing that may affect the judge’s exercise of discretion within the range of punishments authorized by the jury’s verdict. But it does mean that a jury must find any facts that trigger a new mandatory minimum prison term.[5] This logic respects not only our precedents, but the original meaning of the jury trial right they seek to protect. The Constitution seeks to safeguard the people’s control over the business of judicial punishments by ensuring that any accusation triggering a new and addi- tional punishment is proven to the satisfaction of a jury beyond a reasonable doubt. By contrast, the view the government and dissent espouse would demote the jury from its historic role as “circuitbreaker in the State’s machinery of justice,” Blakely, 542 U. S., at 306, to “ ‘low-level gatekeeping,’ ” Booker, 543 U. S., at 230. If the government and dissent were correct, Congress could require anyone convicted of even a modest crime to serve a sentence of supervised release for the rest of his life. At that point, a judge could try and convict him of any violation of the terms of his release under a preponderance of the evidence standard, and then sentence him to pretty much anything. At oral argument, the government even con- ceded that, under its theory, a defendant on supervised re- lease would have no Sixth Amendment right to a jury trial when charged with an infraction carrying the death penalty. We continue to doubt whether even Apprendi’s fiercest critics “would advocate” such an “absurd result.” Blakely, 542 U. S., at 306.[6] B Where it previously suggested that Mr. Haymond’s supervised release revocation proceeding was entirely divorced from his criminal prosecution, the government next turns around and suggests that Mr. Haymond’s sentence for violating the terms of his supervised release was actually fully authorized by the jury’s verdict. See also post, at 7–8 (Alito, J., dissenting) (proposing a similar theory). After all, the government observes, on the strength of the jury’s findings the judge was entitled to impose as punishment a term of supervised release; and, in turn, that term of supervised release was from the outset always subject to the possibility of judicial revocation and §3583(k)’s mandatory prison sentence. Presto: Sixth Amendment problem solved. But we have been down this road too. In Apprendi and Alleyne, the jury’s verdict triggered a statute that authorized a judge at sentencing to increase the defendant’s term of imprisonment based on judge-found facts. This Court had no difficulty rejecting that scheme as an impermissible evasion of the historic rule that a jury must find all of the facts necessary to authorize a judicial punishment. See Alleyne, 570 U. S., at 117; Apprendi, 530 U. S., at 483. And what was true there can be no less true here: A mandatory minimum 5-year sentence that comes into play only as a result of additional judicial factual findings by a preponderance of the evidence cannot stand. This Court’s observation that “postrevocation sanctions” are “treat[ed] . . . as part of the penalty for the initial offense,” Johnson, 529 U. S., at 700, only highlights the constitutional infirmity of §3583(k): Treating Mr. Haymond’s 5-year mandatory minimum prison term as part of his sentence for his original offense makes clear that it mirrors the unconstitutional sentencing enhancement in Alleyne. See supra, at 12–13. Notice, too, that following the government down this road would lead to the same destination as the last: If the government were right, a jury’s conviction on one crime would (again) permit perpetual supervised release and allow the government to evade the need for another jury trial on any other offense the defendant might commit, no matter how grave the punishment. And if there’s any doubt about the incentives such a rule would create, consider this case. Instead of seeking a revocation of supervised release, the government could have chosen to prosecute Mr. Haymond under a statute mandating a term of imprisonment of 10 to 20 years for repeat child-pornography offenders. 18 U. S. C. §2252(b)(2). But why bother with an old-fashioned jury trial for a new crime when a quick-and-easy “supervised release revocation hearing” before a judge carries a penalty of five years to life? This displacement of the jury’s traditional supervi- sory role, under cover of a welter of new labels, exemplifies the “Framers’ fears that the jury right could be lost not only by gross denial, but by erosion.” Apprendi, 530 U. S., at 483 (internal quotation marks omitted). C Pivoting once more, the government and the dissent seem to accept for argument’s sake that “postjudgment sentence-administration proceedings” can implicate the Fifth and Sixth Amendments. See post, at 6–11. But, they contend, §3583(k)’s supervised release revocation procedures are practically identical to historic parole and probation revocation procedures. See, e.g., Gagnon v. Scarpelli, 411 U.S. 778 (1973); Morrissey v. Brewer, 408 U.S. 471 (1972). And, because those other procedures have usually been understood to comport with the Fifth and Sixth Amendments, they submit, §3583(k)’s procedures must do so as well. But this argument, too, rests on a faulty premise, overlooking a critical difference between §3583(k) and traditional parole and probation practices. Before the Sentencing Reform Act of 1984, a federal criminal defendant could serve as little as a third of his assigned prison term before becoming eligible for release on parole. See 18 U. S. C. §4205(a) (1982 ed.). Or he might avoid prison altogether in favor of probation. See §3561 (1982 ed.). If the defendant violated the terms of his parole or probation, a judge could send him to prison. But either way and as we’ve seen, a judge generally could sentence the defendant to serve only the remaining prison term authorized by statute for his original crime of conviction. See supra, at 7; Morrissey, 408 U. S., at 477 (“The essence of parole is release from prison, before the completion of sentence” (emphasis added)). Thus, a judge could not imprison a defendant for any longer than the jury’s factual findings allowed—a result entirely harmonious with the Fifth and Sixth Amendments. See Apprendi, 530 U. S., at 498 (Scalia, J., concurring); Blakely, 542 U. S., at 309. All that changed beginning in 1984. That year, Congress overhauled federal sentencing procedures to make prison terms more determinate and abolish the practice of parole. Now, when a defendant is sentenced to prison he generally must serve the great bulk of his assigned term. In parole’s place, Congress established the system of supervised release. But “[u]nlike parole,” supervised release wasn’t introduced to replace a portion of the defendant’s prison term, only to encourage rehabilitation after the completion of his prison term. United States Sentencing Commission, Guidelines Manual ch. 7, pt. A(2)(b) (Nov. 2012); see Doherty, Indeterminate Sentencing Returns: The Invention of Supervised Release, 88 N. Y. U. L. Rev. 958, 1024 (2013). In this case, that structural difference bears constitutional consequences. Where parole and probation violations generally exposed a defendant only to the remaining prison term authorized for his crime of conviction, as found by a unanimous jury under the reasonable doubt standard, supervised release violations subject to §3583(k) can, at least as applied in cases like ours, expose a defendant to an additional mandatory minimum prison term well beyond that authorized by the jury’s verdict—all based on facts found by a judge by a mere preponderance of the evidence. In fact, §3583(k) differs in this critical respect not only from parole and probation; it also represents a break from the supervised release practices that Congress authorized in §3583(e)(3) and that govern most federal criminal proceedings today. Unlike all those procedures, §3583(k) alone requires a substantial increase in the minimum sentence to which a defendant may be exposed based only on judge-found facts under a preponderance standard. And, as we explained in Alleyne and reaffirm today, that offends the Fifth and Sixth Amendments’ ancient protections.[7] D The dissent suggests an analogy between revocation under §3583(k) and prison disciplinary procedures that do not normally require the involvement of a jury. Post, at 19–20. But the analogy is a strained one: While the Sixth Amendment surely does not require a jury to find every fact that the government relies on to adjust the terms of a prisoner’s confinement (say, by reducing some of his privileges as a sanction for violating the prison rules), that does not mean the government can send a free man back to prison for years based on judge-found facts. Again, practice in the early Republic confirms this. At that time, a term of imprisonment may have been understood as encompassing a degree of summary discipline for alleged infractions of prison regulations without the involvement of a jury. See F. Gray, Prison Discipline in America 22–23, 48–49 (1848). But that does not mean any sanction, no matter how serious, would have been considered part and parcel of the original punishment. On the contrary, the few courts that grappled with this issue seem to have recognized that “infamous” punishments, such as a substantial additional term in prison, might implicate the right to trial by jury. See, e.g., Gross v. Rice, 71 Me. 241, 246–252 (1880); In re Edwards, 43 N. J. L. 555, 557–558 (1881). What’s more, a tradition of summary process in prison, where administrators face the “formidable task” of controlling a large group of potentially unruly prisoners, does not necessarily support the use of such summary process outside the prison walls. O’Lone v. Estate of Shabazz, 482 U.S. 342, 353 (1987); cf. Morrissey, 408 U. S., at 482. We have long held that prison regulations that impinge on the constitutional rights inmates would enjoy outside of prison must be “reasonably related to legitimate penological interests” in managing the prison. Turner v. Safley, 482 U.S. 78, 89 (1987). That approach, we have said, ensures that corrections officials can “ ‘anticipate security problems’ ” and address “ ‘the intractable problems of prison administration.’ ” O’Lone, 482 U. S., at 349; see also Dahne v. Richey, 587 U. S. ___, ___ (2019) (Alito, J., dissent- ing from denial of certiorari) (slip op., at 2) (“To maintain order, prison authorities may insist on compliance with rules that would not be permitted in the outside world”). Whether or not the Turner test applies to prisoners’ jury trial rights, we certainly have never extended it to the jury rights of persons out in the world who retain the core attributes of liberty. Cf. Griffin v. Wisconsin, 483 U.S. 868, 874, n. 2 (1987) (reserving question whether Turner applies to probation). Even the government has not asked us to do so today.[8] E Finally, much of the dissent is consumed by what it calls the “potentially revolutionary” consequences of our opinion. Post, at 1; see also post, at 15, 25 (calling our opinion “inexcusable,” “unpardonabl[e],” and “dangerous”); post, at 4 (our opinion threatens to bring “the whole concept of supervised release . . . crashing down”); post, at 9 (under our opinion, “the whole system of supervised release would be like a 40–ton truck speeding down a steep mountain road with no brakes”). But what agitates the dissent so much is an issue not presented here: whether all supervised release proceedings comport with Apprendi. As we have emphasized, our decision is limited to §3583(k)—an unusual provision enacted little more than a decade ago—and the Alleyne problem raised by its 5-year mandatory minimum term of imprisonment. See n. 7, supra. Section §3583(e), which governs supervised release revocation proceedings generally, does not contain any similar mandatory minimum triggered by judge-found facts. Besides, even if our opinion could be read to cast doubts on §3583(e) and its consistency with Apprendi, the practical consequences of a holding to that effect would not come close to fulfilling the dissent’s apocalyptic prophecy. In most cases (including this one), combining a defendant’s initial and post-revocation sentences issued under §3583(e) will not yield a term of imprisonment that exceeds the statutory maximum term of imprisonment the jury has authorized for the original crime of conviction. That’s because “courts rarely sentence defendants to the statutory maxima,” United States v. Caso, 723 F.3d 215, 224–225 (CADC 2013) (citing Sentencing Commission data indicating that only about 1% of defendants receive the maximum), and revocation penalties under §3583(e)(3) are only a small fraction of those available under §3583(k). So even if §3583(e)(3) turns out to raise Sixth Amendment issues in a small set of cases, it hardly follows that “as a practical matter supervised-release revocation proceedings cannot be held” or that “the whole idea of supervised release must fall.” Post, at 4–5. Indeed, the vast majority of supervised release revocation proceedings under subsec- tion (e)(3) would likely be unaffected. In the end, the dissent is left only to echo an age-old criticism: Jury trials are inconvenient for the government. Yet like much else in our Constitution, the jury system isn’t designed to promote efficiency but to protect liberty. In what now seems a prescient passage, Blackstone warned that the true threat to trial by jury would come less from “open attacks,” which “none will be so hardy as to make,” as from subtle “machinations, which may sap and undermine i[t] by introducing new and arbitrary methods.” 4 Blackstone 343. This Court has repeatedly sought to guard the historic role of the jury against such incursions. For “however convenient these may appear at first, (as doubtless all arbitrary powers, well executed, are the most convenient) yet let it be again remembered, that delays, and little inconveniences in the forms of justice, are the price that all free nations must pay for their lib- erty in more substantial matters.” Id., at 344.[9] IV Having concluded that the application of §3583(k)’s mandatory minimum in this case violated Mr. Haymond’s right to trial by jury, we face the question of remedy. Recall that the Tenth Circuit declared the last two sentences of §3583(k) “unconstitutional and unenforceable.” Those two sentences provide in relevant part that “[i]f a defendant required to register under [SORNA]” commits certain specified offenses, “the court shall revoke the term of supervised release and require the defendant to serve a term of imprisonment [of] not . . . less than 5 years.” Before us, the government suggests that the Tenth Circuit erred in declaring those two sentences “unenforceable.” That remedy, the government says, sweeps too broadly. In the government’s view, any constitutional infirmity can be cured simply by requiring juries acting under the reasonable doubt standard, rather than judges proceeding under the preponderance of the evidence standard, to find the facts necessary to trigger §3583(k)’s mandatory minimum. This remedy would be consistent with the statute’s terms, the government assures us, because “the court” authorized to revoke a term of supervised release in §3583(k) can and should be construed as embracing not only judges but also juries. And, the government insists, that means we should direct the court of appeals to send this case back to the district court so a jury may be empaneled to decide whether Mr. Haymond violated §3583(k). Unsurprisingly, Mr. Haymond contests all of this vigorously. We decline to tangle with the parties’ competing remedial arguments today. The Tenth Circuit did not address these arguments; it appears the government did not even discuss the possibility of empaneling a jury in its brief to that court; and this Court normally proceeds as a “court of review, not of first view,” Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7 (2005). Given all this, we believe the wiser course lies in returning the case to the court of appeals for it to have the opportunity to address the government’s remedial argument in the first instance, including any question concerning whether that argument was adequately preserved in this case. * The judgment of the court of appeals is vacated, and the case is remanded for further proceedings. It is so ordered. Notes 1 Section 3583(e)(3) states in pertinent part: “The court may, after considering the factors set forth in section 3553(a)(1), . . . (3) revoke a term of supervised release, and require the defendant to serve in prison all or part of the term of supervised release authorized by statute for the offense that resulted in such term of supervised release without credit for time previously served on postrelease supervision, if the court . . . finds by a preponderance of the evidence that the defendant violated a condition of supervised release, except that a defendant whose term is revoked under this paragraph may not be required to serve on any such revocation more than 5 years in prison if the offense that resulted in the term of supervised release is a class A felony, more than 3 years in prison if such offense is a class B felony, more than 2 years in prison if such offense is a class C or D felony, or more than one year in any other case . . . .” 2 Section 3583(k) provides: “Notwithstanding subsection (b), the authorized term of supervised release for any offense under section 1201 involving a minor victim, and for any offense under section 1591, 1594(c), 2241, 2242, 2243, 2244, 2245, 2250, 2251, 2251A, 2252, 2252A, 2260, 2421, 2422, 2423, or 2425, is any term of years not less than 5, or life. If a defendant required to register under the Sex Offender Registration and Notification Act (SORNA) commits any criminal offense under chapter 109A, 110, or 117, or section 1201 or 1591, for which imprisonment for a term longer than 1 year can be imposed, the court shall revoke the term of supervised release and require the defendant to serve a term of imprisonment under subsection (e)(3) without regard to the exception contained therein. Such term shall be not less than5 years.” 3 The Court has recognized two narrow exceptions to Apprendi’s general rule, neither of which is implicated here: Prosecutors need not prove to a jury the fact of a defendant’s prior conviction, Almendarez-Torres v. United States, 523 U.S. 224 (1998), or facts that affect whether a defendant with multiple sentences serves them concurrently or consecutively, Oregon v. Ice, 555 U.S. 160 (2009). 4 Because we hold that this mandatory minimum rendered Mr. Haymond’s sentence unconstitutional in violation of Alleyne v. United States, 570 U.S. 99 (2013), we need not address the constitutionality of the statute’s effect on his maximum sentence under Apprendi v. New Jersey, 530 U.S. 466 (2000). 5 The dissent asserts that “a sentence is ‘imposed’ at final judgment, not again and again every time a convicted criminal . . . violates a condition of his release.” Post, at 17 (opinion of Alito, J.) (citation omitted). But saying it does not make it so. As Johnson recognized, when a defendant is penalized for violating the terms of his supervised release, what the court is really doing is adjusting the defendant’s sentence for his original crime. Even the dissent recognizes that the sword of Damocles hangs over a defendant “every time [he] wakes up to serve a day of supervised release.” Post, at 17. 6 But perhaps we underestimate their fervor. While not openly embracing that result, the dissent fails to articulate any meaningful limiting principle to avoid it. If, as the dissent suggests, a term of supervised release is interchangeable with whatever sanction is prescribed for a violation, why stop at life in prison? The dissent replies that we might discover some relevant limitation in the Eighth Amendment, which does not mention jury trials, but is unwilling to find that limitation in the Sixth Amendment, which does. Post, at 8, n. 4. 7 Just as we have no occasion to decide whether §3583(k) implicates Apprendi by raising the ceiling of permissible punishments beyond those authorized by the jury’s verdict, see n. 4, supra, we do not pass judgment one way or the other on §3583(e)’s consistency with Apprendi. Nor do we express a view on the mandatory revocation provision for certain drug and gun violations in §3583(g), which requires courts to impose “a term of imprisonment” of unspecified length. 8 Contrary to the dissent’s characterization, we do not suggest that any prison discipline that is “too harsh” triggers the right to a jury trial. Post, at 20, n. 9 (emphasis deleted). Instead, we distinguish between altering a prisoner’s conditions of confinement, which gener-ally does not require a jury trial, and sentencing a free man to substantial additional time in prison, which generally does. 9 Justice Breyer agrees that a jury was required here for three reasons “considered in combination.” Post, at 2 (opinion concurring in judgment). Two of the reasons seem to amount to the same thing—a worry that §3583(k) imposes a new mandatory minimum sentence without a jury. And for the reasons we’ve already given, we can agree that this is indeed a problem under Alleyne. But Justice Breyer’s remaining reason is another story. He stresses that §3583(k)’s mandatory minimum applies only to a “discrete set of federal criminal of-fenses.” Post, at 2. But why should that matter? Whether the Sixth Amendment is violated in “discrete” instances or vast numbers, our duty to enforce the Constitution remains the same. Besides, any attempt to draw lines based on when an erosion of the jury trial right goes “too far” would prove inherently subjective and depend on judges’ intuitions about the proper role of the juries that are supposed to supervise them. As we have previously explained, “[w]hether the Sixth Amendment incorporates [such a] manipulable standard rather than Apprendi’s bright-line rule depends on the plausibility of the claim that the Framers would have left definition of the scope of jury power up to judges’ intuitive sense of how far is too far.” Blakely v. Washington, 542 U.S. 296, 308 (2004). And we continue to think that claim is “not plausible at all, because the very reason the Framers put a jury-trial guarantee in the Constitution” was to ensure the jury trial right would limit the power of judges and not be ground down to nothing through a balancing of interests by judges themselves. Ibid.

586.US.2018_17-765

Respondents Victor J. Stitt and Jason Daniel Sims were each convicted in federal court of unlawfully possessing a firearm, in violation of 18 U. S. C. §922(g)(1). The sentencing judge in each case imposed the mandatory minimum 15-year prison term that the Armed Career Criminal Act requires for §922(g)(1) offenders who have at least three previous convictions for certain “violent” or drug-related felonies, §924(e)(1). The Act defines “violent felony” to mean, among other things, “any crime punishable by imprisonment for a term exceeding one year . . . that . . . is burglary.” §924(e)(2)(B). Respondents’ prior convictions were for violations of state burglary statutes—a Tennessee statute in Stitt’s case and an Arkansas statute in Sims’ case—that prohibit burglary of a structure or vehicle that has been adapted or is customarily used for overnight accommodation. In both cases, the District Courts found that the state statutory crimes fell within the scope of the federal Act’s term “burglary.” The relevant Court of Appeals in each case disagreed, vacated the sentence, and remanded for resentencing. Held: 1. The term “burglary” in the Armed Career Criminal Act includes burglary of a structure or vehicle that has been adapted or is customarily used for overnight accommodation. Pp. 4–8. (a) In deciding whether an offense qualifies as a violent felony under the Act, the categorical approach first adopted in Taylor v. United States, 495 U. S. 575, requires courts to evaluate a prior state conviction by reference to the elements of the state offense, rather than to the defendant’s behavior on a particular occasion. A prior state conviction does not qualify as generic burglary under the Act where “the elements of [the relevant state statute] are broader than those of generic burglary.” Mathis v. United States, 579 U. S. ___, ___. Taylor, which specifically considered the statutory term “burglary” and defined the elements of generic burglary as “an unlawful or unprivileged entry into, or remaining in, a building or other structure, with intent to commit a crime,” 495 U. S., at 598, governs and determines the outcome here. Pp. 4–5. (b) The state statutes at issue here fall within the scope of Taylor’s definition of generic burglary. Congress intended that definition to reflect “the generic sense in which the term [was] used in the criminal codes of most States” when the Act was passed. 495 U. S., at 598. And at that time, a majority of state burglary statutes covered vehicles adapted or customarily used for lodging. Congress also viewed burglary as an inherently dangerous crime that “creates the possibility of a violent confrontation” between the offender and an occupant or someone who comes to investigate. Id., at 588. An offender who breaks into a mobile home, an RV, a camping tent, or another structure or vehicle that is adapted or customarily used for lodging creates a similar or greater risk of violent confrontation. Although the risk of violence is diminished if the vehicle is only used for lodging part of the time, the Court finds no reason to believe that Congress intended to make a part-time/full-time distinction. Respondents also argue that the vehicles covered here are analogous to the nontypical structures and vehicles that Taylor, Mathis, and other cases described as falling outside the scope of generic burglary, but none of those prior cases presented the question whether generic burglary includes structures or vehicles that are adapted or customarily used for overnight use. Pp. 5–8. 2. Sims’ case is remanded for further proceedings. His argument that Arkansas’ residential burglary statute is too broad to count as generic burglary because it also covers burglary of “a vehicle . . . [w]here any person lives,” Ark. Code Ann. §5–39–101(1)(A), rests in part upon state law, and the lower courts have not considered it. Those courts remain free to determine whether Sims properly presented that argument and, if so, to decide the merits. Pp. 8–9. No. 17–765, 860 F. 3d 854, reversed; No. 17–766, 854 F. 3d 1037, vacated and remanded. Breyer, J., delivered the opinion for a unanimous Court. Notes 1 Together with No. 17–766, United States v. Sims, on certiorari to the United States Court of Appeals for the Eighth Circuit.

The Armed Career Criminal Act requires a federal sentencing judge to impose upon certain persons convicted of unlawfully possessing a firearm a 15-year minimum prison term. The judge is to impose that special sentence if the offender also has three prior convictions for certain violent or drug-related crimes. 18 U. S. C. §924(e). Those prior convictions include convictions for “burglary.” §924(e)(2)(B)(ii). And the question here is whether the statutory term “burglary” includes burglary of a structure or vehicle that has been adapted or is customarily used for overnight accommodation. We hold that it does. I The consolidated cases before us involve two defendants, each of whom was convicted in a federal court of unlaw- fully possessing a firearm in violation of §922(g)(1). The maximum punishment for this offense is typically 10 years in prison. §924(a)(2). Each offender, however, had prior state burglary convictions sufficient, at least potentially, to require the sentencing judge to impose a mandatory 15-year minimum prison term under the Armed Career Criminal Act. That Act, as we have just said, requires an enhanced sentence for offenders who have at least three previous convictions for certain “violent” or drug-related felonies. §924(e)(1). Those prior felonies include “any crime” that is “punishable by imprisonment for a term exceeding one year” and that also “(i) has as an element the use, attempted use, or threatened use of physical force against the person of another; or “(ii) is burglary, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another.” §924(e)(2)(B) (emphasis added). The question here concerns the scope of the statutory word “burglary.” The relevant prior convictions of one of the unlawful firearms offenders, Victor J. Stitt, were for violations of a Tennessee statute that defines “[a]ggravated burglary” as “burglary of a habitation.” Tenn. Code Ann. §39–14–403(a) (1997). It further defines “[h]abitation” to include: (1) “any structure, including . . . mobile homes, trailers, and tents, which is designed or adapted for the overnight accommodation of persons,” and (2) any “self-propelled vehicle that is designed or adapted for the overnight accommodation of persons and is actually occupied at the time of initial entry by the defendant.” §§39–14–401(1)(A), (B) (emphasis added). The relevant prior convictions of the other unlawful firearms offender, Jason Daniel Sims, were for violations of an Arkansas statute that prohibits burglary of a “residential occupiable structure.” Ark. Code Ann. §5–39–201(a)(1) (Michie 1997). The statute defines “[r]esidential occupiable structure” to include: “a vehicle, building, or other structure: “(A)[w]here any person lives; or “(B)[w]hich is customarily used for overnight accommodation of persons whether or not a person is actually present.” §5–39–101(1) (emphasis added). In both cases, the District Courts found that the state statutory crimes fell within the scope of the word “bur- glary” in the Armed Career Criminal Act and consequently imposed that statute’s mandatory sentence enhancement. In both cases, the relevant Federal Court of Appeals held that the statutory crimes did not fall within the scope of the word “burglary,” vacated the sentence, and remanded for resentencing. See 860 F. 3d 854 (CA6 2017) (en banc) (reversing panel decision to the contrary); 854 F. 3d 1037 (CA8 2017). The Government asked us to grant certiorari to consider the question “[w]hether burglary of a nonpermanent or mobile structure that is adapted or used for overnight accommodation can qualify as ‘burglary’ under the Armed Career Criminal Act.” Pet. for Cert. in No. 17–765, p. i; Pet. for Cert. in No. 17–766, p. i. And, in light of uncertainty about the scope of the term “burglary” in the lower courts, we granted the Government’s request. Compare 860 F. 3d, at 862–863; 854 F. 3d, at 1040; United States v. White, 836 F. 3d 437, 446 (CA4 2016); United States v. Grisel, 488 F. 3d 844 (CA9 2007) (en banc), with Smith v. United States, 877 F. 3d 720, 724 (CA7 2017), cert. pending, No. 17–7517; United States v. Spring, 80 F. 3d 1450, 1462 (CA10 1996). II A The word “burglary,” like the word “crime” itself, is ambiguous. It might refer to a kind of crime, a generic crime, as set forth in a statute (“a burglary consists of behavior that . . . ”), or it might refer to the way in which an individual offender acted on a particular occasion (“on January 25, Jones committed a burglary on Oak Street in South San Francisco”). We have held that the words in the Armed Career Criminal Act do the first. Accordingly, we have held that the Act requires us to evaluate a prior state conviction “in terms of how the law defines the offense and not in terms of how an individual offender might have committed it on a particular occasion.” Begay v. United States, 553 U. S. 137, 141 (2008). A prior state conviction, we have said, does not qualify as generic burglary under the Act where “the elements of [the relevant state statute] are broader than those of generic burglary.” Mathis v. United States, 579 U. S. ___, ___ (2016) (slip op., at 19). The case in which we first adopted this “categorical approach” is Taylor v. United States, 495 U. S. 575 (1990). That case, which specifically considered the statutory term “burglary,” governs here and determines the outcome. In Taylor, we did more than hold that the word “bur- glary” refers to a kind of generic crime rather than to the defendant’s behavior on a particular occasion. We also explained, after examining the Act’s history and purpose, that Congress intended a “uniform definition of burglary [to] be applied to all cases in which the Government seeks” an enhanced sentence under the Act. Id., at 580–592. We held that this uniform definition includes “at least the ‘classic’ common-law definition,” namely, breaking and entering a dwelling at night with intent to commit a fel- ony. Id., at 593. But we added that it must include more. The classic definition, by excluding all places other than dwellings, we said, has “little relevance to modern law enforcement concerns.” Ibid. Perhaps for that reason, by the time the Act was passed in 1986, most States had expanded the meaning of burglary to include “structures other than dwellings.” Ibid. (citing W. LaFave & A. Scott, Substantive Criminal Law §§8.13(a)–(f) (1986)). In addition, the statute’s purpose, revealed by its language, ruled out limiting the scope of “burglary” to especially serious burglaries, e.g., those having elements that created a particularly serious risk of physical harm. If that had been Congress’s intent, adding the word “bur- glary” would have been unnecessary, since the (now-invalid) residual clause “already include[d] any crime that ‘involves conduct that presents a serious potential risk of physical injury to another.’ ” Taylor, 495 U. S., at 597 (quoting 18 U. S. C. §924(e)(2)(B)(ii)); see Johnson v. United States, 576 U. S. ___, ___–___ (2015) (slip op., at 5–10) (holding residual clause unconstitutionally vague). We concluded that the Act’s term “burglary” must include “ordinary,” “run-of-the-mill” burglaries as well as aggravated ones. Taylor, 495 U. S., at 597. And we defined the elements of generic “burglary” as “an unlawful or unprivileged entry into, or remaining in, a building or other structure, with intent to commit a crime.” Id., at 598. B The relevant language of the Tennessee and Arkansas statutes falls within the scope of generic burglary’s definition as set forth in Taylor. For one thing, we made clear in Taylor that Congress intended the definition of “bur- glary” to reflect “the generic sense in which the term [was] used in the criminal codes of most States” at the time the Act was passed. Ibid. In 1986, a majority of state bur- glary statutes covered vehicles adapted or customarily used for lodging—either explicitly or by defining “building” or “structure” to include those vehicles. See, e.g., N. H. Rev. Stat. Ann. §635:1 (1974) (prohibiting burglary of an “[o]ccupied structure,” defined to include “any structure, vehicle, boat or place adapted for overnight accommodation of persons”); Ore. Rev. Stat. §§164.205, 164.215, 164.225 (1985) (prohibiting burglary of a “building,” defined to include “any booth, vehicle, boat, aircraft or other structure adapted for overnight accommodation of persons”); see also ALI, Model Penal Code §§220.0(1), 221.1(1) (1980) (defining “ ‘occupied structure’ ” for purposes of burglary as “any structure, vehicle or place adapted for overnight accommodation of persons, or for carrying on business therein, whether or not a person is actually present”); Appendix, infra (collecting burglary statutes from 1986 or earlier that covered either vehicles adapted or customarily used for overnight accommodation or a broader class of vehicles). For another thing, Congress, as we said in Taylor, viewed burglary as an inherently dangerous crime because burglary “creates the possibility of a violent confrontation between the offender and an occupant, caretaker, or some other person who comes to investigate.” 495 U. S., at 588; see also James v. United States, 550 U. S. 192, 203 (2007). An offender who breaks into a mobile home, an RV, a camping tent, a vehicle, or another structure that is adapted for or customarily used for lodging runs a similar or greater risk of violent confrontation. See Spring, 80 F. 3d, at 1462 (noting the greater risk of confrontation in a mobile home or camper, where “it is more difficult for the burglar to enter or escape unnoticed”). Although, as respondents point out, the risk of violence is diminished if, for example, a vehicle is only used for lodging part of the time, we have no reason to believe that Congress intended to make a part-time/full-time distinction. After all, a burglary is no less a burglary because it took place at a summer home during the winter, or a commercial building during a holiday. Cf. Model Penal Code §221.1, Comment 3(b), p. 72 (burglary should cover places with the “apparent potential for regular occupancy”). Respondents make several additional arguments. Respondent Stitt argues that the Tennessee statute is too broad even under the Government’s definition of generic burglary. That is so, Stitt contends, because the statute covers the burglary of a “structure appurtenant to or connected with” a covered structure or vehicle, a provision that Stitt reads to include the burglary of even ordinary vehicles that are plugged in or otherwise appurtenant to covered structures. Tenn. Code Ann. §39–14–401(1)(C). Stitt’s interpretation, however, ignores that the “appurtenant to” provision extends only to “structure[s],” not to the separate statutory term “vehicle[s].” Ibid. We therefore disagree with Stitt’s argument that the “appurtenant to” provision sweeps more broadly than generic burglary, as defined in Taylor, 495 U. S., at 598. Respondents also point out that in Taylor, Mathis, and other cases, we said that burglary of certain nontypical structures and vehicles fell outside the scope of the federal Act’s statutory word “burglary.” See, e.g., Taylor, 495 U. S., at 599 (noting that some States “define burglary more broadly” than generic burglary by, for example, “including places, such as automobiles and vending machines, other than buildings”). And they argue that the vehicles covered here are analogous to the nontypical structures and vehicles to which the Court referred in those cases. Our examination of those cases, however, convinces us that we did not decide in either case the question now before us. In Taylor, for example, we referred to a Missouri breaking and entering statute that among other things criminalized breaking and entering “any boat or vessel, or railroad car.” Ibid. (citing Mo. Rev. Stat. §560.070 (1969); emphasis added). We did say that that particular provision was beyond the scope of the federal Act. But the statute used the word “any”; it referred to ordinary boats and vessels often at sea (and railroad cars often filled with cargo, not people), nowhere restricting its coverage, as here, to vehicles or structures customarily used or adapted for overnight accommodation. The statutes before us, by using these latter words, more clearly focus upon circumstances where burglary is likely to present a serious risk of violence. In Mathis, we considered an Iowa statute that covered “any building, structure, . . . land, water or air vehicle, or similar place adapted for overnight accommodation of persons [or used] for the storage or safekeeping of anything of value.” Iowa Code §702.12 (2013). Courts have construed that statute to cover ordinary vehicles because they can be used for storage or safekeeping. See State v. Buss, 325 N. W. 2d 384 (Iowa 1982); Weaver v. Iowa, 949 F. 2d 1049 (CA8 1991). That is presumably why, as we wrote in our opinion, “all parties agree[d]” that Iowa’s burglary statute “covers more conduct than generic burglary does.” Mathis, 579 U. S., at ___ (slip op., at 5). The question before us was whether federal generic “burglary” includes within its scope a burglary statute that lists multiple, alternative means of satisfying one element, some of which fall within Taylor’s generic definition and some of which fall outside it. We held, in light of the parties’ agreement that the Iowa statute covered some “outside” behavior (i.e., ordinary vehicles), that the statute did not count as a generic burglary statute. But for present purposes, what matters is that the Court in Mathis did not decide the question now before us—that is, whether coverage of vehicles designed or adapted for overnight use takes the statute outside the generic burglary definition. We now decide that latter question, and, for the reasons we have stated, we hold that it does not. III Respondent Sims argues that Arkansas’ residential burglary statute is too broad to count as generic burglary for a different reason, namely, because it also covers burglary of “a vehicle . . . [i]n which any person lives.” See supra, at 3. Sims adds that these words might cover a car in which a homeless person occasionally sleeps. Sims’ argument rests in part upon state law, and the lower courts have not considered it. As “we are a court of review, not of first view,” Cutter v. Wilkinson, 544 U. S. 709, 718, n. 7 (2005), we remand the Arkansas case to the lower courts for further proceedings. Those courts remain free to determine whether Sims properly presented the argument and to decide the merits, if appropriate. We reverse the judgment of the Sixth Circuit Court of Appeals. We vacate the judgment of the Eighth Circuit Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. APPENDIX Alaska Stat. §§11.46.300, 11.46.310, 11.81.900(b)(3) (1989) (effective 1978); Ariz. Rev. Stat. Ann. §§13–1501(7)–(8), 13–1507, 13–1508 (1978); Ark. Code Ann. §§41–2001(1), 41–2002 (Michie 1977); Cal. Penal Code Ann. §§459, 460 (West 1970); Colo. Rev. Stat. §§18–4–101(1)–(2), 18–4–202, 18–4–203 (1978); Conn. Gen. Stat. Ann. §§53a–100(a), 53a–101, 53a–103 (1985 Cum. Supp.); Del. Code Ann., Tit. 11, §§222(1), 824, 825 (1979); Fla. Stat. Ann. §§810.011(2), 810.02 (1976); Ga. Code Ann. §16–7–1(a) (1984); Idaho Code Ann. §18–1401 (1979); Ill. Comp. Stat., ch. 38, §19–1 (West 1985); Iowa Code §§702.12, 713.1 (1985); Kan. Stat. Ann. §§21–3715, 21–3716 (1988) (effective 1970); La. Rev. Stat. Ann. §14:62 (West 1974 Cum. Supp.); Me. Rev. Stat. Ann., Tit. 17–A, §§2(10), 2(24), 401 (1983); Mass. Gen. Laws Ann., ch. 266, §16A (West 1970); Mont. Code Ann. §§45–2–101(40), 45–6–204 (1983); Nev. Rev. Stat. Ann. §205.060 (1986); N. H. Rev. Stat. Ann. §635:1 (1974); N. J. Stat. Ann. §§2C:18–1, 2C:18–2 (West 1982); N. M. Stat. Ann. §§30–16–3, 30–16–4 (2018) (effective 1978); Ohio Rev. Code Ann. §§2909.01, 2911.11, 2911.12 (Lexis 1982); Okla. Stat., Tit. 21, §1435 (1983); Ore. Rev. Stat. §§164.205, 164.215, 164.225 (1985); Pa. Stat. Ann. Tit. 18, §§3501, 3502 (Purdon 1973); S. D. Codified Laws §§22–1–2(49), 22–32–1, 22–32–3, 22–32–8 (1988) (effective 1976); Tenn. Code Ann. §39–3–406 (1982); Tex. Penal Code Ann. §§30.01, 30.02 (West 1989) (effective 1974); Utah Code Ann. §§76–6–201(1), 76–6–202 (1978); W. Va. Code Ann. §61–3–11 (Lexis 1984); Wisc. Stat. Ann. §943.10(1) (West 1982).

587.US.2018_18-281

After the 2010 census, Virginia redrew legislative districts for the State’s Senate and House of Delegates. Voters in 12 impacted House districts sued two state agencies and four election officials (collectively, State Defendants), charging that the redrawn districts were racially gerrymandered in violation of the Fourteenth Amendment’s Equal Protection Clause. The House of Delegates and its Speaker (collectively, the House) intervened as defendants, participating in the bench trial, on appeal to this Court, and at a second bench trial, where a three-judge District Court held that 11 of the districts were unconstitutionally drawn, enjoined Virginia from conducting elections for those districts before adoption of a new plan, and gave the General Assembly several months to adopt that plan. Virginia’s Attorney General announced that the State would not pursue an appeal to this Court. The House, however, did file an appeal. Held: The House lacks standing, either to represent the State’s interests or in its own right. Pp. 3–12. (a) To cross the standing threshold, a litigant must show (1) a concrete and particularized injury, that (2) is fairly traceable to the challenged conduct, and (3) is likely to be redressed by a favorable decision. Hollingsworth v. Perry, 570 U.S. 693, 704. Standing must be met at every stage of the litigation, including on appeal. Arizonans for Official English v. Arizona, 520 U.S. 43, 64. And as a jurisdictional requirement, standing cannot be waived or forfeited. To appeal a decision that the primary party does not challenge, an intervenor must independently demonstrate standing. Wittman v. Personhuballah, 578 U. S. ___, ___. Pp. 3–4. (b) The House lacks standing to represent the State’s interests. The State itself had standing to press this appeal, see Diamond v. Charles, 476 U.S. 54, 62, and could have designated agents to do so, Hollingsworth, 570 U. S., at 710. However, the State did not designate the House to represent its interests here. Under Virginia law, authority and responsibility for representing the State’s interests in civil litigation rest exclusively with the State’s Attorney General. Virginia state courts permitted the House to intervene to defend legislation in Vesilind v. Virginia State Bd. of Elections, 295 Va. 427, 813 S.E.2d 739, but the House’s participation in Vesilind occurred in the same defensive posture as did the House’s participation in earlier phases of this case, when the House did not need to establish standing. Moreover, the House pointed to nothing in the Vesilind litigation suggesting that the Virginia courts understood the House to be representing the interests of the State itself. Karcher v. May, 484 U.S. 72, distinguished. Throughout this litigation, the House has purported to represent only its own interests. The House thus lacks authority to displace Virginia’s Attorney General as the State’s representative. Pp. 4–7. (c) The House also lacks standing to pursue this appeal in its own right. This Court has never held that a judicial decision invalidating a state law as unconstitutional inflicts a discrete, cognizable injury on each organ of government that participated in the law’s passage. Virginia’s Constitution allocates redistricting authority to the “General Assembly,” of which the House constitutes only a part. That fact distinguishes this case from Arizona State Legislature v. Arizona Independent Redistricting Comm’n, 576 U. S. ___, where Arizona’s House and Senate—acting together—had standing to challenge the constitutionality of a referendum that gave redistricting authority exclusively to an independent commission. The Arizona referendum was also assailed on the ground that it permanently deprived the legislative plaintiffs of their role in the redistricting process, while the order challenged here does not alter the General Assembly’s dominant initiating and ongoing redistricting role. Coleman v. Miller, 307 U.S. 433, also does not aid the House here, where the issue is the constitutionality of a concededly enacted redistricting plan, not the results of a legislative chamber’s poll or the validity of any counted or uncounted vote. Redrawing district lines indeed may affect the chamber’s membership, but the House as an institution has no cognizable interest in the identity of its members. The House has no prerogative to select its own members. It is a representative body composed of members chosen by the people. Changes in its membership brought about by the voting public thus inflict no cognizable injury on the House. Sixty-seventh Minnesota State Senate v. Beens, 406 U.S. 187, distinguished. Nor does a court order causing legislators to seek reelection in districts different from those they currently represent affect the House’s representational nature. Legislative districts change frequently, and the Virginia Constitution guards against representational confusion by providing that delegates continue to represent the districts that elected them, even if their reelection campaigns will be waged in different districts. In short, the State of Virginia would rather stop than fight on. One House of its bicameral legislature cannot alone continue the litigation against the will of its partners in the legislative process. Pp. 7–12. Appeal dismissed. Reported below: 326 F. Supp. 3d 128. Ginsburg, J., delivered the opinion of the Court, in which Thomas, Sotomayor, Kagan, and Gorsuch, JJ., joined. Alito, J., filed a dissenting opinion, in which Roberts, C. J., and Breyer and Kavanaugh, JJ., joined.

The Court resolves in this opinion a question of standing to appeal. In 2011, after the 2010 census, Virginia redrew legislative districts for the State’s Senate and House of Delegates. Voters in 12 of the impacted House districts sued two Virginia state agencies and four election officials (collectively, State Defendants) charging that the redrawn districts were racially gerrymandered in violation of the Fourteenth Amendment’s Equal Protection Clause. The Virginia House of Delegates and its Speaker (collectively, the House) intervened as defendants and carried the laboring oar in urging the constitutionality of the challenged districts at a bench trial, see Bethune-Hill v. Virginia State Bd. of Elections, 141 F. Supp. 3d 505 (ED Va. 2015), on appeal to this Court, see Bethune-Hill v. Virginia State Bd. of Elections, 580 U. S. ___ (2017), and at a second bench trial. In June 2018, after the second bench trial, a three-judge District Court in the Eastern District of Virginia, dividing 2 to 1, held that in 11 of the districts “the [S]tate ha[d] [unconstitutionally] sorted voters . . . based on the color of their skin.” Bethune-Hill v. Virginia State Bd. of Elections, 326 F. Supp. 3d 128, 180 (2018). The court therefore enjoined Virginia “from conducting any elections . . . for the office of Delegate . . . in the Challenged Districts until a new redistricting plan is adopted.” Id., at 227. Recognizing the General Assembly’s “primary jurisdiction” over redistricting, the District Court gave the General Assembly approximately four months to “adop[t] a new redistricting plan that eliminate[d] the constitutional infirmity.” Ibid. A few weeks after the three-judge District Court’s ruling, Virginia’s Attorney General announced, both publicly and in a filing with the District Court, that the State would not pursue an appeal to this Court. Continuing the litigation, the Attorney General concluded, “would not be in the best interest of the Commonwealth or its citizens.” Defendants’ Opposition to Intervenor-Defendants’ Motion to Stay Injunction Pending Appeal Under 28 U. S. C. §1253 in No. 3:14–cv–852 (ED Va.), Doc. 246, p. 1. The House, however, filed an appeal to this Court, App. to Juris. Statement 357–358, which the State Defendants moved to dismiss for want of standing. We postponed probable jurisdiction, 586 U. S. ___ (2018), and now grant the State Defendants’ motion. The House, we hold, lacks authority to displace Virginia’s Attorney General as representative of the State. We further hold that the House, as a single chamber of a bicameral legislature, has no standing to appeal the invalidation of the redistricting plan separately from the State of which it is a part.[1] I To reach the merits of a case, an Article III court must have jurisdiction. “One essential aspect of this requirement is that any person invoking the power of a federal court must demonstrate standing to do so.” Hollingsworth v. Perry, 570 U.S. 693, 704 (2013). The three elements of standing, this Court has reiterated, are (1) a concrete and particularized injury, that (2) is fairly traceable to the challenged conduct, and (3) is likely to be redressed by a favorable decision. Ibid. (citing Lujan v. Defenders of Wildlife, 504 U.S. 555, 560–561 (1992)). Although rulings on standing often turn on a plaintiff’s stake in initially filing suit, “Article III demands that an ‘actual contro- versy’ persist throughout all stages of litigation.” Hollingsworth, 570 U. S., at 705 (quoting Already, LLC v. Nike, Inc., 568 U.S. 85, 90–91 (2013)). The standing requirement therefore “must be met by persons seeking appellate review, just as it must be met by persons appearing in courts of first instance.” Arizonans for Official English v. Arizona, 520 U.S. 43, 64 (1997). As a jurisdictional requirement, standing to litigate cannot be waived or forfeited. And when standing is questioned by a court or an opposing party, the litigant invoking the court’s jurisdiction must do more than simply allege a nonobvious harm. See Wittman v. Personhuballah, 578 U. S. ___, ___–___ (2016) (slip op., at 5–6). To cross the standing threshold, the litigant must explain how the elements essential to standing are met. Before the District Court, the House participated in both bench trials as an intervenor in support of the State Defendants. And in the prior appeal to this Court, the House participated as an appellee. Because neither role entailed invoking a court’s jurisdiction, it was not previously incumbent on the House to demonstrate its standing. That situation changed when the House alone endeavored to appeal from the District Court’s order holding 11 districts unconstitutional, thereby seeking to invoke this Court’s jurisdiction. As the Court has repeatedly recognized, to appeal a decision that the primary party does not challenge, an intervenor must independently demonstrate standing. Wittman, 578 U. S. ___; Diamond v. Charles, 476 U.S. 54 (1986). We find unconvincing the House’s arguments that it has standing, either to represent the State’s interests or in its own right. II A The House urges first that it has standing to represent the State’s interests. Of course, “a State has standing to defend the constitutionality of its statute.” Id., at 62. No doubt, then, the State itself could press this appeal. And, as this Court has held, “a State must be able to designate agents to represent it in federal court.” Hollingsworth, 570 U. S., at 710. So if the State had designated the House to represent its interests, and if the House had in fact carried out that mission, we would agree that the House could stand in for the State. Neither precondition, however, is met here. To begin with, the House has not identified any legal basis for its claimed authority to litigate on the State’s behalf. Authority and responsibility for representing the State’s interests in civil litigation, Virginia law prescribes, rest exclusively with the State’s Attorney General: “All legal service in civil matters for the Commonwealth, the Governor, and every state department, institution, division, commission, board, bureau, agency, entity, official, court, or judge . . . shall be rendered and performed by the Attorney General, except as provided in this chapter and except for [certain judicial misconduct proceedings].” Va. Code Ann. §2.2–507(A) (2017).[2] Virginia has thus chosen to speak as a sovereign entity with a single voice. In this regard, the State has adopted an approach resembling that of the Federal Government, which “centraliz[es]” the decision whether to seek certiorari by “reserving litigation in this Court to the Attorney General and the Solicitor General.” United States v. Providence Journal Co., 485 U.S. 693, 706 (1988) (dismissing a writ of certiorari sought by a special prosecutor without authorization from the Solicitor General); see 28 U. S. C. §518(a); 28 CFR §0.20(a) (2018). Virginia, had it so chosen, could have authorized the House to litigate on the State’s behalf, either generally or in a defined class of cases. Hollingsworth, 570 U. S., at 710. Some States have done just that. Indiana, for example, empowers “[t]he House of Representatives and Senate of the Indiana General Assembly . . . to employ attorneys other than the Attorney General to defend any law enacted creating legislative or congressional districts for the State of Indiana.” Ind. Code §2–3–8–1 (2011). But the choice belongs to Virginia, and the House’s argument that it has authority to represent the State’s interests is foreclosed by the State’s contrary decision. The House observes that Virginia state courts have permitted it to intervene to defend legislation. But the sole case the House cites on this point—Vesilind v. Virginia State Bd. of Elections, 295 Va. 427, 813 S.E.2d 739 (2018)—does not bear the weight the House would place upon it. In Vesilind, the House intervened in support of defendants in the trial court, and continued to defend the trial court’s favorable judgment on appeal. Id., at 433–434, 813 S. E. 2d, at 742. The House’s participation in Vesilind thus occurred in the same defensive posture as did the House’s participation in earlier phases of this case, when the House did not need to establish standing. Moreover, the House has pointed to nothing in the Virginia courts’ decisions in the Vesilind litigation suggesting that the courts understood the House to be representing the interests of the State itself. Nonetheless, the House insists, this Court’s decision in Karcher v. May, 484 U.S. 72 (1987), dictates that we treat Vesilind as establishing conclusively the House’s authority to litigate on the State’s behalf. True, in Karcher, the Court noted a record, similar to that in Vesilind, of litigation by state legislative bodies in state court, and concluded without extensive explanation that “the New Jersey Legislature had authority under state law to represent the State’s interests . . . .” 484 U. S., at 82. Of crucial significance, however, the Court in Karcher noted no New Jersey statutory provision akin to Virginia’s law vesting the Attorney General with exclusive authority to speak for the Commonwealth in civil litigation. Karcher therefore scarcely impels the conclusion that, despite Virginia’s clear enactment making the Attorney General the State’s sole representative in civil litigation, Virginia has designated the House as its agent to assert the State’s interests in this Court. Moreover, even if, contrary to the governing statute, we indulged the assumption that Virginia had authorized the House to represent the State’s interests, as a factual matter the House never indicated in the District Court that it was appearing in that capacity. Throughout this litigation, the House has purported to represent its own interests. Thus, in its motion to intervene, the House observed that it was “the legislative body that actually drew the redistricting plan at issue,” and argued that the existing parties—including the State Defendants—could not adequately protect its interests. App. 2965–2967. Nowhere in its motion did the House suggest it was intervening as agent of the State. That silence undermines the House’s attempt to proceed before us on behalf of the State. As another portion of the Court’s Karcher decision clarifies, a party may not wear on appeal a hat different from the one it wore at trial. 484 U. S., at 78 (parties may not appeal in particular capacities “unless the record shows that they participated in those capacities below”).[3] B The House also maintains that, even if it lacks standing to pursue this appeal as the State’s agent, it has standing in its own right. To support standing, an injury must be “legally and judicially cognizable.” Raines v. Byrd, 521 U.S. 811, 819 (1997). This Court has never held that a judicial decision invalidating a state law as unconstitutional inflicts a discrete, cognizable injury on each organ of government that participated in the law’s passage. The Court’s precedent thus lends no support for the notion that one House of a bicameral legislature, resting solely on its role in the legislative process, may appeal on its own behalf a judgment invalidating a state enactment. Seeking to demonstrate its asserted injury, the House emphasizes its role in enacting redistricting legislation in particular. The House observes that, under Virginia law, “members of the Senate and of the House of Delegates of the General Assembly shall be elected from electoral districts established by the General Assembly.” Va. Const., Art. 2, §6. The House has standing, it contends, because it is “the legislative body that actually drew the redistricting plan,” and because, the House asserts, any remedial order will transfer redistricting authority from it to the District Court. Brief for Appellants 23, 26–28 (internal quotation marks omitted). But the Virginia constitutional provision the House cites allocates redistricting authority to the “General Assembly,” of which the House constitutes only a part. That fact distinguishes this case from Arizona State Legislature v. Arizona Independent Redistricting Comm’n, 576 U. S. ___ (2015), in which the Court recognized the standing of the Arizona House and Senate—acting to- gether—to challenge a referendum that gave redistricting authority exclusively to an independent commission, thereby allegedly usurping the legislature’s authority under the Federal Constitution over congressional redistricting. In contrast to this case, in Arizona State Legislature there was no mismatch between the body seeking to litigate and the body to which the relevant constitutional provision allegedly assigned exclusive redistricting authority. See 576 U. S., at ___–___ (slip op., at 11–12). Just as individual members lack standing to assert the institutional interests of a legislature, see Raines, 521 U. S., at 829,[4] a single House of a bicameral legislature lacks capacity to assert interests belonging to the legislature as a whole. Moreover, in Arizona State Legislature, the challenged referendum was assailed on the ground that it permanently deprived the legislative plaintiffs of their role in the redistricting process. Here, by contrast, the challenged order does not alter the General Assembly’s dominant initiating and ongoing role in redistricting. Compare Arizona State Legislature, 576 U. S., at ___ (slip op., at 14) (allegation of nullification of “any vote by the Legislature, now or in the future, purporting to adopt a redistricting plan” (internal quotation marks omitted)), with 326 F. Supp. 3d, at 227 (recognizing the General Assembly’s “primary jurisdiction” over redistricting and giving the General Assembly first crack at enacting a revised redistricting plan).[5] Nor does Coleman v. Miller, 307 U.S. 433 (1939), aid the House. There, the Court recognized the standing of 20 state legislators who voted against a resolution ratifying the proposed Child Labor Amendment to the Federal Constitution. Id., at 446. The resolution passed, the opposing legislators stated, only because the Lieutenant Governor cast a tie-breaking vote—a procedure the legislators argued was impermissible under Article V of the Federal Constitution. See Arizona State Legislature, 576 U. S., at ___–___ (slip op., at 13–14) (citing Coleman, 307 U. S., at 446). As the Court has since observed, Coleman stands “at most” “for the proposition that legislators whose votes would have been sufficient to defeat (or enact) a specific legislative Act have standing to sue if that legislative action goes into effect (or does not go into effect), on the ground that their votes have been completely nullified.” Raines, 521 U. S., at 823. Nothing of that sort happened here. Unlike Coleman, this case does not concern the results of a legislative chamber’s poll or the validity of any counted or uncounted vote. At issue here, instead, is the constitutionality of a concededly enacted redistricting plan. As we have already explained, a single House of a bicameral legislature generally lacks standing to appeal in cases of this order. Aside from its role in enacting the invalidated redistricting plan, the House, echoed by the dissent, see post, at 1–5, asserts that the House has standing because altered district boundaries may affect its composition. For support, the House and the dissent rely on Sixty-seventh Minnesota State Senate v. Beens, 406 U.S. 187 (1972) (per curiam), in which this Court allowed the Minnesota Senate to challenge a District Court malapportionment litigation order that reduced the Senate’s size from 67 to 35 members. The Court said in Beens: “[C]ertainly the [Minnesota Senate] is directly affected by the District Court’s orders,” rendering the Senate “an appropriate legal entity for purpose of intervention and, as a consequence, of an appeal in a case of this kind.” Id., at 194. Beens predated this Court’s decisions in Diamond v. Charles and other cases holding that intervenor status alone is insufficient to establish standing to appeal. Whether Beens established law on the question of standing, as distinct from intervention, is thus less than pellucid. But even assuming, arguendo, that Beens was, and remains, binding precedent on standing, the order there at issue injured the Minnesota Senate in a way the order challenged here does not injure the Virginia House. Cutting the size of a legislative chamber in half would necessarily alter its day-to-day operations. Among other things, leadership selection, committee structures, and voting rules would likely require alteration. By contrast, al- though redrawing district lines indeed may affect the membership of the chamber, the House as an institution has no cognizable interest in the identity of its members.[6] Although the House urges that changes to district lines will “profoundly disrupt its day-to-day operations,” Reply Brief 3, it is scarcely obvious how or why that is so. As the party invoking this Court’s jurisdiction, the House bears the burden of doing more than “simply alleg[ing] a nonobvious harm.” Wittman, 578 U. S., at ___ (slip op., at 6). Analogizing to “group[s] other than a legislative body,” the dissent insists that the House has suffered an “obvious” injury. Post, at 3. But groups like the string quartet and basketball team posited by the dissent select their own members. Similarly, the political parties involved in the cases the dissent cites, see post, at 3, n. 1 (citing New York State Bd. of Elections v. Lopez Torres, 552 U.S. 196, 202 (2008), and Eu v. San Francisco County Democratic Central Comm., 489 U.S. 214, 229–230 (1989)), select their own leadership and candidates. In stark contrast, the House does not select its own members. Instead, it is a representative body composed of members chosen by the people. Changes to its membership brought about by the voting public thus inflict no cognizable injury on the House.[7] The House additionally asserts injury from the creation of what it calls “divided constituencies,” suggesting that a court order causing legislators to seek reelection in districts different from those they currently represent affects the House’s representational nature. But legislative districts change frequently—indeed, after every decennial census—and the Virginia Constitution resolves any confusion over which district is being represented. It provides that delegates continue to represent the districts that elected them, even if their reelection campaigns will be waged in different districts. Va. Const., Art. 2, §6 (“A member in office at the time that a decennial redistricting law is enacted shall complete his term of office and shall continue to represent the district from which he was elected for the duration of such term of office . . . .”). We see little reason why the same would not hold true after districting changes caused by judicial decisions, and we thus foresee no representational confusion. And if harms centered on costlier or more difficult election campaigns are cognizable—a question that, as in Wittman, 578 U. S., at ___–___ (slip op., at 5–6), we need not decide today—those harms would be suffered by individual legislators or candidates, not by the House as a body. In short, Virginia would rather stop than fight on. One House of its bicameral legislature cannot alone continue the litigation against the will of its partners in the legislative process. * * * For the reasons stated, we dismiss the House’s appeal for lack of jurisdiction. It is so ordered. Notes 1 After the General Assembly failed to enact a new redistricting plan within the four months allowed by the District Court, that court entered a remedial order delineating districts for the 2019 election. The House has noticed an appeal to this Court from that order as well, and the State Defendants have moved to dismiss the follow-on appeal for lack of standing. See Virginia House of Delegates v. Bethune-Hill,No. 18–1134. In the appeal from the remedial order, the House and the State Defendants largely repeat the arguments on standing earlier advanced in this appeal. The House’s claim to standing to pursue an appeal from the remedial order fares no better than its assertion of standing here. See post, p. ___. 2 The exceptions referenced in the statute’s text are inapposite here. They include circumstances where, “in the opinion of the Attorney General, it is impracticable or uneconomical for [the] legal service to be rendered by him or one of his assistants,” or where the Virginia Supreme Court or any of its justices are litigating matters “arising out of [that court’s] official duties.” §2.2–507(C). 3 Nor can we give ear to the House’s assertion that forfeiture or acquiescence bar the State Defendants from contesting the House’s authority to represent the State’s interests. See Brief for Appellants 29–30. As earlier observed, standing to sue (or appeal) is a nonwaiv-able jurisdictional requirement. See supra, at 3. Moreover, even if forfeiture were not beyond the pale, the State Defendants here could hardly be held to have relinquished an objection to the House’s participation in a capacity—on behalf of the State itself—in which the House was not participating in the District Court. 4 Raines held that individual Members of Congress lacked standing to challenge the Line Item Veto Act. 5 Misplaced for similar reasons is the House’s reliance on this Court’s statements in INS v. Chadha, 462 U.S. 919, 929–931, and nn. 5–6, 939–940 (1983), that the United States House and Senate were “proper parties” or “adverse parties.” First, it is far from clear that the Court meant those terms to refer to standing, as opposed to the simple fact that both Houses of Congress had intervened. In any event, the statute at issue in Chadha granted each Chamber of Congress an ongoing power—to veto certain Executive Branch decisions—that each House could exercise independent of any other body. 6 The dissent urges that changes to district lines will alter the House’s future legislative output. See post, at 1–5. A legislative chamber as an institution, however, suffers no legally cognizable injury from changes to the content of legislation its future members may elect to enact. By contrast, the House has an obvious institutional interest in the manner in which it goes about its business. 7 The dissent further suggests that “we must assume that the districting plan enacted by the legislature embodies the House’s judgment” regarding the best way to select its members. Post, at 4. For the reasons explained supra, at 7–10, however, the House’s role in the legislative process does not give it standing to pursue this appeal.

587.US.2018_16-1275

Petitioner Virginia Uranium, Inc., wants to mine raw uranium ore from a site near Coles Hill, Virginia, but Virginia law flatly prohibits uranium mining in the Commonwealth. The company filed suit, alleging that, under the Constitution’s Supremacy Clause, the Atomic Energy Act (AEA) preempts state uranium mining laws like Virginia’s and ensconces the Nuclear Regulatory Commission (NRC) as the lone regulator in the field. Both the District Court and the Fourth Circuit rejected the company’s argument, finding that while the AEA affords the NRC considerable authority over the nuclear fuel life cycle, it offers no hint that Congress sought to strip States of their traditional power to regulate mining on private lands within their borders. Held: The judgment is affirmed. 848 F.3d 590, affirmed. Justice Gorsuch, joined by Justice Thomas and Justice Kavanaugh, concluded that the AEA does not preempt Virginia’s law banning uranium mining. Pp. 3–17. (a) Virginia Uranium claims that the AEA is best read to reserve to the NRC alone the regulation of uranium mining based on nuclear safety concerns. But the AEA contains no provision expressly preempting state law. More pointedly, it grants the NRC extensive and sometimes exclusive authority to regulate nearly every aspect of the nuclear fuel life cycle except mining, expressly stating that the NRC’s regulatory powers arise only “after [uranium’s] removal from its place of deposit in nature,” 42 U. S. C. §2092. And statutory context confirms this reading: If the federal government wants to control uranium mining on private land, it must purchase or seize the land by eminent domain and make it federal land, §2096, indicating that state authority remains untouched. Later amendments to the AEA point to the same conclusion. Section 2021 allows the NRC to devolve certain of its regulatory powers to the States but does nothing to extend the NRC’s power to activities, like mining, historically beyond its reach. And §2021(k) explains that States remain free to regulate the activities discussed in §2021 for purposes other than nuclear safety without the NRC’s consent. Virginia Uranium contends instead that subsection (k) greatly expands the AEA’s preemptive effect by demanding the displacement of any state law enacted for the purpose of protection the public against “radiation hazards.” But subsection (k) merely clarifies that nothing in §2021 limits States’ ability to regulate the activities subject to NRC control for other purposes. In addition, the company’s reading would prohibit not only the States from regulating uranium mining to protect against radiation hazards but the federal government as well, since the AEA affords it no authority to regulate uranium mining on private land. Pp. 4–7. (b) Virginia Uranium also submits that preemption may be found in this Court’s precedents, pointing to Pacific Gas & Elec. Co. v. State Energy Resources Conservation and Development Comm’n, 461 U.S. 190, which rejected a preemption challenge to a state law prohibiting the construction of new nuclear power plants after the Court observed that it was enacted out of concern with economic development, not for the purpose of addressing radiation safety hazards. But Pacific Gas concerned a state moratorium on construction of new nuclear power plants, and nuclear plant construction has always been an area exclusively regulated by the federal government. It is one thing to inquire exactingly into state legislative purposes when state law comes close to trenching on core federal powers; it is another thing altogether to insist on the same exacting scrutiny for state laws far removed from core NRC powers. Later cases confirm the propriety of restraint in this area. See, e.g., Silkwood v. Kerr-McGee Corp., 464 U.S. 238; English v. General Elec. Co., 496 U.S. 72. This Court has generally treated field preemption as depending on what the State did, not why it did it. See, e.g., Arizona v. United States, 567 U.S. 387. And because inquiries into legislative purpose both invite well-known conceptual and practical problems and pose risks to federalism and individual liberty, this Court has long warned against undertaking potential misadventures into hidden state legislative intentions without a clear statutory mandate for the project, see, e.g., Shady Grove Orthopedic Associates, P. A. v. Allstate Ins. Co., 559 U.S. 393, 404–405. Pp. 7–14. (c) Virginia Uranium alternatively suggests that that the AEA displaces state law through so-called conflict preemption—in particular, that Virginia’s mining law stands as an impermissible “obstacle to the accomplishment and execution of the full purposes and objectives of Congress,” Hines v. Davidowitz, 312 U.S. 52, 67. But any “[e]vidence of pre-emptive purpose,” whether express or implied, must be “sought in the [statute’s] text and structure.” CSX Transp., Inc. v. Easterwood, 507 U.S. 658, 664. Efforts to ascribe unenacted purposes and objectives to a federal statute face many of the same challenges as inquiries into state legislative intent. The only thing a court can be sure of is what can be found in the law itself. And the compromise that Congress actually struck in the AEA leaves mining regulation on private land to the States and grants the NRC regulatory authority only after uranium is removed from the earth. It is also unclear whether laws like Virginia’s might have a meaningful impact on the development of nuclear power in this country given the other available foreign and domestic sources of uranium. Pp. 14–17. Justice Ginsburg, joined by Justice Sotomayor and Justice Kagan, agreed with Justice Gorsuch that the Commonwealth’s mining ban is not preempted but concluded that his discussion of the perils of inquiring into legislative motive sweeps well beyond the confines of this case. Further, Virginia Uranium’s obstacle preemption arguments fail under existing doctrine, so there is little reason to question whether that doctrine should be retained. Pp. 1–14. (a) The Commonwealth has forbidden conventional uranium mining on private land. The AEA leaves that activity unregulated. State law on the subject is therefore not preempted, whatever the reason for the law’s enactment. Pp. 7–8. (b) Section 2021(k) lends no support for Virginia Uranium’s cause. That provision is most sensibly read to clarify that the door newly opened for state regulation of certain activities for nuclear safety purposes left in place pre-existing state authority to regulate activities for nonradiological purposes. House and Senate Reports endorse this reading of §2021(k). Pp. 8–9. (c) Virginia Uranium leans heavily on a statement in Pacific Gas & Elec. Co. v. State Energy Resources Conservation and Development Comm’n, 461 U.S. 190, that “the Federal Government has occupied the entire field of nuclear safety concerns.” Id., at 212. But neither in that case nor in later decisions in its wake—Silkwood v. Kerr-McGee Corp., 464 U.S. 238; English v. General Elec. Co., 496 U. S. 72—did the Court rest preemption on the purposes for which state laws were enacted. Indeed, in all three, the Court held that the laws at issue were not preempted. Moreover, the state law involved in Pacific Gas addressed an activity—construction of nuclear power plants—closely regulated by the AEA. Inquiry into why the state law at issue in that case was enacted was therefore proper under §2021(k). The Commonwealth’s mining ban, in contrast, governs an activity not regulated by the AEA. Pp. 9–10. (d) The Solicitor General’s argument—that the Commonwealth’s mining ban is preempted because it is a pretext for regulating the radiological safety hazards of milling and tailings storage—is unpersuasive. To the degree the AEA preempts state laws based on the purposes for which they were enacted, §2021(k) stakes out the boundaries of the preempted field. National Meat Assn. v. Harris, 565 U.S. 452, distinguished. Pp. 10–11. (e) Virginia Uranium and the United States also fail to show that the mining ban creates an “unacceptable ‘obstacle to the accomplishment and execution of the full purposes and objectives of Congress.’ ” Wyeth v. Levine, 555 U.S. 555, 563–564. Pp. 12–14. Gorsuch, J., announced the judgment of the Court and delivered an opinion, in which Thomas and Kavanaugh, JJ., joined. Ginsburg, J., filed an opinion concurring in the judgment, in which Sotomayor and Kagan, JJ., joined. Roberts, C. J., filed a dissenting opinion, in which Breyer and Alito, JJ., joined.

in which Justice Thomas and Justice Kavanaugh join. Virginia Uranium insists that the federal Atomic Energy Act preempts a state law banning uranium mining, but we do not see it. True, the AEA gives the Nuclear Regulatory Commission significant authority over the milling, transfer, use, and disposal of uranium, as well as the construction and operation of nuclear power plants. But Congress conspicuously chose to leave untouched the States’ historic authority over the regulation of mining activities on private lands within their borders. Nor do we see anything to suggest that the enforcement of Virginia’s law would frustrate the AEA’s purposes and objectives. And we are hardly free to extend a federal statute to a sphere Congress was well aware of but chose to leave alone. In this, as in any field of statutory interpretation, it is our duty to respect not only what Congress wrote but, as importantly, what it didn’t write. I Virginia Uranium thought its plan was pretty straightforward. First, the company wanted to use conventional mining techniques to extract raw uranium ore from a site near Coles Hill, Virginia. Next, it intended to mill that ore into a usable form. Typically performed at the mine site, milling involves grinding the ore into sand-sized grains and then exposing it to a chemical solution that leaches out pure uranium. Once dried, the resulting mixture forms a solid “yellowcake,” which the company planned to sell to enrichment facilities that produce fuel for nuclear reactors. Finally, because the leaching process does not remove all of the uranium from the ore, the company expected to store the leftover “tailings” near the mine to reduce the chances of contaminating the air or water. But putting the plan into action didn’t prove so simple. Pursuant to the AEA, ch. 724, 60Stat. 755, 42 U. S. C. §2011 et seq., the NRC regulates milling and tailing storage activities nationwide, and it has issued an array of rules on these subjects. See, e.g., 10 CFR §40 et seq. (2018). None of those, though, proved the real problem for Virginia Uranium. The company hit a roadblock even before it could get to the point where the NRC’s rules kick in: State law flatly prohibits uranium mining in Virginia. See Va. Code Ann. §§45.1–161.292:30, 45.1–283 (2013); 848 F.3d 590, 593–594 (CA4 2017). To overcome that obstacle, Virginia Uranium filed this lawsuit. The company alleged that, under the Constitution’s Supremacy Clause, the AEA preempts state uranium mining laws like Virginia’s and ensconces the NRC as the lone regulator in the field. And because the NRC’s regulations say nothing about uranium mining, the company continued, it remains free to mine as it will in Virginia or elsewhere. Both the district court and a divided panel of the Fourth Circuit rejected the company’s argument. The courts acknowledged that the AEA affords the NRC considerable authority over the nuclear fuel life cycle. But both courts found missing from the AEA any hint that Congress sought to strip States of their traditional power to regulate mining on private lands within their borders. Given the significance of the question presented, we granted review. 584 U. S. ___ (2018). II The Supremacy Clause supplies a rule of priority. It provides that the “Constitution, and the Laws of the United States which shall be made in Pursuance thereof,” are “the supreme Law of the Land . . . any Thing in the Constitution or Laws of any state to the Contrary notwithstanding.” Art. VI, cl. 2. This Court has sometimes used different labels to describe the different ways in which federal statutes may displace state laws—speaking, for example, of express, field, and conflict preemption. But these categories “are not rigidly distinct.” Crosby v. National Foreign Trade Council, 530 U.S. 363, 372, n. 6 (2000) (internal quotation marks omitted). And at least one feature unites them: Invoking some brooding federal interest or appealing to a judicial policy preference should never be enough to win preemption of a state law; a litigant must point specifically to “a constitutional text or a federal statute” that does the displacing or conflicts with state law. Puerto Rico Dept. of Consumer Affairs v. ISLA Petroleum Corp., 485 U.S. 495, 503 (1988); see also 3 J. Story, Commentaries on the Constitution of the United States §1831, p. 694 (1st ed. 1833) (“the supremacy of the laws is attached to those only, which are made in pursuance of the constitution”). Before us, Virginia Uranium contends that the AEA (and only the AEA) unseats state uranium mining regulations and that it does so under the doctrines of both field and conflict preemption. We examine these arguments about the AEA’s preemptive effect much as we would any other about statutory meaning, looking to the text and context of the law in question and guided by the traditional tools of statutory interpretation. Here, no more than in any statutory interpretation dispute, is it enough for any party or court to rest on a supposition (or wish) that “it must be in there somewhere.” A We begin with the company’s claim that the text and structure of the AEA reserve the regulation of uranium mining for the purpose of addressing nuclear safety concerns to the NRC alone—and almost immediately problems emerge. Unlike many federal statutes,[1] the AEA contains no provision preempting state law in so many words. Even more pointedly, the statute grants the NRC extensive and sometimes exclusive authority to regulate nearly every aspect of the nuclear fuel life cycle except mining. Companies like Virginia Uranium must abide the NRC’s rules and regulations if they wish to handle enriched uranium, to mill uranium ore or store tailings, or to build or run a nuclear power plant. See 42 U. S. C. §§2111(a), 2113(a), 2073. But when it comes to mining, the statute speaks very differently, expressly stating that the NRC’s regulatory powers arise only “after [uranium’s] removal from its place of deposit in nature.” §2092 (emphasis added). As the government itself has conceded, this means that “uranium mining” lies “outside the NRC’s jurisdiction,” Brief for United States as Amicus Curiae 14, and the agency’s grip takes hold only “at the mill, rather than at the mine,” In re Hydro Resources, Inc., 63 N. R. C. 510, 512 (2006). What the text states, context confirms. After announcing a general rule that mining regulation lies outside the NRC’s jurisdiction, the AEA carves out a notably narrow exception. On federal lands, the statute says, the NRC may regulate uranium mining. §2097. And if the federal government wants to control mining of uranium on private land, the AEA tells the NRC exactly what to do: It may purchase or seize the land by eminent domain and make it federal land. §2096. Congress thus has spoken directly to the question of uranium mining on private land, and every bit of what it’s said indicates that state authority remains untouched. Later amendments to the AEA point to the same conclusion. Some years after the statute’s passage, Congress added a provision, currently codified in §2021, allowing the NRC to devolve certain of its regulatory powers to the States. Unsurprisingly, Congress indicated that the NRC must maintain regulatory control over especially sensitive activities like the construction of nuclear power plants. §2021(c). But under §2021(b) the NRC may now, by agreement, pass to the States some of its preexisting authorities to regulate various nuclear materials “for the protection of the public health and safety from radiation hazards.” Out of apparent concern that courts might (mis)read these new provisions as prohibiting States from regulating any activity even tangentially related to nuclear power without first reaching an agreement with the NRC, Congress added subsection (k): “Nothing in this section [that is, §2021] shall be construed to affect the authority of any State or local agency to regulate activities for purposes other than protection against radiation hazards.” Section 2021, thus, did nothing to extend the NRC’s power to activities, like mining, historically beyond its reach. Instead, it served only to allow the NRC to share with the States some of the powers previously reserved to the federal government. Even then, the statute explained in subsection (k) that States remain free to regulate the activities discussed in §2021 for purposes other than nuclear safety without the NRC’s consent. Indeed, if anything, subsection (k) might be described as a non-preemption clause. Virginia Uranium’s case hinges on a very different construction of subsection (k). The company suggests that, properly read, the provision greatly expands the preemptive effect of the AEA and demands the displacement of any state law (touching on mining or any other subject) if that law was enacted for the purpose of protecting the public against “radiation hazards.” And, the company adds, Virginia’s law bears just such an impermissible purpose. In our view, this reading nearly turns the provision on its head. Subsection (k) does not displace traditional state regulation over mining or otherwise extend the NRC’s grasp to matters previously beyond its control. It does not expose every state law on every subject to a searching judicial inquiry into its latent purposes. Instead and much more modestly, it clarifies that “nothing in this [new] section [2021]”—a section allowing for the devolution-by-agreement of federal regulatory authority—should be construed to curtail the States’ ability to regulate the activities discussed in that same section for purposes other than protecting against radiation hazards. So only state laws that seek to regulate the activities discussed in §2021 without an NRC agreement—activities like the construction of nuclear power plants—may be scrutinized to ensure their purposes aim at something other than regulating nuclear safety. Really, to accomplish all it wants, Virginia Uranium would have to persuade us to read 13 words out of the statute and add 2 more: Nothing in this section shall be construed to affect the authority of any State or local agency to may regulate activities only for purposes other than protection against radiation hazards. That may be a statute some would prefer, but it is not the statute we have. Just consider what would follow from Virginia Uranium’s interpretation. Not only would States be prohibited from regulating uranium mining to protect against radiation hazards; the federal government likely would be barred from doing so as well. After all, the NRC has long believed, and still maintains, that the AEA affords it no authority to regulate uranium mining on private land. Nor does Virginia Uranium dispute the federal government’s understanding. Admittedly, if Virginia Uranium were to prevail here, the NRC might respond by changing course and seeking to regulate uranium mining for the first time. But given the statute’s terms, the prospects that it might do so successfully in the face of a legal challenge appear gloomy. Admittedly, as well, federal air and water and other regulations might apply at a uranium mine much as at any other workplace. But the possibility that both state and federal authorities would be left un- able to regulate the unique risks posed by an activity as potentially hazardous as uranium mining seems more than a little unlikely, and quite a lot to find buried deep in subsection (k). Talk about squeezing elephants into mouseholes. See Whitman v. American Trucking Assns., Inc., 531 U.S. 457, 468 (2001). B If the best reading of the AEA doesn’t require us to hold the state law before us preempted, Virginia Uranium takes another swing in the same direction. Only this time, the company submits, our precedents have adopted a different, even if maybe doubtful, reading of the AEA that we must follow. Most prominently, Virginia Uranium points to this Court’s decision in Pacific Gas & Elec. Co. v. State Energy Resources Conservation and Development Comm’n, 461 U.S. 190 (1983). But here, too, problems quickly appear. Pacific Gas rejected a preemption challenge to a state law prohibiting the construction of new nuclear power plants. Along the way, the Court expressly dismissed the notion that §2021 establishes the federal government as “the sole regulator of all matters nuclear.” Id., at 205. The Court observed that subsection (k) addresses itself only to “the preemptive effect of ‘this section,’ that is [§2021].” Id., at 210. And the Court acknowledged that subsection (k) does not “cut back on pre-existing state authority outside the NRC’s jurisdiction,” a field that surely includes uranium mining. Id., at 209. None of this remotely helps Virginia Uranium’s cause. Still, Virginia Uranium seeks to make the best of a bad situation. The company points out that Pacific Gas upheld the state law at issue there only after observing that it was enacted out of concern with economic development, not for the purpose of addressing radiation safety hazards. Id., at 205. From this, the company reasons, we should infer that any state law enacted with the purpose of addressing nuclear hazards must fall thanks to our precedent. But even that much does not follow. Since the passage of the AEA, the NRC has always played a significant role in regulating the construction of nuclear power plants. Indeed, under §2021(c) this remains one area where the NRC generally cannot devolve its responsibilities to the States. See id., at 197–198, 206–207. And because §2021 classifies the construction of nuclear power plants as one of the core remaining areas of special federal concern, any state law regulating that activity risks being subjected to an inquiry into its purposes under subsection (k). But the activity Virginia’s law regulates—mining on private land—isn’t one the AEA has ever addressed, and it isn’t one §2021 discusses, so subsection (k) does not authorize any judicial inquiry into state legislative purpose in this case. Admittedly, there is a wrinkle here. Pacific Gas seemed to accept California’s argument that its law addressed whether new power plants may be built, while the NRC’s regulatory power under §2021(c) extends only to the question how such plants are constructed and operated. Id., at 212. And accepting (without granting) these premises, it would appear that California’s law did not implicate an activity addressed by §2021, so an inquiry into state legislative purpose under subsection (k) was not statutorily authorized. Yet Pacific Gas inquired anyway, perhaps on the unstated belief that the state law just came “too close” to a core power §2021(c) reserves to the federal government. Does that mean we must do the same? Certainly Virginia Uranium sees it that way. We do not. Just because Pacific Gas may have made more of state legislative purposes than the terms of the AEA allow does not mean we must make more of them yet. It is one thing to do as Pacific Gas did and inquire exactingly into state legislative purposes when state law prohibits a regulated activity like the construction of a nuclear plant, and thus comes close to trenching on core federal powers reserved to the federal government by the AEA. It is another thing to do as Virginia Uranium wishes and impose the same exacting scrutiny on state laws prohibiting an activity like mining far removed from the NRC’s historic powers. And without some clearer congressional mandate suggesting an inquiry like that would be appropriate, we decline to undertake it on our own authority. The preemption of state laws represents “a serious intrusion into state sovereignty.” Medtronic, Inc. v. Lohr, 518 U.S. 470, 488 (1996) (plurality opinion). And to order preemption based not on the strength of a clear congressional command, or even on the strength of a judicial gloss requiring that much of us, but based only on a doubtful extension of a questionable judicial gloss would represent not only a significant federal intrusion into state sovereignty. It would also represent a significant judicial intrusion into Congress’s authority to delimit the preemptive effect of its laws. Being in for a dime doesn’t mean we have to be in for a dollar. This Court’s later cases confirm the propriety of restraint in this area. In a decision issued just a year after Pacific Gas (and by the same author), this Court considered whether the AEA preempted state tort remedies for radiation injuries after a nuclear plant accident. Silkwood v. Kerr-McGee Corp., 464 U.S. 238 (1984). In doing so, the Court did not inquire into state legislative purposes, apparently because it thought state tort law (unlike a law prohibiting the construction of a nuclear power plant) fell beyond any fair understanding of the NRC’s reach under the AEA. Id., at 251. Exactly the same, as we have seen, can be said of Virginia’s mining law. In fact, if the Silkwood Court had inquired into state legislative purposes, the law there might well have been harder to sustain than the one now before us. State tort laws, after all, plainly intend to regulate public safety. And as applied in Silkwood, state tort law sought to regulate the safety of a nuclear plant’s operations, an area of special federal interest under §2021(c). Id., at 256. Nothing comparable, of course, can be said of the mining regulations before us. Some years later, this Court in English v. General Elec. Co., 496 U.S. 72 (1990), went further still, casting doubt on whether an inquiry into state legislative purposes had been either necessary or appropriate in Pacific Gas itself. 496 U. S., at 84–85, n. 7 (“Whether the suggestion of the majority in Pacific Gas that legislative purpose is relevant to the definition of the pre-empted field is part of the holding of that case is not an issue before us today” (emphasis added)). If Pacific Gas and its progeny alone marked our path, this case might be a close one, as our dissenting colleagues suggest. Post, at 3–5 (opinion of Roberts, C. J.). But for us any lingering doubt dissipates when we consult other cases in this area and this Court’s traditional tools of statutory interpretation.[2] Start with the fact that this Court has generally treated field preemption inquiries like this one as depending on what the State did, not why it did it. Indeed, this Court has analyzed most every other modern field preemption doctrine dispute in this way—from immigration, Arizona v. United States, 567 U.S. 387 (2012), to arbitration, AT&T Mobility LLC v. Concepcion, 563 U.S. 333 (2011), to foreign affairs, Crosby v. National Foreign Trade Council, 530 U.S. 363 (2000), to railroads, Kurns v. Railroad Friction Products Corp., 565 U.S. 625 (2012), to energy, Hughes v. Talen Energy Marketing, LLC, 578 U. S. ___ (2016), to civil procedure, Shady Grove Orthopedic Associates, P. A. v. Allstate Ins. Co., 559 U.S. 393 (2010). It is unclear why we would proceed differently here without some clear congressional instruction requiring it.[3] Our field preemption cases proceed as they do, more- over, for good reasons. Consider just some of the costs to cooperative federalism and individual liberty we would invite by inquiring into state legislative purpose too precipitately. The natural tendency of regular federal judicial inquiries into state legislative intentions would be to stifle deliberation in state legislatures and encourage resort to secrecy and subterfuge. That would inhibit the sort of open and vigorous legislative debate that our Constitution recognizes as vital to testing ideas and improving laws. In Virginia Uranium’s vision as well, federal courts would have to allow depositions of state legislators and governors, and perhaps hale them into court for cross-examination at trial about their subjective motivations in passing a mining statute. And at the end of it all, federal courts would risk subjecting similarly situated persons to radically different legal rules as judges uphold and strike down materially identical state regulations based only on the happenstance of judicial assessments of the “true” intentions lurking behind them. In light of all this, it can surprise no one that our precedents have long warned against undertaking potential misadventures into hidden state legislative intentions without a clear statutory mandate for the project. See, e.g., Shady Grove, 559 U. S., at 404–405; Rowe v. New Hampshire Motor Transp. Assn., 552 U.S. 364, 373–374 (2008); Palmer v. Thompson, 403 U.S. 217, 225 (1971); Arizona v. California, 283 U.S. 423, 455, n. 7 (1931) (collecting cases). To be sure, Virginia Uranium insists that we don’t need to worry about concerns like these in this case. We don’t, the company says, because Virginia has admitted that it enacted its law with the (impermissible) purpose of pro- tecting the public from nuclear safety hazards. But the Commonwealth denies making any such admission. Instead, it says it has merely accepted as true the allegations in the company’s complaint about the intentions animating state law for purposes of the Commonwealth’s own motion to dismiss this suit under Federal Rule of Civil Procedure 12(b)(6). If the case were to proceed beyond the pleadings stage, Virginia insists, a more searching judicial inquiry into the law’s motivation would be inevitable. Whoever may be right about the status of Virginia’s admissions in this case, though, the point remains that following Virginia Uranium’s lead would require serious intrusions into state legislative processes in future cases. Beyond these concerns, as well, lie well-known conceptual and practical ones this Court has also advised against inviting unnecessarily. State legislatures are composed of individuals who often pursue legislation for multiple and unexpressed purposes, so what legal rules should determine when and how to ascribe a particular intention to a particular legislator? What if an impermissible intention existed but wasn’t necessary to her vote? And what percentage of the legislature must harbor the impermissible intention before we can impute it to the collective institution? Putting all that aside, how are courts supposed to conduct a reasonable inquiry into these questions when recorded state legislative history materials are often not as readily available or complete as their federal counterparts? And if trying to peer inside legislators’ skulls is too fraught an enterprise, shouldn’t we limit ourselves to trying to glean legislative purposes from the statutory text where we began? Even Pacific Gas warned future courts against too hastily accepting a litigant’s invitation to “become embroiled in attempting to ascertain” state legislative “motive[s],” acknowledging that such inquiries “often” prove “unsatisfactory venture[s]. What motivates one legislator to vote for a statute is not necessarily what motivates scores of others to enact it.” 461 U. S., at 216 (citation omitted). See also Shady Grove, 559 U. S., at 403–404, n. 6; Palmer, 403 U. S., at 225; Edwards v. Aguillard, 482 U.S. 578, 636–639 (1987) (Scalia, J., dissenting). Cf. Oncale v. Sundowner Offshore Services, Inc., 523 U.S. 75, 79 (1998). We think these warnings wise, and we heed them today. C If the AEA doesn’t occupy the field of radiation safety in uranium mining, Virginia Uranium suggests the statute still displaces state law through what’s sometimes called conflict preemption. In particular, the company suggests, Virginia’s mining law stands as an impermissible “obstacle to the accomplishment and execution of the full purposes and objectives of Congress.” Hines v. Davidowitz, 312 U.S. 52, 67 (1941). On Virginia Uranium’s account, Congress sought to capture the benefits of developing nuclear power while mitigating its safety and environmental costs. And, the company contends, Virginia’s moratorium disrupts the delicate “balance” Congress sought to achieve between these benefits and costs. Maybe the text of the AEA doesn’t touch on mining in so many words, but its authority to regulate later stages of the nuclear fuel life cycle would be effectively undermined if mining laws like Virginia’s were allowed. A sound preemption analysis cannot be as simplistic as that. No more than in field preemption can the Supremacy Clause be deployed here to elevate abstract and unenacted legislative desires above state law; only federal laws “made in pursuance of” the Constitution, through its prescribed processes of bicameralism and presentment, are entitled to preemptive effect. Art. VI, cl. 2; ISLA Petroleum, 485 U. S., at 503. So any “[e]vidence of pre-emptive purpose,” whether express or implied, must therefore be “sought in the text and structure of the statute at issue.” CSX Transp., Inc. v. Easterwood, 507 U.S. 658, 664 (1993). Sound and well-documented reasons underlie this rule too. Efforts to ascribe unenacted purposes and objectives to a federal statute face many of the same challenges as inquiries into state legislative intent. Trying to discern what motivates legislators individually and collectively invites speculation and risks overlooking the reality that individual Members of Congress often pursue multiple and competing purposes, many of which are compromised to secure a law’s passage and few of which are fully realized in the final product. Hefty inferences may be required, as well, when trying to estimate whether Congress would have wanted to prohibit States from pursuing regulations that may happen to touch, in various degrees and different ways, on unenacted federal purposes and objectives. Worse yet, in piling inference upon inference about hidden legislative wishes we risk displacing the legislative compromises actually reflected in the statutory text—compromises that sometimes may seem irrational to an outsider coming to the statute cold, but whose genius lies in having won the broad support our Constitution demands of any new law. In disregarding these legislative compromises, we may only wind up displacing perfectly legitimate state laws on the strength of “purposes” that only we can see, that may seem perfectly logical to us, but that lack the democratic provenance the Constitution demands before a federal law may be declared supreme. See, e.g., Pacific Gas, 461 U. S., at 222 (acknowledging that under the AEA “the promotion of nuclear power is not to be accomplished ‘at all costs’ ”); Cyan, Inc. v. Beaver County Employees Retirement Fund, 583 U. S. ___, ___–___ (2018) (slip op., at 14–15); Aguillard, 482 U. S., at 636–639 (Scalia, J., dissenting); United States v. O’Brien, 391 U.S. 367, 382–384 (1968); Fletcher v. Peck, 6 Cranch 87, 130 (1810). So it may be that Congress meant the AEA to promote the development of nuclear power. It may be that Congress meant the AEA to balance that goal against various safety concerns. But it also may be that Members of Congress held many other disparate or conflicting goals in mind when they voted to enact and amend the AEA, and many different views on exactly how to manage the competing costs and benefits. If polled, they might have reached very different assessments, as well, about the consistency of Virginia’s law with their own purposes and objectives. The only thing a court can be sure of is what can be found in the law itself. And every indication in the law before us suggests that Congress elected to leave mining regulation on private land to the States and grant the NRC regulatory authority only after uranium is removed from the earth. That compromise may not be the only permissible or even the most rationally attractive one, but it is surely both permissible and rational to think that Congress might have chosen to regulate the more novel aspects of nuclear power while leaving to States their traditional function of regulating mining activities on private lands within their boundaries.[4] As an alternative to proceeding down the purposes-and-objectives branch of conflict preemption, Virginia Uranium might have pursued another. Our cases have held that we can sometimes infer a congressional intent to displace a state law that makes compliance with a federal statute impossible. English, 496 U. S., at 79. But Virginia Uranium hasn’t pursued an argument along any of these lines, and understandably so. Not only can Virginia Uranium comply with both state and federal laws; it is also unclear whether laws like Virginia’s might have a meaningful impact on the development of nuclear power in this country. Some estimate that the United States currently imports over 90 percent of the uranium used in this country. App. to Pet. for Cert. 19a. Domestic uranium mines currently exist on federal lands as well and are thus beyond the reach of state authorities. Ibid. And if the federal government concludes that development of the Coles Hill deposit or any other like it is crucial, it may always purchase the site (or seize it through eminent domain) under the powers Congress has supplied. 42 U. S. C. §2096. All this may be done without even amending the AEA, itself another course which Congress is always free to pursue—but which this Court should never be tempted into pursuing on its own. * The judgment of the court of appeals is Affirmed. Notes 1 See, e.g., Chamber of Commerce of United States of America v. Whiting, 563 U.S. 582, 594–595 (2011); Geier v. American Honda Motor Co., 529 U.S. 861, 867 (2000). 2 Far from “sweep[ing] well beyond the confines of this case,” as our concurring colleagues suggest, see post, at 1 (Ginsburg, J., concurring in judgment), these considerations are, to us, essential to its resolution. 3 Certainly the dissent’s case, National Meat Assn. v. Harris, 565 U.S. 452 (2012), doesn’t command a different result. There, the Court merely enforced an express statutory preemption clause that prohibited States from setting standards for handling non-ambulatory pigs that differed from federal standards. As we’ve seen, the AEA contains no comparable preemption clause forbidding Virginia to regulate mining in any way. Admittedly, National Meat went on to say that a State could not enforce a preempted animal-handling standard indirectly by banning the sale of meat from non-ambulatory pigs if its law “function[ed] as a command to slaughterhouses to structure their operations in the exact way” state regulators desired rather than as federal standards required. Id., at 464. But here, by contrast, no one sug-gests that Virginia’s mining law requires anyone to disregard NRC regulations. 4 The concurrence takes a slightly different tack. It seems to accept the premise that the Court can divine the unenacted “purposes” and “objectives” underlying the AEA and weigh them against Virginia’s mining law. But in rejecting Virginia Uranium’s argument, it winds up emphasizing repeatedly that the text of the AEA does not address mining. See post, at 12–14. That may not fully address Virginia Uranium’s assertion that state mining regulations interfere with a latent statutory purpose lying beyond the text, but it does highlight the propriety of confining our inquiries to the statute’s terms.

586.US.2018_16-1498

The State of Washington taxes “motor vehicle fuel importer[s]” who bring large quantities of fuel into the State by “ground transportation.” Wash. Rev. Code §§82.36.010(4), (12), (16). Respondent Cougar Den, Inc., a wholesale fuel importer owned by a member of the Yakama Nation, imports fuel from Oregon over Washington’s public highways to the Yakama Reservation to sell to Yakama-owned retail gas stations located within the reservation. In 2013, the Washington State Department of Licensing assessed Cougar Den $3.6 million in taxes, penalties, and licensing fees for importing motor vehicle fuel into the State. Cougar Den appealed, arguing that the Washington tax, as applied to its activities, is pre-empted by an 1855 treaty between the United States and the Yakama Nation that, among other things, reserves the Yakamas’ “right, in common with citizens of the United States, to travel upon all public highways,” 12Stat. 953. A Washington Superior Court held that the tax was pre-empted, and the Washington Supreme Court affirmed. Held: The judgment is affirmed. 188 Wash. 2d 55, 392 P.3d 1014, affirmed. Justice Breyer, joined by Justice Sotomayor and Justice Kagan, concluded that the 1855 treaty between the United States and the Yakama Nation pre-empts the State of Washington’s fuel tax as applied to Cougar Den’s importation of fuel by public highway. Pp. 4–18. (a) The Washington statute at issue here taxes the importation of fuel by public highway. The Washington Supreme Court construed the statute that way in the decision below. That court wrote that the statute “taxes the importation of fuel, which is the transportation of fuel.” 188 Wash. 2d 55, 69, 392 P.3d 1014, 1020. It added that “travel on public highways is directly at issue because the tax [is] an importation tax.” Id., at 67, 392 P. 3d, at 1019. The incidence of a tax is a question of state law, Oklahoma Tax Comm’n v. Chickasaw Nation, 515 U.S. 450, 461, and this Court is bound by the Washington Supreme Court’s interpretation of Washington law, Johnson v. United States, 559 U.S. 133, 138. Nor is there any reason to doubt that the Washington Supreme Court meant what it said when it interpreted the statute. In the statute’s own words, Washington “impose[s] upon motor vehicle fuel licensees,” including “licensed importer[s],” a tax for “each gallon of motor vehicle fuel” that “enters into this state,” but only “if . . . entry is” by means of “a railcar, trailer, truck, or other equipment suitable for ground transportation.” Wash. Rev. Code §§82.36.010(4), 82.36.020(1), (2), 82.36.026(3). Thus, Cougar Den owed the tax because Cougar Den traveled with fuel by public highway. See App. 10a–26a; App. to Pet. for Cert. 55a. Pp. 4–10. (b) The State of Washington’s application of the tax to Cougar Den’s importation of fuel is pre-empted by the Yakama Nation’s reservation of “the right, in common with citizens of the United States, to travel upon all public highways.” This conclusion rests upon three considerations taken together. First, this Court has considered this treaty four times previously; each time it has considered language very similar to the language now before the Court; and each time it has stressed that the language of the treaty should be understood as bearing the meaning that the Yakamas understood it to have in 1855. See United States v. Winans, 198 U.S. 371, 380–381; Seufert Brothers Co. v. United States, 249 U.S. 194, 196–198; Tulee v. Washington, 315 U.S. 681, 683–685; Washington v. Washington State Commercial Passenger Fishing Vessel Assn., 443 U.S. 658, 677–678. Thus, al- though the words “in common with” on their face could be read to permit application to the Yakamas of general legislation (like the legislation at issue here) that applies to all citizens, this Court has refused to read “in common with” in this way because that is not what the Yakamas understood the words to mean in 1855. See Winans, 198 U. S., at 379, 381; Seufert Brothers, 249 U. S., at 198–199; Tulee, 315 U. S., at 684; Fishing Vessel, 443 U. S., at 679, 684–685. Second, the historical record adopted by the agency and the courts below indicates that the treaty negotiations and the United States’ representatives’ statements to the Yakamas would have led the Yakamas to understand that the treaty’s protection of the right to travel on the public highways included the right to travel with goods for purposes of trade. Third, to impose a tax upon traveling with certain goods burdens that travel. And the right to travel on the public highways without such burdens is just what the treaty protects. Therefore, precedent tells the Court that the tax must be pre-empted. In Tulee, for example, the fishing right reserved by the Yakamas in the treaty was held to pre-empt the application to the Yakamas of a state law requiring fishermen to buy fishing licenses. 315 U. S., at 684. The Court concluded that “such exaction of fees as a prerequisite to the enjoyment of” a right reserved in the treaty “cannot be reconciled with a fair construction of the treaty.” Id., at 685. If the cost of a fishing license interferes with the right to fish, so must a tax imposed on travel with goods (here fuel) interfere with the right to travel. Pp. 10–18. Justice Gorsuch, joined by Justice Ginsburg, concluded that the 1855 treaty guarantees tribal members the right to move their goods, including fuel, to and from market freely. When dealing with a tribal treaty, a court must “give effect to the terms as the Indians themselves would have understood them.” Minnesota v. Mille Lacs Band of Chippewa Indians, 526 U.S. 172, 196. The Yakamas’ understanding of the terms of the 1855 treaty can be found in a set of unchallenged factual findings in Yakama Indian Nation v. Flores, 955 F. Supp. 1229, which are binding here and sufficient to resolve this case. They provide “no evidence [suggesting] that the term ‘in common with’ placed Indians in the same category as non-Indians with respect to any tax or fee the latter must bear with respect to public roads.” Id., at 1247. Instead, they suggest that the Yakamas understood the treaty’s right-to-travel provision to provide them “with the right to travel on all public highways without being subject to any licensing and permitting fees related to the exercise of that right while engaged in the transportation of tribal goods.” Id., at 1262. A wealth of historical evidence confirms this understanding. “Far-reaching travel was an intrinsic ingredient in virtually every aspect of Yakama culture,” and travel for purposes of trade was so important to their “way of life that they could not have performed and functioned as a distinct culture” without it. Id., at 1238. Everyone then understood that the treaty would protect the Yakamas’ preexisting right to take goods to and from market freely throughout its traditional trading area. The State reads the treaty only as a promise to tribal members of the right to venture out of their reservation and use the public highways like everyone else. But the record shows that the consideration the Yakamas supplied—millions of acres desperately wanted by the United States to settle the Washington Territory—was worth far more than an abject promise they would not be made prisoners on their reservation. This Court’s cases interpreting the treaty’s neighboring and parallel right-to-fish provision further confirm this understanding. See, e.g., United States v. Winans, 198 U.S. 371. Pp. 1–11. Breyer, J., announced the judgment of the Court and delivered an opinion, in which Sotomayor and Kagan, JJ., joined. Gorsuch, J., filed an opinion concurring in the judgment, in which Ginsburg, J., joined. Roberts, C. J., filed a dissenting opinion, in which Thomas, Alito, and Kavanaugh, JJ., joined. Kavanaugh, J., filed a dissenting opinion, in which Thomas, J., joined.

in which Justice Sotomayor and Justice Kagan join. The State of Washington imposes a tax upon fuel importers who travel by public highway. The question before us is whether an 1855 treaty between the United States and the Yakama Nation forbids the State of Washington to impose that tax upon fuel importers who are members of the Yakama Nation. We conclude that it does, and we affirm the Washington Supreme Court’s similar decision. I A A Washington statute applies to “motor vehicle fuel importer[s]” who bring large quantities of fuel into the State by “ground transportation” such as a “railcar, trailer, [or] truck.” Wash. Rev. Code §§82.36.010(4), (12), (16) (2012). The statute requires each fuel importer to obtain a license, and it says that a fuel tax will be “levied and imposed upon motor vehicle fuel licensees” for “each gallon of motor vehicle fuel” that the licensee brings into the State. §§82.36.020(1), (2)(c). Licensed fuel importers who import fuel by ground transportation become liable to pay the tax as of the time the “fuel enters into this [S]tate.” §82.36.020(2)(c); see also §§82.38.020(4), (12), (15), (26), 82.38.030(1), (7)(c)(ii) (equivalent regulation of diesel fuel importers). But only those licensed fuel importers who import fuel by ground transportation are liable to pay the tax. §§82.36.026(3), 82.36.020(2)(c). For example, if a licensed fuel importer brings fuel into the State by pipeline, that fuel importer need not pay the tax. §§82.36.026(3), 82.36.020(2)(c)(ii), 82.36.010(3). Similarly, if a licensed fuel importer brings fuel into the State by vessel, that fuel importer need not pay the tax. §§82.36.026(3), 82.36.020(2)(c)(ii), 82.36.010(3). Instead, in each of those instances, the next purchaser or possessor of the fuel will pay the tax. §§82.36.020(2)(a), (b), (d). The only licensed fuel importers who must pay this tax are the fuel importers who bring fuel into the State by means of ground transportation. B The relevant treaty provides for the purchase by the United States of Yakama land. See Treaty Between the United States and the Yakama Nation of Indians, June 9, 1855, 12Stat. 951. Under the treaty, the Yakamas granted to the United States approximately 10 million acres of land in what is now the State of Washington, i.e., about one-fourth of the land that makes up the State today. Art. I, id., at 951–952; see also Brief for Respondent 4, 9. In return for this land, the United States paid the Yakamas $200,000, made improvements to the remaining Yakama land, such as building a hospital and schools for the Yakamas to use, and agreed to respect the Yakamas’ reservation of certain rights. Arts. III–V, 12Stat. 952–953. Those reserved rights include “the right, in common with citizens of the United States, to travel upon all public highways,” “the right of taking fish at all usual and accustomed places, in common with citizens of the Territory,” and other rights, such as the right to hunt, to gather roots and berries, and to pasture cattle on open and unclaimed land. Art. III, id., at 953. C Cougar Den, Inc., the respondent, is a wholesale fuel importer owned by a member of the Yakama Nation, incorporated under Yakama law, and designated by the Yakama Nation as its agent to obtain fuel for members of the Tribe. App. to Pet. for Cert. 63a–64a; App. 99a. Cougar Den buys fuel in Oregon, trucks the fuel over public highways to the Yakama Reservation in Washington, and then sells the fuel to Yakama-owned retail gas stations located within the reservation. App. to Pet. for Cert. 50a, 55a. Cougar Den believes that Washington’s fuel import tax, as applied to Cougar Den’s activities, is pre-empted by the treaty. App. 15a. In particular, Cougar Den believes that requiring it to pay the tax would infringe the Yakamas’ reserved “right, in common with citizens of the United States, to travel upon all public highways.” Art. III, 12Stat. 953. In December 2013, the Washington State Department of Licensing (Department), believing that the state tax was not pre-empted by the treaty, assessed Cougar Den $3.6 million in taxes, penalties, and licensing fees. App. to Pet. for Cert. 65a; App. 10a. Cougar Den appealed the assessment to higher authorities within the state agency. App. 15a. An Administrative Law Judge agreed with Cougar Den that the tax was pre-empted. App. to Brief in Opposition 14a. The Department’s Director, however, disagreed and overturned the ALJ’s order. App. to Pet. for Cert. 59a. A Washington Superior Court in turn disagreed with the director and held that the tax was pre-empted. Id., at 34a. The director appealed to the Washington Supreme Court. 188 Wash. 2d 55, 58, 392 P.3d 1014, 1015 (2017). And that court, agreeing with Cougar Den, upheld the Superior Court’s determination of pre-emption. Id., at 69, 392 P. 3d, at 1020. The Department filed a petition for certiorari asking us to review the State Supreme Court’s determination. And we agreed to do so. II A The Washington statute at issue here taxes the importation of fuel by public highway. The Washington Supreme Court construed the statute that way in the decision below. That court wrote that the statute “taxes the importation of fuel, which is the transportation of fuel.” Ibid. It added that “travel on public highways is directly at issue because the tax [is] an importation tax.” Id., at 67, 392 P. 3d, at 1019. Nor is there any reason to doubt that the Washington Supreme Court means what it said when it interpreted the Washington statute. We read the statute the same way. In the statute’s own words, Washington “impose[s] upon motor vehicle fuel licensees,” including “licensed importer[s],” a tax for “each gallon of motor vehicle fuel” that “enters into this state,” but only “if . . . entry is” by means of “a railcar, trailer, truck, or other equipment suitable for ground transportation.” Wash. Rev. Code §§82.36.010(4), 82.36.020(1), (2), 82.36.026(3). As is true of most tax laws, the statute is long and complex, and it is easy to stumble over this technical language. But if you are able to walk slowly through its provisions, the statute is easily followed. We need take only five steps. We start our journey at the beginning of the statute which first declares that “[t]here is hereby levied and imposed upon motor vehicle fuel licensees, other than motor vehicle fuel distributors, a tax at the rate . . . provided in [the statute] on each gallon of motor vehicle fuel.” §82.36.020(1). That is simple enough. Washington imposes a tax on a group of persons called “motor vehicle fuel licensees” for “each gallon of motor vehicle fuel.” Who are the “motor vehicle fuel licensees” that Washington taxes? We take a second step to find out. As the definitions section of the statute explains, the “motor vehicle fuel licensees” upon whom the tax is imposed are “person[s] holding a . . . motor vehicle fuel importer, motor vehicle fuel exporter, motor vehicle fuel blender, motor vehicle distributor, or international fuel tax agreement license.” §82.36.010(12). This, too, is easy to grasp. Not everyone who possesses motor vehicle fuel owes the tax. Instead, only motor vehicle fuel importers (and other similar movers and shakers within the motor vehicle fuel industry) who are licensed by the State to deal in fuel, must pay the tax. But must each of these motor vehicle fuel licensees pay the tax, so that the fuel is taxed as it passes from blender, to importer, to exporter, and so on? We take a third step, and learn that the answer is “no.” As the statute explains, “the tax shall be imposed at the time and place of the first taxable event and upon the first taxable person within this state.” §82.36.022. Reading that, we understand that only the first licensee who can be taxed, will be taxed. So, we ask, who is the first taxable licensee? Who must actually pay this tax? We take a fourth step to find out. Logic tells us that the first licensee who can be taxed will likely be the licensee who brings fuel into the State. But, the statute tells us that a “licensed importer” is “liable for and [must] pay tax to the department” when “[m]otor vehicle fuel enters into this state if . . . [t]he entry is not by bulk transfer.” §§82.36.020(2)(c), 82.36.026(3) (emphasis added). That is, a licensed importer can only be the first taxable licensee (and therefore the licensee that must pay the tax) if the importer brings fuel into the State by a method other than “bulk transfer.” But what is “bulk transfer”? What does it mean to say that licensed fuel importers need only pay the tax if they do not bring in fuel by “bulk transfer”? We take a fifth, and final, step to find out. “[B]ulk transfer,” the definitions section explains, “means a transfer of motor vehicle fuel by pipeline or vessel,” as opposed to “railcar, trailer, truck, or other equipment suitable for ground transportation.” §§82.36.010(3), (4). So, we learn that if the licensed fuel importer brings fuel into the State by ground transportation, then the fuel importer owes the tax. But if the licensed fuel importer brings fuel into the State by pipeline or vessel, then the importer will not be the first tax- able person to possess the fuel, and he will not owe the tax. In sum, Washington taxes travel by ground transportation with fuel. That feature sets the Washington statute apart from other statutes with which we are more familiar. It is not a tax on possession or importation. A statute that taxes possession would ordinarily require all people who own a good to pay the tax. A good example of that would be a State’s real estate property tax. That statute would require all homeowners to pay the tax, every year, regardless of the specifics of their situation. And a statute that taxes importation would ordinarily require all people who bring a good into the State to pay a tax. A good example of that would be a federal tax on newly manufactured cars. That statute would ordinarily require all people who bring a new car into the country to pay a tax. But Washington’s statute is different because it singles out ground transportation. That is, Washington does not just tax possession of fuel, or even importation of fuel, but instead taxes importation by ground transportation. The facts of this case provide a good example of the tax in operation. Each of the assessment orders that the Department sent to Cougar Den explained that Cougar Den owed the tax because Cougar Den traveled by highway. See App. 10a–26a; App. to Pet. for Cert. 55a. As the director explained, Cougar Den owed the tax because Cougar Den had caused fuel to enter “into this [S]tate at the Washington-Oregon boundary on the Highway 97 bridge” by means of a “tank truck” destined for “the Yakama Reservation.” Ibid. The director offers this explanation in addition to quoting the quantity of fuel that Cougar Den possessed because the element of travel by ground transportation is a necessary prerequisite to the imposition of the tax. Put another way, the State must prove that Cougar Den traveled by highway in order to apply its tax. B We are not convinced by the arguments raised to the contrary. The Department claims, and The Chief Justice agrees, that the state tax has little or nothing to do with the treaty because it is not a tax on travel with fuel but rather a tax on the possession of fuel. See Brief for Petitioner 26–28; post, at 5 (dissenting opinion). We cannot accept that characterization of the tax, however, for the Washington Supreme Court has authoritatively held that the statute is a tax on travel. The Washington Supreme Court held that the Washington law at issue here “taxes the importation of fuel, which is the transportation of fuel.” 188 Wash. 2d, at 69, 392 P. 3d, at 1020. It added that “travel on public highways is directly at issue because the tax [is] an importation tax.” Id., at 67, 392 P. 3d, at 1019. In so doing, the State Supreme Court heard, considered, and rejected the construction of the fuel tax that the Department advances here. See ibid., 392 P. 3d, at 1019 (“The Department argues, and the director agreed, that the taxes are assessed based on incidents of ownership or possession of fuel, and not incident to use of or travel on the roads or highways. . . . The Department’s argument is unpersuasive. . . . Here, travel on public highways is directly at issue because the tax was an importation tax”). The incidence of a tax is a question of state law, Oklahoma Tax Comm’n v. Chickasaw Nation, 515 U.S. 450, 461 (1995), and this Court is bound by the Washington Supreme Court’s interpretation of Washington law, Johnson v. United States, 559 U.S. 133, 138 (2010). We decline the Department’s invitation to overstep the bounds of our authority and construe the tax to mean what the Washington Supreme Court has said it does not. Nor would it make sense to construe the tax’s incidence differently. The Washington Supreme Court’s conclusion follows directly from its (and our) interpretation of how the tax operates. See supra, at 4–7. To be sure, it is generally true that fuel imported into the State by trucks driving the public highways can also be described as fuel that is possessed for the first time in the State. But to call the Washington statute a tax on “first possession” would give the law an over-inclusive label. As explained at length above, there are several ways in which a company could be a “first possessor” of fuel without incurring the tax. See ibid. For example, Cougar Den would not owe the tax had Cougar Den “first possessed” fuel by piping fuel from out of State into a Washington refinery. First possession is not taxed if the fuel is brought into the State by pipeline and bound for a refinery. §§82.36.026(3), 82.36.020(2)(c)(ii), 82.36.010(3). Similarly, Cougar Den would not owe the tax had Cougar Den “first possessed” fuel by bringing fuel into Washington through its waterways rather than its highways. First possession is not taxed if the fuel is brought into the State by vessel. §§82.36.026(3), 82.36.020(2)(c)(ii), 82.36.010(3). Thus, it seems rather clear that the tax cannot accurately be described as a tax on the first possession of fuel. But even if the contrary were true, the tax would still have the practical effect of burdening the Yakamas’ travel. Here, the Yakamas’ lone off-reservation act within the State is traveling along a public highway with fuel. The tax thus operates on the Yakamas exactly like a tax on transportation would: It falls upon them only because they happened to transport goods on a highway while en route to their reservation. And it is the practical effect of the state law that we have said makes the difference. We held, for instance, that the fishing rights reserved in the treaty pre-empted the State’s enforcement of a trespass law against Yakama fishermen crossing private land to access the river. See, e.g., United States v. Winans, 198 U.S. 371, 381 (1905). That was so even though the trespass law was not limited to those who trespass in order to fish but applied more broadly to any trespasser. Put another way, it mattered not that the tax was “on” trespassing rather than fishing because the tax operated upon the Yakamas when they were exercising their treaty-protected right. Ibid.; see also Tulee v. Washington, 315 U.S. 681, 685 (1942) (holding that the fishing rights reserved in the treaty pre-empted the State’s application of a fishing licensing fee to a Yakama fisherman, even though the fee also applied to types of fishing not practiced by the Yakamas). And this approach makes sense. When the Yakamas bargained in the treaty to protect their right to travel, they could only have cared about preventing the State from burdening their exercise of that right. To the Yakamas, it is thus irrelevant whether the State’s tax might apply to other activities beyond transportation. The only relevant question is whether the tax “act[ed] upon the Indians as a charge for exercising the very right their ancestors intended to reserve.” Tulee, 315 U. S., at 685. And the State’s tax here acted upon Cougar Den in exactly that way. For the same reason, we are unpersuaded by the Department’s insistence that it adopted this tax after a District Court, applying this Court’s decision in Chickasaw Nation, barred the State from taxing the sale of fuel products on tribal land. See Brief for Petitioner 6–7; Squaxin Island Tribe v. Stephens, 400 F. Supp. 2d 1250, 1262 (WD Wash. 2005). Although a State “generally is free to amend its law to shift the tax’s legal incidence,” Chickasaw Nation, 515 U. S., at 460, it may not burden a treaty-protected right in the process, as the State has done here. Thus, we must turn to the question whether this fuel tax, falling as it does upon members of the Tribe who travel on the public highways, violates the treaty. III A In our view, the State of Washington’s application of the fuel tax to Cougar Den’s importation of fuel is pre-empted by the treaty’s reservation to the Yakama Nation of “the right, in common with citizens of the United States, to travel upon all public highways.” We rest this conclusion upon three considerations taken together. First, this Court has considered this treaty four times previously; each time it has considered language very similar to the language before us; and each time it has stressed that the language of the treaty should be understood as bearing the meaning that the Yakamas understood it to have in 1855. See Winans, 198 U. S., at 380–381; Seufert Brothers Co. v. United States, 249 U.S. 194, 196–198 (1919); Tulee, 315 U. S., at 683–685; Washington v. Washington State Commercial Passenger Fishing Vessel Assn., 443 U.S. 658, 677–678 (1979). The treaty language at issue in each of the four cases is similar, though not identical, to the language before us. The cases focus upon language that guarantees to the Yakamas “the right of taking fish at all usual and accustomed places, in common with citizens of the Territory.” Art. III, para. 2, 12Stat. 953. Here, the language guarantees to the Yakamas “the right, in common with citizens of the United States, to travel upon all public highways.” Art. III, para. 1, ibid. The words “in common with” on their face could be read to permit application to the Yakamas of general legislation (like the legislation before us) that applies to all citizens, Yakama and non-Yakama alike. But this Court concluded the contrary because that is not what the Yakamas understood the words to mean in 1855. See Winans, 198 U. S., at 379, 381; Seufert Brothers, 249 U. S., at 198–199; Tulee, 315 U. S., at 684; Fishing Vessel, 443 U. S., at 679, 684–685. The cases base their reasoning in part upon the fact that the treaty negotiations were conducted in, and the treaty was written in, languages that put the Yakamas at a significant disadvantage. See, e.g., Winans, 198 U. S., at 380; Seufert Brothers, 249 U. S., at 198; Fishing Vessel, 443 U. S., at 667, n. 10. The parties negotiated the treaty in Chinook jargon, a trading language of about 300 words that no Tribe used as a primary language. App. 65a; Fishing Vessel, 443 U. S., at 667, n. 10. The parties memorialized the treaty in English, a language that the Yakamas could neither read nor write. And many of the representations that the United States made about the treaty had no adequate translation in the Yakamas’ own language. App. 68a–69a. Thus, in the year 1905, in Winans, this Court wrote that, to interpret the treaty, courts must focus upon the historical context in which it was written and signed. 198 U. S., at 381; see also Tulee, 315 U. S., at 684 (“It is our responsibility to see that the terms of the treaty are carried out, so far as possible, in accordance with the meaning they were understood to have by the tribal representatives at the council”); cf. Water Splash, Inc. v. Menon, 581 U. S. ___, ___ (2017) (slip op., at 8) (noting that, to ascertain the meaning of a treaty, courts “may look beyond the written words to the history of the treaty, the negotiations, and the practical construction adopted by the parties”) (internal quotation marks omitted). The Court added, in light of the Yakamas’ understanding in respect to the reservation of fishing rights, the treaty words “in common with” do not limit the reservation’s scope to a right against discrimination. Winans, 198 U. S., at 380–381. Instead, as we explained in Tulee, Winans held that “Article III [of the treaty] conferred upon the Yakimas continuing rights, beyond those which other citizens may enjoy, to fish at their ‘usual and accustomed places’ in the ceded area.” Tulee, 315 U. S., at 684 (citing Winans, 198 U.S. 371; emphasis added). Also compare, e.g., Fishing Vessel, 443 U. S., at 677, n. 22 (“Whatever opportunities the treaties assure Indians with respect to fish are admittedly not ‘equal’ to, but are to some extent greater than, those afforded other citizens” (emphasis added)), with post, at 4 (Kavanaugh, J., dissenting) (citing this same footnote in Fishing Vessel as support for the argument that the treaty guarantees the Yakamas only a right against discrimination). Construing the treaty as giving the Yakamas only antidiscrimination rights, rights that any inhabitant of the territory would have, would amount to “an impotent outcome to negotiations and a convention, which seemed to promise more and give the word of the Nation for more.” Winans, 198 U. S., at 380. Second, the historical record adopted by the agency and the courts below indicates that the right to travel includes a right to travel with goods for sale or distribution. See App. to Pet. for Cert. 33a; App. 56a–74a. When the United States and the Yakamas negotiated the treaty, both sides emphasized that the Yakamas needed to protect their freedom to travel so that they could continue to fish, to hunt, to gather food, and to trade. App. 65a–66a. The Yakamas maintained fisheries on the Columbia River, following the salmon runs as the fish moved through Yakama territory. Id., at 62a–63a. The Yakamas traveled to the nearby plains region to hunt buffalo. Id., at 61a. They traveled to the mountains to gather berries and roots. Ibid. The Yakamas’ religion and culture also depended on certain goods, such as buffalo byproducts and shellfish, which they could often obtain only through trade. Id., at 61a–62a. Indeed, the Yakamas formed part of a great trading network that stretched from the Indian tribes on the Northwest coast of North America to the plains tribes to the east. Ibid. The United States’ representatives at the treaty negotiations well understood these facts, including the importance of travel and trade to the Yakamas. Id., at 63a. They repeatedly assured the Yakamas that under the treaty the Yakamas would be able to travel outside their reservation on the roads that the United States built. Id., at 66a–67a; see also, e.g., id., at 66a (“ ‘[W]e give you the privilege of traveling over roads’ ”). And the United States repeatedly assured the Yakamas that they could travel along the roads for trading purposes. Id., at 65a–67a. Isaac Stevens, the Governor of the Washington Territory, told the Yakamas, for example, that, under the terms of the treaty, “You will be allowed to go on the roads, to take your things to market, your horses and cattle.” App. to Brief for Confederated Tribes and Bands of the Yakama Nation as Amicus Curiae 68a (record of the treaty proceedings). He added that the Yakamas “will be allowed to go to the usual fishing places and fish in common with the whites, and to get roots and berries and to kill game on land not occupied by the whites; all this outside the Reservation.” Ibid. Governor Stevens further urged the Yakamas to accept the United States’ proposals for reservation boundaries in part because the proposal put the Yakama Reservation in close proximity to public highways that would facilitate trade. He said, “ ‘You will be near the great road and can take your horses and your cattle down the river and to the [Puget] Sound to market.’ ” App. 66a. In a word, the treaty negotiations and the United States’ representatives’ statements to the Yakamas would have led the Yakamas to understand that the treaty’s protection of the right to travel on the public highways included the right to travel with goods for purposes of trade. We consequently so construe the relevant treaty provision. Third, to impose a tax upon traveling with certain goods burdens that travel. And the right to travel on the public highways without such burdens is, as we have said, just what the treaty protects. Therefore, our precedents tell us that the tax must be pre-empted. In Tulee, for example, we held that the fishing right reserved by the Yakamas in the treaty pre-empted the application to the Yakamas of a state law requiring fishermen to buy fishing licenses. 315 U. S., at 684. We concluded that “such exaction of fees as a prerequisite to the enjoyment of ” a right reserved in the treaty “cannot be reconciled with a fair construction of the treaty.” Id., at 685. If the cost of a fishing license interferes with the right to fish, so must a tax imposed on travel with goods (here fuel) interfere with the right to travel. We consequently conclude that Washington’s fuel tax “acts upon the Indians as a charge for exercising the very right their ancestors intended to reserve.” Ibid. Washington’s fuel tax cannot lawfully be assessed against Cougar Den on the facts here. Treaties with federally recognized Indian tribes—like the treaty at issue here—constitute federal law that pre-empts conflicting state law as applied to off-reservation activity by Indians. Cf. Mescalero Apache Tribe v. Jones, 411 U.S. 145, 148–149 (1973). B Again, we are not convinced by the arguments raised to the contrary. The Chief Justice concedes that “the right to travel with goods is just an application of the Yakamas’ right to travel.” Post, at 2 (dissenting opinion); see also ibid. (“It ensures that the Yakamas enjoy the same privileges when they travel with goods as when they travel without them.”). But he nevertheless insists that, because of the way in which the Washington statute taxes fuel, the statute does not interfere with the right to travel reserved by the Yakamas in the treaty. Post, at 3. First, The Chief Justice finds it significant that “[t]he tax is calculated per gallon of fuel; not, like a toll, per vehicle or distance traveled.” Ibid., see also ibid. (“The tax before us does not resemble a blockade or a toll”). But that argument fails on its own terms. A toll on highway travel is no less a toll when the toll varies based on the number of axels on a vehicle traveling the highway, or on the number of people traveling in the vehicle. We cannot, therefore, see why the number of gallons of fuel that the vehicle carries should make all the difference. Put another way, the fact that a tax on travel varies based on the features of that travel does not mean that the tax is not a tax on travel. Second, The Chief Justice argues that it “makes no sense,” for example, to hold that “a tax on certain luxury goods” that is assessed the first time the goods are possessed in Washington cannot apply to a Yakama member “who buys” a mink coat “over the state line in Portland and then drives back to the reservation,” but the tax can apply to a Yakama member who “buys a mink coat at an off-reservation store in Washington.” Post, at 4. The short, conclusive answer to this argument is that there is a treaty that forbids taxing Yakama travel on highways with goods (e.g., fuel, or even furs) for market; and there is no treaty that forbids taxing Yakama off-reservation purchases of goods. Indeed, if our precedents supported The Chief Justice’s rule, then our fishing rights cases would have turned on whether Washington also taxed fish purchased in the grocery store. Compare, e.g., Tulee, 315 U. S., at 682, n. 1 (holding that the fishing right reserved by the Yakamas in the treaty pre-empted the application to the Yakamas of a state law which prohibited “ ‘catch[ing] . . . fish for food’ ” without having purchased a license). But in those cases, we did not look to whether fish were taxed elsewhere in Washington. That is because the treaty does not protect the Yakamas from state sales taxes imposed on the off-reservation sale of goods. Instead, the treaty protects the Yakamas’ right to travel the public highways without paying state taxes on that activ- ity, much like the treaty protects the Yakamas’ right to fish without paying state taxes on that activity. Third, The Chief Justice argues that only a law that “punished or charged the Yakamas” for an “integral feature” of a treaty right could be pre-empted by the treaty. Post, at 6. But that is true of the Washington statute at issue here. The treaty protects the right to travel with goods, see supra, at 10–14, and the Washington statute taxes travel with goods, see supra, at 4–7. Therefore, the statute charges the Yakamas for an “integral feature” of a treaty right. But even if the statute indirectly burdened a treaty right, under our precedents, the statute would still be pre-empted. One of the Washington statutes at issue in Winans was not a fishing regulation, but instead a trespassing statute. That trespassing statute indirectly burdened the right to fish by preventing the Yakamas from crossing privately owned land so that the Yakamas could reach their traditional fishing places and camp on that private property during the fishing season. See 198 U. S., at 380–381. It cannot be true that a law prohibiting trespassing imposed a burden on the right to fish that is “integral” enough to be pre-empted by the treaty, while a law taxing goods carried to the reservation on the public highway imposes a burden on the right to travel that is too attenuated to be pre-empted by the treaty. C Although we hold that the treaty protects the right to travel on the public highway with goods, we do not say or imply that the treaty grants protection to carry any and all goods. Nor do we hold that the treaty deprives the State of the power to regulate, say, when necessary for conservation. To the contrary, we stated in Tulee that, although the treaty “forecloses the [S]tate from charging the Indians a fee of the kind in question here,” the State retained the “power to impose on Indians, equally with others, such restrictions of a purely regulatory nature . . . as are necessary for the conservation of fish.” 315 U. S., at 684. Indeed, it was crucial to our decision in Tulee that, although the licensing fees at issue were “regulatory as well as revenue producing,” “their regulatory purpose could be accomplished otherwise,” and “the imposition of license fees [was] not indispensable to the effectiveness of a state conservation program.” Id., at 685. See also Puyallup Tribe v. Department of Game of Wash., 391 U.S. 392, 402, n. 14 (1968) (“As to a ‘regulation’ concerning the time and manner of fishing outside the reservation (as opposed to a ‘tax’), we said that the power of the State was to be measured by whether it was ‘necessary for the conservation of fish’ ” (quoting Tulee, 315 U. S., at 684)). Nor do we hold that the treaty deprives the State of the power to regulate to prevent danger to health or safety occasioned by a tribe member’s exercise of treaty rights. The record of the treaty negotiations may not support the contention that the Yakamas expected to use the roads entirely unconstrained by laws related to health or safety. See App. to Brief for Confederated Tribes and Bands of the Yakama Nation as Amicus Curiae 20a–21a, 31a–32a. Governor Stevens explained, at length, the United States’ awareness of crimes committed by United States citizens who settled amongst the Yakamas, and the United States’ intention to enact laws that would restrain both the United States citizens and the Yakamas alike for the safety of both groups. See id., at 31a. Nor do we here interpret the treaty as barring the State from collecting revenue through sales or use taxes (applied outside the reservation). Unlike the tax at issue here, which applies explicitly to transport by “railcar, trailer, truck, or other equipment suitable for ground transportation,” see supra, at 6, a sales or use tax normally applies irrespective of transport or its means. Here, however, we deal with a tax applicable simply to importation by ground transportation. Moreover, it is a tax designed to secure revenue that, as far as the record shows here, the State might obtain in other ways. IV To summarize, our holding rests upon three propositions: First, a state law that burdens a treaty-protected right is pre-empted by the treaty. See supra, at 14–18. Second, the treaty protects the Yakamas’ right to travel on the public highway with goods for sale. See supra, at 10–14. Third, the Washington statute at issue here taxes the Yakamas for traveling with fuel by public highway. See supra, at 4–10. For these three reasons, Washington’s fuel tax cannot lawfully be assessed against Cougar Den on the facts here. Therefore, the judgment of the Supreme Court of Washington is affirmed. It is so ordered.

586.US.2018_17-71

The Fish and Wildlife Service administers the Endangered Species Act of 1973 on behalf of the Secretary of the Interior. In 2001, the Service listed the dusky gopher frog as an endangered species. See 16 U. S. C. §1533(a)(1). That required the Service to designate “critical habitat” for the frog. The Service proposed designating as part of that critical habitat a site in St. Tammany Parish, Louisiana, which the Service dubbed “Unit 1.” The frog had once lived in Unit 1, but the land had long been used as a commercial timber plantation, and no frogs had been spotted there for decades. The Service concluded that Unit 1 met the statutory definition of unoccupied critical habitat because its rare, high-quality breeding ponds and distance from existing frog populations made it essential for the species’ conservation. §1532(5)(A)(ii). The Service then commissioned a report on the probable economic impact of its proposed critical-habitat designation. §1533(b)(2). With regard to Unit 1, the report found that designation might bar future development of the site, depriving the owners of up to $33.9 million. The Service nonetheless concluded that the potential costs were not disproportionate to the conservation benefits and proceeded to designate Unit 1 as critical habitat for the dusky gopher frog. Unit 1 is owned by petitioner Weyerhaeuser and a group of family landowners. The owners of Unit 1 sued, contending that the closed-canopy timber plantation on Unit 1 could not be critical habitat for the dusky gopher frog, which lives in open-canopy forests. The District Court upheld the designation. The landowners also challenged the Service’s decision not to exclude Unit 1 from the frog’s critical habitat, arguing that the Service had failed to adequately weigh the benefits of designating Unit 1 against the economic impact, had used an unreasonable methodology for estimating economic impact, and had failed to consider several categories of costs. The District Court approved the Service’s methodology and declined to consider the challenge to the Service’s decision not to exclude Unit 1. The Fifth Circuit affirmed, rejecting the suggestion that the “critical habitat” definition contains any habitability requirement and concluding that the Service’s decision not to exclude Unit 1 was committed to agency discretion by law and was therefore unreviewable. Held: 1. An area is eligible for designation as critical habitat under §1533(a)(3)(A)(i) only if it is habitat for the species. That provision, the sole source of authority for critical-habit designations, states that when the Secretary lists a species as endangered he must also “designate any habitat of such species which is then considered to be critical habitat.” It does not authorize the Secretary to designate the area as critical habitat unless it is also habitat for the species. The definition allows the Secretary to identify a subset of habitat that is critical, but leaves the larger category of habitat undefined. The Service does not now dispute that critical habitat must be habitat, but argues that habitat can include areas that, like Unit 1, would require some degree of modification to support a sustainable population of a given species. Weyerhaeuser urges that habitat cannot include areas where the species could not currently survive. The Service, in turn, disputes the premise that the administrative record shows that the frog could not survive in Unit 1. The Court of Appeals, which had no occasion to interpret the term “habitat” in §1533(a)(3)(A)(i) or to assess the Service’s administrative findings regarding Unit 1, should address these questions in the first instance. Pp. 8–10. 2. The Secretary’s decision not to exclude an area from critical habitat under §1533(b)(2) is subject to judicial review. The Administrative Procedure Act creates a “basic presumption of judicial review” of agency action. Abbott Laboratories v. Gardner, 387 U. S. 136, 140. The Service contends that the presumption is rebutted here because the action is “committed to agency discretion by law,” 5 U. S. C. §701(a)(2), because §1533(b)(2) is one of those rare provisions “drawn so that a court would have no meaningful standard against which to judge the agency’s exercise of discretion,” Lincoln v. Vigil, 508 U. S. 182, 191. Section 1533(b)(2) describes a unified process for weighing the impact of designating an area as critical habitat. The provision’s first sentence requires the Secretary to “tak[e] into consideration” economic and other impacts before designation, and the second sentence authorizes the Secretary to act on his consideration by providing that he “may exclude any area from critical habitat if he determines that the benefits of such exclusion outweigh the benefits of ” designation. The word “may” certainly confers discretion on the Secretary, but it does not segregate his discretionary decision not to exclude from the mandated procedure to consider the economic and other impacts of designation when making his exclusion decisions. The statute is, therefore, not “drawn so that a court would have no meaningful standard against which to judge the [Secretary’s] exercise of [his] discretion” not to exclude. Lincoln, 508 U. S., at 191. Weyerhaeuser’s claim—that the agency did not appropriately consider all the relevant statutory factors meant to guide the agency in the exercise of its discretion—is the sort of claim that federal courts routinely assess when determining whether to set aside an agency decision as an abuse of discretion. The Court of Appeals should consider in the first instance the question whether the Service’s assessment of the costs and benefits of designation and resulting decision not to exclude Unit 1 was arbitrary, capricious, or an abuse of discretion. Pp. 10–15. 827 F. 3d 452, vacated and remanded. Roberts, C. J., delivered the opinion of the Court, in which all other Members joined, except Kavanaugh, J., who took no part in the consideration or decision of the case.

The Endangered Species Act directs the Secretary of the Interior, upon listing a species as endangered, to also designate the “critical habitat” of the species. A group of landowners whose property was designated as critical habitat for an endangered frog challenged the designation. The landowners urge that their land cannot be critical habitat because it is not habitat, which they contend refers only to areas where the frog could currently survive. The court below ruled that the Act imposed no such limitation on the scope of critical habitat. The Act also authorizes the Secretary to exclude an area that would otherwise be included as critical habitat, if the benefits of exclusion outweigh the benefits of designation. The landowners challenged the decision of the Secretary not to exclude their property, but the court below held that the Secretary’s action was not subject to judicial review. We granted certiorari to review both rulings. I A The amphibian Rana sevosa is popularly known as the “dusky gopher frog”—“dusky” because of its dark coloring and “gopher” because it lives underground. The dusky gopher frog is about three inches long, with a large head, plump body, and short legs. Warts dot its back, and dark spots cover its entire body. Final Rule To List the Missisippi Gopher Frog Distinct Population Segment of Dusky Gopher Frog as Endangered, 66 Fed. Reg. 62993 (2001) (Final Listing). It is noted for covering its eyes with its front legs when it feels threatened, peeking out periodi- cally until danger passes. Markle Interests, LLC v. United States Fish and Wildlife Serv., 827 F. 3d 452, 458, n. 2 (CA5 2016). Less endearingly, it also secretes a bitter, milky substance to deter would-be diners. Brief for Intervenor-Respondents 6, n. 1. The frog spends most of its time in burrows and stump holes located in upland longleaf pine forests. In such forests, frequent fires help maintain an open canopy, which in turn allows vegetation to grow on the forest floor. The vegetation supports the small insects that the frog eats and provides a place for the frog’s eggs to attach when it breeds. The frog breeds in “ephemeral” ponds that are dry for part of the year. Such ponds are safe for tadpoles because predatory fish cannot live in them. Designation of Critical Habitat for Dusky Gopher Frog, 77 Fed. Reg. 35129–35131 (2012) (Designation). The dusky gopher frog once lived throughout coastal Alabama, Louisiana, and Mississippi, in the longleaf pine forests that used to cover the southeast. But more than 98% of those forests have been removed to make way for urban development, agriculture, and timber plantations. The timber plantations consist of fast-growing loblolly pines planted as close together as possible, resulting in a closed-canopy forest inhospitable to the frog. The near eradication of the frog’s habitat sent the species into severe decline. By 2001, the known wild population of the dusky gopher frog had dwindled to a group of 100 at a single pond in southern Mississippi. That year, the Fish and Wildlife Service, which administers the Endangered Species Act of 1973 on behalf of the Secretary of the Interior, listed the dusky gopher frog as an endangered species. Final Listing 62993–62995; see 87Stat. 886, 16 U. S. C. §1533(a)(1). B When the Secretary lists a species as endangered, he must also designate the critical habitat of that species. §1533(a)(3)(A)(i). The ESA defines “critical habitat” as: “(i) the specific areas within the geographical area occupied by the species . . . on which are found those physical or biological features (I) essential to the conservation of the species and (II) which may require special management considerations or protection; and “(ii) specific areas outside the geographical area occupied by the species . . . upon a determination by the Secretary that such areas are essential for the conservation of the species.” §1532(5)(A). Before the Secretary may designate an area as critical habitat, the ESA requires him to “tak[e] into consideration the economic impact” and other relevant impacts of the designation. §1533(b)(2). The statute goes on to authorize him to “exclude any area from critical habitat if he determines that the benefits of such exclusion outweigh the benefits of [designation],” unless exclusion would result in extinction of the species. Ibid. A critical-habitat designation does not directly limit the rights of private landowners. It instead places conditions on the Federal Government’s authority to effect any physical changes to the designated area, whether through activities of its own or by facilitating private development. Section 7 of the ESA requires all federal agencies to consult with the Secretary to “[e]nsure that any action authorized, funded, or carried out by such agency” is not likely to adversely affect a listed species’ critical habitat. 16 U. S. C. §1536(a)(2). If the Secretary determines that an agency action, such as issuing a permit, would harm critical habitat, then the agency must terminate the action, implement an alternative proposed by the Secretary, or seek an exemption from the Cabinet-level Endangered Species Committee. See National Assn. of Home Builders v. Defenders of Wildlife, 551 U. S. 644, 652 (2007); 50 CFR 402.15 (2017). Due to resource constraints, the Service did not designate the frog’s critical habitat in 2001, when it listed the frog as endangered. Designation, at 35118–35119. In the following years, the Service discovered two additional naturally occurring populations and established another population through translocation. The first population nonetheless remains the only stable one and by far the largest. Dept. of Interior, U. S. Fish and Wildlife Serv., Dusky Gopher Frog (Rana sevosa) Recovery Plan iv, 6–7 (2015). In 2010, in response to litigation by the Center for Biological Diversity, the Service published a proposed critical-habitat designation. Designation, at 35119. The Service proposed to designate as occupied critical habitat all four areas with existing dusky gopher frog populations. The Service found that each of those areas possessed the three features that the Service considered “essential to the conservation” of the frog and that required special protection: ephemeral ponds; upland open-canopy forest containing the holes and burrows in which the frog could live; and open-canopy forest connecting the two. But the Service also determined that designating only those four sites would not adequately ensure the frog’s conservation. Because the existing dusky gopher frog populations were all located in two adjacent counties on the Gulf Coast of Mississippi, local events such as extreme weather or an outbreak of an infectious disease could jeopardize the entire species. Designation of Critical Habitat for Mississippi Gopher Frog, 75 Fed. Reg. 31394 (2010) (proposed 50 CFR Part 17). To protect against that risk, the Service proposed to designate as unoccupied critical habitat a 1,544-acre site in St. Tammany Parish, Louisiana. The site, dubbed “Unit 1” by the Service, had been home to the last known population of dusky gopher frogs outside of Mississippi. The frog had not been seen in Unit 1 since 1965, and a closed-canopy timber plantation occupied much of the site. But the Service found that the site retained five ephem- eral ponds “of remarkable quality,” and determined that an open-canopy forest could be restored on the surrounding uplands “with reasonable effort.” Although the uplands in Unit 1 lacked the open-canopy forests (and, of course, the frogs) necessary for designation as occupied critical habitat, the Service concluded that the site met the statutory definition of unoccupied critical habitat because its rare, high-quality breeding ponds and its distance from existing frog populations made it essential for the conservation of the species. Designation, at 35118, 35124, 35133, 35135. After issuing its proposal, the Service commissioned a report on the probable economic impact of designating each area, including Unit 1, as critical habitat for the dusky gopher frog. See 16 U. S. C. §1533(b)(2); App. 63. Petitioner Weyerhaeuser Company, a timber company, owns part of Unit 1 and leases the remainder from a group of family landowners. Brief for Petitioner 16. While the critical-habitat designation has no direct effect on the timber operations, St. Tammany Parish is a fast-growing part of the New Orleans metropolitan area, and the landowners have already invested in plans to more profitably develop the site. App. 80–83. The report recognized that anyone developing the area may need to obtain Clean Water Act permits from the Army Corps of Engineers before filling any wetlands on Unit 1. 33 U. S. C. §1344(a). Because Unit 1 is designated as critical habitat, Section 7 of the ESA would require the Corps to consult with the Service before issuing any permits. According to the report, that consultation process could result in one of three outcomes. First, it could turn out that the wetlands in Unit 1 are not subject to the Clean Water Act permitting requirements, in which case the landowners could proceed with their plans unimpeded. Second, the Service could ask the Corps not to issue permits to the landowners to fill some of the wetlands on the site, in effect prohibiting development on 60% of Unit 1. The report estimated that this would deprive the owners of $20.4 million in development value. Third, by asking the Corps to deny even more of the permit requests, the Service could bar all development of Unit 1, costing the owners $33.9 million. The Service concluded that those potential costs were not “disproportionate” to the conservation benefits of designation. “Consequently,” the Service announced, it would not “exercis[e] [its] discretion to exclude” Unit 1 from the dusky gopher frog’s critical habitat. App. 188–190. C Weyerhaeuser and the family landowners sought to vacate the designation in Federal District Court. They contended that Unit 1 could not be critical habitat for the dusky gopher frog because the frog could not survive there: Survival would require replacing the closed-canopy timber plantation encircling the ponds with an open-canopy longleaf pine forest. The District Court nonetheless upheld the designation. Markle Interests, LLC v. United States Fish and Wildlife Serv., 40 F. Supp. 3d 744 (ED La. 2014). The court determined that Unit 1 satisfied the statutory definition of unoccupied critical habitat, which requires only that the Service deem the land “essential for the conservation [of] the species.” Id., at 760. Weyerhaeuser also challenged the Service’s decision not to exclude Unit 1 from the dusky gopher frog’s critical habitat, arguing that the Service had failed to adequately weigh the benefits of designating Unit 1 against the economic impact. In addition, Weyerhaeuser argued that the Service had used an unreasonable methodology for estimating economic impact and, regardless of methodology, had failed to consider several categories of costs. Id., at 759. The court approved the Service’s methodology and declined to consider Weyerhaeuser’s challenge to the decision not to exclude. See id., at 763–767, and n. 29. The Fifth Circuit affirmed. 827 F. 3d 452. The Court of Appeals rejected the suggestion that the definition of critical habitat contains any “habitability requirement.” Id., at 468. The court also concluded that the Service’s decision not to exclude Unit 1 was committed to agency discretion by law and was therefore unreviewable. Id., at 473–475. Judge Owen dissented. She wrote that Unit 1 could not be “essential for the conservation of the species” because it lacked the open-canopy forest that the Service itself had determined was “essential to the conservation” of the frog. Id., at 480–481. The Fifth Circuit denied rehearing en banc. Markle Interests, LLC v. United States Fish and Wildlife Serv., 848 F. 3d 635 (2017). Judge Jones dissented, joined by Judges Jolly, Smith, Clement, Owen, and Elrod. They reasoned that critical habitat must first be habitat, and Unit 1 in its present state could not be habitat for the dusky gopher frog. Id., at 641. The dissenting judges also concluded that the Service’s decision not to exclude Unit 1 was reviewable for abuse of discretion. Id., at 654, and n. 21. We granted certiorari to consider two questions: (1) whether “critical habitat” under the ESA must also be habitat; and (2) whether a federal court may review an agency decision not to exclude a certain area from critical habitat because of the economic impact of such a designation. 583 U. S. ___ (2018).[1] II A Our analysis starts with the phrase “critical habitat.” According to the ordinary understanding of how adjectives work, “critical habitat” must also be “habitat.” Adjectives modify nouns—they pick out a subset of a category that possesses a certain quality. It follows that “critical habitat” is the subset of “habitat” that is “critical” to the conservation of an endangered species. Of course, “[s]tatutory language cannot be construed in a vacuum,” Sturgeon v. Frost, 577 U. S. ___, ___ (2016) (slip op., at 12) (internal quotation marks omitted), and so we must also consider “critical habitat” in its statutory context. Section 4(a)(3)(A)(i), which the lower courts did not analyze, is the sole source of authority for critical-habitat designations. That provision states that when the Secretary lists a species as endangered he must also “designate any habitat of such species which is then considered to be critical habitat.” 16 U. S. C. §1533(a)(3)(A)(i) (emphasis added). Only the “habitat” of the endangered species is eligible for designation as critical habitat. Even if an area otherwise meets the statutory definition of unoccupied critical habitat because the Secretary finds the area essential for the conservation of the species, Section 4(a)(3)(A)(i) does not authorize the Secretary to designate the area as critical habitat unless it is also habitat for the species. The Center for Biological Diversity contends that the statutory definition of critical habitat is complete in itself and does not require any independent inquiry into the meaning of the term “habitat,” which the statute leaves undefined. Brief for Intervenor-Respondents 43–49. But the statutory definition of “critical habitat” tells us what makes habitat “critical,” not what makes it “habitat.” Under the statutory definition, critical habitat comprises areas occupied by the species “on which are found those physical or biological features (I) essential to the conservation of the species and (II) which may require special management considerations or protection,” as well as unoccupied areas that the Secretary determines to be “essential for the conservation of the species.” 16 U. S. C. §1532(5)(A). That is no baseline definition of habitat—it identifies only certain areas that are indispensable to the conservation of the endangered species. The definition allows the Secretary to identify the subset of habitat that is critical, but leaves the larger category of habitat undefined. The Service does not now dispute that critical habitat must be habitat, see Brief for Federal Respondents 23, although it made no such concession below. Instead, the Service argues that habitat includes areas that, like Unit 1, would require some degree of modification to support a sustainable population of a given species. Id., at 27. Weyerhaeuser, for its part, urges that habitat cannot include areas where the species could not currently survive. Brief for Petitioner 25. (Habitat can, of course, include areas where the species does not currently live, given that the statute defines critical habitat to include unoccupied areas.) The Service in turn disputes Weyerhaeuser’s premise that the administrative record shows that the frog could not survive in Unit 1. Brief for Federal Respondents 22, n. 4. The Court of Appeals concluded that “critical habitat” designations under the statute were not limited to areas that qualified as habitat. See 827 F. 3d, at 468 (“There is no habitability requirement in the text of the ESA or the implementing regulations.”). The court therefore had no occasion to interpret the term “habitat” in Section 4(a)(3)(A)(i) or to assess the Service’s administrative findings regarding Unit 1. Accordingly, we vacate the judgment below and remand to the Court of Appeals to con- sider these questions in the first instance.[2] B Weyerhaeuser also contends that, even if Unit 1 could be properly classified as critical habitat for the dusky gopher frog, the Service should have excluded it from designation under Section 4(b)(2) of the ESA. That provision requires the Secretary to “tak[e] into consideration the economic impact . . . of specifying any particular area as critical habitat” and authorizes him to “exclude any area from critical habitat if he determines that the benefits of such exclusion outweigh the benefits of specifying such area as part of the critical habitat.” 16 U. S. C. §1533(b)(2). To satisfy its obligation to consider economic impact, the Service commissioned a report estimating the costs of its proposed critical-habitat designation. The Service concluded that the costs of designating the proposed areas, including Unit 1, were not “disproportionate” to the conservation benefits and, “[c]onsequently,” declined to make any exclusions. Weyerhaeuser claims that the Service’s conclusion rested on a faulty assessment of the costs and benefits of designation and that the resulting decision not to exclude should be set aside. Specifically, Weyerhaeuser contends that the Service improperly weighed the costs of designating Unit 1 against the benefits of designating all proposed critical habitat, rather than the benefits of designating Unit 1 in particular. Weyerhaeuser also argues that the Service did not fully account for the economic impact of designating Unit 1 because it ignored, among other things, the costs of replacing timber trees with longleaf pines, maintaining an open canopy through controlled burning, and the tax revenue that St. Tammany Parish would lose if Unit 1 were never developed. Brief for Petitioner 53–54. The Court of Appeals did not consider Weyerhaeuser’s claim because it concluded that a decision not to exclude a certain area from critical habitat is unreviewable. The Administrative Procedure Act creates a “basic presumption of judicial review [for] one ‘suffering legal wrong because of agency action.’ ” Abbott Laboratories v. Gardner, 387 U. S. 136, 140 (1967) (quoting 5 U. S. C. §702). As we explained recently, “legal lapses and violations occur, and especially so when they have no consequence. That is why this Court has so long applied a strong presumption favoring judicial review of administrative action.” Mach Mining, LLC v. EEOC, 575 U. S. ___, ___–___ (2015) (slip op., at 7–8). The presumption may be rebutted only if the relevant statute precludes review, 5 U. S. C. §701(a)(1), or if the action is “committed to agency discretion by law,” §701(a)(2). The Service contends, and the lower courts agreed, that Section 4(b)(2) of the ESA commits to the Secretary’s discretion decisions not to exclude an area from critical habitat. This Court has noted the “tension” between the prohibition of judicial review for actions “committed to agency discretion” and the command in §706(2)(A) that courts set aside any agency action that is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” Heckler v. Chaney, 470 U. S. 821, 829 (1985). A court could never determine that an agency abused its discretion if all matters committed to agency discretion were unreviewable. To give effect to §706(2)(A) and to honor the presumption of review, we have read the exception in §701(a)(2) quite narrowly, restricting it to “those rare circumstances where the relevant statute is drawn so that a court would have no meaningful standard against which to judge the agency’s exercise of discretion.” Lincoln v. Vigil, 508 U. S. 182, 191 (1993). The Service contends that Section 4(b)(2) of the ESA is one of those rare statutory provisions. There is, at the outset, reason to be skeptical of the Service’s position. The few cases in which we have applied the §701(a)(2) exception involved agency decisions that courts have traditionally regarded as unreviewable, such as the allocation of funds from a lump-sum appropriation, Lincoln, 508 U. S., at 191, or a decision not to reconsider a final action, ICC v. Locomotive Engineers, 482 U. S. 270, 282 (1987). By contrast, this case involves the sort of routine dispute that federal courts regularly review: An agency issues an order affecting the rights of a private party, and the private party objects that the agency did not properly justify its determination under a standard set forth in the statute. Section 4(b)(2) states that the Secretary “shall designate critical habitat . . . after taking into consideration the economic impact, the impact on national security, and any other relevant impact, of specifying any particular area as critical habitat. The Secretary may exclude any area from critical habitat if he determines that the benefits of such exclusion outweigh the benefits of specifying such area . . . unless he determines . . . that the failure to designate such area as critical habitat will result in the extinction of the species concerned.” 16 U. S. C. §1533(b)(2). Although the text meanders a bit, we recognized in Bennett v. Spear, 520 U. S. 154 (1997), that the provision describes a unified process for weighing the impact of designating an area as critical habitat. The first sentence of Section 4(b)(2) imposes a “categorical requirement” that the Secretary “tak[e] into consideration” economic and other impacts before such a designation. Id., at 172 (emphasis deleted). The second sentence authorizes the Secretary to act on his consideration by providing that he may exclude an area from critical habitat if he determines that the benefits of exclusion outweigh the benefits of designation. The Service followed that procedure here (albeit in a flawed manner, according to Weyerhaeuser). It commissioned a report to estimate the costs of designating the proposed critical habitat, concluded that those costs were not “disproportionate” to the benefits of designation, and “[c]onsequently” declined to “exercis[e] [its] discretion to exclude any areas from [the] designation of critical habitat.” App. 190. Bennett explained that the Secretary’s “ultimate decision” to designate or exclude, which he “arriv[es] at” after considering economic and other impacts, is reviewable “for abuse of discretion.” 520 U. S., at 172. The Service dismisses that language as a “passing reference . . . not necessarily inconsistent with the Service’s understanding,” which is that the Secretary’s decision not to exclude an area is wholly discretionary and therefore unreviewable. Brief for Federal Respondents 50. The Service bases its understanding on the second sentence of Section 4(b)(2), which states that the Secretary “may exclude [an] area from critical habitat if he determines that the benefits of such exclusion outweigh the benefits of [designation].” The use of the word “may” certainly confers discretion on the Secretary. That does not, however, segregate his discretionary decision not to exclude from the procedure mandated by Section 4(b)(2), which directs the Secretary to consider the economic and other impacts of designation when making his exclusion decisions. Weyerhaeuser’s claim is the familiar one in administrative law that the agency did not appropriately consider all of the relevant factors that the statute sets forth to guide the agency in the exercise of its discretion. Specifically, Weyerhaeuser contends that the Service ignored some costs and conflated the benefits of designating Unit 1 with the benefits of designating all of the proposed critical habitat. This is the sort of claim that federal courts routinely assess when determining whether to set aside an agency decision as an abuse of discretion under §706(2)(A). See Judulang v. Holder, 565 U. S. 42, 53 (2011) (“When reviewing an agency action, we must assess . . . whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.” (internal quotation marks omitted)). Section 4(b)(2) requires the Secretary to consider economic impact and relative benefits before deciding whether to exclude an area from critical habitat or to proceed with designation. The statute is, therefore, not “drawn so that a court would have no meaningful standard against which to judge the [Secretary’s] exercise of [his] discretion” not to exclude. Lincoln, 508 U. S., at 191. Because it determined that the Service’s decisions not to exclude were committed to agency discretion and therefore unreviewable, the Court of Appeals did not consider whether the Service’s assessment of the costs and benefits of designation was flawed in a way that rendered the resulting decision not to exclude Unit 1 arbitrary, capricious, or an abuse of discretion. Accordingly, we remand to the Court of Appeals to consider that question, if necessary, in the first instance. * * * The judgment of the Court of Appeals for the Fifth Circuit is vacated, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Justice Kavanaugh took no part in the consideration or decision of this case. Notes 1 Intervenor Center for Biological Diversity raises an additional question in its brief, arguing that Weyerhaeuser lacks standing to challenge the critical-habitat designation because it has not suffered an injury in fact. We agree with the lower courts that the decrease in the market value of Weyerhaeuser’s land as a result of the designation is a sufficiently concrete injury for Article III purposes. See Village of Euclid v. Ambler Realty Co., 272 U. S. 365, 386 (1926) (holding that a zoning ordinance that “greatly . . . reduce[d] the value of appellee’s lands and destroy[ed] their marketability for industrial, commercial and residential uses” constituted a “present invasion of appellee’s property rights”). 2 Because we hold that an area is eligible for designation as critical habitat under Section 4(a)(3)(A)(i) only if it is habitat for the species, it is not necessary to consider the landowners’ argument that land cannot be “essential for the conservation of the species,” and thus cannot satisfy the statutory definition of unoccupied critical habitat, if it is not habitat for the species. See Brief for Petitioner 27–28; Brief for Respondent Markle Interests, LLC, et al. in Support of Petitioner 28–31.

589.US.2019_18-877

In 1996, a marine salvage company named Intersal, Inc., discovered the shipwreck of the Queen Anne’s Revenge off the North Carolina coast. North Carolina, the shipwreck’s legal owner, contracted with Intersal to conduct recovery operations. Intersal, in turn, hired videographer Frederick Allen to document the efforts. Allen recorded videos and took photos of the recovery for more than a decade. He registered copyrights in all of his works. When North Carolina published some of Allen’s videos and photos online, Allen sued for copyright infringement. North Carolina moved to dismiss the lawsuit on the ground of state sovereign immunity. Allen countered that the Copyright Remedy Clarification Act of 1990 (CRCA) removed the States’ sovereign immunity in copyright infringement cases. The District Court agreed with Allen, finding in the CRCA’s text a clear congressional intent to abrogate state sovereign immunity and a proper constitutional basis for that abrogation. The court acknowledged that Florida Prepaid Postsecondary Ed. Expense Bd. v. College Savings Bank, 527 U.S. 627, precluded Congress from using its Article I powers—including its authority over copyrights—to deprive States of sovereign immunity. But the court held that Congress could accomplish its objective under Section 5 of the Fourteenth Amendment. The Fourth Circuit reversed, reading Florida Prepaid to prevent recourse to both Article I and Section 5. Held: Congress lacked authority to abrogate the States’ immunity from copyright infringement suits in the CRCA. Pp. 4–17. (a) In general, a federal court may not hear a suit brought by any person against a nonconsenting State. But such suits are permitted if Congress has enacted “unequivocal statutory language” abrogating the States’ immunity from suit, Seminole Tribe of Fla. v. Florida, 517 U.S. 44, 56, and some constitutional provision allows Congress to have thus encroached on the States’ sovereignty. Congress used clear language to abrogate the States’ immunity from copyright infringement suits in the CRCA. Allen contends that Congress’s constitutional power to do so arises either from the Intellectual Property Clause, Art. I, §8, cl. 8, or from Section 5 of the Fourteenth Amendment, which authorizes Congress to “enforce” the commands of the Due Process Clause. Each contention is foreclosed by precedent. Pp. 4–6. (b) The Intellectual Property Clause enables Congress to grant both copyrights and patents. In Allen’s view, Congress’s authority to abrogate sovereign immunity from copyright suits naturally follows, in order to “secur[e]” a copyright holder’s “exclusive Right” as against a State’s intrusion. But that theory was rejected in Florida Prepaid. That case considered the constitutionality of the Patent Remedy Act, which, like the CRCA, attempted to put “States on the same footing as private parties” in patent infringement lawsuits. 527 U. S., at 647, 648. Florida Prepaid acknowledged that Congress’s goal of providing uniform remedies in infringement cases was a “proper Article I concern,” but held that Seminole Tribe precluded Congress from using its Article I powers “to circumvent” the limits sovereign immunity “place[s] upon federal jurisdiction,” 517 U. S., at 73. For the same reason, Article I cannot support the CRCA. Allen reads Central Va. Community College v. Katz, 546 U.S. 356 to have replaced Seminole Tribe’s general rule with a clause-by-clause approach to evaluating whether a particular constitutional provision allows the abrogation of sovereign immunity. But Katz rested on the unique history of the Bankruptcy Clause. 546 U. S., at 369, n. 9. And even if the limits of Katz’s holding were not so clear, Florida Prepaid, together with stare decisis, would doom Allen’s argument. Overruling Florida Prepaid would require a “special justification,” over and above the belief “that the precedent was wrongly decided,” Halliburton Co. v. Erica P. John Fund, Inc., 573 U.S. 258, 266, which Allen does not offer. Pp. 6–10. (c) Section 5 of the Fourteenth Amendment allows Congress to abrogate the States’ immunity as part of its power “to enforce” the Amendment’s substantive prohibitions. City of Boerne v. Flores, 521 U.S. 507, 519. For Congress’s action to fall within its Section 5 authority, “[t]here must be a congruence and proportionality between the injury to be prevented or remedied and the means adopted to that end.” Id., at 520. This test requires courts to consider the nature and extent of state conduct violating the Fourteenth Amendment and to examine the scope of Congress’s response to that injury. Florida Prepaid again serves as the critical precedent. There, the Court defined the scope of unconstitutional patent infringement as intentional conduct for which there is no adequate state remedy. 527 U. S., at 642–643, 645. Because Congress failed to identify a pattern of unconstitutional patent infringement when it enacted the Patent Remedy Act, the Court held that the Act swept too far. Given the identical scope of the CRCA and Patent Remedy Act, this case could be decided differently only if the CRCA responded to materially stronger evidence of unconstitutional infringement. But as in Florida Prepaid, the legislative record contains thin evidence of infringement. Because this record cannot support Congress’s choice to strip the States of their sovereign immunity in all copyright infringement cases, the CRCA fails the “congruence and proportionality” test. Pp. 10–16. 895 F.3d 337, affirmed. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Alito, Sotomayor, Gorsuch, and Kavanaugh, JJ., joined, and in which Thomas, J., joined except for the final paragraph in Part II–A and the final paragraph in Part II–B. Thomas, J., filed an opinion concurring in part and concurring in the judgment. Breyer, J., filed an opinion concurring in the judgment, in which Ginsburg, J., joined.

In two basically identical statutes passed in the early 1990s, Congress sought to strip the States of their sovereign immunity from patent and copyright infringement suits. Not long after, this Court held in Florida Prepaid Postsecondary Ed. Expense Bd. v. College Savings Bank, 527 U.S. 627 (1999), that the patent statute lacked a valid constitutional basis. Today, we take up the copyright statute. We find that our decision in Florida Prepaid compels the same conclusion. I In 1717, the pirate Edward Teach, better known as Blackbeard, captured a French slave ship in the West Indies and renamed her Queen Anne’s Revenge. The vessel became his flagship. Carrying some 40 cannons and 300 men, the Revenge took many prizes as she sailed around the Caribbean and up the North American coast. But her reign over those seas was short-lived. In 1718, the ship ran aground on a sandbar a mile off Beaufort, North Carolina. Blackbeard and most of his crew escaped without harm. Not so the Revenge. She sank beneath the waters, where she lay undisturbed for nearly 300 years. In 1996, a marine salvage company named Intersal, Inc., discovered the shipwreck. Under federal and state law, the wreck belongs to North Carolina. See 102Stat. 433, 43 U. S. C. §2105(c); N. C. Gen. Stat. Ann. §121–22 (2019). But the State contracted with Intersal to take charge of the recovery activities. Intersal in turn retained petitioner Frederick Allen, a local videographer, to document the operation. For over a decade, Allen created videos and photos of divers’ efforts to salvage the Revenge’s guns, anchors, and other remains. He registered copyrights in all those works. This suit arises from North Carolina’s publication of some of Allen’s videos and photos. Allen first protested in 2013 that the State was infringing his copyrights by uploading his work to its website without permission. To address that allegation, North Carolina agreed to a settlement paying Allen $15,000 and laying out the parties’ respective rights to the materials. But Allen and the State soon found themselves embroiled in another dispute. Allen complained that North Carolina had impermissibly posted five of his videos online and used one of his photos in a newsletter. When the State declined to admit wrongdoing, Allen filed this action in Federal District Court. It charges the State with copyright infringement (call it a modern form of piracy) and seeks money damages. North Carolina moved to dismiss the suit on the ground of sovereign immunity. It invoked the general rule that federal courts cannot hear suits brought by individuals against nonconsenting States. See State Defendants’ Memorandum in No. 15–627 (EDNC), Doc. 50, p. 7. But Allen responded that an exception to the rule applied because Congress had abrogated the States’ sovereign immunity from suits like his. See Plaintiffs’ Response, Doc. 57, p. 7. The Copyright Remedy Clarification Act of 1990 (CRCA or Act) provides that a State “shall not be immune, under the Eleventh Amendment [or] any other doctrine of sovereign immunity, from suit in Federal court” for copyright infringement. 17 U. S. C. §511(a). And the Act specifies that in such a suit a State will be liable, and subject to remedies, “in the same manner and to the same extent as” a private party. §501(a); see §511(b).[1] That meant, Allen contended, that his suit against North Carolina could go forward. The District Court agreed. Quoting the CRCA’s text, the court first found that “Congress has stated clearly its intent to abrogate sovereign immunity for copyright claims against a state.” 244 F. Supp. 3d 525, 533 (EDNC 2017). And that abrogation, the court next held, had a proper constitutional basis. Florida Prepaid and other precedent, the District Court acknowledged, precluded Congress from using its Article I powers—including its authority over copyrights—to take away a State’s sovereign immunity. See 244 F. Supp. 3d, at 534. But in the court’s view, Florida Prepaid left open an alternative route to abrogation. Given the States’ “pattern” of “abus[ive]” copyright infringement, the court held, Congress could accomplish its object under Section 5 of the Fourteenth Amendment. 244 F. Supp. 3d, at 535. On interlocutory appeal, the Court of Appeals for the Fourth Circuit reversed. It read Florida Prepaid to prevent recourse to Section 5 no less than to Article I. A Section 5 abrogation, the Fourth Circuit explained, must be “congruent and proportional” to the Fourteenth Amendment injury it seeks to remedy. 895 F.3d 337, 350 (2018). Florida Prepaid had applied that principle to reject Congress’s attempt, in the Patent Remedy Act, to abolish the States’ immunity from patent infringement suits. See 527 U. S., at 630. In the Fourth Circuit’s view, nothing distinguished the CRCA. That abrogation, the court reasoned, was “equally broad” and rested on a “similar legislative record” of constitutional harm. 895 F. 3d, at 352. So Section 5 could not save the law. Because the Court of Appeals held a federal statute invalid, this Court granted certiorari. 587 U. S. ___ (2019). We now affirm. II In our constitutional scheme, a federal court generally may not hear a suit brought by any person against a nonconsenting State. That bar is nowhere explicitly set out in the Constitution. The text of the Eleventh Amendment (the single most relevant provision) applies only if the plaintiff is not a citizen of the defendant State.[2] But this Court has long understood that Amendment to “stand not so much for what it says” as for the broader “presupposition of our constitutional structure which it confirms.” Blatchford v. Native Village of Noatak, 501 U.S. 775, 779 (1991). That premise, the Court has explained, has several parts. First, “each State is a sovereign entity in our federal system.” Seminole Tribe of Fla. v. Florida, 517 U.S. 44, 54 (1996). Next, “[i]t is inherent in the nature of sovereignty not to be amenable to [a] suit” absent consent. Id., at 54, n. 13 (quoting The Federalist No. 81, p. 487 (C. Rossiter ed. 1961) (A. Hamilton)). And last, that fundamental aspect of sovereignty constrains federal “judicial authority.” Blatchford, 501 U. S., at 779. But not entirely. This Court has permitted a federal court to entertain a suit against a nonconsenting State on two conditions. First, Congress must have enacted “un- equivocal statutory language” abrogating the States’ immunity from the suit. Seminole Tribe, 517 U. S., at 56 (internal quotation marks omitted); see Dellmuth v. Muth, 491 U.S. 223, 228 (1989) (requiring Congress to “mak[e] its intention unmistakably clear”). And second, some constitutional provision must allow Congress to have thus encroached on the States’ sovereignty. Not even the most crystalline abrogation can take effect unless it is “a valid exercise of constitutional authority.” Kimel v. Florida Bd. of Regents, 528 U.S. 62, 78 (2000). No one here disputes that Congress used clear enough language to abrogate the States’ immunity from copyright infringement suits. As described above, the CRCA provides that States “shall not be immune” from those actions in federal court. §511(a); see supra, at 2–3. And the Act specifies that a State stands in the identical position as a private defendant—exposed to liability and remedies “in the same manner and to the same extent.” §501(a); see §511(b). So there is no doubt what Congress meant to accomplish. Indeed, this Court held in Florida Prepaid that the essentially verbatim provisions of the Patent Remedy Act “could not have [made] any clearer” Congress’s intent to remove the States’ immunity. 527 U. S., at 635. The contested question is whether Congress had authority to take that step. Allen maintains that it did, under either of two constitutional provisions. He first points to the clause in Article I empowering Congress to provide copyright protection. If that fails, he invokes Section 5 of the Fourteenth Amendment, which authorizes Congress to “enforce” the commands of the Due Process Clause. Neither contention can succeed. The slate on which we write today is anything but clean. Florida Prepaid, along with other precedent, forecloses each of Allen’s arguments. A Congress has power under Article I “[t]o promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” §8, cl. 8. That provision—call it the Intellectual Property Clause—enables Congress to grant both copyrights and patents. And the monopoly rights so given impose a corresponding duty (i.e., not to infringe) on States no less than private parties. See Goldstein v. California, 412 U.S. 546, 560 (1973). In Allen’s view, Congress’s authority to abrogate sovereign immunity from copyright suits naturally follows. Abrogation is the single best—or maybe, he says, the only—way for Congress to “secur[e]” a copyright holder’s “exclusive Right[s]” as against a State’s intrusion. See Brief for Petitioners 20 (quoting Art. I, §8, cl. 8). So, Allen contends, the authority to take that step must fall within the Article I grant of power to protect intellectual property. The problem for Allen is that this Court has already rejected his theory. The Intellectual Property Clause, as just noted, covers copyrights and patents alike. So it was the first place the Florida Prepaid Court looked when deciding whether the Patent Remedy Act validly stripped the States of immunity from infringement suits. In doing so, we acknowledged the reason for Congress to put “States on the same footing as private parties” in patent litigation. 527 U. S., at 647. It was, just as Allen says here, to ensure “uniform, surefire protection” of intellectual property. Reply Brief 10. That was a “proper Article I concern,” we allowed. 527 U. S., at 648. But still, we said, Congress could not use its Article I power over patents to remove the States’ immunity. We based that conclusion on Seminole Tribe v. Florida, decided three years earlier. There, the Court had held that “Article I cannot be used to circumvent” the limits sovereign immunity “place[s] upon federal jurisdiction.” 517 U. S., at 73. That proscription ended the matter. Because Congress could not “abrogate state sovereign immunity [under] Article I,” Florida Prepaid explained, the Intellectual Property Clause could not support the Patent Remedy Act. 527 U. S., at 636. And to extend the point to this case: if not the Patent Remedy Act, not its copyright equivalent either, and for the same reason. Here too, the power to “secur[e ]” an intellectual property owner’s “exclusive Right” under Article I stops when it runs into sovereign immunity. §8, cl. 8. Allen claims, however, that a later case offers an exit ramp from Florida Prepaid. In Central Va. Community College v. Katz, 546 U.S. 356, 359 (2006), we held that Article I’s Bankruptcy Clause enables Congress to subject nonconsenting States to bankruptcy proceedings (there, to recover a preferential transfer). We thus exempted the Bankruptcy Clause from Seminole Tribe’s general rule that Article I cannot justify haling a State into federal court. In bankruptcy, we decided, sovereign immunity has no place. But if that is true, Allen asks, why not say the same thing here? Allen reads Katz as “adopt[ing] a clause-by-clause approach to evaluating whether a particular clause of Article I” allows the abrogation of sovereign immunity. Brief for Petitioners 20. And he claims that the Intellectual Property Clause “supplies singular warrant” for Congress to take that step. Ibid. That is so, Allen reiterates, because “Congress could not ‘secur[e]’ authors’ ‘exclusive Right’ to their works if [it] were powerless” to make States pay for infringing conduct. Ibid. But everything in Katz is about and limited to the Bankruptcy Clause; the opinion reflects what might be called bankruptcy exceptionalism. In part, Katz rested on the “singular nature” of bankruptcy jurisdiction. 546 U. S., at 369, n. 9. That jurisdiction is, and was at the Founding, “principally in rem”—meaning that it is “premised on the debtor and his estate, and not on the creditors” (including a State). Id., at 369–370 (internal quotation marks omitted). For that reason, we thought, “it does not implicate States’ sovereignty to nearly the same degree as other kinds of jurisdiction.” Id., at 362. In remaining part, Katz focused on the Bankruptcy Clause’s “unique history.” Id., at 369, n. 9. The Clause emerged from a felt need to curb the States’ authority. The States, we explained, “had wildly divergent schemes” for discharging debt, and often “refus[ed] to respect one another’s discharge orders.” Id., at 365, 377. “[T]he Framers’ primary goal” in adopting the Clause was to address that problem—to stop “competing sovereigns[ ]” from interfering with a debtor’s discharge. Id., at 373. And in that project, the Framers intended federal courts to play a leading role. The nation’s first Bankruptcy Act, for example, empowered those courts to order that States release people they were holding in debtors’ prisons. See id., at 374. So through and through, we thought, the Bankruptcy Clause embraced the idea that federal courts could impose on state sovereignty. In that, it was sui generis—again, “unique”—among Article I’s grants of authority. Id., at 369, n. 9. Indeed, Katz’s view of the Bankruptcy Clause had a yet more striking aspect, which further separates it from any other. The Court might have concluded from its analysis that the Clause allows Congress to abrogate the States’ sovereign immunity (as Allen argues the Intellectual Property Clause does). But it did not; it instead went further. Relying on the above account of the Framers’ intentions, the Court found that the Bankruptcy Clause itself did the abrogating. Id., at 379 (“[T]he relevant ‘abrogation’ is the one effected in the plan of the [Constitutional] Convention”). Or stated another way, we decided that no congressional abrogation was needed because the States had already “agreed in the plan of the Convention not to assert any sovereign immunity defense” in bankruptcy proceedings. Id., at 377. We therefore discarded our usual rule—which Allen accepts as applying here—that Congress must speak, and indeed speak unequivocally, to abrogate sovereign immunity. Compare id., at 378–379 (“[O]ur decision today” does not “rest[ ] on any statement Congress ha[s] made on the subject of state sovereign immunity”), with supra, at 5 (our ordinary rule). Our decision, in short, viewed bankruptcy as on a different plane, governed by principles all its own. Nothing in that understanding invites the kind of general, “clause-by-clause” reexamination of Article I that Allen proposes. See supra, at 7. To the contrary, it points to a good-for-one-clause-only holding. And even if Katz’s confines were not so clear, Florida Prepaid, together with stare decisis, would still doom Allen’s argument. As Allen recognizes, if the Intellectual Property Clause permits the CRCA’s abrogation, it also would permit the Patent Remedy Act’s. See Tr. of Oral Arg. 9 (predicting that if his position prevailed, “ultimately, the Patent Remedy Act would be revisited and properly upheld as a valid exercise of Congress’s Article I power”). Again, there is no difference between copyrights and patents under the Clause, nor any material difference between the two statutes’ provisions. See supra, at 3, and n. 1, 6. So we would have to overrule Florida Prepaid if we were to decide this case Allen’s way. But stare decisis, this Court has understood, is a “foundation stone of the rule of law.” Michigan v. Bay Mills Indian Community, 572 U.S. 782, 798 (2014). To reverse a decision, we demand a “special justification,” over and above the belief “that the precedent was wrongly decided.” Halliburton Co. v. Erica P. John Fund, Inc., 573 U.S. 258, 266 (2014). Allen offers us nothing special at all; he contends only that if the Court were to use a clause-by-clause approach, it would discover that Florida Prepaid was wrong (because, he says again, the decision misjudged Congress’s authority under the Intellectual Property Clause). See Brief for Petitioners 37; supra, at 6–7. And with that charge of error alone, Allen cannot overcome stare decisis. B Section 5 of the Fourteenth Amendment, unlike almost all of Article I, can authorize Congress to strip the States of immunity. The Fourteenth Amendment “fundamentally altered the balance of state and federal power” that the original Constitution and the Eleventh Amendment struck. Seminole Tribe, 517 U. S., at 59. Its first section imposes prohibitions on the States, including (as relevant here) that none may “deprive any person of life, liberty, or property, without due process of law.” Section 5 then gives Congress the “power to enforce, by appropriate legislation,” those limitations on the States’ authority. That power, the Court has long held, may enable Congress to abrogate the States’ immunity and thus subject them to suit in federal court. See Fitzpatrick v. Bitzer, 427 U.S. 445, 456 (1976). For an abrogation statute to be “appropriate” under Section 5, it must be tailored to “remedy or prevent” conduct infringing the Fourteenth Amendment’s substantive prohibitions. City of Boerne v. Flores, 521 U.S. 507, 519 (1997). Congress can permit suits against States for actual violations of the rights guaranteed in Section 1. See Fitzpatrick, 427 U. S., at 456. And to deter those violations, it can allow suits against States for “a somewhat broader swath of conduct,” including acts constitutional in themselves. Kimel, 528 U. S., at 81. But Congress cannot use its “power to enforce” the Fourteenth Amendment to alter what that Amendment bars. See id., at 88 (prohibiting Congress from “substantively redefin[ing]” the Fourteenth Amendment’s requirements). That means a congressional abrogation is valid under Section 5 only if it sufficiently connects to conduct courts have held Section 1 to proscribe. To decide whether a law passes muster, this Court has framed a type of means-end test. For Congress’s action to fall within its Section 5 authority, we have said, “[t]here must be a congruence and proportionality between the injury to be prevented or remedied and the means adopted to that end.” Boerne, 521 U. S., at 520. On the one hand, courts are to consider the constitutional problem Congress faced—both the nature and the extent of state conduct violating the Fourteenth Amendment. That assessment usually (though not inevitably) focuses on the legislative record, which shows the evidence Congress had before it of a constitutional wrong. See Florida Prepaid, 527 U. S., at 646. On the other hand, courts are to examine the scope of the response Congress chose to address that injury. Here, a critical question is how far, and for what reasons, Congress has gone beyond redressing actual constitutional violations. Hard problems often require forceful responses and, as noted above, Section 5 allows Congress to “enact[ ] reasonably prophylactic legislation” to deter constitutional harm. Kimel, 528 U. S., at 88; Boerne, 521 U. S., at 536 (Congress’s conclusions on that score are “entitled to much deference”); supra, at 10. But “[s]trong measures appropriate to address one harm may be an unwarranted response to another, lesser one.” Boerne, 521 U. S., at 530. Always, what Congress has done must be in keeping with the Fourteenth Amendment rules it has the power to “enforce.” All this raises the question: When does the Fourteenth Amendment care about copyright infringement? Sometimes, no doubt. Copyrights are a form of property. See Fox Film Corp. v. Doyal, 286 U.S. 123, 128 (1932). And the Fourteenth Amendment bars the States from “depriv[ing]” a person of property “without due process of law.” But even if sometimes, by no means always. Under our precedent, a merely negligent act does not “deprive” a person of property. See Daniels v. Williams, 474 U.S. 327, 328 (1986). So an infringement must be intentional, or at least reckless, to come within the reach of the Due Process Clause. See id., at 334, n. 3 (reserving whether reckless conduct suffices). And more: A State cannot violate that Clause unless it fails to offer an adequate remedy for an infringement, because such a remedy itself satisfies the demand of “due process.” See Hudson v. Palmer, 468 U.S. 517, 533 (1984). That means within the broader world of state copyright infringement is a smaller one where the Due Process Clause comes into play. Because the same is true of patent infringement, Florida Prepaid again serves as the critical precedent. That decision defined the scope of unconstitutional infringement in line with the caselaw cited above—as intentional conduct for which there is no adequate state remedy. See 527 U. S., at 642–643, 645. It then searched for evidence of that sort of infringement in the legislative record of the Patent Remedy Act. And it determined that the statute’s abrogation of immunity—again, the equivalent of the CRCA’s—was out of all proportion to what it found. That analysis is the starting point of our inquiry here. And indeed, it must be the ending point too unless the evidence of unconstitutional infringement is materially different for copyrights than patents. Consider once more, then, Florida Prepaid, now not on Article I but on Section 5. In enacting the Patent Remedy Act, Florida Prepaid found, Congress did not identify a pattern of unconstitutional patent infringement. To begin with, we explained, there was only thin evidence of States infringing patents at all—putting aside whether those actions violated due process. The House Report, recognizing that “many states comply with patent law,” offered just two examples of patent infringement suits against the States. Id., at 640 (quoting H. R. Rep. No. 101–960, pt. 1, p. 38 (1990)). The appellate court below, boasting some greater research prowess, discovered another seven in the century-plus between 1880 and 1990. See 527 U. S., at 640. Even the bill’s House sponsor conceded the lack of “any evidence” of “widespread violation of patent laws.” Id., at 641 (quoting statement of Rep. Kastenmeier). What was more, there was no evidence that any instance of infringement by States crossed constitutional lines. Congress, we observed, “did not focus” on intentional or reckless conduct; to the contrary, the legislative record suggested that “most state infringement was innocent or at worst negligent.” Id., at 645. And similarly, Congress “barely considered the availability of state remedies for patent infringement.” Id., at 643. So, we concluded, nothing could support the idea that States were more than sporadically (if that) “depriving patent owners of property without due process of law.” Id., at 646. Given that absence of evidence, Florida Prepaid held, the Patent Remedy Act swept too far. Recall what the Patent Remedy Act did—and did not. It abrogated sovereign immunity for any and every patent suit, thereby “plac[ing] States on the same footing as private parties.” Id., at 647. It did not set any limits. It did not, for example, confine the abrogation to suits alleging “nonnegligent infringement or infringement authorized [by] state policy.” Ibid. Neither did it target States refusing to offer alternative remedies to patent holders. No, it exposed all States to the hilt—on a record that failed to show they had caused any discernible constitutional harm (or, indeed, much harm at all). That imbalance made it impossible to view the legislation “as responsive to, or designed to prevent, unconstitutional behavior.” Id., at 646 (quoting Boerne, 521 U. S., at 532). The statute’s “indiscriminate scope” was too “out of proportion” to any due process problem. 527 U. S., at 646–647. It aimed not to correct such a problem, but to “provide a uniform remedy for patent infringement” writ large. Id., at 647. The Patent Remedy Act, in short, did not “enforce” Section 1 of the Fourteenth Amendment—and so was not “appropriate” under Section 5. Could, then, this case come out differently? Given the identical scope of the CRCA and Patent Remedy Act, that could happen only if the former law responded to materially stronger evidence of infringement, especially of the unconstitutional kind. Allen points to a significant disparity in how Congress created a record for the two statutes. See Brief for Petitioners 7–10, 47–50. Before enacting the CRCA, Congress asked the then-Register of Copyrights, Ralph Oman, to submit a report about the effects of the Eleventh Amendment on copyright enforcement. Oman and his staff conducted a year-long examination, which included a request for public comments eliciting letters from about 40 copyright holders and industry groups. The final 158-page report concluded that “copyright proprietors have demonstrated they will suffer immediate harm if they are unable to sue infringing states in federal court.” Copyright Office, Copyright Liability of States and the Eleventh Amendment 103 (1988) (Oman Report). Is that report enough, as Allen claims, to flip Florida Prepaid’s outcome when it comes to copyright cases against the States? It is not. Behind the headline-grabbing conclusion, nothing in the Oman Report, or the rest of the legislative record, cures the problems we identified in Florida Prepaid. As an initial matter, the concrete evidence of States infringing copyrights (even ignoring whether those acts violate due process) is scarcely more impressive than what the Florida Prepaid Court saw. Despite undertaking an exhaustive search, Oman came up with only a dozen possible examples of state infringement. He listed seven court cases brought against States (with another two dismissed on the merits) and five anecdotes taken from public comments (but not further corroborated). See Oman Report, at 7–9, 90–97. In testifying about the report, Oman acknowledged that state infringement is “not widespread” and “the States are not going to get involved in wholesale violation of the copyright laws.” Hearings on H. R. 1131 before the Subcommittee on Courts, Intellectual Property, and the Administration of Justice, 101st Cong., 1st Sess., 53 (1989) (House Hearings). Indeed, he opined: “They are all respectful of the copyright law” and “will continue to respect the law”; what State, after all, would “want[ ] to get a reputation as a copyright pirate?” Id., at 8. The bill’s House and Senate sponsors got the point. The former admitted that “there have not been any significant number” of copyright violations by States. Id., at 48 (Rep. Kastenmeier). And the latter conceded he could not currently see “a big problem.” Hearings on S. 497 before the Subcommittee on Patents, Copyrights and Trademarks, 101st Cong., 1st Sess., 130 (1989) (Sen. DeConcini). This is not, to put the matter charitably, the stuff from which Section 5 legislation ordinarily arises. And it gets only worse. Neither the Oman Report nor any other part of the legislative record shows concern with whether the States’ copyright infringements (however few and far between) violated the Due Process Clause. Of the 12 infringements listed in the report, only two appear intentional, as they must be to raise a constitutional issue. See Oman Report, at 7–8, 91 (describing a judicial finding of “willful” infringement and a public comment charging continued infringement after a copyright owner complained). As Oman testified, the far greater problem was the frequency of “honest mistakes” or “innocent” misunderstandings; the benefit of the bill, he therefore thought, would be to “guard against sloppiness.” House Hearings, at 9. Likewise, the legislative record contains no informa- tion about the availability of state-law remedies for copyright infringement (such as contract or unjust enrichment suits)—even though they might themselves satisfy due process. Those deficiencies in the record match the ones Florida Prepaid emphasized. See 527 U. S., at 643–645. Here no less than there, they signal an absence of constitutional harm. Under Florida Prepaid, the CRCA thus must fail our “congruence and proportionality” test. Boerne, 521 U. S., at 520. As just shown, the evidence of Fourteenth Amendment injury supporting the CRCA and the Patent Remedy Act is equivalent—for both, that is, exceedingly slight. And the scope of the two statutes is identical—extending to every infringement case against a State. It follows that the balance the laws strike between constitutional wrong and statutory remedy is correspondingly askew. In this case, as in Florida Prepaid, the law’s “indiscriminate scope” is “out of proportion” to any due process problem. 527 U. S., at 646–647; see supra, at 13. In this case, as in that one, the statute aims to “provide a uniform remedy” for statutory infringement, rather than to redress or prevent unconstitutional conduct. 527 U. S., at 647; see supra, at 13. And so in this case, as in that one, the law is invalid under Section 5. That conclusion, however, need not prevent Congress from passing a valid copyright abrogation law in the future. In doing so, Congress would presumably approach the issue differently than when it passed the CRCA. At that time, the Court had not yet decided Seminole Tribe, so Congress probably thought that Article I could support its all-out abrogation of immunity. See supra, at 6. And to the extent it relied on Section 5, Congress acted before this Court created the “congruence and proportionality” test. See supra, at 11. For that reason, Congress likely did not appreciate the importance of linking the scope of its abrogation to the redress or prevention of unconstitutional injuries—and of creating a legislative record to back up that connection. But going forward, Congress will know those rules. And under them, if it detects violations of due process, then it may enact a proportionate response. That kind of tailored statute can effectively stop States from behaving as copyright pirates. Even while respecting constitutional limits, it can bring digital Blackbeards to justice. III Florida Prepaid all but prewrote our decision today. That precedent made clear that Article I’s Intellectual Property Clause could not provide the basis for an abrogation of sovereign immunity. And it held that Section 5 of the Fourteenth Amendment could not support an abrogation on a legislative record like the one here. For both those reasons, we affirm the judgment below. It is so ordered. Notes 1 The CRCA served as the model for the Patent and Plant Variety Protection Clarification Act (Patent Remedy Act), passed two years later (and repudiated by this Court in Florida Prepaid, see supra, at 1). Using the same language, the latter statute provided that a State “shall not be immune, under the [E]leventh [A]mendment [or] any other doctrine of sovereign immunity, from suit in Federal court” for patent infringement. §2, 106Stat. 4230. And so too, the statute specified that in such a suit, a State will be liable, and subject to remedies, “in the same manner and to the same extent as” a private party. Ibid. 2 The Eleventh Amendment reads: “The Judicial Power of the United States shall not be construed to extend to any suit in law or equity, commenced or prosecuted against one of the United States by Citizens of another State, or by Citizens or Subjects of any Foreign State.”

590.US.2019_17-1498

The Comprehensive Environmental Response, Compensation, and Liability Act, 42 U. S. C. §9601 et seq., also known as the Superfund statute, promotes “the timely cleanup of hazardous waste sites and [ensures] that the costs of such cleanup efforts [are] borne by those responsible for the contamination,” CTS Corp. v. Waldburger, 573 U.S. 1, 4 (internal quotation marks omitted). The Act directs the Environmental Protection Agency to compile and annually revise a prioritized list of contaminated sites for cleanup, known as Superfund sites, and makes responsible parties liable for the cost of the cleanup. Before a cleanup plan is selected, a remedial investigation and feasibility study is conducted to assess the contamination and evaluate cleanup options. Once that study begins, §122(e)(6) of the Act provides, “no potentially responsible party may undertake any remedial action” at the site without EPA approval. To insulate cleanup plans from collateral attack, §113(b) provides federal district courts with “exclusive original jurisdiction over all controversies arising under” the Act, and §113(h) then strips those courts of jurisdiction “to review any challenges to removal or remedial action,” except in five limited circumstances. For nearly a century, the Anaconda Copper Smelter in Butte, Montana contaminated an area of over 300 square miles with arsenic and lead. Over the past 35 years, EPA has worked with the current owner of the now-closed smelter, Atlantic Richfield Company, to implement a cleanup plan for a remediation expected to continue through 2025. A group of 98 landowners sued Atlantic Richfield in Montana state court for common law nuisance, trespass, and strict liability, seeking restoration damages, which Montana law requires to be spent on property rehabilitation. The landowners’ proposed plan exceeds the measures found necessary to protect human health and the environment by EPA. The trial court granted summary judgment to the landowners on the issue of whether the Act precluded their restoration damages claim and allowed the lawsuit to continue. After granting a writ of supervisory control, the Montana Supreme Court affirmed, rejecting Atlantic Richfield’s argument that §113 stripped the Montana courts of jurisdiction over the landowners’ claim and concluding that the landowners were not potentially responsible parties (or PRPs) prohibited from taking remedial action without EPA approval under §122(e)(6). Held: 1. This Court has jurisdiction to review the Montana Supreme Court’s decision. To qualify as a final judgment subject to review under 28 U. S. C. §1257(a), a state court judgment must be “an effective determination of the litigation and not of merely interlocutory or intermediate steps therein.” Jefferson v. City of Tarrant, 522 U.S. 75, 81. Under Montana law, a supervisory writ proceeding is a self-contained case, not an interlocutory appeal. Mont. Const., Art. VII, §§2(1)–(2); Mont. Rules App. Proc. 6(6), 14(1), 14(3). Thus, the writ issued in this case is a “final judgment” within this Court’s jurisdiction. Fisher v. District Court of Sixteenth Judicial Dist. of Mont., 424 U.S. 382, 385, n. 7. P. 8. 2. The Act does not strip the Montana courts of jurisdiction over this lawsuit. Section 113(b) of the Act provides that “the United States district courts shall have exclusive original jurisdiction over all controversies arising under this chapter,” so state courts lack jurisdiction over such actions. The use of “arising under” in §113(b) echoes Congress’s more familiar use of that phrase in granting federal courts jurisdiction over “all civil actions arising under the Constitution, laws, or treaties of the United States.” 28 U. S. C. §1331. In the mine run of cases, “[a] suit arises under the law that creates the cause of action.” American Well Works Co. v. Layne & Bowler Co., 241 U.S. 257, 260. The landowners’ common law nuisance, trespass, and strict liability claims arise under Montana law and not under the Act. Atlantic Richfield mistakenly argues that §113(h)—which states that “[n]o Federal court shall have jurisdiction under Federal law . . . to review any challenges to removal or remedial action” selected under the Act—implicitly broadens the scope of actions precluded from state court jurisdiction under §113(b). But §113(h) speaks of “Federal court[s],” not state courts. There is no textual basis for Atlantic Richfield’s argument that Congress precluded state courts from hearing a category of cases in §113(b) by stripping federal courts of jurisdiction over those cases in §113(h). Often the simplest explanation is the best: Section 113(b) deprives state courts of jurisdiction over cases “arising under” the Act—just as it says—while §113(h) deprives federal courts of jurisdiction over certain “challenges” to Superfund remedial actions—just as it says. Pp. 8–13. 3. The Montana Supreme Court erred by holding that the landowners were not potentially responsible parties under the Act and thus did not need EPA approval to take remedial action. To determine who is a potentially responsible party, the Court looks to the list of “covered persons” in §107, the Act’s liability section, which includes any “owner” of “a facility.” “Facility” in turn is defined to include “any site or area where a hazardous substance has been deposited, stored, disposed of, or placed, or otherwise come to be located.” 42 U. S. C. §9601(9)(B). Because arsenic and lead are hazardous substances that have “come to be located” on the landowners’ properties, the landowners are potentially responsible parties. The landowners argue they are no longer potentially responsible parties because the Act’s six-year limitations period for recovery of remedial costs has run, and thus they could not be held liable in a hypothetical lawsuit. But even “ ‘innocent’ . . . landowner[s] whose land has been contaminated by another,” and who are thus shielded from liability by §107(b)(3)’s so-called “innocent landowner” or “third party” defense, “may fall within the broad definitions of PRPs in §§107(a)(1)–(4).” United States v. Atlantic Research Corp., 551 U.S. 128, 136. The same principle holds true for parties facing no liability because of the Act’s limitations period. Interpreting “potentially responsible parties” to include owners of polluted property reflects the Act’s objective to develop a “Comprehensive Environmental Response” to hazardous waste pollution. Section 122(e)(6) is one of several tools in the Act that ensure the careful development of a single EPA-led cleanup effort rather than tens of thousands of competing individual ones. Yet under the landowners’ interpretation, property owners would be free to dig up arsenic-infected soil and build trenches to redirect lead-contaminated groundwater without even notifying EPA, so long as they have not been sued within six years of commencement of the cleanup. Congress did not provide such a fragile remedy for such a serious problem. The landowners alternatively argue that they are not potentially responsible parties because they did not receive the notice of settlement negotiations required by §122(e)(1). EPA has a policy of not suing innocent homeowners for pollution they did not cause, so it did not include the landowners in settlement negotiations. But EPA’s nonenforcement policy does not alter the landowners’ status as potentially responsible parties. Section 107(a) unambiguously defines potentially responsible parties, and EPA does not have authority to alter that definition. The landowners also argue that §122(e)(6) cannot carry the weight ascribed to it because it is located in the section on settlement negotiations. Settlements, however, are the heart of the Superfund statute. Section 122(a) of the Act commands EPA to proceed by settlement “[w]henever practicable and in the public interest . . . in order to expedite effective remedial actions and minimize litigation.” And EPA’s efforts to negotiate settlement agreements and issue orders for cleanups account for approximately 69% of all cleanup work currently underway. Pp. 13–21. 390 Mont. 76, 408 P.3d 515, affirmed in part, vacated in part, and remanded. Roberts, C. J., delivered the opinion of the Court, Parts I and II–A of which were unanimous, Part II–B of which was joined by Thomas, Ginsburg, Breyer, Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., and Part III of which was joined by Ginsburg, Breyer, Alito, Sotomayor, Kagan, and Kavanaugh, JJ. Alito, J., filed an opinion concurring in part and dissenting in part. Gorsuch, J., filed an opinion concurring part and dissenting in part, in which Thomas, J., joined.

For nearly a century, the Anaconda Copper Smelter in Butte, Montana contaminated an area of over 300 square miles with arsenic and lead. Over the past 35 years, the Environmental Protection Agency has worked with the current owner of the smelter, Atlantic Richfield Company, to implement a cleanup plan under the Comprehensive Environmental Response, Compensation, and Liability Act of 1980. EPA projects that the cleanup will continue through 2025. A group of 98 landowners sued Atlantic Richfield in Montana state court for common law nuisance, trespass, and strict liability. Among other remedies, the landowners sought restoration damages, which under Montana law must be spent on rehabilitation of the property. The landowners’ proposed restoration plan includes measures beyond those the agency found necessary to protect human health and the environment. We consider whether the Act strips the Montana courts of jurisdiction over the landowners’ claim for restoration damages and, if not, whether the Act requires the land- owners to seek EPA approval for their restoration plan. I A In 1980, Congress enacted the Comprehensive Environmental Response, Compensation, and Liability Act, 94Stat. 2767, as amended, 42 U. S. C. §9601 et seq., also known as the Superfund statute, to address “the serious environmental and health risks posed by industrial pollution,” Burlington N. & S. F. R. Co. v. United States, 556 U.S. 599, 602 (2009). The Act seeks “to promote the timely cleanup of hazardous waste sites and to ensure that the costs of such cleanup efforts [are] borne by those responsible for the contamination.” CTS Corp. v. Waldburger, 573 U.S. 1, 4 (2014) (internal quotation marks omitted). The Act directs EPA to compile and annually revise a prioritized list of contaminated sites for cleanup, commonly known as Superfund sites. 42 U. S. C. §9605.[1] EPA may clean those sites itself or compel responsible parties to perform the cleanup. §§9604, 9606, 9615. If the Government performs the cleanup, it may recover its costs from responsible parties. §9607(a)(4)(A). Responsible parties are jointly and severally liable for the full cost of the cleanup, but may seek contribution from other responsible parties. §9613(f )(1). Prior to selecting a cleanup plan, EPA conducts (or orders a private party to conduct) a remedial investigation and feasibility study to assess the contamination and evaluate cleanup options. 40 CFR §300.430 (2019). Section 122(e)(6) of the Act provides that, once the study begins, “no potentially responsible party may undertake any remedial action” at the site without EPA approval. 42 U. S. C. §9622(e)(6). The Act prescribes extensive public consultation while a cleanup plan is being developed. It requires an opportunity for public notice and comment on proposed cleanup plans. §§9613(k), 9617. It requires “substantial and mean- ingful involvement by each State in initiation, development, and selection” of cleanup actions in that State. §9621(f )(1). And, in most instances, it requires that remedial action comply with “legally applicable or relevant and appropriate” requirements of state environmental law. §9621(d)(2)(A). But once a plan is selected, the time for debate ends and the time for action begins. To insulate cleanup plans from collateral attack, §113(b) of the Act provides federal district courts with “exclusive original jurisdiction over all controversies arising under” the Act, and §113(h) then strips such courts of jurisdiction “to review any challenges to removal or remedial action,” except in five limited circumstances. §§9613(b), (h). B Between 1884 and 1902, the Anaconda Copper Mining Company built three copper smelters 26 miles west of the mining town of Butte, Montana. The largest one, the Washoe Smelter, featured a 585-foot smoke stack, taller than the Washington Monument. The structure still towers over the area today, as part of the Anaconda Smoke Stack State Park. Together, the three smelters refined tens of millions of pounds of copper ore mined in Butte, the “Richest Hill on Earth,” to feed burgeoning demand for telephone wires and power lines. M. Malone, The Battle for Butte 34 (1981). “It was hot. It was dirty. It was dangerous. But it was a job for thousands.” Dunlap, A Dangerous Job That Gave Life to a Town: A Look Back at the Anaconda Smelter, Montana Standard (Aug. 8, 2018). From 1912 to 1973, Anaconda Company payrolls totaled over $2.5 billion, compensating around three-quarters of Montana’s work force. Bust followed boom. By the 1970s, the falling price of copper, an ongoing energy crisis, and the nationalization of Anaconda’s copper mines in Chile and Mexico squeezed Anaconda. But what others saw as an ailing relic, Atlantic Richfield saw as a turnaround opportunity, purchasing the Anaconda Company for the discount price of $700 million. Unfortunately, Atlantic Richfield was unable to revive Anaconda’s fortunes. By 1980 Atlantic Richfield had closed the facility for good, and by 1984 Fortune had dubbed the purchase one of the “Decade’s Worst Mergers.” Fisher, The Decade’s Worst Mergers, Fortune, Apr. 30, 1984, p. 262. Atlantic Richfield’s troubles were just beginning. After Congress passed the Superfund statute in 1980, Atlantic Richfield faced strict and retroactive liability for the many tons of arsenic and lead that Anaconda had spewed across the area over the previous century. In 1983, EPA designated an area of more than 300 square miles around the smelters as one of the inaugural Superfund sites. 48 Fed. Reg. 40667. In the 35 years since, EPA has managed an extensive cleanup at the site, working with Atlantic Richfield to remediate more than 800 residential and commercial properties; remove 10 million cubic yards of tailings, mine waste, and contaminated soil; cap in place 500 million cubic yards of waste over 5,000 acres; and reclaim 12,500 acres of land. EPA, Superfund Priority “Anaconda” 9 (Apr. 2018), https://semspub.epa.gov/work/08/100003986.pdf. To date, Atlantic Richfield estimates that it has spent roughly $450 million implementing EPA’s orders. More work remains. As of 2015, EPA’s plan anticipated cleanup of more than 1,000 additional residential yards, revegetation of 7,000 acres of uplands, removal of several waste areas, and closure of contaminated stream banks and railroad beds. Brief for United States as Amicus Curiae 7–8 (citing EPA, Fifth Five-Year Review Report: Anaconda Smelter Superfund Site, Anaconda-Deer Lodge County, Montana, Table 10–1 (Sept. 25, 2015), https://semspub.epa. gov/work/08/1549381.pdf ). EPA projects that remedial work will continue through 2025. Id., Table 10–7; Tr. of Oral Arg. 30. C In 2008, a group of 98 owners of property within the Superfund site filed this lawsuit against Atlantic Richfield in Montana state court, asserting trespass, nuisance, and strict liability claims under state common law. The landowners sought restoration damages, among other forms of relief. Under Montana law, property damages are generally measured by the “difference between the value of the property before and after the injury, or the diminution in value.” Sunburst School Dist. No. 2 v. Texaco, Inc., 338 Mont. 259, 269, 165 P.3d 1079, 1086 (2007). But “when the damaged property serves as a private residence and the plaintiff has an interest in having the property restored, diminution in value will not return the plaintiff to the same position as before the tort.” Id., at 270, 165 P. 3d, at 1087. In that circumstance, the plaintiff may seek restoration damages, even if they exceed the property’s diminution in value. See ibid.; Restatement (Second) of Torts §929, and Comment b (1977). To collect restoration damages, a plaintiff must demonstrate that he has “reasons personal” for restoring the property and that his injury is temporary and abatable, meaning “[t]he ability to repair [the] injury must be more than a theoretical possibility.” Sunburst School Dist. No. 2, 338 Mont., at 269, 165 P. 3d, at 1086–1087. The injured party must “establish that the award actually will be used for restoration.” Lampi v. Speed, 362 Mont. 122, 130, 261 P.3d 1000, 1006 (2011). The landowners here propose a restoration plan that goes beyond EPA’s own cleanup plan, which the agency had found “protective of human health and the environment.” EPA, Community Soils Operable Unit, Record of Decision (1996), App. 62. See also 42 U. S. C. §9621(d)(1). For example, the landowners propose a maximum soil contamination level of 15 parts per million of arsenic, rather than the 250 parts per million level set by EPA. And the landowners seek to excavate offending soil within residential yards to a depth of two feet rather than EPA’s chosen depth of one. The landowners also seek to capture and treat shallow groundwater through an 8,000-foot long, 15-foot deep, and 3-foot wide underground permeable barrier, a plan the agency rejected as costly and unnecessary to secure safe drinking water. The landowners estimate that their cleanup would cost Atlantic Richfield $50 to $58 million. Atlantic Richfield would place that amount in a trust and the trustee would release funds only for restoration work. In the trial court, Atlantic Richfield and the landowners filed competing motions for summary judgment on whether the Act precluded the landowners’ claim for restoration damages.[2] The court granted judgment for the landowners on that issue and allowed the lawsuit to continue. After granting a writ of supervisory control, the Montana Supreme Court affirmed. Atlantic Richfield Co. v. Montana Second Jud. Dist. Ct., 390 Mont. 76, 408 P.3d 515 (2017). The Montana Supreme Court rejected Atlantic Richfield’s argument that §113 stripped the Montana courts of jurisdiction over the landowners’ claim for restoration damages. The court recognized that §113 strips federal courts (and, it was willing to assume, state courts) of jurisdiction to review challenges to EPA cleanup plans. But the Montana Supreme Court reasoned that the landowners’ plan was not such a challenge because it would not “stop, delay, or change the work EPA is doing.” Id., at 83, 408 P. 3d, at 520. The landowners were “simply asking to be allowed to present their own plan to restore their own private property to a jury of twelve Montanans who will then assess the merits of that plan.” Id., at 84, 408 P. 3d, at 521. The Montana Supreme Court also rejected Atlantic Richfield’s argument that the landowners were potentially responsible parties (sometimes called PRPs) prohibited from taking remedial action without EPA approval under §122(e)(6) of the Act. The Court observed that the landowners had “never been treated as PRPs for any purpose—by either EPA or [Atlantic Richfield]—during the entire thirty-plus years” since the designation of the Superfund site, and that the statute of limitations for a claim against the landowners had run. Id., at 86, 408 P. 3d, at 522. “Put simply, the PRP horse left the barn decades ago.” Ibid. Justice Baker concurred, stressing that on remand Atlantic Richfield could potentially defeat the request for restoration damages on the merits by proving that the restoration plan conflicted with EPA’s cleanup plan. Id., at 87–90, 408 P. 3d, at 523–525. Justice McKinnon dissented. She argued that the landowners’ restoration plan did conflict with the Superfund cleanup and thus constituted a challenge under §113(h) of the Act, over which Montana courts lacked jurisdiction. Id., at 90–101, 408 P. 3d, at 525–532. We granted certiorari. 587 U. S. ___ (2019). II We begin with two threshold questions: whether this Court has jurisdiction to review the decision of the Montana Supreme Court and, if so, whether the Montana courts have jurisdiction over the landowners’ claim for restoration damages. A Congress has authorized this Court to review “[f]inal judgments or decrees rendered by the highest court of a State.” 28 U. S. C. §1257(a). To qualify as final, a state court judgment must be “an effective determination of the litigation and not of merely interlocutory or intermediate steps therein.” Jefferson v. City of Tarrant, 522 U.S. 75, 81 (1997). The landowners contend that, because the Montana Supreme Court allowed the case to proceed to trial, its judgment was not final and we lack jurisdiction. But the Montana Supreme Court exercised review in this case through a writ of supervisory control. Under Montana law, a supervisory writ proceeding is a self-contained case, not an interlocutory appeal. Mont. Const., Art. VII, §§2(1)–(2); Mont. Rules App. Proc. 6(6), 14(1), 14(3) (2019). Thus we have held that a “writ of supervisory control issued by the Montana Supreme Court is a final judgment within our jurisdiction.” Fisher v. District Court of Sixteenth Judicial Dist. of Mont., 424 U.S. 382, 385, n. 7 (1976) (per curiam). The landowners protest that our precedents only support reviewing supervisory writ proceedings that are limited to jurisdictional questions. But the scope of our jurisdiction to review supervisory writ proceedings is not so restricted. When the Montana Supreme Court issues a writ of supervisory control, it initiates a separate lawsuit. It is the nature of the Montana proceeding, not the issues the state court reviewed, that establishes our jurisdiction. B We likewise find that the Act does not strip the Montana courts of jurisdiction over this lawsuit. It deprives state courts of jurisdiction over claims brought under the Act. But it does not displace state court jurisdiction over claims brought under other sources of law.[3] Section 113(b) of the Act provides that “the United States district courts shall have exclusive original jurisdiction over all controversies arising under this chapter,” so state courts lack jurisdiction over such actions. 42 U. S. C. §9613(b). This case, however, does not “arise under” the Act. The use of “arising under” in §113(b) echoes Congress’s more familiar use of that phrase in granting federal courts jurisdiction over “all civil actions arising under the Constitution, laws, or treaties of the United States.” 28 U. S. C. §1331. In the mine run of cases, “[a] suit arises under the law that creates the cause of action.” American Well Works Co. v. Layne & Bowler Co., 241 U.S. 257, 260 (1916).[4] The landowners’ common law claims for nuisance, trespass, and strict liability therefore arise under Montana law and not under the Act. As a result, the Montana courts retain jurisdiction over this lawsuit, notwithstanding the channeling of Superfund claims to federal courts in §113(b).[5] Atlantic Richfield takes a different view, arguing that §113(h) implicitly broadens the scope of actions precluded from state court jurisdiction under §113(b). Section 113(h) states that “[n]o Federal court shall have jurisdiction under Federal law other than under section 1332 of title 28 (relating to diversity of citizenship jurisdiction) . . . to review any challenges to removal or remedial action” selected under the Act. 42 U. S. C. §9613(h). The company’s argument proceeds in five steps. Step one: Section 113(h) removes federal court jurisdiction over all cleanup challenges, regardless of whether they originate in federal or state law (except for when the court is sitting in diversity). Step two: Section 113(h) can only remove jurisdiction that §113(b) provides in the first place. Step three: Section 113(b) thus provides federal courts jurisdiction over all cleanup challenges, whether brought under federal or state law. Step four: The grant of jurisdiction to federal courts in §113(b) is exclusive to federal courts. Step five: State courts thus do not have jurisdiction over cleanup challenges. This interpretation faces several insurmountable obstacles. First, by its own terms, §113(h) speaks of “Federal court[s],” not state courts. There is no textual basis for Atlantic Richfield’s argument that Congress precluded state courts from hearing a category of cases in §113(b) by stripping federal courts of jurisdiction over those cases in §113(h). And if that were Congress’s goal, it would be hard to imagine a more oblique way of achieving it. Often the simplest explanation is the best: Section 113(b) deprives state courts of jurisdiction over cases “arising under” the Act—just as it says—while §113(h) deprives federal courts of jurisdiction over certain “challenges” to Superfund remedial actions—just as it says. Second, the company’s argument does not account for the exception in §113(h) for federal courts sitting in diversity. Section 113(h) permits federal courts in diversity cases to entertain state law claims regardless of whether they are challenges to cleanup plans. See DePue v. Exxon Mobil Corp., 537 F.3d 775, 784 (CA7 2008). But Atlantic Richfield does not even try to explain why the Act would permit such state law claims to proceed in federal court, but not in state court. The Act permits federal courts and state courts alike to entertain state law claims, including challenges to cleanups. That leads us to the third difficulty with Atlantic Richfield’s argument. We have recognized a “deeply rooted presumption in favor of concurrent state court jurisdiction” over federal claims. Tafflin v. Levitt, 493 U.S. 455, 458–459 (1990). Only an “explicit statutory directive,” an “unmistakable implication from legislative history,” or “a clear incompatibility between state-court jurisdiction and federal interests” can displace this presumption. Id., at 460. Explicit, unmistakable, and clear are not words that describe Atlantic Richfield’s knotty interpretation of §§113(b) and (h). It would be one thing for Atlantic Richfield to try to surmount the clear statement rule that applies to the uncommon, but not unprecedented, step of stripping state courts of jurisdiction over federal claims. But Atlantic Richfield’s position requires a more ambitious step: Congress stripping state courts of jurisdiction to hear their own state claims. We would not expect Congress to take such an extraordinary step by implication. Yet the only provision Atlantic Richfield invokes addresses “[f]ederal court[s]” without even mentioning state courts, let alone stripping those courts of jurisdiction to hear state law claims. 42 U. S. C. §9613(h). Finally, the Government, supporting Atlantic Richfield, emphasizes that the opening clause of §113(b) excepts §113(h) from its application. See 42 U. S. C. §9613(b) (“Except as provided in subsections (a) and (h) of this section . . . .”). According to the Government, because “exceptions must by definition be narrower than the corresponding rule,” all challenges to remedial plans under §113(h)—whether based in federal or state law—must “arise under” the Act for purposes of §113(b). Brief for United States as Amicus Curiae 25. We reject the premise and with it the conclusion. “Thousands of statutory provisions use the phrase ‘except as provided in . . . ’ followed by a cross-reference in order to indicate that one rule should prevail over another in any circumstance in which the two conflict.” Cyan, Inc. v. Beaver County Employees Retirement Fund, 583 U. S. ___, ___ (2018) (slip op., at 9). Such clauses explain what happens in the case of a clash, but they do not otherwise expand or contract the scope of either provision by implication. Cf. NLRB v. SW General, Inc., 580 U. S. ___, ___ (2017) (slip op., at 11) (explaining the same principle for “notwithstanding” clauses). The actions referred to in §113(h) do not fall entirely within §113(b). Challenges to remedial actions under federal statutes other than the Act, for example, are precluded by §113(h) but do not fall within §113(b). To cite another example, §113(h) addresses state law challenges to cleanup plans in federal court, although those actions also do not fall within §113(b).[6] At the same time, §113(b) is not subsumed by §113(h). Many claims brought under the Act, such as those to recover cleanup costs under §107, are not challenges to cleanup plans. Sections 113(b) and 113(h) thus each do work independent of one another. The two provisions overlap in a particular type of case: challenges to cleanup plans in federal court that arise under the Act. In such cases, the exceptions clause in §113(b) instructs that the limitation of §113(h) prevails. It does nothing more. III Although the Montana Supreme Court answered the jurisdictional question correctly, the Court erred by holding that the landowners were not potentially responsible parties under the Act and therefore did not need EPA approval to take remedial action. Section 122(e)(6), titled “Inconsistent response action,” provides that “[w]hen either the President, or a potentially responsible party . . . has initiated a remedial investigation and feasibility study for a particular facility under this chapter, no potentially responsible party may undertake any remedial action at the facility unless such remedial action has been authorized by the President.” 42 U. S. C. §9622(e)(6). Both parties agree that this provision would require the landowners to obtain EPA approval for their restoration plan if the landowners qualify as potentially responsible parties. To determine who is a potentially responsible party, we look to the list of “covered persons” in §107, the liability section of the Act. §9607(a). “Section 107(a) lists four classes of potentially responsible persons (PRPs) and provides that they ‘shall be liable’ for, among other things, ‘all costs of removal or remedial action incurred by the United States Government.’ ” Cooper Industries, Inc. v. Aviall Services, Inc., 543 U.S. 157, 161 (2004) (quoting §9607(a)(4)(A)). The first category under §107(a) includes any “owner” of “a facility.” §9607(a)(1). “Facility” is defined to include “any site or area where a hazardous substance has been deposited, stored, disposed of, or placed, or otherwise come to be located.” §9601(9)(B). Arsenic and lead are hazardous substances. 40 CFR §302.4, Table 302.4. Because those pollutants have “come to be located” on the landowners’ properties, the landowners are potentially responsible parties. The landowners and Justice Gorsuch argue that even if the landowners were once potentially responsible parties, they are no longer because the Act’s six-year limitations period for recovery of remedial costs has run, and thus they could not be held liable in a hypothetical lawsuit. 42 U. S. C. §9613(g)(2)(B). This argument collapses status as a potentially responsible party with liability for the payment of response costs. A property owner can be a potentially responsible party even if he is no longer subject to suit in court. As we have said, “[E]ven parties not responsible for contamination may fall within the broad definitions of PRPs in §§107(a)(1)–(4).” United States v. Atlantic Research Corp., 551 U.S. 128, 136 (2007). That includes “ ‘innocent’ . . . landowner[s] whose land has been contaminated by another,” who would be shielded from liability by the Act’s so-called “innocent landowner” or “third party” defense in §107(b)(3). Ibid. See also 42 U. S. C. §9607(b)(3). The same principle holds true for parties that face no liability because of the Act’s limitations period. Interpreting “potentially responsible parties” to include owners of polluted property reflects the Act’s objective to develop, as its name suggests, a “Comprehensive Environmental Response” to hazardous waste pollution. Section 122(e)(6) is one of several tools in the Act that ensure the careful development of a single EPA-led cleanup effort rather than tens of thousands of competing individual ones. Yet under the landowners’ interpretation, property owners would be free to dig up arsenic-infected soil and build trenches to redirect lead-contaminated groundwater without even notifying EPA, so long as they have not been sued within six years of commencement of the cleanup.[7] We doubt Congress provided such a fragile remedy for such a serious problem. And we suspect most other landowners would not be too pleased if Congress required EPA to sue each and every one of them just to ensure an orderly cleanup of toxic waste in their neighborhood. A straight- forward reading of the text avoids such anomalies. Justice Gorsuch argues that equating “potentially responsible parties” with “covered persons” overlooks the fact that the terms “use different language, appear in different statutory sections, and address different matters.” Post, at 7 (opinion concurring in part and dissenting in part). He contends that “potentially responsible party” as used in §122(e)(6) should be read as limited to the settlement context, and that if Congress intended the phrase to have broader reach—to refer more generally to those potentially liable under §107(a)—then Congress would have used the term “covered person.” Post, at 7–8. But there is no reason to think Congress used these phrases to refer to two distinct groups of persons. Neither phrase appears among the Act’s list of over 50 defined terms. 42 U. S. C. §9601. “Covered persons,” in fact, appears in the caption to §107(a) and nowhere else. Meanwhile, “potentially responsible parties” are referenced not just in the section on settlements, but also in the Act’s sections regarding EPA response authority, cleanup standards and procedures, cleanup contractors, Superfund moneys, Federal Government cleanup sites, and civil proceedings. §§9604, 9605, 9611, 9613, 9619, 9620, 9622. Across the statute “potentially responsible parties” refers to what it says: parties that may be held accountable for hazardous waste in particular circumstances. The only place in the Act that identifies such persons is the list of “Covered persons” in §107(a). Congress therefore must have intended “potentially responsible party” in §122(e)(6) (as elsewhere in the Act) to refer to “Covered persons” in §107(a). Turning from text to consequences, the landowners warn that our interpretation of §122(e)(6) creates a permanent easement on their land, forever requiring them “to get permission from EPA in Washington if they want to dig out part of their backyard to put in a sandbox for their grandchildren.” Tr. of Oral Arg. 62. The grandchildren of Montana can rest easy: The Act does nothing of the sort. Section 122(e)(6) refers only to “remedial action,” a defined term in the Act encompassing technical actions like “storage, confinement, perimeter protection using dikes, trenches, or ditches, clay cover, neutralization, cleanup of released hazardous substances and associated contaminated materials,” and so forth. 42 U. S. C. §9601(24). While broad, the Act’s definition of remedial action does not reach so far as to cover planting a garden, installing a lawn sprinkler, or digging a sandbox. In addition, §122(e)(6) applies only to sites on the Superfund list. The Act re- quires EPA to annually review and reissue that list. §9605(a)(8)(B). EPA delists Superfund sites once responsible parties have taken all appropriate remedial action and the pollutant no longer poses a significant threat to public health or the environment. See 40 CFR §300.425(e). The landowners and Justice Gorsuch alternatively argue that the landowners are not potentially responsible parties because they did not receive the notice of settlement negotiations required by §122(e)(1). Under a policy dating back to 1991, EPA does not seek to recover costs from residential landowners who are not responsible for contamination and do not interfere with the agency’s remedy. EPA, Policy Towards Owners of Residential Property at Superfund Sites, OSWER Directive #9834.6 (July 3, 1991), https://www.epa.gov/sites/production/files/documents/policy-owner-rpt.pdf. EPA views this policy as an exercise of its “enforcement discretion in pursuing potentially responsible parties.” Id., at 3. Because EPA has a policy of not suing innocent homeowners for pollution they did not cause, it did not include the landowners in settlement negotiations. But EPA’s nonenforcement policy does not alter the landowners’ status as potentially responsible parties. Section 107(a) unambiguously defines potentially responsible parties and EPA does not have authority to alter that definition. See, e.g., Sturgeon v. Frost, 587 U. S. ___, ___, n. 3 (2019) (slip op., at 16, n. 3). Section 122(e)(1) requires notification of settlement negotiations to all potentially responsible parties. To say that provision determines who is a potentially responsible party in the first instance would render the Act circular. Even the Government does not claim that its decisions whether to send notices of settlement negotiations carry such authority. In short, even if EPA ran afoul of §122(e)(1) by not providing the landowners notice of settlement negotiations, that does not change the landowners’ status as potentially responsible parties. The landowners relatedly argue that the limitation in §122(e)(6) on remedial action by potentially responsible parties cannot carry the weight we assign to it because it is located in the Act’s section on settlement negotiations. Congress, we are reminded, does not “hide elephants in mouseholes.” Whitman v. American Trucking Assns., Inc., 531 U.S. 457, 468 (2001). We take no issue with characterizing §122(e)(6) as an elephant. It is, after all, one of the Act’s crucial tools for ensuring an orderly cleanup of toxic waste. But §122 of the Act is, at the risk of the tired metaphor spinning out of control, less a mousehole and more a watering hole—exactly the sort of place we would expect to find this elephant. Settlements are the heart of the Superfund statute. EPA’s efforts to negotiate settlement agreements and issue orders for cleanups account for approximately 69% of all cleanup work currently underway. EPA, Superfund Site Cleanup Work Through Enforcement Agreements and Orders, https://www.epa.gov/enforcement/superfund-site-cleanup-work-through-enforcement-agreements-and-orders. The Act commands EPA to proceed by settlement “[w]henever practicable and in the public interest . . . in order to expedite effective remedial actions and minimize litigation.” 42 U. S. C. §9622(a). EPA, for its part, “prefers to reach an agreement with a potentially responsible party (PRP) to clean up a Superfund site instead of issuing an order or paying for it and recovering the cleanup costs later.” EPA, Negotiating Superfund Settlements, https://www.epa. gov/enforcement/negotiating-superfund-settlements. The Act encourages potentially responsible parties to enter into such agreements by authorizing EPA to include a “covenant not to sue,” which caps the parties’ liability to the Government. §9622(c)(1). The Act also protects settling parties from contribution claims by other potentially responsible parties. §9613(f )(2). Once finalized, the terms of a settlement become legally binding administrative orders, subject to civil penalties of up to $25,000 a day. §§9609(a)(1)(E), 9622(l). Moreover, subsection (e) is an important component of §122. It establishes a reticulated scheme of notices, proposals, and counterproposals for the settlement negotiation process. §9622(e). And the subsection places a moratorium on EPA remedial actions while negotiations are under way. §9622(e)(2)(A). It is far from surprising to find an analogous provision restricting potentially responsible parties from taking remedial actions in the same subsection. Justice Gorsuch also contends that our interpretation violates the Act’s “saving clauses,” which provide that the Act does not preempt liability or requirements under state law. Post, at 3–4. But we have long rejected interpretations of sweeping saving clauses that prove “absolutely inconsistent with the provisions of the act” in which they are found. American Telephone & Telegraph Co. v. Central Office Telephone, Inc., 524 U.S. 214, 228 (1998) (quoting Texas & Pacific R. Co. v. Abilene Cotton Oil Co., 204 U.S. 426, 446 (1907)). Interpreting the Act’s saving clauses to erase the clear mandate of §122(e)(6) would allow the Act “to destroy itself.” Ibid. What is more, Atlantic Richfield remains potentially liable under state law for compensatory damages, including loss of use and enjoyment of property, diminution of value, incidental and consequential damages, and annoyance and discomfort. The damages issue before the Court is whether Atlantic Richfield is also liable for the landowners’ own remediation beyond that required under the Act. Even then, the answer is yes—so long as the landowners first obtain EPA approval for the remedial work they seek to carry out. We likewise resist Justice Gorsuch’s evocative claim that our reading of the Act endorses “paternalistic central planning” and turns a cold shoulder to “state law efforts to restore state lands.” Post, at 10. Such a charge fails to appreciate that cleanup plans generally must comply with “legally applicable or relevant and appropriate” standards of state environmental law. 42 U. S. C. §9621(d)(2)(A)(ii). Or that States must be afforded opportunities for “substantial and meaningful involvement” in initiating, developing, and selecting cleanup plans. §9621(f )(1). Or that EPA usually must defer initiating a cleanup at a contaminated site that a State is already remediating. §9605(h). It is not “paternalistic central planning” but instead the “spirit of cooperative federalism [that] run[s] throughout CERCLA and its regulations.” New Mexico v. General Elec. Corp., 467 F.3d 1223, 1244 (CA10 2006). As a last ditch effort, the landowners contend that, even if §107(a) defines potentially responsible parties, they qualify as contiguous property owners under §107(q), which would pull them outside the scope of §107(a). The landowners are correct that contiguous property owners are not potentially responsible parties. Section 107(q)(1)(A) provides that “[a] person that owns real property that is contiguous to or otherwise similarly situated with respect to, and that is or may be contaminated by a release or threatened release of a hazardous substance from, real property that is not owned by that person shall not be considered” an owner of a facility under §107(a). §9607(q)(1)(A). The problem for the landowners is that there are eight further requirements to qualify as a contiguous property owner. §§9607(q)(1)(A)(i)–(viii). Each landowner individually must “establish by a preponderance of the evidence” that he satisfies the criteria. §9607(q)(1)(B). The landowners cannot clear this high bar. One of the eight requirements is that, at the time the person acquired the property, the person “did not know or have reason to know that the property was or could be contaminated by a release or threatened release of one or more hazardous substances.” §9607(q)(1)(A)(viii)(II). All of the landowners here purchased their property after the Anaconda Company built the Washington Monument sized smelter. Indeed “evidence of public knowledge” of contamination was “almost overwhelming.” Christian v. Atlantic Richfield Co., 380 Mont. 495, 529, 358 P.3d 131, 155 (2015). In the early 1900s, the Anaconda Company actually obtained smoke and tailing easements authorizing the disposition of smelter waste onto many properties now owned by the landowners. Id., at 500–501, 358 P. 3d, at 137–138. The landowners had reason to know their property “could be contaminated by a release or threatened release” of a hazardous substance. 42 U. S. C. §9607(q)(1)(A)(viii)(II). At any rate, contiguous landowners must provide “full cooperation, assistance, and access” to EPA and those carrying out Superfund cleanups in order to maintain that status. §9607(q)(1)(A)(iv). But the Government has represented that the landowners’ restoration plan, if implemented, would interfere with its cleanup by, for example, digging up contaminated soil that has been deliberately capped in place. See Brief for United States as Amicus Curiae 20–21. If that is true, the landowners’ plan would soon trigger a lack of cooperation between EPA and the landowners. At that point, the landowners would no longer qualify as contiguous landowners and we would be back to square one. * * * The Montana Supreme Court erred in holding that the landowners were not potentially responsible parties under §122(e)(6) and therefore did not need to seek EPA approval. Montana law requires that “an award of restoration damages actually . . . be used to repair the damaged property.” Sunburst School Dist. No. 2, 338 Mont., at 273, 165 P. 3d, at 1089. But such action cannot be taken in the absence of EPA approval. That approval process, if pursued, could ameliorate any conflict between the landowners’ restoration plan and EPA’s Superfund cleanup, just as Congress envisioned. In the absence of EPA approval of the current restoration plan, we have no occasion to entertain Atlantic Richfield’s claim that the Act otherwise preempts the plan. The judgment of the Montana Supreme Court is affirmed in part and vacated in part. The case is remanded for further proceedings not inconsistent with this opinion. It is so ordered. Notes 1 The Act vests powers and duties in the President, who has delegated the responsibilities relevant here to the EPA Administrator. See 42 U. S. C. §9615; Exec. Order No. 12580, 3 CFR §193 (1988). 2 Atlantic Richfield concedes that the Act preserves the landowners’ claims for other types of compensatory damages under Montana law, including loss of use and enjoyment of property, diminution of value, incidental and consequential damages, and annoyance and discomfort. See Atlantic Richfield Co. v. Montana Second Jud. Dist. Ct., 390 Mont. 76, 79, 408 P.3d 515, 518 (2017). We therefore consider only the landowners’ claim for restoration damages. 3 Justice Alito argues that this jurisdictional question “may turn out not to matter in this case” because we remand for further proceedings that may end the litigation. Post, at 2 (opinion concurring in part and dissenting in part). But Atlantic Richfield seeks more than a remand. It contends that the lawsuit should be dismissed because the Montana courts lack jurisdiction, and the Federal Government agrees. The difference between outright dismissal and further proceedings matters. We granted review of this issue and both parties have fully briefed and argued it. Simply leaving the question unanswered at this point would leave the parties in a state of uncertainty as to whether the litigation is proceeding in the proper forum. We therefore find it both “necessary” and “prudent” to decide the issue. Post, at 1. 4 There is a “special and small category of cases” that originate in state law yet still arise under federal law for purposes of federal question jurisdiction. Gunn v. Minton, 568 U.S. 251, 258 (2013) (internal quotation marks omitted). To qualify for this narrow exception, a state law claim must “necessarily raise[ ]” a federal issue, among other requirements. Ibid. No element of the landowners’ state common law claims necessarily raises a federal issue. Atlantic Richfield raises the Act as an affirmative defense, but “[f]ederal jurisdiction cannot be predicated on an actual or anticipated defense.” Vaden v. Discover Bank, 556 U.S. 49, 60 (2009). 5 Section 113(b) specifies that federal courts have exclusive jurisdiction “without regard to the citizenship of the parties or the amount in controversy.” 42 U. S. C. §9613(b). This is somewhat redundant because all actions that “arise under” the Act necessarily satisfy federal question jurisdiction. But “[s]ometimes the better overall reading of the statute contains some redundancy.” Rimini Street, Inc. v. Oracle USA, Inc., 586 U. S. ___, ___ (2019) (slip op., at 11). We find it much more likely that Congress employed a belt and suspenders approach to make sure that all CERCLA lawsuits are routed to federal court than that Congress intended the reference to federal courts in §113(h) to affect state courts. 6 Justice Alito argues that our interpretation leaves no meaning for the exceptions in §113(h) for federal courts hearing state law actions while sitting in diversity and federal courts hearing actions invoking state law standards deemed “applicable or relevant and appropriate” by the Act. 42 U. S. C. §9613(h). Because we read §113(b) to cover only federal law claims, Justice Alito assumes that these exceptions in §113(h) would never apply. But as we explained, §113(h) applies to all “challenges to removal or remedial action” that make their way into “[f]ederal court,” whether through §113(b) or some other route. §9613(h). That includes state law challenges arising by way of diversity jurisdiction or supplemental jurisdiction as well as federal law challenges arising under sources of law other than the Act. The exceptions in §113(h) are thus necessary to delineate which of these challenges may proceed in federal court and which may not. 7 EPA does have other tools to address serious environmental harm. Under §106, for example, EPA can initiate an injunctive abatement action if it finds an “imminent and substantial endangerment to the public health or welfare or the environment.” 42 U. S. C. §9606(a). But EPA may have good reasons to preserve the status quo of a cleanup site even absent an imminent threat. More importantly, the landowners’ interpretation would require EPA to monitor tens of thousands of properties across 1,335 Superfund sites nationwide to ensure landowners do not derail an EPA cleanup. EPA, Superfund: National Priorities List (NPL) (Apr. 13, 2020), https://www.epa.gov/superfund/superfund-national-priorities-list-npl. Congress provided a far more effective and efficient solution in §122(e)(6): Landowners at Superfund sites containing hazardous waste must seek EPA approval before initiating their own bespoke cleanups.

589.US.2019_18-882

Petitioner Noris Babb, a clinical pharmacist at a U. S. Department of Veterans Affairs Medical Center, sued the Secretary of Veterans Affairs (hereinafter VA) for, inter alia, age discrimination in various adverse personnel actions. The VA moved for summary judgment, offering nondiscriminatory reasons for the challenged actions. The District Court granted the VA’s motion after finding that Babb had established a prima facie case, that the VA had proffered legitimate reasons for the challenged actions, and that no jury could reasonably conclude that those reasons were pretextual. On appeal, Babb contended the District Court’s requirement that age be a but-for cause of a personnel action was inappropriate under the federal-sector provision of the Age Discrimination in Employment Act of 1967 (ADEA). Because most federal-sector “personnel actions” affecting individuals aged 40 and older must be made “free from any discrimination based on age,” 29 U. S. C. §633a(a), Babb argued, such a personnel action is unlawful if age is a factor in the challenged decision. Thus, even if the VA’s proffered reasons in her case were not pretextual, it would not necessarily follow that age discrimination played no part. The Eleventh Circuit found Babb’s argument foreclosed by Circuit precedent. Held: The plain meaning of §633a(a) demands that personnel actions be untainted by any consideration of age. To obtain reinstatement, damages, or other relief related to the end result of an employment decision, a showing that a personnel action would have been different if age had not been taken into account is necessary, but if age discrimination played a lesser part in the decision, other remedies may be appropriate. Pp. 3–14. (a) The Government argues that the ADEA’s federal-sector provision imposes liability only when age is a but-for cause of an employment decision, while Babb maintains that it prohibits any adverse consideration of age in the decision-making process. The plain meaning of the statutory text shows that age need not be a but-for cause of an employment decision in order for there to be a violation. Pp. 4–7. (1) The ADEA does not define the term “personnel action,” but a statutory provision governing federal employment, 5 U. S. C. §2302(a)(2)(A), defines it to include most employment-related decisions—an interpretation consistent with the term’s general usage. The phrase “free from” means “untainted,” and “any” underscores that phrase’s scope. As for “discrimination,” its “normal definition” is “differential treatment.” Jackson v. Birmingham Bd. of Ed., 544 U.S. 167, 174. And “[i]n common talk, the phrase ‘based on’ indicates a but-for causal relationship,” Safeco Ins. Co. of America v. Burr, 551 U.S. 47, 63, thus indicating that age must be a but-for cause of the discrimination alleged. The remaining phrase—“shall be made”—denotes a duty, emphasizing the importance of avoiding the taint. Pp. 4–5. (2) Two matters of syntax are critical here. First, “based on age” is an adjectival phrase modifying the noun “discrimination,” not the phrase “personnel actions.” Thus, age must be a but-for cause of discrimination but not the personnel action itself. Second, “free from any discrimination” is an adverbial phrase that modifies the verb “made” and describes how a personnel action must be “made,” namely, in a way that is not tainted by differential treatment based on age. Thus, the straightforward meaning of §633a(a)’s terms is that the statute does not require proof that an employment decision would have turned out differently if age had not been taken into account. Instead, if age is a factor in an employment decision, the statute has been violated. The Government has no answer to this parsing of the statutory text. It makes correct points about the meaning of particular words, but draws the unwarranted conclusion that the statutory text requires something more than a federal employer’s mere consideration of age in personnel decisions. The Government’s only other textual argument is that the term “made” refers to a particular moment in time, i.e., the moment when the final employment decision is made. That interpretation, however, does not mean that age must be a but-for cause of the ultimate outcome. Pp. 5–7. (b) Contrary to the Government’s primary argument, this interpretation is not undermined by prior cases interpreting the Fair Credit Reporting Act, 15 U. S. C. §1681m(a), see Safeco Ins. Co. of America, 551 U.S. 47; the ADEA’s private-sector provision, 29 U. S. C. §623(a)(1), see Gross v. FBL Financial Services, Inc., 557 U.S. 167; and Title VII’s anti-retaliation provision, 42 U. S. C. §2000e–3(a), see University of Tex. Southwestern Medical Center v. Nassar, 570 U.S. 338. The language of §633a(a) is markedly different than the language of those statutes; thus the holdings in those cases are entirely consistent with the holding here. And the traditional rule favoring but-for causation does not change the result: §633a(a) requires proof of but-for causation, but the objection of that causation is “discrimination,” not the personnel action. Pp. 8–11. (c) It is not anomalous to hold the Federal Government to a stricter standard than private employers or state and local governments. See §623(a). When Congress expanded the ADEA’s scope beyond private employers, it added state and local governments to the definition of employers in the private-sector provision. But it “deliberately prescribed a distinct statutory scheme applicable only to the federal sector,” Lehman v. Nakshian, 453 U.S. 156, 166, eschewing the private-sector provision language. That Congress would want to hold the Federal Government to a higher standard is not unusual. See, e.g., 5 U. S. C. §2301(b)(2). Regardless, where the statute’s words are unambiguous, the judicial inquiry is complete. Pp. 11–13. (d) But-for causation is nevertheless important in determining the appropriate remedy. Plaintiffs cannot obtain compensatory damages or other forms of relief related to the end result of an employment decision without showing that age discrimination was a but-for cause of the employment outcome. This conclusion is supported by basic principles long employed by this Court, see, e.g., Steel Co. v. Citizens for Better Environment, 523 U.S. 83, 103, and traditional principles of tort and remedies law. Remedies must be tailored to the injury. Plaintiffs who show that age was a but-for cause of differential treatment in an employment decision, but not a but-for cause of the decision itself, can still seek injunctive or other forward-looking relief. Pp. 13–14. 743 Fed. Appx. 280, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Breyer, Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., joined, and in which Ginsburg, J., joined as to all but footnote 3. Sotomayor, J., filed a concurring opinion, in which Ginsburg, J., joined. Thomas, J., filed a dissenting opinion.

o arrive at” a particular conclusion, i.e., to “make a decision”). And holding that §633a(a) is violated when the consideration of age plays no role in the final decision would have startling implications. Consider this example: A decision-maker must decide whether to promote employee A, who is under 40, or employee B, who is over 40. A subordinate recommends employee A and says that the recommendation is based in part on employee B’s age. The decision-maker rebukes this subordinate for taking age into account, disregards the recommendation, and makes the decision independently. Under an interpretation that read “made” expansively to encompass a broader personnel process, §633a(a) would be violated even though age played no role whatsoever in the ultimate decision. Indeed, there might be a violation even if the decision-maker decided to promote employee B. We are aware of no other anti-discrimination statute that imposes liability under such circumstances, and we do not think that §633a(a) should be understood as the first. 5 Moreover, even if “discriminating with respect to compensation, terms, conditions, or privileges of employment” could be read more broadly to encompass things that occur before a final decision is made, the ejusdem generis canon would counsel a court to read that final phrase to refer––like the prior terms––to the final decision. See Christopher v. SmithKline Beecham Corp., 567 U.S. 142, 163, and n. 19 (2012).

590.US.2019_18-6943

Federal Rule of Civil Procedure 59(e) allows a litigant to file a motion to alter or amend a district court’s judgment within 28 days from the entry of judgment, with no possibility of an extension. The Rule enables a district court to “rectify its own mistakes in the period immediately following” its decision, White v. New Hampshire Dept. of Employment Security, 455 U.S. 445, 450, but not to address new arguments or evidence that the moving party could have raised before the decision. A timely filed motion suspends the finality of the original judgment for purposes of appeal, and only the district court’s disposition of the motion restores finality and starts the 30-day appeal clock. If an appeal follows, the ruling on the motion merges with the original determination into a single judgment. Title 28 U. S. C. §2244(b), the so-called gatekeeping provision of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), governs federal habeas proceedings. Under AEDPA, a state prisoner is entitled to one fair opportunity to seek federal habeas relief from his conviction. Section 2244(b), however, sets stringent limits on second or successive habeas applications. Among those restrictions, a prisoner may not reassert any claims “presented in a prior application,” §2244(b)(1), and may bring a new claim only in limited situations. Because habeas proceedings are civil in nature, the Federal Rules of Civil Procedure generally apply, but statutory habeas restrictions, including §2244(b), trump any “inconsistent” Rule. §2254 Rule 12. Petitioner Gregory Banister was convicted by a Texas court of aggravated assault and sentenced to 30 years in prison. After exhausting his state remedies, he filed for federal habeas relief, which the District Court denied. Banister timely filed a Rule 59(e) motion, which the District Court also denied. He then filed a notice of appeal in accordance with the timeline for appealing a judgment after the denial of a Rule 59(e) motion. But the Fifth Circuit construed Banister’s Rule 59(e) motion as a successive habeas petition and dismissed his appeal as untimely. Held: Because a Rule 59(e) motion to alter or amend a habeas court’s judgment is not a second or successive habeas petition under 28 U. S. C. §2244(b), Banister’s appeal was timely. Pp. 5–16. (a) The phrase “second or successive application” is a term of art and does not “simply ‘refe[r]’ ” to all habeas filings made “ ‘second or successively in time,’ ” following an initial application. Magwood v. Patterson, 561 U.S. 320, 332. In addressing what qualifies as second or successive, this Court has looked to historical habeas doctrine and practice and AEDPA’s purposes. Here, both point toward permitting Rule 59(e) motions in habeas proceedings. Prior to AEDPA, the Court held in Browder v. Director, Dept. of Corrections of Ill., 434 U.S. 257, that Rule 59(e) applied in habeas proceedings. The Rule, the Court recounted, derived from courts’ common-law power “to alter or amend [their] own judgments during[ ] the term of court in which [they were] rendered,” prior to any appeal, including “in habeas corpus cases.” Id., at 270. Although the drafters of the Federal Rules eventually replaced the “term of court” power with Rule 59(e), the Court concluded that this did nothing to narrow the set of judgments amenable to alteration. The record of judicial decisions accords with that view. Pre-AEDPA, habeas courts were to dismiss repetitive applications except in “rare case[s].” Kuhlmann v. Wilson, 477 U.S. 436, 451. Yet in the half century from Rule 59(e)’s adoption through Browder to AEDPA’s enactment, there exists only one dismissal of a Rule 59(e) motion as impermissibly successive. In all other cases, the district courts resolved Rule 59(e) motions on the merits. Congress passed AEDPA against this backdrop, and gave no indication that it meant to change what qualifies as a successive application. Nor do AEDPA’s purposes of reducing delay, conserving judicial resources, and promoting finality suggest any different result. Rule 59(e) offers a narrow, 28-day window to ask for relief; limits requests for reconsideration to matters properly raised in the challenged judgment; and consolidates proceedings by producing a single final judgment for appeal. Indeed, the Rule may make habeas proceedings more efficient by enabling a district court to reverse a mistaken judgment or to clarify its reasoning so as to make an appeal unnecessary. Pp. 5–12. (b) Gonzalez v. Crosby, 545 U.S. 524, which held that a Rule 60(b) motion counts as a second or successive habeas application if it “attacks the federal court’s previous resolution of a claim on the merits,” id., at 532, does not alter that conclusion. Rule 60(b) differs from Rule 59(e) in just about every way that matters here. Whereas Rule 59(e) derives from a common-law court’s plenary power to revise its judgment before anyone could appeal, Rule 60(b) codifies various writs used to collaterally attack a court’s already completed judgment. That distinction was not lost on pre-AEDPA habeas courts, which routinely dismissed Rule 60(b) motions for raising repetitive claims. Next, the Rules’ modern-day operations also diverge, with only Rule 60(b) undermining AEDPA’s scheme to prevent delay and protect finality. That is because a Rule 60(b) motion, which can arise long after the denial of a prisoner’s initial petition, generally goes beyond pointing out alleged errors in the just-issued decision. Still more, a Rule 60(b) motion “does not affect the [original] judgment’s finality or suspend its operation” and is appealable as “a separate final order.” Stone v. INS, 514 U.S. 386, 401. Left unchecked, a Rule 60(b) motion threatens serial habeas litigation, while a Rule 59(e) motion is a one-time effort to point out alleged errors in a just-issued decision before taking a single appeal. Pp. 12–16. Reversed and remanded. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Sotomayor, Gorsuch, and Kavanaugh, JJ., joined. Alito, J., filed a dissenting opinion, in which Thomas, J., joined.

A state prisoner is entitled to one fair opportunity to seek federal habeas relief from his conviction. But he may not usually make a “second or successive habeas corpus application.” 28 U. S. C. §2244(b). The question here is whether a motion brought under Federal Rule of Civil Procedure 59(e) to alter or amend a habeas court’s judgment qualifies as such a successive petition. We hold it does not. A Rule 59(e) motion is instead part and parcel of the first habeas proceeding. I This case is about two procedural rules. First, Rule 59(e) applies in federal civil litigation generally. (Habeas proceedings, for those new to the area, are civil in nature. See Fisher v. Baker, 203 U.S. 174, 181 (1906).) The Rule enables a party to request that a district court reconsider a just-issued judgment. Second, the so-called gatekeeping provision of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), codified at 28 U. S. C. §2244(b), governs federal habeas proceedings. It sets stringent limits on second or successive habeas applications. We say a few words about each before describing how the courts below applied them here. A Rule 59(e) allows a litigant to file a “motion to alter or amend a judgment.”[1] The time for doing so is short—28 days from entry of the judgment, with no possibility of an extension. See Fed. Rule Civ. Proc. 6(b)(2) (prohibiting extensions to Rule 59(e)’s deadline). The Rule gives a district court the chance “to rectify its own mistakes in the period immediately following” its decision. White v. New Hampshire Dept. of Employment Security, 455 U.S. 445, 450 (1982). In keeping with that corrective function, “federal courts generally have [used] Rule 59(e) only” to “reconsider[ ] matters properly encompassed in a decision on the merits.” Id., at 451. In particular, courts will not address new arguments or evidence that the moving party could have raised before the decision issued. See 11 C. Wright, A. Miller, & M. Kane, Federal Practice and Procedure §2810.1, pp. 163–164 (3d ed. 2012) (Wright & Miller); accord, Exxon Shipping Co. v. Baker, 554 U.S. 471, 485–486, n. 5 (2008) (quoting prior edition).[2] The motion is therefore tightly tied to the underlying judgment. The filing of a Rule 59(e) motion within the 28-day period “suspends the finality of the original judgment” for purposes of an appeal. FCC v. League of Women Voters of Cal., 468 U.S. 364, 373, n. 10 (1984) (internal quotation marks and alterations omitted). Without such a motion, a litigant must take an appeal no later than 30 days from the district court’s entry of judgment. See Fed. Rule App. Proc. (FRAP) 4(a)(1)(A). But if he timely submits a Rule 59(e) motion, there is no longer a final judgment to appeal from. See Osterneck v. Ernst & Whinney, 489 U.S. 169, 174 (1989). Only the disposition of that motion “restores th[e] finality” of the original judgment, thus starting the 30-day appeal clock. League of Women Voters, 468 U. S., at 373, n. 10 (internal quotation marks omitted); see FRAP 4(a)(4)(A)(iv) (A party’s “time to file an appeal runs” from “the entry of the order disposing of the [Rule 59(e)] motion”). And if an appeal follows, the ruling on the Rule 59(e) motion merges with the prior determination, so that the reviewing court takes up only one judgment. See 11 Wright & Miller §2818, at 246; Foman v. Davis, 371 U.S. 178, 181 (1962). The court thus addresses any attack on the Rule 59(e) ruling as part of its review of the underlying decision. Now turn to §2244(b)’s restrictions on second or successive habeas petitions. Under AEDPA, a state prisoner always gets one chance to bring a federal habeas challenge to his conviction. See Magwood v. Patterson, 561 U.S. 320, 333–334 (2010). But after that, the road gets rockier. To file a second or successive application in a district court, a prisoner must first obtain leave from the court of appeals based on a “prima facie showing” that his petition satisfies the statute’s gatekeeping requirements. 28 U. S. C. §2244(b)(3)(C). Under those provisions, which bind the district court even when leave is given, a prisoner may not reassert any claims “presented in a prior application.” §2244(b)(1). And he may bring a new claim only if it falls within one of two narrow categories—roughly speaking, if it relies on a new and retroactive rule of constitutional law or if it alleges previously undiscoverable facts that would establish his innocence. See §2244(b)(2). Still more: Those restrictions, like all statutes and rules pertaining to habeas, trump any “inconsistent” Federal Rule of Civil Procedure otherwise applicable to habeas proceedings. 28 U. S. C. §2254 Rule 12. B This case began when, nearly two decades ago, petitioner Gregory Banister struck and killed a bicyclist while driving a car. Texas charged him with the crime of aggravated assault with a deadly weapon. A jury found him guilty, and he was sentenced to 30 years in prison. State courts upheld the conviction on direct appeal and in collateral proceedings. Banister then turned to federal district court for habeas relief. Although raising many claims, his petition mainly argued that his trial and appellate counsel provided him with constitutionally ineffective assistance. The District Court disagreed and entered judgment denying the application. At that point, Banister timely filed a Rule 59(e) motion asking the District Court to alter its judgment. Consistent with the Rule’s corrective purpose, Banister urged the court to fix what he saw as “manifest errors of law and fact.” App. 219. Five days later and without requiring a response from the State, the court issued a one-paragraph order explaining that it had reviewed all relevant materials and stood by its decision. See id., at 254. In accordance with the timeline for appealing a judgment after the denial of a Rule 59(e) motion, see supra, at 3, Banister then filed a notice of appeal (along with a request for a certificate of appealability) to challenge the District Court’s rejection of his habeas application. Yet the Court of Appeals for the Fifth Circuit dismissed the appeal as untimely. That ruling rested on the view that Banister’s Rule 59(e) motion, although captioned as such, was not really a Rule 59(e) motion at all. Because it “attack[ed] the federal court’s previous resolution of [his] claim on the merits,” the Fifth Circuit held that the motion must be “construed as a successive habeas petition.” App. 305 (internal quotation marks omitted). In any future case, that holding would prohibit a habeas court from considering claims made in a self-styled Rule 59(e) motion except in rare circumstances—that is, when a court of appeals gave permission and the claim fell within one of §2244(b)’s two slender categories. See supra, at 3. In Banister’s own case, that bar was of no moment because the District Court had already addressed his motion’s merits. But viewing a Rule 59(e) motion as a successive habeas petition also had another consequence, and this one would affect him. Unlike a Rule 59(e) motion, the Court of Appeals noted, a successive habeas application does not postpone the time to file an appeal. That meant the clock started ticking when the District Court denied Banister’s habeas application (rather than his subsequent motion)—and so Banister’s appeal was several weeks late. We granted certiorari to resolve a Circuit split about whether a Rule 59(e) motion to alter or amend a habeas court’s judgment counts as a second or successive habeas application. 588 U. S. ___ (2019). We hold it does not, and reverse. II This case requires us to choose between two rules—more specifically, to decide whether AEDPA’s §2244(b) displaces Rule 59(e) in federal habeas litigation. The Federal Rules of Civil Procedure generally govern habeas proceedings. See Fed. Rule Civ. Proc. 81(a)(4). They give way, however, if and to the extent “inconsistent with any statutory provisions or [habeas-specific] rules.” 28 U. S. C. §2254 Rule 12; see supra, at 3–4. Here, the Fifth Circuit concluded and Texas now contends that AEDPA’s limitation of repetitive habeas applications conflicts with Rule 59(e)’s ordinary operation. That argument in turn hinges on viewing a Rule 59(e) motion in a habeas case as a “second or successive application.” §2244(b); see Brief for Respondent 10. If such a motion constitutes a second or successive petition, then all of §2244(b)’s restrictions kick in—limiting the filings Rule 59(e) would allow. But if a Rule 59(e) motion is not so understood—if it is instead part of resolving a prisoner’s first habeas application—then §2244(b)’s requirements never come into the picture. The phrase “second or successive application,” on which all this rides, is a “term of art,” which “is not self-defining.” Slack v. McDaniel, 529 U.S. 473, 486 (2000); Panetti v. Quarterman, 551 U.S. 930, 943 (2007). We have often made clear that it does not “simply ‘refer’ ” to all habeas filings made “ ‘second or successively in time,’ ” following an initial application. Magwood, 561 U. S., at 332 (quoting Panetti, 551 U. S., at 944 (alteration omitted)). For example, the courts of appeals agree (as do both parties) that an amended petition, filed after the initial one but before judgment, is not second or successive. See 2 R. Hertz & J. Liebman, Federal Habeas Corpus Practice and Procedure §28.1, pp. 1656–1657, n. 4 (7th ed. 2017) (collecting cases); Brief for Petitioner 20–21; Brief for Respondent 16. So too, appeals from the habeas court’s judgment (or still later petitions to this Court) are not second or successive; rather, they are further iterations of the first habeas application.[3] Chronology here is by no means all. In addressing what qualifies as second or successive, this Court has looked for guidance in two main places. First, we have explored historical habeas doctrine and practice. The phrase “second or successive application,” we have explained, is “given substance in our prior habeas corpus cases,” including those “predating [AEDPA’s] enactment.” Slack, 529 U. S., at 486; Panetti, 551 U. S., at 944; see id., at 943 (stating that the phrase “takes its full meaning from our case law”). In particular, we have asked whether a type of later-in-time filing would have “constituted an abuse of the writ, as that concept is explained in our [pre-AEDPA] cases.” Id., at 947. If so, it is successive; if not, likely not. Second, we have considered AEDPA’s own purposes. The point of §2244(b)’s restrictions, we have stated, is to “conserve judicial resources, reduc[e] piecemeal litigation,” and “lend[ ] finality to state court judgments within a reasonable time.” Id., at 945–946 (internal quotation marks omitted). With those goals in mind, we have considered “the implications for habeas practice” of allowing a type of filing, to assess whether Congress would have viewed it as successive. Stewart v. Martinez-Villareal, 523 U.S. 637, 644 (1998). Here, both historical precedents and statutory aims point in the same direction—toward permitting Rule 59(e) motions in habeas proceedings. And nothing cuts the opposite way. A This Court has already held that history supports a habeas court’s consideration of a Rule 59(e) motion. In Browder v. Director, Dept. of Corrections of Ill., 434 U.S. 257 (1978), we addressed prior to AEDPA “the applicability of Federal Rule [59(e)] in habeas corpus proceedings.” Id., at 258. In deciding that the Rule applied in habeas—that “a prompt motion for reconsideration” was “thoroughly consistent” with habeas law and “well suited to the special problems and character of [habeas] proceedings”—we mainly looked to historical practice. Id., at 271 (internal quotation marks omitted). Rule 59(e), we recounted, derived from a court’s common-law power “to alter or amend its own judgments during[] the term of court in which [they were] rendered,” prior to any appeal. Id., at 270; see Zimmern v. United States, 298 U.S. 167, 169–170 (1936) (“The judge had plenary power while the term was in existence to modify his judgment [or] revoke it altogether”).[4] Courts exercised that authority, we explained, “in habeas corpus cases” just as “in other civil proceedings.” Browder, 434 U. S., at 270. In 1946, the drafters of the Federal Rules replaced the “term of court” power with Rule 59(e), thus prescribing a set number of days (then 10, now 28) in which a party could move to amend a judgment. See id., at 271. But in our view, that change did nothing to narrow the set of judgments amenable to alteration. See id., at 270–271. After Rule 59(e), just as before, a district court could “reconsider the grant or denial of habeas corpus relief ” in the same way it could review any other decision. Id., at 270; see id., at 271. A timely Rule 59(e) motion, we held, “suspend[ed] the finality” of any judgment, including one in habeas—thus enabling a district court to address the matter again. Id., at 267 (internal quotation marks omitted).[5] The record of judicial decisions accords with Browder’s view of the use of Rule 59(e) in habeas practice. Before AEDPA, “abuse-of-the-writ principles limit[ed] a [habeas applicant’s] ability to file repetitive petitions.” McCleskey v. Zant, 499 U.S. 467, 483 (1991). That doctrine was more forgiving than AEDPA’s gatekeeping provision—for example, enabling courts to hear a second or successive petition if the “ends of justice” warranted doing so. Id., at 485. But the rule against repetitive litigation still had plenty of bite. It demanded the dismissal of successive applications except in “rare case[s].” Kuhlmann v. Wilson, 477 U.S. 436, 451 (1986) (plurality opinion). So if courts had viewed Rule 59(e) motions as successive, there should be lots of decisions dismissing them on that basis. But nothing of the kind exists. In the half century from Rule 59(e)’s adoption (1946) through Browder (1978) to AEDPA’s enactment (1996), we (and the parties) have found only one such dismissal. See Bannister v. Armontrout, 4 F.3d 1434, 1445 (CA8 1993). In every other case, courts resolved Rule 59(e) motions on the merits—and without any comment about repetitive litigation. Mostly, courts denied the motions and adhered to their original judgments. See, e.g., Gajewski v. Stevens, 346 F.2d 1000, 1001 (CA8 1965) (per curiam). Occasionally, courts decided they had erred in those decisions. See, e.g., York v. Tate, 858 F.2d 322, 325 (CA6 1988) (per curiam). The win-loss rate is for this point irrelevant. What matters is that they all (but one) treated Rule 59(e) motions not as successive, but as attendant on the initial habeas application. Congress passed AEDPA against this legal backdrop, and did nothing to change it. AEDPA of course made the limits on entertaining second or successive habeas applications more stringent than before. See supra, at 3. But the statute did not redefine what qualifies as a successive petition, much less place Rule 59(e) motions in that category. Cf. Magwood, 561 U. S., at 336–337 (distinguishing between two questions: “§2244(b)’s threshold inquiry into whether an application is ‘second or successive’ and its subsequent inquiry into whether [to dismiss] a successive application”). When Congress “intends to effect a change” in existing law—in particular, a holding of this Court—it usually provides a clear statement of that objective. TC Heartland LLC v. Kraft Foods Group Brands LLC, 581 U. S. ___, ___ (2017) (slip op., at 8). AEDPA offers no such indication that Congress meant to change the historical practice Browder endorsed of applying Rule 59(e) in habeas proceedings. Nor do AEDPA’s purposes demand a change in that tradition. As explained earlier, AEDPA aimed to prevent serial challenges to a judgment of conviction, in the interest of reducing delay, conserving judicial resources, and promoting finality. See supra, at 7. Nothing in Rule 59(e)—a rule Browder described as itself “based on an interest in speedy disposition and finality,” 434 U. S., at 271 (internal quotation marks omitted)—conflicts with those goals. Recall everything said above about the Rule’s operation. See supra, at 2–3. To begin with, Rule 59(e) gives a prisoner only a narrow window to ask for relief—28 days, with no extensions. Next, a prisoner may invoke the rule only to request “reconsideration of matters properly encompassed” in the challenged judgment. White, 455 U. S., at 451. And “reconsideration” means just that: Courts will not entertain arguments that could have been but were not raised before the just-issued decision. A Rule 59(e) motion is therefore backward-looking; and because that is so, it maintains a prisoner’s incentives to consolidate all of his claims in his initial application. Yet more, the Rule consolidates appellate proceedings. A Rule 59(e) motion briefly suspends finality to enable a district court to fix any mistakes and thereby perfect its judgment before a possible appeal. The motion’s disposition then merges into the final judgment that the prisoner may take to the next level. In that way, the Rule avoids “piecemeal appellate review.” Osterneck, 489 U. S., at 177. Its operation, rather than allowing repeated attacks on a decision, helps produce a single final judgment for appeal. Indeed, the availability of Rule 59(e) may make habeas proceedings more efficient. Most obviously, the Rule enables a district court to reverse a mistaken judgment, and so make an appeal altogether unnecessary. See United States v. Ibarra, 502 U.S. 1, 5 (1991) (per curiam) (noting that giving district courts a short time to correct their own errors “prevents unnecessary burdens being placed on the courts of appeals”). Of course, Rule 59(e) motions seldom change judicial outcomes. But even when they do not, they give habeas courts the chance to clarify their reasoning or address arguments (often made in less-than-limpid pro se petitions) passed over or misunderstood before. See Brief for National Association of Criminal Defense Lawyers as Amicus Curiae 12–20 (describing examples). That opportunity, too, promotes an economic and effective appellate process, as the reviewing court gets “the benefit of the district court’s plenary findings.” Osterneck, 489 U. S., at 177. And when a district court sees no need to change a decision, the costs of permitting a Rule 59(e) motion are typically slight. A judge familiar with a habeas applicant’s claims can usually make quick work of a meritless motion. This case may well provide an example: The District Court declined to make the State respond to Banister’s motion and decided it within five days. Nothing in such a process conflicts with AEDPA’s goal of streamlining habeas cases. The upshot, after AEDPA as before, is that Rule 59(e) motions are not second or successive petitions, but instead a part of a prisoner’s first habeas proceeding. In timing and substance, a Rule 59(e) motion hews closely to the initial application; and the habeas court’s disposition of the former fuses with its decision on the latter. Such a motion does not enable a prisoner to abuse the habeas process by stringing out his claims over the years. It instead gives the court a brief chance to fix mistakes before its (single) judgment on a (single) habeas application becomes final and thereby triggers the time for appeal. No surprise, then, that habeas courts historically entertained Rule 59(e) motions, rather than dismiss them as successive. Or that Congress said not a word about changing that familiar practice even when enacting other habeas restrictions. B Texas (along with the dissent) resists this conclusion on one main ground: this Court’s prior decision in Gonzalez v. Crosby, 545 U.S. 524 (2005). The question there was whether a Rule 60(b) motion for “relie[f] from a final judgment” denying habeas relief counts as a second or successive habeas application. Fed. Rule Civ. Proc. 60(b).[6] We said that it does, so long as the motion “attacks the federal court’s previous resolution of a claim on the merits.” 545 U. S., at 532 (emphasis deleted).[7] Texas thinks the “Gonzalez principle applies with equal force to Rule 59(e) motions.” Brief for Respondent 8. After all, the State argues, both Rule 59(e) and Rule 60(b) provide “vehicles for asserting habeas claims” after a district court has entered judgment denying relief. Id., at 2. And if Gonzalez does apply, Texas concludes, Banister must lose because (as everyone agrees) his Rule 59(e) motion pressed only merits-based claims. But Rule 60(b) differs from Rule 59(e) in just about every way that matters to the inquiry here. (Contra the dissent’s refrain, see post, at 1, 3, 5, 6, 10, 14, the variance goes far beyond their “labels.”) Begin, again, with history. Recall that Rule 59(e) derives from a common-law court’s plenary power to revise its judgment during a single term of court, before anyone could appeal. See supra, at 7–8. By contrast, Rule 60(b) codifies various writs used to seek relief from a judgment at any time after the term’s expiration—even after an appeal had (long since) concluded. Those mechanisms did not (as the term rule did) aid the trial court to get its decision right in the first instance; rather, they served to collaterally attack its already completed judgment. See Advisory Committee’s 1946 Notes on Amendments to Fed. Rule Civ. Proc. 60; Mann, Note, History and Interpretation of Federal Rule 60(b), 25 Temp. L. Q. 77, 78 (1951). And that distinction was not lost on pre-AEDPA habeas courts applying the two rules. As discussed earlier, it is practically impossible to find a case dismissing a Rule 59(e) motion for raising repetitive claims. See supra, at 9. But decisions abound dismissing Rule 60(b) motions for that reason. See, e.g., Williamson v. Rison, 1993 WL 262632 (CA9, July 9, 1993); see also Brewer v. Ward, 1996 WL 194830, *1 (CA10, Apr. 22, 1996) (collecting cases from multiple Circuits). That is because those courts recognized Rule 60(b)—as contrasted to Rule 59(e)—as threatening an already final judgment with successive litigation.[8] The modern-day operation of the two Rules also diverge, with only Rule 60(b) undermining AEDPA’s scheme to prevent delay and protect finality. Unlike Rule 59(e) motions with their fixed 28-day window, Rule 60(b) motions can arise long after the denial of a prisoner’s initial petition—depending on the reason given for relief, within either a year or a more open-ended “reasonable time.” Fed. Rule Civ. Proc. 60(c)(1). In Gonzalez itself, the prisoner made his motion nearly three years after the habeas court’s denial of relief, and more than one year after his appeal ended. See 545 U. S., at 527. Given that extended timespan, Rule 60(b) inevitably elicits motions that go beyond Rule 59(e)’s mission of pointing out the alleged errors in the habeas court’s decision. See, e.g., Lopez v. Douglas, 141 F.3d 974, 975 (CA10 1998) (per curiam) (seeking relief in light of a Supreme Court decision issued a decade after judgment); Tyler v. Anderson, 749 F.3d 499, 504–505 (CA6 2014) (seeking to raise claims that former counsel had neglected in a years-old habeas application). Still more, the appeal of a Rule 60(b) denial is independent of the appeal of the original petition. Recall that a Rule 59(e) motion suspends the finality of the habeas judgment, and a decision on the former merges into the latter for appellate review. See supra, at 2–3, 10-11. By contrast, a Rule 60(b) motion “does not affect the [original] judgment’s finality or suspend its operation.” Fed. Rule Civ. Proc. 60(c)(2). And an appeal from the denial of Rule 60(b) relief “does not bring up the underlying judgment for review.” Browder, 434 U. S., at 263, n. 7. Instead, that denial is appealed as “a separate final order.” Stone v. INS, 514 U.S. 386, 401 (1995).[9] In short, a Rule 60(b) motion differs from a Rule 59(e) motion in its remove from the initial habeas proceeding. A Rule 60(b) motion—often distant in time and scope and always giving rise to a separate appeal—attacks an already completed judgment. Its availability threatens serial habeas litigation; indeed, without rules suppressing abuse, a prisoner could bring such a motion endlessly. By contrast, a Rule 59(e) motion is a one-time effort to bring alleged errors in a just-issued decision to a habeas court’s attention, before taking a single appeal. It is a limited continuation of the original proceeding—indeed, a part of producing the final judgment granting or denying habeas relief. For those reasons, Gonzalez does not govern here. A Rule 59(e) motion, unlike a Rule 60(b) motion, does not count as a second or successive habeas application. III Our holding means that the Court of Appeals should not have dismissed Banister’s appeal as untimely. Banister properly brought a Rule 59(e) motion in the District Court. As noted earlier, the 30-day appeals clock runs from the disposition of such a motion, rather than from the initial entry of judgment. See supra, at 3. And Banister filed his notice of appeal within that time. The Fifth Circuit reached a contrary conclusion because it thought that Banister’s motion was really a second or successive habeas application, and so did not reset the appeals clock. For all the reasons we have given, that understanding of a Rule 59(e) motion is wrong. We therefore reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 The complete text of the Rule reads: “A motion to alter or amend a judgment must be filed no later than 28 days after the entry of the judgment.” 2 By contrast, courts may consider new arguments based on an “intervening change in controlling law” and “newly discovered or previously unavailable evidence.” 11 Wright & Miller §2810.1, at 161–162 (3d ed. 2012). But it is rare for such arguments or evidence to emerge within Rule 59(e)’s strict 28-day timeframe. 3 For additional examples, see Slack v. McDaniel, 529 U.S. 473, 487 (2000) (allowing a prisoner to file a second-in-time, post-judgment application to assert claims earlier dismissed for failure to exhaust) and Stewart v. Martinez-Villareal, 523 U.S. 637, 643–644 (1998) (permitting a prisoner to file a second-in-time, post-judgment application to argue that he was incompetent to be executed). 4 A term of court in those days was simply a period in which a court was open for business. A statute or rule set the date of its commencement, and the court itself determined the date to adjourn. See United States v. Pitman, 147 U.S. 669, 670–671 (1893). 5 The dissent’s attempt to dismiss Browder is impossible to square with the opinion. Mostly, the dissent claims that Browder is just a case about “time limits.” Post, at 10 (opinion of Alito, J.). But Browder is about time limits only in the sense that this case is about time limits: There, as here, the timeliness of a motion depended on the broader question whether Rule 59(e) applied in habeas proceedings. See 434 U. S., at 258 (“In order to resolve th[e] question” whether the “appeal was untimely,” “we must consider the applicability of Federal Rule[ ] 59 in habeas corpus proceedings”). The dissent also intimates that Browder was different because there the prison warden rather than the prisoner moved for reconsideration of the habeas ruling. See post, at 10, and n. 2. But the Court’s decision explicitly addressed “motion[s ] to reconsider the grant or denial of habeas corpus relief.” 434 U. S., at 270 (emphasis added). In other words, the identity of the movant—whether warden or prisoner—was irrelevant. 6 Under Rule 60(b), a court may relieve a party in civil litigation from a final judgment if the party can show (1) mistake, inadvertence, surprise, or excusable neglect; (2) certain newly discovered evidence; (3) fraud, misrepresentation, or misconduct by an opposing party; (4) voidness of the judgment; (5) certain events that would cast doubt on the validity or equity of continuing to apply the judgment; or (6) “any other reason that justifies relief.” Fed. Rule Civ. Proc. 60(b)(1)–(6). 7 By contrast, Gonzalez held, a Rule 60(b) motion that attacks “some defect in the integrity of the federal habeas proceedings”—like the mistaken application of a statute of limitations—does not count as a habeas petition at all, and so can proceed. 545 U. S., at 532. Texas concedes that if Gonzalez controls Rule 59(e) motions, that decision’s distinction between merits-based motions and integrity-based motions would have to apply. See Brief for Respondent 37. The need for a habeas court to make that not-always-easy threshold determination further undermines the notion—already on shaky ground, see supra, at 10–11—that Texas’s position would lead to any efficiency gains. 8 The dissent’s alternative explanation for this disparity does not pass muster. According to the dissent, habeas courts “might have been more inclined” to rule on the merits of Rule 59(e) motions because doing so was easier: after all, they (but not Rule 60(b) motions) always challenge a just-issued decision. Post, at 12. But another course would have been easier still: throwing out the motion for raising repetitive claims. And even more to the point, that course would usually have been required if the dissent were right that Rule 59(e) motions counted as successive. Although pre-AEDPA courts had some discretion around the edges, the consideration of successive petitions was supposed to be “rare.” Kuhlmann v. Wilson, 477 U.S. 436, 451 (1986) (plurality opinion); see supra, at 9. It is a “tall order,” post, at 12, then, to think that a half century’s worth of habeas courts would have resolved Rule 59(e) motions on the merits if they thought of those motions as successive. The only plausible account of their actions is that they did not. 9 Texas objects that if a Rule 60(b) motion is filed within 28 days, it too suspends the finality of the underlying judgment so that the denial of the motion merges with that judgment on appeal. See Brief for Respondent 25, 28. But that is only because courts of appeals have long treated Rule 60(b) motions filed within 28 days as . . . Rule 59(e) motions. See, e.g., Skagerberg v. Oklahoma, 797 F.2d 881, 882–883 (CA10 1986) (per curiam) (“A post-judgment motion made within [28] days of the entry of judgment that questions the correctness of a judgment,” however denominated, “is properly construed as a motion to alter or amend judgment under [Rule] 59(e)”); see also Fed. Rule App. Proc. 4(a)(4)(A)(vi) (codifying that approach by setting the same appeals clock for self-styled Rule 60(b) motions filed within 28 days as for Rule 59(e) motions).

590.US.2019_18-725

When a lawful permanent resident commits certain serious crimes, the Government may initiate removal proceedings before an immigration judge. 8 U. S. C. §1229a. If the lawful permanent resident is found removable, the immigration judge may cancel removal, but only if the lawful permanent resident meets strict statutory eligibility requirements. §§1229b(a), 1229b(d)(1)(B). Over the span of 12 years, lawful permanent resident Andre Barton was convicted of state crimes, including a firearms offense, drug offenses, and aggravated assault offenses. An Immigration Judge found him removable based on his state firearms and drug offenses. Barton applied for cancellation of removal. Among the eligibility requirements, a lawful permanent resident must have “resided in the United States continuously for 7 years after having been admitted in any status.” §1229b(a)(2). Another provision, the so-called stop-time rule, provides that a continuous period of residence “shall be deemed to end” when the lawful permanent resident commits “an offense referred to in section 1182(a)(2) . . . that renders the alien inadmissible to the United States under section 1182(a)(2).” §1229b(d)(1)(B). Because Barton’s aggravated assault offenses were committed within his first seven years of admission and were covered by §1182(a)(2), the Immigration Judge concluded that Barton was not eligible for cancellation of removal. The Board of Immigration Appeals and the Eleventh Circuit agreed. Held: For purposes of cancellation-of-removal eligibility, a §1182(a)(2) offense committed during the initial seven years of residence does not need to be one of the offenses of removal. Pp. 6–17. (a) The cancellation-of-removal statute functions like a traditional recidivist sentencing statute, making a noncitizen’s prior crimes relevant to eligibility for cancellation of removal. The statute’s text clarifies two points relevant here. First, cancellation of removal is precluded when, during the initial seven years of residence, the noncitizen “committed an offense referred to in section 1182(a)(2),” even if (as in Barton’s case) the conviction occurred after the seven years elapsed. Second, the offense must “rende[r] the alien inadmissible” as a result. For crimes involving moral turpitude, the relevant category here, §1182(a)(2) provides that a noncitizen is rendered “inadmissible” when he is convicted of or admits the offense. §1182(a)(2)(A)(i). As a matter of statutory text and structure, the analysis here is straightforward. Barton’s aggravated assault offenses were crimes involving moral turpitude and therefore “referred to in section 1182(a)(2).” He committed the offenses during his initial seven years of residence and was later convicted of the offenses, thereby rendering him “inadmissible.” Barton was, therefore, ineligible for cancellation of removal. Pp. 6–10. (b) Barton’s counterarguments are unpersuasive. First, he claims that the statute’s structure supports an “offense of removal” approach. But §1227(a)(2) offenses—not §1182(a)(2) offenses—are ordinarily the basis for removal of lawful permanent residents. Therefore, Barton’s structural argument falls apart. If he were correct, the statute presumably would specify offenses “referred to in section 1182(a)(2) or section 1227(a)(2).” By contrast, some other immigration law provisions do focus only on the offense of removal, and their statutory text and context support that limitation. See, e.g., §§1226(a), (c)(1)(A), 1252(a)(2)(C). Second, seizing on the statutory phrase “committed an offense referred to in section 1182(a)(2) . . . that renders the alien inadmissible to the United States under section 1182(a)(2),” §1229b(d)(1)(B), Barton argues that a noncitizen is rendered “inadmissible” when actually adjudicated as inadmissible and denied admission to the United States, something that usually cannot happen to a lawfully admitted noncitizen. But the statutory text employs the term “inadmissibility” as a status that can result from, e.g., a noncitizen’s (including a lawfully admitted noncitizen’s) commission of certain offenses listed in §1182(a)(2). See, e.g., §§1182(a)(2)(A)(i), (B). And Congress has made that status relevant in several statutory contexts that apply to lawfully admitted noncitizens such as Barton. In those contexts, a noncitizen faces immigration consequences from being convicted of a §1182(a)(2) offense even though the noncitizen is lawfully admitted and is not necessarily removable solely because of that offense. See, e.g., §§1160(a)(1)(C), (a)(3)(B)(ii). Such examples pose a major hurdle for Barton’s textual argument, and Barton has no persuasive answer. Third, Barton argues that the Government’s interpretation treats as surplusage the phrase “or removable from the United States under section 1227(a)(2) or 1227(a)(4).” But redundancies are common in statutory drafting. The Court has often recognized that sometimes the better overall reading of a statute contains some redundancy. And redundancy in one portion of a statute is not a license to rewrite or eviscerate another portion of the statute contrary to its text. Finally, Barton argues alternatively that, even if inadmissibility is a status, and even if the offense that precludes cancellation of removal need not be one of the offenses of removal, a noncitizen must at least have been capable of being charged with a §1182(a)(2) inadmissibility offense as the basis for removal. Because the cancellation-of-removal statute is a recidivist statute, however, whether the offense that precludes cancellation of removal was charged or could have been charged as one of the offenses of removal is irrelevant. Pp. 10–17. 904 F.3d 1294, affirmed. Kavanaugh, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Alito, and Gorsuch, JJ., joined. Sotomayor, J., filed a dissenting opinion, in which Ginsburg, Breyer, and Kagan, JJ., joined.

Under the immigration laws, a noncitizen who is authorized to live permanently in the United States is a lawful permanent resident—also commonly known as a green-card holder. But unlike a U. S. citizen, a lawful permanent resident who commits a serious crime may be removed from the United States. Andre Barton is a Jamaican national and a longtime lawful permanent resident of the United States. During his time in the United States, Barton has been convicted of state crimes on three separate occasions spanning 12 years. The crimes include a firearms offense, drug offenses, and aggravated assault offenses. By law, the firearms offense and the drug offenses each independently rendered Barton eligible for removal from the United States. In September 2016, the U. S. Government sought to remove Barton, and a U. S. Immigration Judge determined that Barton was removable. Barton applied for cancellation of removal, a form of relief that allows a noncitizen to remain in the United States despite being found removable. The immigration laws authorize an immigration judge to cancel removal, but Congress has established strict eligibility requirements. See 8 U. S. C. §§1229b(a), (d)(1)(B). For a lawful permanent resident such as Barton, the applicant for cancellation of removal (1) must have been a lawful permanent resident for at least five years; (2) must have continuously resided in the United States for at least seven years after lawful admission; (3) must not have been convicted of an aggravated felony as defined in the immigration laws; and (4) during the initial seven years of continuous residence, must not have committed certain other offenses listed in 8 U. S. C. §1182(a)(2). If a lawful permanent resident meets those eligibility requirements, the immigration judge has discretion to (but is not required to) cancel removal and allow the lawful permanent resident to remain in the United States. Under the cancellation-of-removal statute, the immigration judge examines the applicant’s prior crimes, as well as the offense that triggered his removal. If a lawful permanent resident has ever been convicted of an aggravated felony, or has committed an offense listed in §1182(a)(2) during the initial seven years of residence, that criminal record will preclude cancellation of removal.[1] In that way, the statute operates like traditional criminal recidivist laws, which ordinarily authorize or impose greater sanctions on offenders who have committed prior crimes. In this case, after finding Barton removable based on his state firearms and drug offenses, the Immigration Judge and the Board of Immigration Appeals (BIA) concluded that Barton was not eligible for cancellation of removal. Barton had committed offenses listed in §1182(a)(2) during his initial seven years of residence—namely, his state aggravated assault offenses in 1996. Barton’s 1996 aggravated assault offenses were not the offenses that triggered his removal. But according to the BIA, and contrary to Barton’s argument, the offense that precludes cancellation of removal need not be one of the offenses of removal. In re Jurado-Delgado, 24 I. & N. Dec. 29, 31 (BIA 2006). The U. S. Court of Appeals for the Eleventh Circuit agreed with the BIA’s reading of the statute and concluded that Barton was not eligible for cancellation of removal. The Second, Third, and Fifth Circuits have similarly construed the statute; only the Ninth Circuit has disagreed. Barton argues that the BIA and the Eleventh Circuit misinterpreted the statute. He contends that the §1182(a)(2) offense that precludes cancellation of removal must be one of the offenses of removal. We disagree with Barton, and we affirm the judgment of the U. S. Court of Appeals for the Eleventh Circuit. I Federal immigration law governs the admission of noncitizens to the United States and the deportation of noncitizens previously admitted. See 8 U. S. C. §§1182(a), 1227(a), 1229a.[2] The umbrella statutory term for being inadmissible or deportable is “removable.” §1229a(e)(2). A noncitizen who is authorized to live permanently in the United States is a lawful permanent resident, often known as a green-card holder. When a lawful permanent resident commits a crime and is determined by an immigration judge to be removable because of that crime, the Attorney General (usually acting through an immigration judge) may cancel removal. §1229b(a). But the comprehensive immigration law that Congress passed and President Clinton signed in 1996 tightly cabins eligibility for cancellation of removal. See Illegal Immigration Reform and Immigrant Responsibility Act of 1996, 110Stat. 3009–546, 8 U. S. C. §1101 note. For a lawful permanent resident, the cancellation-of- removal statute provides that an immigration judge “may cancel removal in the case of an alien who is inadmissible or deportable from the United States if the alien—(1) has been an alien lawfully admitted for permanent residence for not less than 5 years, (2) has resided in the United States continuously for 7 years after having been admitted in any status, and (3) has not been convicted of any aggravated felony.” §1229b(a).[3] The statute imposes one other requirement known as the “stop-time rule.” As relevant here, the statute provides that a lawful permanent resident, during the initial seven years of residence, also cannot have committed “an offense referred to in section 1182(a)(2) of this title that renders the alien inadmissible to the United States under section 1182(a)(2) of this title or removable from the United States under section 1227(a)(2) or 1227(a)(4) of this title.” §1229b(d)(1)(B). Andre Barton is a Jamaican national and a lawful permanent resident of the United States. In 1996, he was convicted in a Georgia court of a firearms offense stemming from an incident where Barton and a friend shot up the house of Barton’s ex-girlfriend. In separate proceedings in 2007 and 2008, he was convicted in Georgia courts of state drug offenses. One case involved methamphetamine, and the other involved cocaine and marijuana. In 2016, the U. S. Government charged Barton with deportability under 8 U. S. C. §1227(a)(2) based on the 1996 firearms offense and the 2007 and 2008 drug crimes. See §§1227(a)(2)(B)(i), (C). Barton conceded that he was removable based on his criminal co nvictions for the firearms offense and drug offenses, and an Immigration Judge found him removable. Barton applied for cancellation of removal. All agree that Barton meets two of the eligibility requirements for cancellation of removal. He has been a lawful permanent resident for more than five years. And he has not been convicted of an “aggravated felony,” as defined by the immigration laws. The Immigration Judge concluded, however, that Barton had committed an offense listed in §1182(a)(2) during his initial seven years of residence. In 1996, 61∕2 years after his admission to this country, Barton committed aggravated assault offenses for which he was later convicted in a Georgia court. The Immigration Judge concluded that those aggravated assault offenses were covered by §1182(a)(2) and that Barton was therefore not eligible for cancellation of removal. The Board of Immigration Appeals and the U. S. Court of Appeals for the Eleventh Circuit likewise concluded that Barton was not eligible for cancellation of removal. Barton v. United States Atty. Gen., 904 F.3d 1294, 1302 (2018). The key question was whether the offense that precludes cancellation of removal (here, Barton’s 1996 aggravated assault offenses) must also be one of the offenses of removal.[4] The Board of Immigration Appeals has long interpreted the statute to mean that “an alien need not actually be charged and found inadmissible or removable on the applicable ground in order for the criminal conduct in question to terminate continuous residence in this country” and preclude cancellation of removal. Jurado-Delgado, 24 I. & N. Dec., at 31. In this case, the Eleventh Circuit likewise indicated that the §1182(a)(2) offense that precludes cancellation of removal need not be one of the offenses of removal. 904 F. 3d, at 1299–1300. And the Second, Third, and Fifth Circuits have similarly construed the statute. See Heredia v. Sessions, 865 F.3d 60, 68 (CA2 2017); Ardon v. Attorney General of United States, 449 Fed. Appx. 116, 118 (CA3 2011); Calix v. Lynch, 784 F.3d 1000, 1011 (CA5 2015). But in 2018, the Ninth Circuit disagreed with those courts and with the BIA. The Ninth Circuit ruled that a lawful permanent resident’s commission of an offense listed in §1182(a)(2) makes the noncitizen ineligible for cancellation of removal only if that offense was one of the offenses of removal. Nguyen v. Sessions, 901 F.3d 1093, 1097 (2018). Under the Ninth Circuit’s approach, Barton would have been eligible for cancellation of removal because his §1182(a)(2) offenses (his 1996 aggravated assault offenses) were not among the offenses of removal (his 1996 firearms offense and his 2007 and 2008 drug crimes). In light of the division in the Courts of Appeals over how to interpret this statute, we granted certiorari. 587 U. S. ___ (2019). II A Under the immigration laws, when a noncitizen has committed a serious crime, the U. S. Government may seek to remove that noncitizen by initiating removal proceedings before an immigration judge. If the immigration judge determines that the noncitizen is removable, the immigration judge nonetheless has discretion to cancel removal. But the immigration laws impose strict eligibility requirements for cancellation of removal. To reiterate, a lawful permanent resident such as Barton who has been found removable because of criminal activity is eligible for cancellation of removal “if the alien—(1) has been an alien lawfully admitted for permanent residence for not less than 5 years, (2) has resided in the United States continuously for 7 years after having been admitted in any status, and (3) has not been convicted of any aggravated felony.” §1229b(a). To be eligible for cancellation of removal, the lawful permanent resident, during the initial seven years of residence after admission, also must not have committed “an offense referred to in section 1182(a)(2) of this title that renders the alien inadmissible to the United States under section 1182(a)(2) of this title or removable from the United States under section 1227(a)(2) or 1227(a)(4) of this title.” §1229b(d)(1)(B). The law therefore fashions two distinct ways in which a lawful permanent resident’s prior crimes may preclude cancellation of removal. The law precludes cancellation of removal if the lawful permanent resident has been convicted of an “aggravated felony” at any time. The statutory list of aggravated felonies is long: murder, rape, drug trafficking, firearms trafficking, obstruction of justice, treason, gambling, human trafficking, and tax evasion, among many other crimes. §§1101(a)(43)(A)–(U). In addition, the law precludes cancellation of removal if the lawful permanent resident committed certain other serious crimes during the initial seven years of residence. The law defines those offenses by cross-referencing §1182(a)(2), which specifies the offenses that can render a noncitizen “inadmissible” to the United States. Section 1182(a)(2) includes “crime[s] involving moral turpitude,” which is a general category that covers a wide variety of crimes. Section 1182(a)(2) also expressly encompasses various violations of drug laws, prostitution, money laundering, and certain DUIs involving personal injury, among other crimes. §§1182(a)(2)(A)(i), (C), (D), (E), (I); see §1101(h). In specifying when cancellation of removal would be precluded because of prior criminal activity, Congress struck a balance that considers both the nature of the prior crime and the length of time that the noncitizen has resided in the United States. If a lawful permanent resident has been convicted at any time of certain crimes (what the immigration laws refer to as an “aggravated felony”), then the noncitizen is not eligible for cancellation of removal. If during the initial 7-year period of residence, a lawful permanent resident committed certain other offenses referred to in §1182(a)(2), then the noncitizen likewise is not eligible for cancellation of removal. In providing that a noncitizen’s prior crimes (in addition to the offense of removal) can render him ineligible for cancellation of removal, the cancellation-of-removal statute functions like a traditional recidivist sentencing statute. In an ordinary criminal case, a defendant may be convicted of a particular criminal offense. And at sentencing, the defendant’s other criminal offenses may be relevant. So too in the immigration removal context. A noncitizen may be found removable based on a certain criminal offense. In applying for cancellation of removal, the noncitizen must detail his entire criminal record on Form EOIR–42A. An immigration judge then must determine whether the noncitizen has been convicted of an aggravated felony at any time or has committed a §1182(a)(2) offense during the initial seven years of residence. It is entirely ordinary to look beyond the offense of conviction at criminal sentencing, and it is likewise entirely ordinary to look beyond the offense of removal at the cancellation-of-removal stage in immigration cases.[5] It is not surprising, moreover, that Congress required immigration judges considering cancellation of removal to look in part at whether the noncitizen has committed any offenses listed in §1182(a)(2). The offenses listed in §1182(a)(2) help determine whether a noncitizen should be admitted to the United States. Under the cancellation-of-removal statute, immigration judges must look at that same category of offenses to determine whether, after a previously admitted noncitizen has been determined to be deportable, the noncitizen should nonetheless be allowed to remain in the United States. If a crime is serious enough to deny admission to a noncitizen, the crime can also be serious enough to preclude cancellation of removal, at least if committed during the initial seven years of residence. Importantly, the text of the cancellation-of-removal statute does not simply say that cancellation of removal is precluded when, during the initial seven years of residence, the noncitizen was convicted of an offense referred to in §1182(a)(2). Rather, the text says that cancellation of removal is precluded when, during the initial seven years of residence, the noncitizen “committed an offense referred to in section 1182(a)(2) . . . that renders the alien inadmissible.” §1229b(d)(1)(B). That language clarifies two points of relevance here. First, cancellation of removal is precluded if a noncitizen committed a §1182(a)(2) offense during the initial seven years of residence, even if (as in Barton’s case) the conviction occurred after the seven years elapsed. In other words, as Congress specified in the statute and as the BIA and the Courts of Appeals have recognized, the date of commission of the offense is the key date for purposes of calculating whether the noncitizen committed a §1182(a)(2) offense during the initial seven years of residence. See In re Perez, 22 I. & N. Dec. 689, 693–694 (BIA 1999) (date of commission is controlling date); see also Heredia, 865 F. 3d, at 70–71 (“the date of the commission of the offense governs the computation of a lawful permanent resident’s continuous residency in the United States”); Calix, 784 F. 3d, at 1012 (“Once he was convicted of the offense” referred to in §1182(a)(2), “he was rendered inadmissible to the United States. His accrual of continuous residence was halted as of the date he committed that offense”). Second, the text of the law requires that the noncitizen be rendered “inadmissible” as a result of the offense. For crimes involving moral turpitude, which is the relevant category of §1182(a)(2) offenses here, §1182(a)(2) provides that a noncitizen is rendered “inadmissible” when he is convicted of or admits the offense. §1182(a)(2)(A)(i). As the Eleventh Circuit explained, “while only commission is required at step one, conviction (or admission) is required at step two.” 904 F. 3d, at 1301. In this case, Barton’s 1996 state aggravated assault offenses were crimes involving moral turpitude and therefore “referred to in section 1182(a)(2).” Barton committed those offenses during his initial seven years of residence. He was later convicted of the offenses in a Georgia court and thereby rendered “inadmissible.” Therefore, Barton was ineligible for cancellation of removal. As a matter of statutory text and structure, that analysis is straightforward. The Board of Immigration Appeals has long interpreted the statute that way. See Jurado-Delgado, 24 I. & N. Dec., at 31. And except for the Ninth Circuit, all of the Courts of Appeals to consider the question have interpreted the statute that way. B Barton pushes back on that straightforward statutory interpretation and the longstanding position of the Board of Immigration Appeals. Barton says that he may not be denied cancellation of removal based on his 1996 aggravated assault offenses because those offenses were not among the offenses of removal found by the Immigration Judge in Barton’s removal proceeding. Rather, his 1996 firearms offense and his 2007 and 2008 drug offenses were the offenses of removal. To succinctly summarize the parties’ different positions (with the difference highlighted in italics below): The Government would preclude cancellation of removal under this provision if the lawful permanent resident committed a §1182(a)(2) offense during the initial seven years of residence. Barton would preclude cancellation of removal under this provision if the lawful permanent resident committed a §1182(a)(2) offense during the initial seven years of residence and if that §1182(a)(2) offense was one of the offenses of removal in the noncitizen’s removal proceeding. To support his “offense of removal” approach, Barton advances three different arguments. A caution to the reader: These arguments are not easy to unpack. First, according to Barton, the statute’s overall structure with respect to removal proceedings demonstrates that a §1182(a)(2) offense may preclude cancellation of removal only if that §1182(a)(2) offense was one of the offenses of removal. We disagree. In removal proceedings, a lawful permanent resident (such as Barton) may be found “deportable” based on deportability offenses listed in §1227(a)(2). A noncitizen who has not previously been admitted may be found “inadmissible” based on inadmissibility offenses listed in §1182(a)(2). See §§1182(a), 1227(a), 1229a(e)(2). Importantly, then, §1227(a)(2) offenses—not §1182(a)(2) offenses—are typically the basis for removal of lawful permanent residents. Because the offense of removal for lawful permanent residents is ordinarily a §1227(a)(2) offense, Barton’s structural argument falls apart. If Barton were correct that this aspect of the cancellation-of-removal statute focused only on the offense of removal, the statute presumably would specify offenses “referred to in section 1182(a)(2) or section 1227(a)(2).” So why does the statute identify only offenses “referred to in section 1182(a)(2)”? Barton has no good answer. At oral argument, when directly asked that question, Barton’s able counsel forthrightly acknowledged: “It’s a little hard to explain.” Tr. of Oral Arg. 27. This point is the Achilles’ heel of Barton’s structural argument. As we see it, Barton cannot explain the omission of §1227(a)(2) offenses in the “referred to” clause for a simple reason: Barton’s interpretation of the statute is incorrect. Properly read, this is not simply an “offense of removal” statute that looks only at whether the offense of removal was committed during the initial seven years of residence. Rather, this is a recidivist statute that uses §1182(a)(2) offenses as a shorthand cross-reference for a category of offenses that will preclude cancellation of removal if committed during the initial seven years of residence. By contrast to this cancellation-of-removal provision, some other provisions of the immigration laws do focus only on the offense of removal—for example, provisions governing mandatory detention and jurisdiction. See §§1226(a), (c)(1)(A), (B), 1252(a)(2)(C). But the statutory text and context of those provisions support that limitation. Those provisions use the phrase “inadmissible by reason of ” a §1182(a)(2) offense, “deportable by reason of ” a §1227(a)(2) offense, or “removable by reason of ” a §1182(a)(2) or §1227(a)(2) offense. And the provisions make contextual sense only if the offense justifying detention or denying jurisdiction is one of the offenses of removal. The cancellation-of-removal statute does not employ similar language. Second, moving from overall structure to precise text, Barton seizes on the statutory phrase “committed an offense referred to in section 1182(a)(2) . . . that renders the alien inadmissible to the United States under section 1182(a)(2).” §1229b(d)(1)(B) (emphasis added). According to Barton, conviction of an offense listed in §1182(a)(2)—for example, conviction in state court of a crime involving moral turpitude—does not itself render the noncitizen “inadmissible.” He argues that a noncitizen is not rendered “inadmissible” unless and until the noncitizen is actually adjudicated as inadmissible and denied admission to the United States. And he further contends that a lawfully admitted noncitizen usually cannot be removed from the United States on the basis of inadmissibility. As Barton puts it (and the dissent echoes the point), how can a lawfully admitted noncitizen be found inadmissible when he has already been lawfully admitted? As a matter of common parlance alone, that argument would of course carry some force. But the argument fails because it disregards the statutory text, which employs the term “inadmissibility” as a status that can result from, for example, a noncitizen’s (including a lawfully admitted noncitizen’s) commission of certain offenses listed in §1182(a)(2). For example, as relevant here, §1182(a)(2) flatly says that a noncitizen such as Barton who commits a crime involving moral turpitude and is convicted of that offense “is inadmissible.” §1182(a)(2)(A)(i). Full stop. Similarly, a noncitizen who has two or more convictions, together resulting in aggregate sentences of at least five years, “is inadmissible.” §1182(a)(2)(B). A noncitizen who a consular officer or the Attorney General knows or has reason to believe is a drug trafficker “is inadmissible.” §1182(a)(2)(C)(i). A noncitizen who receives the proceeds of prostitution within 10 years of applying for admission “is inadmissible.” §1182(a)(2)(D)(ii). The list goes on. See, e.g., §§1182(a)(2)(C)(ii)–(E), (G)–(I). Those provisions do not say that a noncitizen will become inadmissible if the noncitizen is found inadmissible in a subsequent immigration removal proceeding. Instead, those provisions say that the noncitizen “is inadmissible.” Congress has in turn made that status—inadmissibility because of conviction or other proof of commission of §1182(a)(2) offenses—relevant in several statutory contexts that apply to lawfully admitted noncitizens such as Barton. Those contexts include adjustment to permanent resident status; protection from removal because of temporary protected status; termination of temporary resident status; and here cancellation of removal. See, e.g., §§1160(a)(1)(C), (a)(3)(B)(ii), 1254a(a)(1)(A), (c)(1)(A)(iii), 1255(a), (l)(2). In those contexts, the noncitizen faces immigration consequences from being convicted of a §1182(a)(2) offense even though the noncitizen is lawfully admitted and is not necessarily removable solely because of that offense. Consider how those other proceedings work. A lawfully admitted noncitizen who is convicted of an offense listed in §1182(a)(2) is typically not removable from the United States on that basis (recall that a lawfully admitted noncitizen is ordinarily removable only for commission of a §1227(a)(2) offense). But the noncitizen is “inadmissible” because of the §1182(a)(2) offense and for that reason may not be able to obtain adjustment to permanent resident status. §§1255(a), (l)(2). So too, a lawfully admitted noncitizen who is convicted of an offense listed in §1182(a)(2) is “inadmissible” and for that reason may not be able to obtain temporary protected status. §§1254a(a)(1)(A), (c)(1)(A)(iii). A lawfully admitted noncitizen who is a temporary resident and is convicted of a §1182(a)(2) offense is “inadmissible” and for that reason may lose temporary resident status. §§1160(a)(1)(C), (a)(3)(B)(ii). Those statutory examples pose a major hurdle for Barton’s textual argument. The examples demonstrate that Congress has employed the concept of “inadmissibility” as a status in a variety of statutes similar to the cancellation-of-removal statute, including for lawfully admitted noncitizens. Barton has no persuasive answer to those examples. Barton tries to say that some of those other statutes involve a noncitizen who, although already admitted to the United States, is nonetheless seeking “constructive admission.” Reply Brief 12; Tr. of Oral Arg. 11. But that ginned-up label does not avoid the problem. Put simply, those other statutes show that lawfully admitted noncitizens who are, for example, convicted of §1182(a)(2) crimes are “inadmissible” and in turn may suffer certain immigration consequences, even though those lawfully admitted noncitizens cannot necessarily be removed solely because of those §1182(a)(2) offenses. The same is true here. A lawfully admitted noncitizen who was convicted of a crime involving moral turpitude during his initial seven years of residence is “inadmissible” and for that reason is ineligible for cancellation of removal. In advancing his structural and textual arguments, Barton insists that his interpretation of the statute reflects congressional intent regarding cancellation of removal. But if Congress intended that only the offense of removal would preclude cancellation of removal under the 7-year residence provision, it is unlikely that Congress would have employed such a convoluted way to express that intent. Barton cannot explain why, if his view of Congress’ intent is correct, the statute does not simply say something like: “The alien is not eligible for cancellation of removal if the offense of removal was committed during the alien’s initial seven years of residence.” Third, on a different textual tack, Barton argues that the Government’s interpretation cannot be correct because the Government would treat as surplusage the phrase “or removable from the United States under section 1227(a)(2) or 1227(a)(4) of this title.” Recall that the statute, as relevant here, provides that a lawful permanent resident is not eligible for cancellation of removal if, during the initial seven years of residence, he committed “an offense referred to in section 1182(a)(2) of this title that renders the alien inadmissible to the United States under section 1182(a)(2) of this title or removable from the United States under section 1227(a)(2) or 1227(a)(4) of this title.” §1229b(d)(1)(B) (emphasis added). To begin with, all agree that under either side’s interpretation, the reference to §1227(a)(4)—as distinct from §1227(a)(2)—is redundant surplusage. See §1229b(c)(4); Brief for Petitioner 32–33 & n. 7. Under the Government’s interpretation, it is true that the reference to §1227(a)(2) also appears to be redundant surplusage. Any offense that is both referred to in §1182(a)(2) and an offense that would render the noncitizen deportable under §1227(a)(2) would also render the noncitizen inadmissible under §1182(a)(2). But redundancies are common in statutory drafting—sometimes in a congressional effort to be doubly sure, sometimes because of congressional inadvertence or lack of foresight, or sometimes simply because of the shortcomings of human communication. The Court has often recognized: “Sometimes the better overall reading of the statute contains some redundancy.” Rimini Street, Inc. v. Oracle USA, Inc., 586 U. S. ___, ___ (2019) (slip op., at 11); see Wisconsin Central Ltd. v. United States, 585 U. S. ___, ___ (2018) (slip op., at 7); Marx v. General Revenue Corp., 568 U.S. 371, 385 (2013); Lamie v. United States Trustee, 540 U.S. 526, 536 (2004). So it is here. Most importantly for present purposes, we do not see why the redundant statutory reference to §1227(a)(2) should cause us to entirely rewrite §1229b so that a noncitizen’s commission of an offense referred to in §1182(a)(2) would preclude cancellation of removal only if it is also the offense of removal. Redundancy in one portion of a statute is not a license to rewrite or eviscerate another portion of the statute contrary to its text, as Barton would have us do. One final point: Barton argues in the alternative that even if inadmissibility is a status, and even if the offense that precludes cancellation of removal need not be one of the offenses of removal, the noncitizen must at least have been capable of being charged with a §1182(a)(2) inadmissibility offense as the basis for removal. The dissent seizes on this argument as well. But as we have explained, this cancellation-of-removal statute is a recidivist statute that precludes cancellation of removal if the noncitizen has committed an offense listed in §1182(a)(2) during the initial seven years of residence. Whether the offense that precludes cancellation of removal was charged or could have been charged as one of the offenses of removal is irrelevant to that analysis. * * * Removal of a lawful permanent resident from the United States is a wrenching process, especially in light of the consequences for family members. Removal is particularly difficult when it involves someone such as Barton who has spent most of his life in the United States. Congress made a choice, however, to authorize removal of noncitizens—even lawful permanent residents—who have committed certain serious crimes. And Congress also made a choice to categorically preclude cancellation of removal for noncitizens who have substantial criminal records. Congress may of course amend the law at any time. In the meantime, the Court is constrained to apply the law as enacted by Congress. Here, as the BIA explained in its 2006 Jurado- Delgado decision, and as the Second, Third, Fifth, and Eleventh Circuits have indicated, the immigration laws enacted by Congress do not allow cancellation of removal when a lawful permanent resident has amassed a criminal record of this kind. We affirm the judgment of the U. S. Court of Appeals for the Eleventh Circuit. It is so ordered. Notes 1 As the statute makes clear, and as we discuss below, committing a §1182(a)(2) offense precludes cancellation of removal only if the offense also “renders” the noncitizen inadmissible. See infra, at 10. Section 1182(a)(2) specifies what that means for each of its enumerated offenses. For the offense at issue in this case, the noncitizen must also have been convicted of or admitted the offense. 2 This opinion uses the term “noncitizen” as equivalent to the statutory term “alien.” See 8 U. S. C. §1101(a)(3). 3 The immigration laws impose a similar but even stricter set of eligibility requirements for noncitizens who are not lawful permanent residents. §1229b(b). 4 The term “offense of removal” describes the offense that was the ground on which the immigration judge, at the removal proceeding, found the noncitizen removable. 5 If the offense of removal itself was an aggravated felony or was an offense listed in §1182(a)(2) that was committed during the initial seven years of residence, then the offense of removal alone precludes cancellation of removal, regardless of whether the noncitizen has an additional record of prior crimes.

590.US.2019_17-1618

In each of these cases, an employer allegedly fired a long-time employee simply for being homosexual or transgender. Clayton County, Georgia, fired Gerald Bostock for conduct “unbecoming” a county employee shortly after he began participating in a gay recreational softball league. Altitude Express fired Donald Zarda days after he mentioned being gay. And R. G. & G. R. Harris Funeral Homes fired Aimee Stephens, who presented as a male when she was hired, after she informed her employer that she planned to “live and work full-time as a woman.” Each employee sued, alleging sex discrimination under Title VII of the Civil Rights Act of 1964. The Eleventh Circuit held that Title VII does not prohibit employers from firing employees for being gay and so Mr. Bostock’s suit could be dismissed as a matter of law. The Second and Sixth Circuits, however, allowed the claims of Mr. Zarda and Ms. Stephens, respectively, to proceed. Held: An employer who fires an individual merely for being gay or transgender violates Title VII. Pp. 4–33. (a) Title VII makes it “unlawful . . . for an employer to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual . . . because of such individual’s race, color, religion, sex, or national origin.” 42 U. S. C. §2000e–2(a)(1). The straightforward application of Title VII’s terms interpreted in accord with their ordinary public meaning at the time of their enactment resolves these cases. Pp. 4–12. (1) The parties concede that the term “sex” in 1964 referred to the biological distinctions between male and female. And “the ordinary meaning of ‘because of’ is ‘by reason of’ or ‘on account of,’ ” University of Tex. Southwestern Medical Center v. Nassar, 570 U.S. 338, 350. That term incorporates the but-for causation standard, id., at 346, 360, which, for Title VII, means that a defendant cannot avoid liability just by citing some other factor that contributed to its challenged employment action. The term “discriminate” meant “[t]o make a difference in treatment or favor (of one as compared with others).” Webster’s New International Dictionary 745. In so-called “disparate treatment” cases, this Court has held that the difference in treatment based on sex must be intentional. See, e.g., Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 986. And the statute’s repeated use of the term “individual” means that the focus is on “[a] particular being as distinguished from a class.” Webster’s New International Dictionary, at 1267. Pp. 4–9. (2) These terms generate the following rule: An employer violates Title VII when it intentionally fires an individual employee based in part on sex. It makes no difference if other factors besides the plaintiff’s sex contributed to the decision or that the employer treated women as a group the same when compared to men as a group. A statutory violation occurs if an employer intentionally relies in part on an individual employee’s sex when deciding to discharge the employee. Because discrimination on the basis of homosexuality or transgender status requires an employer to intentionally treat individual employees differently because of their sex, an employer who intentionally penalizes an employee for being homosexual or transgender also violates Title VII. There is no escaping the role intent plays: Just as sex is necessarily a but-for cause when an employer discriminates against homosexual or transgender employees, an employer who discriminates on these grounds inescapably intends to rely on sex in its decisionmaking. Pp. 9–12. (b) Three leading precedents confirm what the statute’s plain terms suggest. In Phillips v. Martin Marietta Corp., 400 U.S. 542, a company was held to have violated Title VII by refusing to hire women with young children, despite the fact that the discrimination also depended on being a parent of young children and the fact that the company favored hiring women over men. In Los Angeles Dept. of Water and Power v. Manhart, 435 U.S. 702, an employer’s policy of requiring women to make larger pension fund contributions than men because women tend to live longer was held to violate Title VII, notwithstanding the policy’s evenhandedness between men and women as groups. And in Oncale v. Sundowner Offshore Services, Inc., 523 U.S. 75, a male plaintiff alleged a triable Title VII claim for sexual harassment by co-workers who were members of the same sex. The lessons these cases hold are instructive here. First, it is irrelevant what an employer might call its discriminatory practice, how others might label it, or what else might motivate it. In Manhart, the employer might have called its rule a “life expectancy” adjustment, and in Phillips, the employer could have accurately spoken of its policy as one based on “motherhood.” But such labels and additional intentions or motivations did not make a difference there, and they cannot make a difference here. When an employer fires an employee for being homosexual or transgender, it necessarily intentionally discriminates against that individual in part because of sex. Second, the plaintiff’s sex need not be the sole or primary cause of the employer’s adverse action. In Phillips, Manhart, and Oncale, the employer easily could have pointed to some other, nonprotected trait and insisted it was the more important factor in the adverse employment outcome. Here, too, it is of no significance if another factor, such as the plaintiff’s attraction to the same sex or presentation as a different sex from the one assigned at birth, might also be at work, or even play a more important role in the employer’s decision. Finally, an employer cannot escape liability by demonstrating that it treats males and females comparably as groups. Manhart is instructive here. An employer who intentionally fires an individual homosexual or transgender employee in part because of that individual’s sex violates the law even if the employer is willing to subject all male and female homosexual or transgender employees to the same rule. Pp. 12–15. (c) The employers do not dispute that they fired their employees for being homosexual or transgender. Rather, they contend that even intentional discrimination against employees based on their homosexual or transgender status is not a basis for Title VII liability. But their statutory text arguments have already been rejected by this Court’s precedents. And none of their other contentions about what they think the law was meant to do, or should do, allow for ignoring the law as it is. Pp. 15–33. (1) The employers assert that it should make a difference that plaintiffs would likely respond in conversation that they were fired for being gay or transgender and not because of sex. But conversational conventions do not control Title VII’s legal analysis, which asks simply whether sex is a but-for cause. Nor is it a defense to insist that intentional discrimination based on homosexuality or transgender status is not intentional discrimination based on sex. An employer who discriminates against homosexual or transgender employees necessarily and intentionally applies sex-based rules. Nor does it make a difference that an employer could refuse to hire a gay or transgender individual without learning that person’s sex. By intentionally setting out a rule that makes hiring turn on sex, the employer violates the law, whatever he might know or not know about individual applicants. The employers also stress that homosexuality and transgender status are distinct concepts from sex, and that if Congress wanted to address these matters in Title VII, it would have referenced them specifically. But when Congress chooses not to include any exceptions to a broad rule, this Court applies the broad rule. Finally, the employers suggest that because the policies at issue have the same adverse consequences for men and women, a stricter causation test should apply. That argument unavoidably comes down to a suggestion that sex must be the sole or primary cause of an adverse employment action under Title VII, a suggestion at odds with the statute. Pp. 16–23. (2) The employers contend that few in 1964 would have expected Title VII to apply to discrimination against homosexual and transgender persons. But legislative history has no bearing here, where no ambiguity exists about how Title VII’s terms apply to the facts. See Milner v. Department of Navy, 562 U.S. 562, 574. While it is possible that a statutory term that means one thing today or in one context might have meant something else at the time of its adoption or might mean something different in another context, the employers do not seek to use historical sources to illustrate that the meaning of any of Title VII’s language has changed since 1964 or that the statute’s terms ordinarily carried some missed message. Instead, they seem to say when a new application is both unexpected and important, even if it is clearly commanded by existing law, the Court should merely point out the question, refer the subject back to Congress, and decline to enforce the law’s plain terms in the meantime. This Court has long rejected that sort of reasoning. And the employers’ new framing may only add new problems and leave the Court with more than a little law to overturn. Finally, the employers turn to naked policy appeals, suggesting that the Court proceed without the law’s guidance to do what it thinks best. That is an invitation that no court should ever take up. Pp. 23–33. No. 17–1618, 723 Fed. Appx. 964, reversed and remanded; No. 17–1623, 883 F.3d 100, and No. 18–107, 884 F.3d 560, affirmed. Gorsuch, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined. Alito, J., filed a dissenting opinion, in which Thomas, J., joined. Kavanaugh, J., filed a dissenting opinion. Notes 1 Together with No. 17–1623, Altitude Express, Inc., et al. v. Zarda et al., as Co-Independent Executors of the Estate of Zarda, on certiorari to the United States Court of Appeals for the Second Circuit, and No. 18–107, R. G. & G. R. Harris Funeral Homes, Inc. v. Equal Employment Opportunity Commission et al., on certiorari to the United States Court of Appeals for the Sixth Circuit.

Sometimes small gestures can have unexpected consequences. Major initiatives practically guarantee them. In our time, few pieces of federal legislation rank in significance with the Civil Rights Act of 1964. There, in Title VII, Congress outlawed discrimination in the workplace on the basis of race, color, religion, sex, or national origin. Today, we must decide whether an employer can fire someone simply for being homosexual or transgender. The answer is clear. An employer who fires an individual for being homosexual or transgender fires that person for traits or actions it would not have questioned in members of a different sex. Sex plays a necessary and undisguisable role in the decision, exactly what Title VII forbids. Those who adopted the Civil Rights Act might not have anticipated their work would lead to this particular result. Likely, they weren’t thinking about many of the Act’s consequences that have become apparent over the years, including its prohibition against discrimination on the basis of motherhood or its ban on the sexual harassment of male employees. But the limits of the drafters’ imagination supply no reason to ignore the law’s demands. When the express terms of a statute give us one answer and extratextual considerations suggest another, it’s no contest. Only the written word is the law, and all persons are entitled to its benefit. I Few facts are needed to appreciate the legal question we face. Each of the three cases before us started the same way: An employer fired a long-time employee shortly after the employee revealed that he or she is homosexual or transgender—and allegedly for no reason other than the employee’s homosexuality or transgender status. Gerald Bostock worked for Clayton County, Georgia, as a child welfare advocate. Under his leadership, the county won national awards for its work. After a decade with the county, Mr. Bostock began participating in a gay recreational softball league. Not long after that, influential members of the community allegedly made disparaging comments about Mr. Bostock’s sexual orientation and participation in the league. Soon, he was fired for conduct “unbecoming” a county employee. Donald Zarda worked as a skydiving instructor at Altitude Express in New York. After several seasons with the company, Mr. Zarda mentioned that he was gay and, days later, was fired. Aimee Stephens worked at R. G. & G. R. Harris Funeral Homes in Garden City, Michigan. When she got the job, Ms. Stephens presented as a male. But two years into her service with the company, she began treatment for despair and loneliness. Ultimately, clinicians diagnosed her with gender dysphoria and recommended that she begin living as a woman. In her sixth year with the company, Ms. Stephens wrote a letter to her employer explaining that she planned to “ live and work full-time as a woman” after she returned from an upcoming vacation. The funeral home fired her before she left, telling her “this is not going to work out.” While these cases began the same way, they ended differently. Each employee brought suit under Title VII alleging unlawful discrimination on the basis of sex. 78Stat. 255, 42 U. S. C. §2000e–2(a)(1). In Mr. Bostock’s case, the Eleventh Circuit held that the law does not prohibit employers from firing employees for being gay and so his suit could be dismissed as a matter of law. 723 Fed. Appx. 964 (2018). Meanwhile, in Mr. Zarda’s case, the Second Circuit concluded that sexual orientation discrimination does violate Title VII and allowed his case to proceed. 883 F.3d 100 (2018). Ms. Stephens’s case has a more complex procedural history, but in the end the Sixth Circuit reached a decision along the same lines as the Second Circuit’s, holding that Title VII bars employers from firing employees because of their transgender status. 884 F.3d 560 (2018). During the course of the proceedings in these long-running disputes, both Mr. Zarda and Ms. Stephens have passed away. But their estates continue to press their causes for the benefit of their heirs. And we granted certiorari in these matters to resolve at last the disagreement among the courts of appeals over the scope of Title VII’s protections for homosexual and transgender persons. 587 U. S. ___ (2019). II This Court normally interprets a statute in accord with the ordinary public meaning of its terms at the time of its enactment. After all, only the words on the page constitute the law adopted by Congress and approved by the President. If judges could add to, remodel, update, or detract from old statutory terms inspired only by extratextual sources and our own imaginations, we would risk amending statutes outside the legislative process reserved for the people’s representatives. And we would deny the people the right to continue relying on the original meaning of the law they have counted on to settle their rights and obligations. See New Prime Inc. v. Oliveira, 586 U. S. ___, ___–___ (2019) (slip op., at 6–7). With this in mind, our task is clear. We must determine the ordinary public meaning of Title VII’s command that it is “unlawful . . . for an employer to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s race, color, religion, sex, or national origin.” §2000e–2(a)(1). To do so, we orient ourselves to the time of the statute’s adoption, here 1964, and begin by examining the key statutory terms in turn before assessing their impact on the cases at hand and then confirming our work against this Court’s precedents. A The only statutorily protected characteristic at issue in today’s cases is “sex”—and that is also the primary term in Title VII whose meaning the parties dispute. Appealing to roughly contemporaneous dictionaries, the employers say that, as used here, the term “sex” in 1964 referred to “status as either male or female [as] determined by reproductive biology.” The employees counter by submitting that, even in 1964, the term bore a broader scope, capturing more than anatomy and reaching at least some norms concerning gender identity and sexual orientation. But because nothing in our approach to these cases turns on the outcome of the parties’ debate, and because the employees concede the point for argument’s sake, we proceed on the assumption that “sex” signified what the employers suggest, referring only to biological distinctions between male and female. Still, that’s just a starting point. The question isn’t just what “sex” meant, but what Title VII says about it. Most notably, the statute prohibits employers from taking certain actions “because of ” sex. And, as this Court has previously explained, “the ordinary meaning of ‘because of ’ is ‘by reason of ’ or ‘on account of.’ ” University of Tex. Southwestern Medical Center v. Nassar, 570 U.S. 338, 350 (2013) (citing Gross v. FBL Financial Services, Inc., 557 U.S. 167, 176 (2009); quotation altered). In the language of law, this means that Title VII’s “because of ” test incorporates the “ ‘simple’ ” and “traditional” standard of but-for causation. Nassar, 570 U. S., at 346, 360. That form of causation is established whenever a particular outcome would not have happened “but for” the purported cause. See Gross, 557 U. S., at 176. In other words, a but-for test directs us to change one thing at a time and see if the outcome changes. If it does, we have found a but-for cause. This can be a sweeping standard. Often, events have multiple but-for causes. So, for example, if a car accident occurred both because the defendant ran a red light and because the plaintiff failed to signal his turn at the intersection, we might call each a but-for cause of the collision. Cf. Burrage v. United States, 571 U.S. 204, 211–212 (2014). When it comes to Title VII, the adoption of the traditional but-for causation standard means a defendant cannot avoid liability just by citing some other factor that contributed to its challenged employment decision. So long as the plaintiff ’s sex was one but-for cause of that decision, that is enough to trigger the law. See ibid.; Nassar, 570 U. S., at 350. No doubt, Congress could have taken a more parsimonious approach. As it has in other statutes, it could have added “solely” to indicate that actions taken “because of ” the confluence of multiple factors do not violate the law. Cf. 11 U. S. C. §525; 16 U. S. C. §511. Or it could have written “primarily because of ” to indicate that the prohibited factor had to be the main cause of the defendant’s challenged employment decision. Cf. 22 U. S. C. §2688. But none of this is the law we have. If anything, Congress has moved in the opposite direction, supplementing Title VII in 1991 to allow a plaintiff to prevail merely by showing that a protected trait like sex was a “motivating factor” in a defendant’s challenged employment practice. Civil Rights Act of 1991, §107, 105Stat. 1075, codified at 42 U. S. C. §2000e–2(m). Under this more forgiving standard, liability can sometimes follow even if sex wasn’t a but-for cause of the employer’s challenged decision. Still, because nothing in our analysis depends on the motivating factor test, we focus on the more traditional but-for causation standard that continues to afford a viable, if no longer exclusive, path to relief under Title VII. §2000e–2(a)(1). As sweeping as even the but-for causation standard can be, Title VII does not concern itself with everything that happens “because of ” sex. The statute imposes liability on employers only when they “fail or refuse to hire,” “discharge,” “or otherwise . . . discriminate against” someone because of a statutorily protected characteristic like sex. Ibid. The employers acknowledge that they discharged the plaintiffs in today’s cases, but assert that the statute’s list of verbs is qualified by the last item on it: “otherwise . . . discriminate against.” By virtue of the word otherwise, the employers suggest, Title VII concerns itself not with every discharge, only with those discharges that involve discrimination. Accepting this point, too, for argument’s sake, the question becomes: What did “discriminate” mean in 1964? As it turns out, it meant then roughly what it means today: “To make a difference in treatment or favor (of one as compared with others).” Webster’s New International Dictionary 745 (2d ed. 1954). To “discriminate against” a person, then, would seem to mean treating that individual worse than others who are similarly situated. See Burlington N. & S. F. R. Co. v. White, 548 U.S. 53, 59 (2006). In so-called “disparate treatment” cases like today’s, this Court has also held that the difference in treatment based on sex must be intentional. See, e.g., Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 986 (1988). So, taken together, an employer who intentionally treats a person worse because of sex—such as by firing the person for actions or attributes it would tolerate in an individual of another sex—discriminates against that person in violation of Title VII. At first glance, another interpretation might seem possible. Discrimination sometimes involves “the act, practice, or an instance of discriminating categorically rather than individually.” Webster’s New Collegiate Dictionary 326 (1975); see also post, at 27–28, n. 22 (Alito, J., dissenting). On that understanding, the statute would require us to consider the employer’s treatment of groups rather than individuals, to see how a policy affects one sex as a whole versus the other as a whole. That idea holds some intuitive appeal too. Maybe the law concerns itself simply with ensuring that employers don’t treat women generally less favorably than they do men. So how can we tell which sense, individual or group, “discriminate” carries in Title VII? The statute answers that question directly. It tells us three times—including immediately after the words “discriminate against”—that our focus should be on individuals, not groups: Employers may not “fail or refuse to hire or . . . discharge any individual, or otherwise . . . discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s . . . sex.” §2000e–2(a)(1) (emphasis added). And the meaning of “individual” was as uncontroversial in 1964 as it is today: “A particular being as distinguished from a class, species, or collection.” Webster’s New International Dictionary, at 1267. Here, again, Congress could have written the law differently. It might have said that “it shall be an unlawful employment practice to prefer one sex to the other in hiring, firing, or the terms or conditions of employment.” It might have said that there should be no “sex discrimination,” perhaps implying a focus on differential treatment between the two sexes as groups. More narrowly still, it could have forbidden only “sexist policies” against women as a class. But, once again, that is not the law we have. The consequences of the law’s focus on individuals rather than groups are anything but academic. Suppose an employer fires a woman for refusing his sexual advances. It’s no defense for the employer to note that, while he treated that individual woman worse than he would have treated a man, he gives preferential treatment to female employees overall. The employer is liable for treating this woman worse in part because of her sex. Nor is it a defense for an employer to say it discriminates against both men and women because of sex. This statute works to protect individuals of both sexes from discrimination, and does so equally. So an employer who fires a woman, Hannah, because she is insufficiently feminine and also fires a man, Bob, for being insufficiently masculine may treat men and women as groups more or less equally. But in both cases the employer fires an individual in part because of sex. Instead of avoiding Title VII exposure, this employer doubles it. B From the ordinary public meaning of the statute’s language at the time of the law’s adoption, a straightforward rule emerges: An employer violates Title VII when it intentionally fires an individual employee based in part on sex. It doesn’t matter if other factors besides the plaintiff ’s sex contributed to the decision. And it doesn’t matter if the employer treated women as a group the same when compared to men as a group. If the employer intentionally relies in part on an individual employee’s sex when deciding to discharge the employee—put differently, if changing the employee’s sex would have yielded a different choice by the employer—a statutory violation has occurred. Title VII’s message is “simple but momentous”: An individual employee’s sex is “not relevant to the selection, evaluation, or compensation of employees.” Price Waterhouse v. Hopkins, 490 U.S. 228, 239 (1989) (plurality opinion). The statute’s message for our cases is equally simple and momentous: An individual’s homosexuality or transgender status is not relevant to employment decisions. That’s because it is impossible to discriminate against a person for being homosexual or transgender without discriminating against that individual based on sex. Consider, for example, an employer with two employees, both of whom are attracted to men. The two individuals are, to the employer’s mind, materially identical in all respects, except that one is a man and the other a woman. If the employer fires the male employee for no reason other than the fact he is attracted to men, the employer discriminates against him for traits or actions it tolerates in his female colleague. Put differently, the employer intentionally singles out an employee to fire based in part on the employee’s sex, and the affected employee’s sex is a but-for cause of his discharge. Or take an employer who fires a transgender person who was identified as a male at birth but who now identifies as a female. If the employer retains an otherwise identical employee who was identified as female at birth, the employer intentionally penalizes a person identified as male at birth for traits or actions that it tolerates in an employee identified as female at birth. Again, the individual employee’s sex plays an unmistakable and impermissible role in the discharge decision. That distinguishes these cases from countless others where Title VII has nothing to say. Take an employer who fires a female employee for tardiness or incompetence or simply supporting the wrong sports team. Assuming the employer would not have tolerated the same trait in a man, Title VII stands silent. But unlike any of these other traits or actions, homosexuality and transgender status are inextricably bound up with sex. Not because homosexuality or transgender status are related to sex in some vague sense or because discrimination on these bases has some disparate impact on one sex or another, but because to discriminate on these grounds requires an employer to intentionally treat individual employees differently because of their sex. Nor does it matter that, when an employer treats one employee worse because of that individual’s sex, other factors may contribute to the decision. Consider an employer with a policy of firing any woman he discovers to be a Yankees fan. Carrying out that rule because an employee is a woman and a fan of the Yankees is a firing “because of sex” if the employer would have tolerated the same allegiance in a male employee. Likewise here. When an employer fires an employee because she is homosexual or transgender, two causal factors may be in play—both the individual’s sex and something else (the sex to which the individual is attracted or with which the individual identifies). But Title VII doesn’t care. If an employer would not have discharged an employee but for that individual’s sex, the statute’s causation standard is met, and liability may attach. Reframing the additional causes in today’s cases as additional intentions can do no more to insulate the employers from liability. Intentionally burning down a neighbor’s house is arson, even if the perpetrator’s ultimate intention (or motivation) is only to improve the view. No less, intentional discrimination based on sex violates Title VII, even if it is intended only as a means to achieving the employer’s ultimate goal of discriminating against homosexual or transgender employees. There is simply no escaping the role intent plays here: Just as sex is necessarily a but-for cause when an employer discriminates against homosexual or transgender employees, an employer who discriminates on these grounds inescapably intends to rely on sex in its decisionmaking. Imagine an employer who has a policy of firing any employee known to be homosexual. The employer hosts an office holiday party and invites employees to bring their spouses. A model employee arrives and introduces a manager to Susan, the employee’s wife. Will that employee be fired? If the policy works as the employer intends, the answer depends entirely on whether the model employee is a man or a woman. To be sure, that employer’s ultimate goal might be to discriminate on the basis of sexual orientation. But to achieve that purpose the employer must, along the way, intentionally treat an employee worse based in part on that individual’s sex. An employer musters no better a defense by responding that it is equally happy to fire male and female employees who are homosexual or transgender. Title VII liability is not limited to employers who, through the sum of all of their employment actions, treat the class of men differently than the class of women. Instead, the law makes each instance of discriminating against an individual employee because of that individual’s sex an independent violation of Title VII. So just as an employer who fires both Hannah and Bob for failing to fulfill traditional sex stereotypes doubles rather than eliminates Title VII liability, an employer who fires both Hannah and Bob for being gay or transgender does the same. At bottom, these cases involve no more than the straightforward application of legal terms with plain and settled meanings. For an employer to discriminate against employees for being homosexual or transgender, the employer must intentionally discriminate against individual men and women in part because of sex. That has always been prohibited by Title VII’s plain terms—and that “should be the end of the analysis.” 883 F. 3d, at 135 (Cabranes, J., concurring in judgment). C If more support for our conclusion were required, there’s no need to look far. All that the statute’s plain terms suggest, this Court’s cases have already confirmed. Consider three of our leading precedents. In Phillips v. Martin Marietta Corp., 400 U.S. 542 (1971) (per curiam), a company allegedly refused to hire women with young children, but did hire men with children the same age. Because its discrimination depended not only on the employee’s sex as a female but also on the presence of another criterion—namely, being a parent of young children—the company contended it hadn’t engaged in discrimination “because of ” sex. The company maintained, too, that it hadn’t violated the law because, as a whole, it tended to favor hiring women over men. Unsurprisingly by now, these submissions did not sway the Court. That an employer discriminates intentionally against an individual only in part because of sex supplies no defense to Title VII. Nor does the fact an employer may happen to favor women as a class. In Los Angeles Dept. of Water and Power v. Manhart, 435 U.S. 702 (1978), an employer required women to make larger pension fund contributions than men. The employer sought to justify its disparate treatment on the ground that women tend to live longer than men, and thus are likely to receive more from the pension fund over time. By everyone’s admission, the employer was not guilty of animosity against women or a “purely habitual assumptio[n] about a woman’s inability to perform certain kinds of work”; instead, it relied on what appeared to be a statistically accurate statement about life expectancy. Id., at 707–708. Even so, the Court recognized, a rule that appears evenhanded at the group level can prove discriminatory at the level of individuals. True, women as a class may live longer than men as a class. But “[t]he statute’s focus on the individual is unambiguous,” and any individual woman might make the larger pension contributions and still die as early as a man. Id., at 708. Likewise, the Court dismissed as irrelevant the employer’s insistence that its actions were motivated by a wish to achieve classwide equality between the sexes: An employer’s intentional discrimination on the basis of sex is no more permissible when it is prompted by some further intention (or motivation), even one as prosaic as seeking to account for actuarial tables. Ibid. The employer violated Title VII because, when its policy worked exactly as planned, it could not “pass the simple test” asking whether an individual female employee would have been treated the same regardless of her sex. Id., at 711. In Oncale v. Sundowner Offshore Services, Inc., 523 U.S. 75 (1998), a male plaintiff alleged that he was singled out by his male co-workers for sexual harassment. The Court held it was immaterial that members of the same sex as the victim committed the alleged discrimination. Nor did the Court concern itself with whether men as a group were subject to discrimination or whether something in addition to sex contributed to the discrimination, like the plaintiff ’s conduct or personal attributes. “[A]ssuredly,” the case didn’t involve “the principal evil Congress was concerned with when it enacted Title VII.” Id., at 79. But, the Court unanimously explained, it is “the provisions of our laws rather than the principal concerns of our legislators by which we are governed.” Ibid. Because the plaintiff alleged that the harassment would not have taken place but for his sex—that is, the plaintiff would not have suffered similar treatment if he were female—a triable Title VII claim existed. The lessons these cases hold for ours are by now familiar. First, it’s irrelevant what an employer might call its discriminatory practice, how others might label it, or what else might motivate it. In Manhart, the employer called its rule requiring women to pay more into the pension fund a “life expectancy” adjustment necessary to achieve sex equality. In Phillips, the employer could have accurately spoken of its policy as one based on “motherhood.” In much the same way, today’s employers might describe their actions as motivated by their employees’ homosexuality or transgender status. But just as labels and additional intentions or motivations didn’t make a difference in Manhart or Phillips, they cannot make a difference here. When an employer fires an employee for being homosexual or transgender, it necessarily and intentionally discriminates against that individual in part because of sex. And that is all Title VII has ever demanded to establish liability. Second, the plaintiff ’s sex need not be the sole or primary cause of the employer’s adverse action. In Phillips, Manhart, and Oncale, the defendant easily could have pointed to some other, nonprotected trait and insisted it was the more important factor in the adverse employment outcome. So, too, it has no significance here if another factor—such as the sex the plaintiff is attracted to or presents as—might also be at work, or even play a more important role in the employer’s decision. Finally, an employer cannot escape liability by demonstrating that it treats males and females comparably as groups. As Manhart teaches, an employer is liable for intentionally requiring an individual female employee to pay more into a pension plan than a male counterpart even if the scheme promotes equality at the group level. Likewise, an employer who intentionally fires an individual homosexual or transgender employee in part because of that individual’s sex violates the law even if the employer is willing to subject all male and female homosexual or transgender employees to the same rule. III What do the employers have to say in reply? For present purposes, they do not dispute that they fired the plaintiffs for being homosexual or transgender. Sorting out the true reasons for an adverse employment decision is often a hard business, but none of that is at issue here. Rather, the employers submit that even intentional discrimination against employees based on their homosexuality or transgender status supplies no basis for liability under Title VII. The employers’ argument proceeds in two stages. Seeking footing in the statutory text, they begin by advancing a number of reasons why discrimination on the basis of homosexuality or transgender status doesn’t involve discrimination because of sex. But each of these arguments turns out only to repackage errors we’ve already seen and this Court’s precedents have already rejected. In the end, the employers are left to retreat beyond the statute’s text, where they fault us for ignoring the legislature’s purposes in enacting Title VII or certain expectations about its operation. They warn, too, about consequences that might follow a ruling for the employees. But none of these contentions about what the employers think the law was meant to do, or should do, allow us to ignore the law as it is. A Maybe most intuitively, the employers assert that discrimination on the basis of homosexuality and transgender status aren’t referred to as sex discrimination in ordinary conversation. If asked by a friend (rather than a judge) why they were fired, even today’s plaintiffs would likely respond that it was because they were gay or transgender, not because of sex. According to the employers, that conversational answer, not the statute’s strict terms, should guide our thinking and suffice to defeat any suggestion that the employees now before us were fired because of sex. Cf. post, at 3 (Alito, J., dissenting); post, at 8–13 (Kavanaugh, J., dissenting). But this submission rests on a mistaken understanding of what kind of cause the law is looking for in a Title VII case. In conversation, a speaker is likely to focus on what seems most relevant or informative to the listener. So an employee who has just been fired is likely to identify the primary or most direct cause rather than list literally every but-for cause. To do otherwise would be tiring at best. But these conversational conventions do not control Title VII’s legal analysis, which asks simply whether sex was a but-for cause. In Phillips, for example, a woman who was not hired under the employer’s policy might have told her friends that her application was rejected because she was a mother, or because she had young children. Given that many women could be hired under the policy, it’s unlikely she would say she was not hired because she was a woman. But the Court did not hesitate to recognize that the employer in Phillips discriminated against the plaintiff because of her sex. Sex wasn’t the only factor, or maybe even the main factor, but it was one but-for cause—and that was enough. You can call the statute’s but-for causation test what you will—expansive, legalistic, the dissents even dismiss it as wooden or literal. But it is the law. Trying another angle, the defendants before us suggest that an employer who discriminates based on homosexuality or transgender status doesn’t intentionally discriminate based on sex, as a disparate treatment claim requires. See post, at 9–12 (Alito, J., dissenting); post, at 12–13 (Kavanaugh, J., dissenting). But, as we’ve seen, an employer who discriminates against homosexual or transgender employees necessarily and intentionally applies sex-based rules. An employer that announces it will not employ anyone who is homosexual, for example, intends to penalize male employees for being attracted to men and female employees for being attracted to women. What, then, do the employers mean when they insist intentional discrimination based on homosexuality or transgender status isn’t intentional discrimination based on sex? Maybe the employers mean they don’t intend to harm one sex or the other as a class. But as should be clear by now, the statute focuses on discrimination against individuals, not groups. Alternatively, the employers may mean that they don’t perceive themselves as motivated by a desire to discriminate based on sex. But nothing in Title VII turns on the employer’s labels or any further intentions (or motivations) for its conduct beyond sex discrimination. In Manhart, the employer intentionally required women to make higher pension contributions only to fulfill the further purpose of making things more equitable between men and women as groups. In Phillips, the employer may have perceived itself as discriminating based on motherhood, not sex, given that its hiring policies as a whole favored women. But in both cases, the Court set all this aside as irrelevant. The employers’ policies involved intentional discrimination because of sex, and Title VII liability necessarily followed. Aren’t these cases different, the employers ask, given that an employer could refuse to hire a gay or transgender individual without ever learning the applicant’s sex? Suppose an employer asked homosexual or transgender applicants to tick a box on its application form. The employer then had someone else redact any information that could be used to discern sex. The resulting applications would disclose which individuals are homosexual or transgender without revealing whether they also happen to be men or women. Doesn’t that possibility indicate that the employer’s discrimination against homosexual or transgender persons cannot be sex discrimination? No, it doesn’t. Even in this example, the individual applicant’s sex still weighs as a factor in the employer’s decision. Change the hypothetical ever so slightly and its flaws become apparent. Suppose an employer’s application form offered a single box to check if the applicant is either black or Catholic. If the employer refuses to hire anyone who checks that box, would we conclude the employer has complied with Title VII, so long as it studiously avoids learning any particular applicant’s race or religion? Of course not: By intentionally setting out a rule that makes hiring turn on race or religion, the employer violates the law, whatever he might know or not know about individual applicants. The same holds here. There is no way for an applicant to decide whether to check the homosexual or transgender box without considering sex. To see why, imagine an applicant doesn’t know what the words homosexual or transgender mean. Then try writing out instructions for who should check the box without using the words man, woman, or sex (or some synonym). It can’t be done. Likewise, there is no way an employer can discriminate against those who check the homosexual or transgender box without discriminating in part because of an applicant’s sex. By discriminating against homosexuals, the employer intentionally penalizes men for being attracted to men and women for being attracted to women. By discriminating against transgender persons, the employer unavoidably discriminates against persons with one sex identified at birth and another today. Any way you slice it, the employer intentionally refuses to hire applicants in part because of the affected individuals’ sex, even if it never learns any applicant’s sex. Next, the employers turn to Title VII’s list of protected characteristics—race, color, religion, sex, and national origin. Because homosexuality and transgender status can’t be found on that list and because they are conceptually distinct from sex, the employers reason, they are implicitly excluded from Title VII’s reach. Put another way, if Congress had wanted to address these matters in Title VII, it would have referenced them specifically. Cf. post, at 7–8 (Alito, J., dissenting); post, at 13–15 (Kavanaugh, J., dissenting). But that much does not follow. We agree that homosexuality and transgender status are distinct concepts from sex. But, as we’ve seen, discrimination based on homosexuality or transgender status necessarily entails discrimination based on sex; the first cannot happen without the second. Nor is there any such thing as a “canon of donut holes,” in which Congress’s failure to speak directly to a specific case that falls within a more general statutory rule creates a tacit exception. Instead, when Congress chooses not to include any exceptions to a broad rule, courts apply the broad rule. And that is exactly how this Court has always approached Title VII. “Sexual harassment” is conceptually distinct from sex discrimination, but it can fall within Title VII’s sweep. Oncale, 523 U. S., at 79–80. Same with “motherhood discrimination.” See Phillips, 400 U. S., at 544. Would the employers have us reverse those cases on the theory that Congress could have spoken to those problems more specifically? Of course not. As enacted, Title VII prohibits all forms of discrimination because of sex, however they may manifest themselves or whatever other labels might attach to them. The employers try the same point another way. Since 1964, they observe, Congress has considered several proposals to add sexual orientation to Title VII’s list of protected characteristics, but no such amendment has become law. Meanwhile, Congress has enacted other statutes addressing other topics that do discuss sexual orientation. This postenactment legislative history, they urge, should tell us something. Cf. post, at 2, 42–43 (Alito, J., dissenting); post, at 4, 15–16 (Kavanaugh, J., dissenting). But what? There’s no authoritative evidence explaining why later Congresses adopted other laws referencing sexual orientation but didn’t amend this one. Maybe some in the later legislatures understood the impact Title VII’s broad language already promised for cases like ours and didn’t think a revision needed. Maybe others knew about its impact but hoped no one else would notice. Maybe still others, occupied by other concerns, didn’t consider the issue at all. All we can know for certain is that speculation about why a later Congress declined to adopt new legislation offers a “particularly dangerous” basis on which to rest an interpretation of an existing law a different and earlier Congress did adopt. Pension Benefit Guaranty Corporation v. LTV Corp., 496 U.S. 633, 650 (1990); see also United States v. Wells, 519 U.S. 482, 496 (1997); Sullivan v. Finkelstein, 496 U.S. 617, 632 (1990) (Scalia, J., concurring) (“Arguments based on subsequent legislative history . . . should not be taken seriously, not even in a footnote”). That leaves the employers to seek a different sort of exception. Maybe the traditional and simple but-for causation test should apply in all other Title VII cases, but it just doesn’t work when it comes to cases involving homosexual and transgender employees. The test is too blunt to capture the nuances here. The employers illustrate their concern with an example. When we apply the simple test to Mr. Bostock—asking whether Mr. Bostock, a man attracted to other men, would have been fired had he been a woman—we don’t just change his sex. Along the way, we change his sexual orientation too (from homosexual to heterosexual). If the aim is to isolate whether a plaintiff ’s sex caused the dismissal, the employers stress, we must hold sexual orientation constant—meaning we need to change both his sex and the sex to which he is attracted. So for Mr. Bostock, the question should be whether he would’ve been fired if he were a woman attracted to women. And because his employer would have been as quick to fire a lesbian as it was a gay man, the employers conclude, no Title VII violation has occurred. While the explanation is new, the mistakes are the same. The employers might be onto something if Title VII only ensured equal treatment between groups of men and women or if the statute applied only when sex is the sole or primary reason for an employer’s challenged adverse employment action. But both of these premises are mistaken. Title VII’s plain terms and our precedents don’t care if an employer treats men and women comparably as groups; an employer who fires both lesbians and gay men equally doesn’t diminish but doubles its liability. Just cast a glance back to Manhart, where it was no defense that the employer sought to equalize pension contributions based on life expectancy. Nor does the statute care if other factors besides sex contribute to an employer’s discharge decision. Mr. Bostock’s employer might have decided to fire him only because of the confluence of two factors, his sex and the sex to which he is attracted. But exactly the same might have been said in Phillips, where motherhood was the added variable. Still, the employers insist, something seems different here. Unlike certain other employment policies this Court has addressed that harmed only women or only men, the employers’ policies in the cases before us have the same adverse consequences for men and women. How could sex be necessary to the result if a member of the opposite sex might face the same outcome from the same policy? What the employers see as unique isn’t even unusual. Often in life and law two but-for factors combine to yield a result that could have also occurred in some other way. Imagine that it’s a nice day outside and your house is too warm, so you decide to open the window. Both the cool temperature outside and the heat inside are but-for causes of your choice to open the window. That doesn’t change just because you also would have opened the window had it been warm outside and cold inside. In either case, no one would deny that the window is open “because of ” the outside temperature. Our cases are much the same. So, for example, when it comes to homosexual employees, male sex and attraction to men are but-for factors that can combine to get them fired. The fact that female sex and attraction to women can also get an employee fired does no more than show the same outcome can be achieved through the combination of different factors. In either case, though, sex plays an essential but-for role. At bottom, the employers’ argument unavoidably comes down to a suggestion that sex must be the sole or primary cause of an adverse employment action for Title VII liability to follow. And, as we’ve seen, that suggestion is at odds with everything we know about the statute. Consider an employer eager to revive the workplace gender roles of the 1950s. He enforces a policy that he will hire only men as mechanics and only women as secretaries. When a qualified woman applies for a mechanic position and is denied, the “simple test” immediately spots the discrimination: A qualified man would have been given the job, so sex was a but-for cause of the employer’s refusal to hire. But like the employers before us today, this employer would say not so fast. By comparing the woman who applied to be a mechanic to a man who applied to be a mechanic, we’ve quietly changed two things: the applicant’s sex and her trait of failing to conform to 1950s gender roles. The “simple test” thus overlooks that it is really the applicant’s bucking of 1950s gender roles, not her sex, doing the work. So we need to hold that second trait constant: Instead of comparing the disappointed female applicant to a man who applied for the same position, the employer would say, we should compare her to a man who applied to be a secretary. And because that jobseeker would be refused too, this must not be sex discrimination. No one thinks that, so the employers must scramble to justify deploying a stricter causation test for use only in cases involving discrimination based on sexual orientation or transgender status. Such a rule would create a curious discontinuity in our case law, to put it mildly. Employer hires based on sexual stereotypes? Simple test. Employer sets pension contributions based on sex? Simple test. Employer fires men who do not behave in a sufficiently masculine way around the office? Simple test. But when that same employer discriminates against women who are attracted to women, or persons identified at birth as women who later identify as men, we suddenly roll out a new and more rigorous standard? Why are these reasons for taking sex into account different from all the rest? Title VII’s text can offer no answer. B Ultimately, the employers are forced to abandon the statutory text and precedent altogether and appeal to assumptions and policy. Most pointedly, they contend that few in 1964 would have expected Title VII to apply to discrimination against homosexual and transgender persons. And whatever the text and our precedent indicate, they say, shouldn’t this fact cause us to pause before recognizing liability? It might be tempting to reject this argument out of hand. This Court has explained many times over many years that, when the meaning of the statute’s terms is plain, our job is at an end. The people are entitled to rely on the law as written, without fearing that courts might disregard its plain terms based on some extratextual consideration. See, e.g., Carcieri v. Salazar, 555 U.S. 379, 387 (2009); Connecticut Nat. Bank v. Germain, 503 U.S. 249, 253–254 (1992); Rubin v. United States, 449 U.S. 424, 430 (1981). Of course, some Members of this Court have consulted legislative history when interpreting ambiguous statutory language. Cf. post, at 40 (Alito, J., dissenting). But that has no bearing here. “Legislative history, for those who take it into account, is meant to clear up ambiguity, not create it.” Milner v. Department of Navy, 562 U.S. 562, 574 (2011). And as we have seen, no ambiguity exists about how Title VII’s terms apply to the facts before us. To be sure, the statute’s application in these cases reaches “beyond the principal evil” legislators may have intended or expected to address. Oncale, 523 U. S., at 79. But “ ‘the fact that [a statute] has been applied in situations not expressly anticipated by Congress’ ” does not demonstrate ambiguity; instead, it simply “ ‘demonstrates [the] breadth’ ” of a legislative command. Sedima, S. P. R. L. v. Imrex Co., 473 U.S. 479, 499 (1985). And “it is ultimately the provisions of ” those legislative commands “rather than the principal concerns of our legislators by which we are governed.” Oncale, 523 U. S., at 79; see also A. Scalia & B. Garner, Reading Law: The Interpretation of Legal Texts 101 (2012) (noting that unexpected applications of broad language reflect only Congress’s “presumed point [to] produce general coverage—not to leave room for courts to recognize ad hoc exceptions”). Still, while legislative history can never defeat unambiguous statutory text, historical sources can be useful for a different purpose: Because the law’s ordinary meaning at the time of enactment usually governs, we must be sensitive to the possibility a statutory term that means one thing today or in one context might have meant something else at the time of its adoption or might mean something different in another context. And we must be attuned to the possibility that a statutory phrase ordinarily bears a different meaning than the terms do when viewed individually or literally. To ferret out such shifts in linguistic usage or subtle distinctions between literal and ordinary meaning, this Court has sometimes consulted the understandings of the law’s drafters as some (not always conclusive) evidence. For example, in the context of the National Motor Vehicle Theft Act, this Court admitted that the term “vehicle” in 1931 could literally mean “a conveyance working on land, water or air.” McBoyle v. United States, 283 U.S. 25, 26 (1931). But given contextual clues and “everyday speech” at the time of the Act’s adoption in 1919, this Court concluded that “vehicles” in that statute included only things “moving on land,” not airplanes too. Ibid. Similarly, in New Prime, we held that, while the term “contracts of employment” today might seem to encompass only contracts with employees, at the time of the statute’s adoption the phrase was ordinarily understood to cover contracts with independent contractors as well. 586 U. S., at ___–___ (slip op., at 6–9). Cf. post, at 7–8 (Kavanaugh, J., dissenting) (providing additional examples). The employers, however, advocate nothing like that here. They do not seek to use historical sources to illustrate that the meaning of any of Title VII’s language has changed since 1964 or that the statute’s terms, whether viewed individually or as a whole, ordinarily carried some message we have missed. To the contrary, as we have seen, the employers agree with our understanding of all the statutory language—“discriminate against any individual . . . because of such individual’s . . . sex.” Nor do the competing dissents offer an alternative account about what these terms mean either when viewed individually or in the aggregate. Rather than suggesting that the statutory language bears some other meaning, the employers and dissents merely suggest that, because few in 1964 expected today’s result, we should not dare to admit that it follows ineluctably from the statutory text. When a new application emerges that is both unexpected and important, they would seemingly have us merely point out the question, refer the subject back to Congress, and decline to enforce the plain terms of the law in the meantime. That is exactly the sort of reasoning this Court has long rejected. Admittedly, the employers take pains to couch their argument in terms of seeking to honor the statute’s “expected applications” rather than vindicate its “legislative intent.” But the concepts are closely related. One could easily contend that legislators only intended expected applications or that a statute’s purpose is limited to achieving applications foreseen at the time of enactment. However framed, the employer’s logic impermissibly seeks to displace the plain meaning of the law in favor of something lying beyond it. If anything, the employers’ new framing may only add new problems. The employers assert that “no one” in 1964 or for some time after would have anticipated today’s result. But is that really true? Not long after the law’s passage, gay and transgender employees began filing Title VII complaints, so at least some people foresaw this potential application. See, e.g., Smith v. Liberty Mut. Ins. Co., 395 F. Supp. 1098, 1099 (ND Ga. 1975) (addressing claim from 1969); Holloway v. Arthur Andersen & Co., 566 F.2d 659, 661 (CA9 1977) (addressing claim from 1974). And less than a decade after Title VII’s passage, during debates over the Equal Rights Amendment, others counseled that its language—which was strikingly similar to Title VII’s—might also protect homosexuals from discrimination. See, e.g., Note, The Legality of Homosexual Marriage, 82 Yale L. J. 573, 583–584 (1973). Why isn’t that enough to demonstrate that today’s result isn’t totally unexpected? How many people have to foresee the application for it to qualify as “expected”? Do we look only at the moment the statute was enacted, or do we allow some time for the implications of a new statute to be worked out? Should we consider the expectations of those who had no reason to give a particular application any thought or only those with reason to think about the question? How do we account for those who change their minds over time, after learning new facts or hearing a new argument? How specifically or generally should we frame the “application” at issue? None of these questions have obvious answers, and the employers don’t propose any. One could also reasonably fear that objections about unexpected applications will not be deployed neutrally. Often lurking just behind such objections resides a cynicism that Congress could not possibly have meant to protect a disfavored group. Take this Court’s encounter with the Americans with Disabilities Act’s directive that no “ ‘public entity’ ” can discriminate against any “ ‘qualified individual with a disability.’ ” Pennsylvania Dept. of Corrections v. Yeskey, 524 U.S. 206, 208 (1998). Congress, of course, didn’t list every public entity the statute would apply to. And no one batted an eye at its application to, say, post offices. But when the statute was applied to prisons, curiously, some demanded a closer look: Pennsylvania argued that “Congress did not ‘envisio[n] that the ADA would be applied to state prisoners.’ ” Id., at 211–212. This Court emphatically rejected that view, explaining that, “in the context of an unambiguous statutory text,” whether a specific application was anticipated by Congress “is irrelevant.” Id., at 212. As Yeskey and today’s cases exemplify, applying protective laws to groups that were politically unpopular at the time of the law’s passage—whether prisoners in the 1990s or homosexual and transgender employees in the 1960s—often may be seen as unexpected. But to refuse enforcement just because of that, because the parties before us happened to be unpopular at the time of the law’s passage, would not only require us to abandon our role as interpreters of statutes; it would tilt the scales of justice in favor of the strong or popular and neglect the promise that all persons are entitled to the benefit of the law’s terms. Cf. post, at 28–35 (Alito, J., dissenting); post, at 21–22 (Kavanaugh, J., dissenting). The employer’s position also proves too much. If we applied Title VII’s plain text only to applications some (yet-to-be-determined) group expected in 1964, we’d have more than a little law to overturn. Start with Oncale. How many people in 1964 could have expected that the law would turn out to protect male employees? Let alone to protect them from harassment by other male employees? As we acknowledged at the time, “male-on-male sexual harassment in the workplace was assuredly not the principal evil Congress was concerned with when it enacted Title VII.” 523 U. S., at 79. Yet the Court did not hesitate to recognize that Title VII’s plain terms forbade it. Under the employer’s logic, it would seem this was a mistake. That’s just the beginning of the law we would have to unravel. As one Equal Employment Opportunity Commission (EEOC) Commissioner observed shortly after the law’s passage, the words of “ ‘the sex provision of Title VII [are] difficult to . . . control.’ ” Franklin, Inventing the “Traditional Concept” of Sex Discrimination, 125 Harv. L. Rev. 1307, 1338 (2012) (quoting Federal Mediation Service To Play Role in Implementing Title VII, [1965–1968 Transfer Binder] CCH Employment Practices ¶8046, p. 6074). The “difficult[y]” may owe something to the initial proponent of the sex discrimination rule in Title VII, Representative Howard Smith. On some accounts, the congressman may have wanted (or at least was indifferent to the possibility of ) broad language with wide-ranging effect. Not necessarily because he was interested in rooting out sex discrimination in all its forms, but because he may have hoped to scuttle the whole Civil Rights Act and thought that adding language covering sex discrimination would serve as a poison pill. See C. Whalen & B. Whalen, The Longest Debate: A Legislative History of the 1964 Civil Rights Act 115–118 (1985). Certainly nothing in the meager legislative history of this provision suggests it was meant to be read narrowly. Whatever his reasons,thanks to the broad language Representative Smith introduced, many, maybe most, applications of Title VII’s sex provision were “unanticipated” at the time of the law’s adoption. In fact, many now-obvious applications met with heated opposition early on, even among those tasked with enforcing the law. In the years immediately following Title VII’s passage, the EEOC officially opined that listing men’s positions and women’s positions separately in job postings was simply helpful rather than discriminatory. Franklin, 125 Harv. L. Rev., at 1340 (citing Press Release, EEOC (Sept. 22, 1965)). Some courts held that Title VII did not prevent an employer from firing an employee for refusing his sexual advances. See, e.g., Barnes v. Train, 1974 WL 10628, *1 (D DC, Aug. 9, 1974). And courts held that a policy against hiring mothers but not fathers of young children wasn’t discrimination because of sex. See Phillips v. Martin Marietta Corp., 411 F.2d 1 (CA5 1969), rev’d, 400 U.S. 542 (1971) (per curiam). Over time, though, the breadth of the statutory language proved too difficult to deny. By the end of the 1960s, the EEOC reversed its stance on sex-segregated job advertising. See Franklin, 125 Harv. L. Rev., at 1345. In 1971, this Court held that treating women with children differently from men with children violated Title VII. Phillips, 400 U. S., at 544. And by the late 1970s, courts began to recognize that sexual harassment can sometimes amount to sex discrimination. See, e.g., Barnes v. Costle, 561 F.2d 983, 990 (CADC 1977). While to the modern eye each of these examples may seem “plainly [to] constitut[e] discrimination because of biological sex,” post, at 38 (Alito, J., dissenting), all were hotly contested for years following Title VII’s enactment. And as with the discrimination we consider today, many federal judges long accepted interpretations of Title VII that excluded these situations. Cf. post, at 21–22 (Kavanaugh, J., dissenting) (highlighting that certain lower courts have rejected Title VII claims based on homosexuality and transgender status). Would the employers have us undo every one of these unexpected applications too? The weighty implications of the employers’ argument from expectations also reveal why they cannot hide behind the no-elephants-in-mouseholes canon. That canon recognizes that Congress “does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisions.” Whitman v. American Trucking Assns., Inc., 531 U.S. 457, 468 (2001). But it has no relevance here. We can’t deny that today’s holding—that employers are prohibited from firing employees on the basis of homosexuality or transgender status—is an elephant. But where’s the mousehole? Title VII’s prohibition of sex discrimination in employment is a major piece of federal civil rights legislation. It is written in starkly broad terms. It has repeatedly produced unexpected applications, at least in the view of those on the receiving end of them. Congress’s key drafting choices—to focus on discrimination against individuals and not merely between groups and to hold employers liable whenever sex is a but-for cause of the plaintiff ’s injuries—virtually guaranteed that unexpected applications would emerge over time. This elephant has never hidden in a mousehole; it has been standing before us all along. With that, the employers are left to abandon their concern for expected applications and fall back to the last line of defense for all failing statutory interpretation arguments: naked policy appeals. If we were to apply the statute’s plain language, they complain, any number of undesirable policy consequences would follow. Cf. post, at 44–54 (Alito, J., dissenting). Gone here is any pretense of statutory interpretation; all that’s left is a suggestion we should proceed without the law’s guidance to do as we think best. But that’s an invitation no court should ever take up. The place to make new legislation, or address unwanted consequences of old legislation, lies in Congress. When it comes to statutory interpretation, our role is limited to applying the law’s demands as faithfully as we can in the cases that come before us. As judges we possess no special expertise or authority to declare for ourselves what a self-governing people should consider just or wise. And the same judicial humility that requires us to refrain from adding to statutes requires us to refrain from diminishing them. What are these consequences anyway? The employers worry that our decision will sweep beyond Title VII to other federal or state laws that prohibit sex discrimination. And, under Title VII itself, they say sex-segregated bathrooms, locker rooms, and dress codes will prove unsustainable after our decision today. But none of these other laws are before us; we have not had the benefit of adversarial testing about the meaning of their terms, and we do not prejudge any such question today. Under Title VII, too, we do not purport to address bathrooms, locker rooms, or anything else of the kind. The only question before us is whether an employer who fires someone simply for being homosexual or transgender has discharged or otherwise discriminated against that individual “because of such individual’s sex.” As used in Title VII, the term “ ‘discriminate against’ ” refers to “distinctions or differences in treatment that injure protected individuals.” Burlington N. & S. F. R., 548 U. S., at 59. Firing employees because of a statutorily protected trait surely counts. Whether other policies and practices might or might not qualify as unlawful discrimination or find justifications under other provisions of Title VII are questions for future cases, not these. Separately, the employers fear that complying with Title VII’s requirement in cases like ours may require some employers to violate their religious convictions. We are also deeply concerned with preserving the promise of the free exercise of religion enshrined in our Constitution; that guarantee lies at the heart of our pluralistic society. But worries about how Title VII may intersect with religious liberties are nothing new; they even predate the statute’s passage. As a result of its deliberations in adopting the law, Congress included an express statutory exception for religious organizations. §2000e–1(a). This Court has also recognized that the First Amendment can bar the application of employment discrimination laws “to claims concerning the employment relationship between a religious institution and its ministers.” Hosanna-Tabor Evangelical Lutheran Church and School v. EEOC, 565 U.S. 171, 188 (2012). And Congress has gone a step further yet in the Religious Freedom Restoration Act of 1993 (RFRA), 107Stat. 1488, codified at 42 U. S. C. §2000bb et seq. That statute prohibits the federal government from substantially burdening a person’s exercise of religion unless it demonstrates that doing so both furthers a compelling governmental interest and represents the least restrictive means of furthering that interest. §2000bb–1. Because RFRA operates as a kind of super statute, displacing the normal operation of other federal laws, it might supersede Title VII’s commands in appropriate cases. See §2000bb–3. But how these doctrines protecting religious liberty interact with Title VII are questions for future cases too. Harris Funeral Homes did unsuccessfully pursue a RFRA-based defense in the proceedings below. In its certiorari petition, however, the company declined to seek review of that adverse decision, and no other religious liberty claim is now before us. So while other employers in other cases may raise free exercise arguments that merit careful consideration, none of the employers before us today represent in this Court that compliance with Title VII will infringe their own religious liberties in any way. * Some of those who supported adding language to Title VII to ban sex discrimination may have hoped it would derail the entire Civil Rights Act. Yet, contrary to those intentions, the bill became law. Since then, Title VII’s effects have unfolded with far-reaching consequences, some likely beyond what many in Congress or elsewhere expected. But none of this helps decide today’s cases. Ours is a society of written laws. Judges are not free to overlook plain statutory commands on the strength of nothing more than suppositions about intentions or guesswork about expectations. In Title VII, Congress adopted broad language making it illegal for an employer to rely on an employee’s sex when deciding to fire that employee. We do not hesitate to recognize today a necessary consequence of that legislative choice: An employer who fires an individual merely for being gay or transgender defies the law. The judgments of the Second and Sixth Circuits in Nos. 17–1623 and 18–107 are affirmed. The judgment of the Eleventh Circuit in No. 17–1618 is reversed, and the case is remanded for further proceedings consistent with this opinion. It is so ordered.

591.US.2019_19-465

When Americans cast ballots for presidential candidates, their votes actually go toward selecting members of the Electoral College, whom each State appoints based on the popular returns. The States have devised mechanisms to ensure that the electors they appoint vote for the presidential candidate their citizens have preferred. With two partial exceptions, every State appoints a slate of electors selected by the political party whose candidate has won the State’s popular vote. Most States also compel electors to pledge to support the nominee of that party. Relevant here, 15 States back up their pledge laws with some kind of sanction. Almost all of these States immediately remove a so-called “faithless elector” from his position, substituting an alternate whose vote the State reports instead. A few States impose a monetary fine on any elector who flouts his pledge. Three Washington electors, Peter Chiafalo, Levi Guerra, and Esther John (the Electors), violated their pledges to support Hillary Clinton in the 2016 presidential election. In response, the State fined the Electors $1,000 apiece for breaking their pledges to support the same candidate its voters had. The Electors challenged their fines in state court, arguing that the Constitution gives members of the Electoral College the right to vote however they please. The Washington Superior Court rejected that claim, and the State Supreme Court affirmed, relying on Ray v. Blair, 343 U.S. 214. In Ray, this Court upheld a pledge requirement—though one without a penalty to back it up. Ray held that pledges were consistent with the Constitution’s text and our Nation’s history, id., at 225–230; but it reserved the question whether a State can enforce that requirement through legal sanctions. Held: A State may enforce an elector’s pledge to support his party’s nominee—and the state voters’ choice—for President. Pp. 8–18. (a) Article II, §1 gives the States the authority to appoint electors “in such Manner as the Legislature thereof may direct.” This Court has described that clause as “conveying the broadest power of determination” over who becomes an elector. McPherson v. Blacker, 146 U.S. 1, 27. And the power to appoint an elector (in any manner) includes power to condition his appointment, absent some other constitutional constraint. A State can require, for example, that an elector live in the State or qualify as a regular voter during the relevant time period. Or more substantively, a State can insist (as Ray allowed) that the elector pledge to cast his Electoral College ballot for his party’s presidential nominee, thus tracking the State’s popular vote. Or—so long as nothing else in the Constitution poses an obstacle—a State can add an associated condition of appointment: It can demand that the elector actually live up to his pledge, on pain of penalty. Which is to say that the State’s appointment power, barring some outside constraint, enables the enforcement of a pledge like Washington’s. Nothing in the Constitution expressly prohibits States from taking away presidential electors’ voting discretion as Washington does. Article II includes only the instruction to each State to appoint electors, and the Twelfth Amendment only sets out the electors’ voting procedures. And while two contemporaneous State Constitutions incorporated language calling for the exercise of elector discretion, no language of that kind made it into the Federal Constitution. Contrary to the Electors’ argument, Article II’s use of the term “electors” and the Twelfth Amendment’s requirement that the electors “vote,” and that they do so “by ballot,” do not establish that electors must have discretion. The Electors and their amici object that the Framers using those words expected the Electors’ votes to reflect their own judgments. But even assuming that outlook was widely shared, it would not be enough. Whether by choice or accident, the Framers did not reduce their thoughts about electors’ discretion to the printed page. Pp. 8–13. (b) “Long settled and established practice” may have “great weight in a proper interpretation of constitutional provisions.” The Pocket Veto Case, 279 U.S. 655, 689. The Electors make an appeal to that kind of practice in asserting their right to independence, but “our whole experience as a Nation” points in the opposite direction. NLRB v. Noel Canning, 573 U.S. 513, 557. From the first elections under the Constitution, States sent electors to the College to vote for pre-selected candidates, rather than to use their own judgment. The electors rapidly settled into that non-discretionary role. See Ray, 343 U. S., at 228–229. Ratified at the start of the 19th century, the Twelfth Amendment both acknowledged and facilitated the Electoral College’s emergence as a mechanism not for deliberation but for party-line voting. Courts and commentators throughout that century recognized the presidential electors as merely acting on other people’s preferences. And state election laws evolved to reinforce that development, ensuring that a State’s electors would vote the same way as its citizens. Washington’s law is only another in the same vein. It reflects a longstanding tradition in which electors are not free agents; they are to vote for the candidate whom the State’s voters have chosen. Pp. 13–17. 193 Wash. 2d 380, 441 P.3d 807, affirmed. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Alito, Sotomayor, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed an opinion concurring in the judgment, in which Gorsuch, J., joined as to Part II.

Every four years, millions of Americans cast a ballot for a presidential candidate. Their votes, though, actually go toward selecting members of the Electoral College, whom each State appoints based on the popular returns. Those few “electors” then choose the President. The States have devised mechanisms to ensure that the electors they appoint vote for the presidential candidate their citizens have preferred. With two partial exceptions, every State appoints a slate of electors selected by the political party whose candidate has won the State’s popular vote. Most States also compel electors to pledge in advance to support the nominee of that party. This Court upheld such a pledge requirement decades ago, rejecting the argument that the Constitution “demands absolute freedom for the elector to vote his own choice.” Ray v. Blair, 343 U.S. 214, 228 (1952). Today, we consider whether a State may also penalize an elector for breaking his pledge and voting for someone other than the presidential candidate who won his State’s popular vote. We hold that a State may do so. I Our Constitution’s method of picking Presidents emerged from an eleventh-hour compromise. The issue, one delegate to the Convention remarked, was “the most difficult of all [that] we have had to decide.” 2 Records of the Federal Convention of 1787, p. 501 (M. Farrand rev. 1966) (Farrand). Despite long debate and many votes, the delegates could not reach an agreement. See generally N. Peirce & L. Longley, The People’s President 19–22 (rev. 1981). In the dying days of summer, they referred the matter to the so-called Committee of Eleven to devise a solution. The Committee returned with a proposal for the Electoral College. Just two days later, the delegates accepted the recommendation with but a few tweaks. James Madison later wrote to a friend that the “difficulty of finding an unexceptionable [selection] process” was “deeply felt by the Convention.” Letter to G. Hay (Aug. 23, 1823), in 3 Farrand 458. Because “the final arrangement of it took place in the latter stage of the Session,” Madison continued, “it was not exempt from a degree of the hurrying influence produced by fatigue and impatience in all such Bodies: tho’ the degree was much less than usually prevails in them.” Ibid. Whether less or not, the delegates soon finished their work and departed for home. The provision they approved about presidential electors is fairly slim. Article II, §1, cl. 2 says: “Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress: but no Senator or Representative, or Person holding an Office of Trust or Profit under the United States, shall be appointed an Elector.” The next clause (but don’t get attached: it will soon be superseded) set out the procedures the electors were to follow in casting their votes. In brief, each member of the College would cast votes for two candidates in the presidential field. The candidate with the greatest number of votes, assuming he had a majority, would become President. The runner-up would become Vice President. If no one had a majority, the House of Representatives would take over and decide the winner. That plan failed to anticipate the rise of political parties, and soon proved unworkable. The Nation’s first contested presidential election occurred in 1796, after George Washington’s retirement. John Adams came in first among the candidates, and Thomas Jefferson second. That meant the leaders of the era’s two warring political parties—the Federalists and the Republicans—became President and Vice President respectively. (One might think of this as fodder for a new season of Veep.) Four years later, a different problem arose. Jefferson and Aaron Burr ran that year as a Republican Party ticket, with the former meant to be President and the latter meant to be Vice. For that plan to succeed, Jefferson had to come in first and Burr just behind him. Instead, Jefferson came in first and Burr . . . did too. Every elector who voted for Jefferson also voted for Burr, producing a tie. That threw the election into the House of Representatives, which took no fewer than 36 ballots to elect Jefferson. (Alexander Hamilton secured his place on the Broadway stage—but possibly in the cemetery too—by lobbying Federalists in the House to tip the election to Jefferson, whom he loathed but viewed as less of an existential threat to the Republic.) By then, everyone had had enough of the Electoral College’s original voting rules. The result was the Twelfth Amendment, whose main part provided that electors would vote separately for President and Vice President. The Amendment, ratified in 1804, says: “The Electors shall meet in their respective states and vote by ballot for President and Vice-President . . .; they shall name in their ballots the person voted for as President, and in distinct ballots the person voted for as Vice-President, and they shall make distinct lists of all persons voted for as President, and of all persons voted for as Vice-President, and of the number of votes for each, which lists they shall sign and certify, and transmit sealed to [Congress, where] the votes shall then be counted.” The Amendment thus brought the Electoral College’s voting procedures into line with the Nation’s new party system. Within a few decades, the party system also became the means of translating popular preferences within each State into Electoral College ballots. In the Nation’s earliest elections, state legislatures mostly picked the electors, with the majority party sending a delegation of its choice to the Electoral College. By 1832, though, all States but one had introduced popular presidential elections. See Peirce & Longley, The People’s President, at 45. At first, citizens voted for a slate of electors put forward by a political party, expecting that the winning slate would vote for its party’s presidential (and vice presidential) nominee in the Electoral College. By the early 20th century, citizens in most States voted for the presidential candidate himself; ballots increasingly did not even list the electors. See Albright, The Presidential Short Ballot, 34 Am. Pol. Sci. Rev. 955, 955–957 (1940). After the popular vote was counted, States appointed the electors chosen by the party whose presidential nominee had won statewide, again expecting that they would vote for that candidate in the Electoral College.[1] In the 20th century, many States enacted statutes meant to guarantee that outcome—that is, to prohibit so-called faithless voting. Rather than just assume that party-picked electors would vote for their party’s winning nominee, those States insist that they do so. As of now, 32 States and the District of Columbia have such statutes on their books. They are typically called pledge laws because most demand that electors take a formal oath or pledge to cast their ballot for their party’s presidential (and vice presidential) candidate. Others merely impose that duty by law. Either way, the statutes work to ensure that the electors vote for the candidate who got the most statewide votes in the presidential election. Most relevant here, States began about 60 years ago to back up their pledge laws with some kind of sanction. By now, 15 States have such a system.[2] Almost all of them immediately remove a faithless elector from his position, substituting an alternate whose vote the State reports instead. A few States impose a monetary fine on any elector who flouts his pledge. Washington is one of the 15 States with a sanctions-backed pledge law designed to keep the State’s electors in line with its voting citizens. As all States now do, Washington requires political parties fielding presidential candidates to nominate a slate of electors. See Wash. Rev. Code §29 A. 56.320(1). On Election Day, the State gives voters a ballot listing only the candidates themselves. See §29 A. 56.320(2). When the vote comes in, Washington moves toward appointing the electors chosen by the party whose candidate won the statewide count. See ibid. But before the appointment can go into effect, each elector must “execute [a] pledge” agreeing to “mark [her] ballots” for the presidential (and vice presidential) candidate of the party nominating her. §29 A. 56.084. And the elector must comply with that pledge, or else face a sanction. At the time relevant here, the punishment was a civil fine of up to $1,000. See §29 A. 56.340 (2016).[3] This case involves three Washington electors who violated their pledges in the 2016 presidential election. That year, Washington’s voters chose Hillary Clinton over Donald Trump for President. The State thus appointed as its electors the nominees of the Washington State Democratic Party. Among those Democratic electors were petitioners Peter Chiafalo, Levi Guerra, and Esther John (the Electors). All three pledged to support Hillary Clinton in the Electoral College. But as that vote approached, they decided to cast their ballots for someone else. The three hoped they could encourage other electors—particularly those from States Donald Trump had carried—to follow their example. The idea was to deprive him of a majority of electoral votes and throw the election into the House of Representatives. So the three Electors voted for Colin Powell for President. But their effort failed. Only seven electors across the Nation cast faithless votes—the most in a century, but well short of the goal. Candidate Trump became President Trump. And, more to the point here, the State fined the Electors $1,000 apiece for breaking their pledges to support the same candidate its voters had. The Electors challenged their fines in state court, arguing that the Constitution gives members of the Electoral College the right to vote however they please. The Washington Superior Court rejected the Electors’ claim in an oral decision, and the State’s Supreme Court affirmed that judgment. See In re Guerra, 193 Wash. 2d 380, 441 P.3d 807 (2019). The court relied heavily on our decision in Ray v. Blair upholding a pledge requirement—though one without a penalty to back it up. See 193 Wash. 2d, at 393–399, 441 P. 3d, at 813–816. In the state court’s view, Washington’s penalty provision made no difference. Article II of the Constitution, the court noted, grants broad authority to the States to appoint electors, and so to impose conditions on their appointments. See id., at 393, 395, 441 P. 3d, at 813, 814. And nothing in the document “suggests that electors have discretion to cast their votes without limitation or restriction by the state legislature.” Id., at 396, 441 P. 3d, at 814. A few months later, the United States Court of Appeals for the Tenth Circuit reached the opposite conclusion in a case involving another faithless elector. See Baca v. Colorado Dept. of State, 935 F.3d 887 (2019). The Circuit Court held that Colorado could not remove the elector, as its pledge law directs, because the Constitution “provide[s] presidential electors the right to cast a vote” for President “with discretion.” Id., at 955. We granted certiorari to resolve the split. 589 U. S. ___ (2020). We now affirm the Washington Supreme Court’s judgment that a State may enforce its pledge law against an elector. II As the state court recognized, this Court has considered elector pledge requirements before. Some seventy years ago Edmund Blair tried to become a presidential elector in Alabama. Like all States, Alabama lodged the authority to pick electors in the political parties fielding presidential candidates. And the Alabama Democratic Party required a pledge phrased much like Washington’s today. No one could get on the party’s slate of electors without agreeing to vote in the Electoral College for the Democratic presidential candidate. Blair challenged the pledge mandate. He argued that the “intention of the Founders was that [presidential] electors should exercise their judgment in voting.” Ray, 343 U. S., at 225. The pledge requirement, he claimed, “interfere[d] with the performance of this constitutional duty to select [a president] according to the best judgment of the elector.” Ibid. Our decision in Ray rejected that challenge. “Neither the language of Art. II, §1, nor that of the Twelfth Amendment,” we explained, prohibits a State from appointing only electors committed to vote for a party’s presidential candidate. Ibid. Nor did the Nation’s history suggest such a bar. To the contrary, “[h]istory teaches that the electors were expected to support the party nominees” as far back as the earliest contested presidential elections. Id., at 228. “[L]ongstanding practice” thus “weigh[ed] heavily” against Blair’s claim. Id., at 228–230. And current voting procedures did too. The Court noted that by then many States did not even put electors’ names on a presidential ballot. See id., at 229. The whole system presupposed that the electors, because of either an “implied” or an “oral pledge,” would vote for the candidate who had won the State’s popular election. Ibid. Ray, however, reserved a question not implicated in the case: Could a State enforce those pledges through legal sanctions? See id., at 230. Or would doing so violate an elector’s “constitutional freedom” to “vote as he may choose” in the Electoral College? Ibid. Today, we take up that question. We uphold Washington’s penalty-backed pledge law for reasons much like those given in Ray. The Constitution’s text and the Nation’s history both support allowing a State to enforce an elector’s pledge to support his party’s nominee—and the state voters’ choice—for President. A Article II, §1’s appointments power gives the States far-reaching authority over presidential electors, absent some other constitutional constraint.[4] As noted earlier, each State may appoint electors “in such Manner as the Legislature thereof may direct.” Art. II, §1, cl. 2; see supra, at 2. This Court has described that clause as “conveying the broadest power of determination” over who becomes an elector. McPherson v. Blacker, 146 U.S. 1, 27 (1892).[5] And the power to appoint an elector (in any manner) includes power to condition his appointment—that is, to say what the elector must do for the appointment to take effect. A State can require, for example, that an elector live in the State or qualify as a regular voter during the relevant time period. Or more substantively, a State can insist (as Ray allowed) that the elector pledge to cast his Electoral College ballot for his party’s presidential nominee, thus tracking the State’s popular vote. See Ray, 343 U. S., at 227 (A pledge requirement “is an exercise of the state’s right to appoint electors in such manner” as it chooses). Or—so long as nothing else in the Constitution poses an obstacle—a State can add, as Washington did, an associated condition of appointment: It can demand that the elector actually live up to his pledge, on pain of penalty. Which is to say that the State’s appointment power, barring some outside constraint, enables the enforcement of a pledge like Washington’s.[6] And nothing in the Constitution expressly prohibits States from taking away presidential electors’ voting discretion as Washington does. The Constitution is barebones about electors. Article II includes only the instruction to each State to appoint, in whatever way it likes, as many electors as it has Senators and Representatives (except that the State may not appoint members of the Federal Government). The Twelfth Amendment then tells electors to meet in their States, to vote for President and Vice President separately, and to transmit lists of all their votes to the President of the United States Senate for counting. Appointments and procedures and . . . that is all. See id., at 225. The Framers could have done it differently; other constitutional drafters of their time did. In the founding era, two States—Maryland and Kentucky—used electoral bodies selected by voters to choose state senators (and in Kentucky’s case, the Governor too). The Constitutions of both States, Maryland’s drafted just before and Kentucky’s just after the U. S. Constitution, incorporated language that would have made this case look quite different. Both state Constitutions required all electors to take an oath “to elect without favour, affection, partiality, or prejudice, such persons for Senators, as they, in their judgment and conscience, believe best qualified for the office.” Md. Declaration of Rights, Art. XVIII (1776); see Ky. Const., Art. I, §14 (1792) (using identical language except adding “[and] for Governor”). The emphasis on independent “judgment and conscience” called for the exercise of elector discretion. But although the Framers knew of Maryland’s Constitution, no language of that kind made it into the document they drafted. See 1 Farrand 218, 289 (showing that Madison and Hamilton referred to the Maryland system at the Convention). The Electors argue that three simple words stand in for more explicit language about discretion. Article II, §1 first names the members of the Electoral College: “electors.” The Twelfth Amendment then says that electors shall “vote” and that they shall do so by “ballot.” The “plain meaning” of those terms, the Electors say, requires electors to have “freedom of choice.” Brief for Petitioners 29, 31. If the States could control their votes, “the electors would not be ‘Electors,’ and their ‘vote by Ballot’ would not be a ‘vote.’ ” Id., at 31. But those words need not always connote independent choice. Suppose a person always votes in the way his spouse, or pastor, or union tells him to. We might question his judgment, but we would have no problem saying that he “votes” or fills in a “ballot.” In those cases, the choice is in someone else’s hands, but the words still apply because they can signify a mechanical act. Or similarly, suppose in a system allowing proxy voting (a common practice in the founding era), the proxy acts on clear instructions from the principal, with no freedom of choice. Still, we might well say that he cast a “ballot” or “voted,” though the preference registered was not his own. For that matter, some elections give the voter no real choice because there is only one name on a ballot (consider an old Soviet election, or even a down-ballot race in this country). Yet if the person in the voting booth goes through the motions, we consider him to have voted. The point of all these examples is to show that although voting and discretion are usually combined, voting is still voting when discretion departs. Maybe most telling, switch from hypotheticals to the members of the Electoral College. For centuries now, as we’ll later show, almost all have considered themselves bound to vote for their party’s (and the state voters’) preference. See infra, at 13–17. Yet there is no better description for what they do in the Electoral College than “vote” by “ballot.” And all these years later, everyone still calls them “electors”—and not wrongly, because even though they vote without discretion, they do indeed elect a President. The Electors and their amici object that the Framers using those words expected the Electors’ votes to reflect their own judgments. See Brief for Petitioners 18–19; Brief for Independence Institute as Amicus Curiae 11–15. Hamilton praised the Constitution for entrusting the Presidency to “men most capable of analyzing the qualities” needed for the office, who would make their choices “under circumstances favorable to deliberation.” The Federalist No. 68, p. 410 (C. Rossiter ed. 1961). So too, John Jay predicted that the Electoral College would “be composed of the most enlightened and respectable citizens,” whose choices would reflect “discretion and discernment.” Id., No. 64, at 389. But even assuming other Framers shared that outlook, it would not be enough. Whether by choice or accident, the Framers did not reduce their thoughts about electors’ discretion to the printed page. All that they put down about the electors was what we have said: that the States would appoint them, and that they would meet and cast ballots to send to the Capitol. Those sparse instructions took no position on how independent from—or how faithful to—party and popular preferences the electors’ votes should be. On that score, the Constitution left much to the future. And the future did not take long in coming. Almost immediately, presidential electors became trusty transmitters of other people’s decisions. B “Long settled and established practice” may have “great weight in a proper interpretation of constitutional provisions.” The Pocket Veto Case, 279 U.S. 655, 689 (1929). As James Madison wrote, “a regular course of practice” can “liquidate & settle the meaning of ” disputed or indeterminate “terms & phrases.” Letter to S. Roane (Sept. 2, 1819), in 8 Writings of James Madison 450 (G. Hunt ed. 1908); see The Federalist No. 37, at 225. The Electors make an appeal to that kind of practice in asserting their right to independence. But “our whole experience as a Nation” points in the opposite direction. NLRB v. Noel Canning, 573 U.S. 513, 557 (2014) (internal quotation marks omitted). Electors have only rarely exercised discretion in casting their ballots for President. From the first, States sent them to the Electoral College—as today Washington does—to vote for pre-selected candidates, rather than to use their own judgment. And electors (or at any rate, almost all of them) rapidly settled into that non-discretionary role. See Ray, 343 U. S., at 228–229. Begin at the beginning—with the Nation’s first contested election in 1796. Would-be electors declared themselves for one or the other party’s presidential candidate. (Recall that in this election Adams led the Federalists against Jefferson’s Republicans. See supra, at 3.) In some States, legislatures chose the electors; in others, ordinary voters did. But in either case, the elector’s declaration of support for a candidate—essentially a pledge—was what mattered. Or said differently, the selectors of an elector knew just what they were getting—not someone who would deliberate in good Hamiltonian fashion, but someone who would vote for their party’s candidate. “[T]he presidential electors,” one historian writes, “were understood to be instruments for expressing the will of those who selected them, not independent agents authorized to exercise their own judgment.” Whittington, Originalism, Constitutional Construction, and the Problem of Faithless Electors, 59 Ariz. L. Rev. 903, 911 (2017). And when the time came to vote in the Electoral College, all but one elector did what everyone expected, faithfully representing their selectors’ choice of presidential candidate.[7] The Twelfth Amendment embraced this new reality—both acknowledging and facilitating the Electoral College’s emergence as a mechanism not for deliberation but for party-line voting. Remember that the Amendment grew out of a pair of fiascos—the election of two then-bitter rivals as President and Vice President, and the tie vote that threw the next election into the House. See supra, at 3. Both had occurred because the Constitution’s original voting procedures gave electors two votes for President, rather than one apiece for President and Vice President. Without the capacity to vote a party ticket for the two offices, the electors had foundered, and could do so again. If the predominant party’s electors used both their votes on their party’s two candidates, they would create a tie (see 1800). If they intentionally cast fewer votes for the intended vice president, they risked the opposite party’s presidential candidate sneaking into the second position (see 1796). By allowing the electors to vote separately for the two offices, the Twelfth Amendment made party-line voting safe. The Amendment thus advanced, rather than resisted, the practice that had arisen in the Nation’s first elections. An elector would promise to legislators or citizens to vote for their party’s presidential and vice presidential candidates—and then follow through on that commitment. Or as the Court wrote in Ray, the new procedure allowed an elector to “vote the regular party ticket” and thereby “carry out the desires of the people” who had sent him to the Electoral College. Ray, 343 U. S., at 224, n. 11. No independent electors need apply. Courts and commentators throughout the 19th century recognized the electors as merely acting on other people’s preferences. Justice Story wrote that “the electors are now chosen wholly with reference to particular candidates,” having either “silently” or “publicly pledge[d]” how they will vote. 3 Commentaries on the Constitution of the United States §1457, p. 321 (1833). “[N]othing is left to the electors,” he continued, “but to register [their] votes, which are already pledged.” Id., at 321–322. Indeed, any “exercise of an independent judgment would be treated[ ] as a political usurpation, dishonourable to the individual, and a fraud upon his constituents.” Id., at 322. Similarly, William Rawle explained how the Electoral College functioned: “[T]he electors do not assemble in their several states for a free exercise of their own judgments, but for the purpose of electing” the nominee of “the predominant political party which has chosen those electors.” A View of the Constitution of the United States of America 57 (2d ed. 1829). Looking back at the close of the century, this Court had no doubt that Story’s and Rawle’s descriptions were right. The electors, the Court noted, were chosen “simply to register the will of the appointing power in respect of a particular candidate.” McPherson, 146 U. S., at 36. State election laws evolved to reinforce that development, ensuring that a State’s electors would vote the same way as its citizens. As noted earlier, state legislatures early dropped out of the picture; by the mid-1800s, ordinary voters chose electors. See supra, at 4. Except that increasingly, they did not do so directly. States listed only presidential candidates on the ballot, on the understanding that electors would do no more than vote for the winner. Usually, the State could ensure that result by appointing electors chosen by the winner’s party. But to remove any doubt, States began in the early 1900s to enact statutes requiring electors to pledge that they would squelch any urge to break ranks with voters. See supra, at 5. Washington’s law, penalizing a pledge’s breach, is only another in the same vein. It reflects a tradition more than two centuries old. In that practice, electors are not free agents; they are to vote for the candidate whom the State’s voters have chosen. The history going the opposite way is one of anomalies only. The Electors stress that since the founding, electors have cast some 180 faithless votes for either President or Vice President. See Brief for Petitioners 7. But that is 180 out of over 23,000. See Brief for Republican National Committee as Amicus Curiae 19. And more than a third of the faithless votes come from 1872, when the Democratic Party’s nominee (Horace Greeley) died just after Election Day.[8] Putting those aside, faithless votes represent just one-half of one percent of the total. Still, the Electors counter, Congress has counted all those votes. See Brief for Petitioners 46. But because faithless votes have never come close to affecting an outcome, only one has ever been challenged. True enough, that one was counted. But the Electors cannot rest a claim of historical tradition on one counted vote in over 200 years. And anyway, the State appointing that elector had no law requiring a pledge or otherwise barring his use of discretion. Congress’s deference to a state decision to tolerate a faithless vote is no ground for rejecting a state decision to penalize one. III The Electors’ constitutional claim has neither text nor history on its side. Article II and the Twelfth Amendment give States broad power over electors, and give electors themselves no rights. Early in our history, States decided to tie electors to the presidential choices of others, whether legislatures or citizens. Except that legislatures no longer play a role, that practice has continued for more than 200 years. Among the devices States have long used to achieve their object are pledge laws, designed to impress on electors their role as agents of others. A State follows in the same tradition if, like Washington, it chooses to sanction an elector for breaching his promise. Then too, the State instructs its electors that they have no ground for reversing the vote of millions of its citizens. That direction accords with the Constitution—as well as with the trust of a Nation that here, We the People rule. The judgment of the Supreme Court of Washington is Affirmed. Notes 1 Maine and Nebraska (which, for simplicity’s sake, we will ignore after this footnote) developed a more complicated system in which two electors go to the winner of the statewide vote and one goes to the winner of each congressional district. See Me. Rev. Stat. Ann., Tit. 21–A, §802 (2006); Neb. Rev. Stat. §32–710 (2016). So, for example, if the Republican candidate wins the popular vote in Nebraska as a whole but loses to the Democratic candidate in one of the State’s three congressional districts, the Republican will get four electors and the Democrat will get one. Here too, though, the States use party slates to pick the electors, in order to reflect the relevant popular preferences (whether in the State or in an individual district). 2 Ariz. Rev. Stat. Ann. §16–212 (2019 Cum. Supp.); Cal. Elec. Code Ann. §§6906, 18002 (West 2019); Colo. Rev. Stat. §1–4–304 (2019); Ind. Code §3–10–4–9 (2019); Mich. Comp. Laws §168.47 (2008); Minn. Stat. §§208.43, 208.46 (2020 Cum. Supp.); Mont. Code Ann. §§13–25–304, 13–25–307 (2019); Neb. Rev. Stat. §§32–713, 32–714; Nev. Rev. Stat. §§298.045, 298.075 (2017); N. M. Stat. Ann. §1–15–9 (Supp. 2011); N. C. Gen. Stat. Ann. §163–212 (2019); Okla. Stat., Tit. 26, §§10–102, 10–109 (2019); S. C. Code Ann. §7–19–80 (2018); Utah Code §20A–13–304 (2020); Wash. Rev. Code §§29 A. 56.084, 29 A. 56.090 (2019). 3 Since the events in this case, Washington has repealed the fine. It now enforces pledges only by removing and replacing faithless electors. See Wash. Rev. Code §29 A. 56.090(3) (2019). 4 Checks on a State’s power to appoint electors, or to impose conditions on an appointment, can theoretically come from anywhere in the Constitution. A State, for example, cannot select its electors in a way that violates the Equal Protection Clause. And if a State adopts a condition on its appointments that effectively imposes new requirements on presidential candidates, the condition may conflict with the Presidential Qualifications Clause, see Art. II, §1, cl. 5. 5 See also U. S. Term Limits, Inc. v. Thornton, 514 U.S. 779, 805 (1995) (describing Article II, §1 as an “express delegation[ ] of power to the States”); but see post, at 2 (Thomas, J., concurring in judgment) (continuing to press the view, taken in the Thornton dissent, that Article II, §1 grants the States no power at all). 6 The concurring opinion would have us make fine distinctions among state laws punishing faithless voting—treating some as conditions of appointment and others not, depending on small semantic differences. See post, at 6–9 (distinguishing, for example, between Oklahoma’s law fining an elector for violating his oath (to vote for his party’s candidate) and Washington’s law fining an elector for not voting for his party’s candidate (whom he took an oath to support)). The Electors themselves raised no such argument, and they were right not to do so. No matter the precise phrasing, a law penalizing faithless voting (like a law merely barring that practice) is an exercise of the State’s power to impose conditions on the appointment of electors. See Ray v. Blair, 343 U.S. 154, 227 (1952). 7 The reaction to even that single elector goes to prove the point that the system was non-discretionary. In the 1796 election, Pennsylvania held a statewide vote for electors under a winner-take-all rule (as all but two States have today). The people voted narrowly for the slate of electors supporting Jefferson. But Federalist chicanery led to the Governor’s inclusion of two Federalist electors in the State’s delegation to the Electoral College. One of them, Samuel Miles, agreed to cast his vote for Jefferson, in line with the winner-take-all expectation on which the race had been run. If he thought other Federalists would forgive him for acting with honor, he was wrong. An irate voter reacted: “[W]hen I voted for the [Federalist] ticket, I voted for John Adams. . . . What! do I chuse Samuel Miles to determine for me whether John Adams or Thomas Jefferson is the fittest man for President of the United States? No—I chuse him to act, not to think.” See Gazette of the United States, Dec. 15, 1796, p. 3, col. 1 (emphasis in original). 8 The Electors contend that elector discretion is needed to deal with the possibility that a future presidential candidate will die between Election Day and the Electoral College vote. See Reply Brief 20–22. We do not dismiss how much turmoil such an event could cause. In recognition of that fact, some States have drafted their pledge laws to give electors voting discretion when their candidate has died. See, e.g., Cal. Elec. Code Ann. §6906; Ind. Code §3–10–4–1.7. And we suspect that in such a case, States without a specific provision would also release electors from their pledge. Still, we note that because the situation is not before us, nothing in this opinion should be taken to permit the States to bind electors to a deceased candidate.

589.US.2019_18-565

Petitioners (collectively CARCO) sub-chartered the oil tanker M/T Athos I from tanker operator Star Tankers, which had chartered the tanker from respondent Frescati Shipping Company. In the final stretch of the tanker’s journey from Venezuela to New Jersey, an abandoned ship anchor punctured the tanker’s hull, causing 264,000 gallons of heavy crude oil to spill into the Delaware River. The Oil Pollution Act of 1990, 33 U. S. C. §2702(a), required Frescati, the vessel’s owner, to cover the cleanup costs in the first instance. Pursuant to the statute, Frescati’s liability was limited to $45 million, and the Oil Spill Liability Trust Fund, operated by the Federal Government (also a respondent here), reimbursed Frescati for an additional $88 million in cleanup costs. Frescati and the United States then sued CARCO to recover their respective portions of the cleanup costs. Both alleged that CARCO was ultimately at fault for the oil spill because CARCO had breached a contractual “safe-berth clause” in the subcharter agreement (“charter party”) between CARCO and Star Tankers. According to Frescati and the United States, that clause obligated CARCO to select a “safe” berth that would allow the vessel to come and go “always safely afloat,” and that obligation amounted to a warranty regarding the safety of the selected berth. After concluding that Frescati was an implied third-party beneficiary of the safe-berth clause, the Third Circuit held that the clause embodied an express warranty of safety made without regard to the charterer’s diligence in selecting the berth. Held: The plain language of the parties’ safe-berth clause establishes a warranty of safety. Pp. 5–16. (a) The Court’s analysis begins and ends with the text of the safe-berth clause. As CARCO acknowledges, the clause imposes on the charterer a duty to select a safe berth. And given the unqualified language of the clause, the charterer’s duty is absolute: The charterer must designate a berth that is “safe” and that allows the vessel to come and go “always” safely afloat. That absolute duty amounts to a warranty of safety. That the safe-berth clause does not expressly invoke the term “warranty” does not alter the charterer’s duty under the safe-berth clause. It is well settled that statements of material fact in a charter party are warranties, regardless of their label. See, e.g., Davison v. Von Lingen, 113 U.S. 40, 49–50. Here, it is plain on the face of the contract that the safe-berth clause sets forth a statement of “material” fact regarding the condition of the berth selected by the charterer. The charterer’s assurance of a safe berth is the entire root of the safe-berth clause, and crucially, it is not subject to qualifications or conditions. CARCO counters that the safe-berth clause merely imposes a duty of due diligence in selecting a safe berth. But as a general rule, tort concepts like due diligence have no place in contract analysis. Under basic precepts of contract law, an obligor is strictly liable for a breach of contract, without regard to fault or diligence. While parties are free to contract for limitations on liability, the parties here contracted for no such thing: There is no language in the safe-berth clause even hinting at due diligence. That omission is particularly notable in context, as the parties expressly contracted for due-diligence limitations on liability elsewhere in the charter party. CARCO’s arguments about other clauses in the charter party do not counsel in favor of a different result. The charter party’s “general exceptions clause,” which limits the charterer’s liability for losses due to “perils of the seas,” does not apply where, as here, another clause expressly provides for liability stemming from the designation of an unsafe berth. Nor does a clause requiring Star Tankers to obtain oil-pollution insurance relieve CARCO of liability under the safe-berth clause. The pollution-insurance clause covers risks beyond those resulting from the selection of an unsafe berth. CARCO’s alternative interpretation of the safe-berth clause, as simply requiring the charterer to pay any expenses resulting from the vessel master’s refusal to enter an unsafe berth, is inapposite. Assuming that the charterer is liable for expenses when the vessel master justifiably refuses to enter an unsafe berth, that does not abate the scope of the charterer’s liability when a vessel in fact enters an unsafe berth. The dissent argues that reading the safe-berth clause to bind the charterer to a warranty of safety would necessarily imply that the safe-berth clause creates contradictory warranties of safety, one on the charterer and one on the vessel master. Because that conflict cannot be, the dissent continues, the safe-berth clause must not bind the charterer to a warranty of safety. The dissent’s conclusion does not follow because the alleged conflict does not exist. Under the safe-berth clause, the charterer has a duty to select a safe berth, while the vessel master has a duty to load and discharge at the chosen safe berth. There is no tension between those two duties. Pp. 5–12. (b) CARCO’s arguments that other authorities have understood safe-berth clauses differently lack foothold in the text of the safe-berth clause and are otherwise unconvincing. For instance, CARCO relies on a leading admiralty treatise that urges that safe-berth clauses ought not be interpreted as establishing a warranty of safety because charterers are not always in the best position to know the dangers attendant to a given berth. But whatever that treatise sought to prevail upon courts to adopt as a prescriptive matter does not alter the plain meaning of the safe-berth clause here. Also unavailing is CARCO’s contention that Atkins v. Disintegrating Co., 18 Wall. 272, determined that safe-berth clauses do not embody a warranty of safety. CARCO relies on a passing statement in Atkins that did not bear on this Court’s ultimate holding that the vessel master in that case had waived the protection of the safe-berth clause. Finally, CARCO points out that the Fifth Circuit has held that a similarly unqualified safe-berth clause merely imposed a duty of due diligence. Orduna S. A. v. Zen-Noh Grain Corp., 913 F.2d 1149. But the Fifth Circuit did not purport to interpret the language of the safe-berth clause at issue in that case and instead relied principally on tort law and policy considerations. The Second Circuit’s long line of decisions interpreting the language of unqualified safe-berth clauses as embodying an express warranty of safety is more consistent with traditional contract analysis. See, e.g., Paragon Oil Co. v. Republic Tankers, S. A., 310 F.2d 169. Pp. 12–16. 886 F.3d 291, affirmed. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a dissenting opinion, in which Alito, J., joined.

In 2004, the M/T Athos I, a 748-foot oil tanker, allided[1] with a nine-ton anchor abandoned on the bed of the Delaware River. The anchor punctured the tanker’s hull, causing 264,000 gallons of heavy crude oil to spill into the river. As required by federal statute, respondents Frescati Shipping Company—the Athos I’s owner—and the United States covered the costs of cleanup. They then sought to reclaim those costs from petitioners CITGO Asphalt Refining Company and others (collectively CARCO), which had chartered the Athos I for the voyage that occasioned the oil spill. According to Frescati and the United States, CARCO had breached a contractual “safe-berth clause” obligating CARCO to select a “safe” berth that would allow the Athos I to come and go “always safely afloat.” The question before us is whether the safe-berth clause is a warranty of safety, imposing liability for an unsafe berth regardless of CARCO’s diligence in selecting the berth. We hold that it is. I A During the relevant period, the Athos I was the subject of a series of contracts involving three parties: Frescati, Star Tankers, and CARCO. Frescati owned the Athos I. Star Tankers, an operator of tanker vessels, contracted with Frescati to charter the Athos I for a period of time. CARCO then contracted with Star Tankers to subcharter the Athos I for the inauspicious voyage resulting in the oil spill. Pertinent here is the subcharter agreement between Star Tankers and CARCO. In admiralty, such contracts to charter a vessel are termed “charter parties.” Like many modern charter parties, the agreement between Star Tankers and CARCO was based on a standard industry form contract. It drew essentially verbatim from a widely used template known as the ASBATANKVOY form, named after the Association of Ship Brokers & Agents (USA) Inc. (ASBA) trade association that publishes it. At the core of the parties’ dispute is a clause in the charter party requiring the charterer, CARCO, to designate a safe berth at which the vessel may load and discharge cargo. This provision, a standard feature of many charter parties, is customarily known as a safe-berth clause. The safe-berth clause here provides, as relevant, that “[t]he vessel shall load and discharge at any safe place or wharf, . . . which shall be designated and procured by the Charterer, provided the Vessel can proceed thereto, lie at, and depart therefrom always safely afloat, any lighterage being at the expense, risk and peril of the Charterer.” Addendum to Brief for Petitioners 8a.[2] The charter party separately requires CARCO to direct the Athos I to a “safe por[t ]” along the Atlantic seaboard of the United States. Id., at 24a. Pursuant to the charter party, CARCO designated as the berth of discharge its asphalt refinery in Paulsboro, New Jersey, on the shore of the Delaware River. In November 2004, the Athos I set out on a 1,900-mile journey from Puerto Miranda, Venezuela, to Paulsboro, New Jersey, carrying a load of heavy crude oil. The vessel was in the final 900-foot stretch of its journey when an abandoned ship anchor in the Delaware River pierced two holes in the vessel’s hull. Much of the Athos I’s freight drained into the river. B After the Exxon-Valdez oil spill in 1989, Congress passed the Oil Pollution Act of 1990 (OPA), 104Stat. 484, 33 U. S. C. §2701 et seq., to promote the prompt cleanup of oil spills. To that end, OPA deems certain entities responsible for the costs of oil-spill cleanups, regardless of fault. §2702(a). It then limits the liability of such “responsible part[ies]” if they (among other things) timely assist with cleanup efforts. §2704. Responsible parties that comply with the statutory conditions receive a reimbursement from the Oil Spill Liability Trust Fund (Fund), operated by the Federal Government, for any cleanup costs exceeding a statutory limit. §2708; see also §2704. Although a statutorily responsible party must pay cleanup costs without regard to fault, it may pursue legal claims against any entity allegedly at fault for an oil spill. §§2710, 2751(e). So may the Fund: By reimbursing a responsible party, the Fund becomes subrogated to the responsible party’s rights (up to the amount reimbursed to the responsible party) against any third party allegedly at fault for the incident. §§2712(f ), 2715(a). As owner of the Athos I, Frescati was deemed a “responsible party” for the oil spill under OPA. Frescati worked with the U. S. Coast Guard in cleanup efforts and covered the costs of the cleanup. As a result, Frescati’s liability was statutorily limited to $45 million, and the Fund reimbursed Frescati for an additional $88 million that Frescati paid in cleanup costs. C Following the cleanup, Frescati and the United States each sought recovery against CARCO: Frescati sought to recover the cleanup costs not reimbursed by the Fund, while the United States sought to recover the amount disbursed by the Fund. As relevant here, both Frescati and the United States claimed that CARCO had breached the safe-berth clause by failing to designate a safe berth, and thus was at fault for the spill. After a complicated series of proceedings—including a 41-day trial, a subsequent 31-day evidentiary hearing, and two appeals—the Court of Appeals for the Third Circuit found for Frescati and the United States. The court first concluded that Frescati was an implied third-party beneficiary of the safe-berth clause in the charter party between CARCO and Star Tankers, thereby allowing the breach-of-contract claims by Frescati and the United States to proceed against CARCO. In re Frescati Shipping Co., 718 F.3d 184, 200 (2013). The court then held that the safe-berth clause embodied an express warranty of safety “made without regard to the amount of diligence taken by the charterer,” and that CARCO was liable to Frescati and the United States for breaching that warranty. Id., at 203; In re Frescati Shipping Co., 886 F.3d 291, 300, 315 (2018) (case below). We granted certiorari, 587 U. S. ___ (2019), to resolve whether the safe-berth clause at issue here merely imposes a duty of diligence, as the Fifth Circuit has held in a similar case, or establishes a warranty of safety, as the Second Circuit has held in other analogous cases. Compare Orduna S. A. v. Zen-Noh Grain Corp., 913 F.2d 1149 (CA5 1990), with, e.g., Paragon Oil Co. v. Republic Tankers, S. A., 310 F.2d 169 (CA2 1962). The former interpretation allows a charterer to avoid liability by exercising due diligence in selecting a berth; the latter imposes liability for an unsafe berth without regard to the care taken by the charterer. Because we find it plain from the language of the safe-berth clause that CARCO warranted the safety of the berth it designated, we affirm the judgment of the Third Circuit. II Maritime contracts “must be construed like any other contracts: by their terms and consistent with the intent of the parties.” Norfolk Southern R. Co. v. James N. Kirby, Pty Ltd., 543 U.S. 14, 31 (2004); see also 2 T. Schoenbaum, Admiralty & Maritime Law §11:2, p. 7 (6th ed. 2018) (“[F]ederal maritime law includes general principles of contract law”). “ ‘Where the words of a contract in writing are clear and unambiguous, its meaning is to be ascertained in accordance with its plainly expressed intent.’ ” M&G Polymers USA, LLC v. Tackett, 574 U.S. 427, 435 (2015) (quoting 11 R. Lord, Williston on Contracts §30:6, p. 108 (4th ed. 2012) (Williston)). In such circumstances, the parties’ intent “can be determined from the face of the agreement” and “the language that they used to memorialize [that] agreement.” 11 Williston §30:6, at 97–98, 112–113. But “[w]hen a written contract is ambiguous, its meaning is a question of fact, requiring a determination of the intent of [the] parties in entering the contract”; that may involve examining “relevant extrinsic evidence of the parties’ intent and the meaning of the words that they used.” Id., §30:7, at 116–119, 124 (footnote omitted). A Our analysis starts and ends with the language of the safe-berth clause. That clause provides, as relevant, that the charterer “shall . . . designat[e] and procur[e]” a “safe place or wharf,” “provided [that] the Vessel can proceed thereto, lie at, and depart therefrom always safely afloat.” Addendum to Brief for Petitioners 8a. As even CARCO acknowledges, the clause plainly imposes on the charterer at least some “duty to select a ‘safe’ berth.” Brief for Petitioners 21. Given the unqualified language of the safe-berth clause, it is similarly plain that this acknowledged duty is absolute. The clause requires the charterer to designate a “safe” berth: That means a berth “free from harm or risk.” Webster’s Collegiate Dictionary 1030 (10th ed. 1994); see also New Oxford American Dictionary 1500 (E. Jewell & F. Abate eds. 2001) (“safe” means “protected from or not exposed to danger or risk”). And the berth must allow the vessel to come and go “always” safely afloat: That means afloat “at all times” and “in any event.” Webster’s Collegiate Dictionary, at 35; see also New Oxford American Dictionary, at 47 (“always” means “at all times; on all occasions”). Selecting a berth that does not satisfy those conditions constitutes a breach. The safe-berth clause, in other words, binds the charterer to a warranty of safety.[3] No matter that the safe-berth clause does not expressly invoke the term “warranty.” It is well settled as a matter of maritime contracts that “[s]tatements of fact contained in a charter party agreement relating to some material matter are called warranties,” regardless of the label ascribed in the charter party. 22 Williston §58.11, at 40–41 (2017); see also Davison v. Von Lingen, 113 U.S. 40, 49–50 (1885) (a stipulation going to “substantive” and “material” parts of a charter party forms “a warranty”); Behn v. Burness, 3 B. & S. 751, 122 Eng. Rep. 281 (K. B. 1863) (“With respect to statements in a [charter party] descriptive of . . . some material incident . . . , if the descriptive statement was intended to be a substantive part of the [charter party], it is to be regarded as a warranty”). What matters, then, is that the safe-berth clause contains a statement of material fact regarding the condition of the berth selected by the charterer. Here, the safety of the selected berth is the entire root of the safe-berth clause: It is the very reason for the clause’s inclusion in the charter party. And crucially, the charterer’s assurance of safety is not subject to qualifications or conditions. Under any conception of materiality and any view of the parties’ intent, the charterer’s assurance surely counts as material. That leaves no doubt that the safe-berth clause establishes a warranty of safety, on equal footing with any other provision of the charter party that invokes express warranty language.[4] CARCO resists this plain reading of the safe-berth clause, arguing instead that the clause contains an implicit limitation: The clause does not impose “strict liability,” says CARCO, or “liability without regard to fault.” Brief for Petitioners 23, 25. In effect, CARCO interprets the safe-berth clause as imposing a mere duty of due diligence in the selection of the berth. See Tr. of Oral Arg. 19–20 (arguing that “[CARCO] did [its] due diligence” in “selecting the port or the berth”); id., at 28 (suggesting that the safe-berth clause is constrained “as a matter of due diligence in tort concepts”); Reply Brief 5, n. 3 (asserting that a charterer’s liability under the safe-berth clause “should be addressed through . . . sources of la[w] such as tort law”). But as a general rule, due diligence and fault-based concepts of tort liability have no place in the contract analysis required here. Under elemental precepts of contract law, an obligor is “liable in damages for breach of contract even if he is without fault.” Restatement (Second) of Contracts, p. 309 (1979) (Restatement (Second)). To put that default contract-law principle in tort-law terms, “Contract liability is strict liability.” Ibid. (emphasis added); see also 23 Willis-ton §63:8, at 499 (2018) (“Liability for a breach of contract is, prima facie, strict liability”). What CARCO thus protests is the straightforward application of contract liability to a breach of contract. Although contract law generally does not, by its own force, limit liability based on tort concepts of fault, parties are of course free to contract for such limitations. See Restatement (Second), at 309 (obligor who wishes to avoid strict liability for breach may “limi[t] his obligation by agreement”). Here, however, the safe-berth clause is clear that the parties contracted for no such thing. CARCO does not identify—nor can we discern—any language in the clause hinting at “due diligence” or related concepts of “fault.” That omission is particularly notable in context: Where the parties intended to limit obligations based on due diligence elsewhere in the charter party, they did so expressly. See Addendum to Brief for Petitioners 4a (providing that the vessel “b[e] seaworthy, and hav[e] all pipes, pumps and heater coils in good working order, . . . so far as the foregoing conditions can be attained by the exercise of due diligence”); id., at 13a (relieving vessel owner of responsibility for certain consequences of any “unseaworthiness existing . . . at the inception of the voyage [that] was discoverable by the exercise of due diligence”); id., at 41a (requiring vessel owner to “exercise due diligence to ensure that [a drug and alcohol] policy [onboard the vessel] is complied with”).[5] That the parties did not do so in the safe-berth clause specifically is further proof that they did not intend for such a liability limitation to inhere impliedly.[6] Unable to identify any liability-limiting language in the safe-berth clause, CARCO points to a separate “general exceptions clause” in the charter party that exempts a charterer from liability for losses due to “perils of the seas.” Id., at 14a. According to CARCO, the “general exceptions clause” demonstrates that the parties did not intend the safe-berth clause to impose liability for a “peri[l] of the seas” like an abandoned anchor. That argument founders on a critical component of the “general exceptions clause”: By its terms, it does not apply when liability is “otherwise . . . expressly provided” in the charter party. Ibid. The safe-berth clause, as explained above, expressly provides for liability stemming from the designation of an unsafe berth. The catchall “general exceptions clause” neither supersedes nor overlays it.[7] Likewise immaterial is another clause of the charter party that requires Star Tankers to obtain oil-pollution insurance. According to CARCO, that clause evidences the parties’ intent to relieve CARCO of oil-spill liability under the safe-berth clause. But the oil-pollution insurance that Star Tankers must obtain covers risks beyond simply those attendant to the selection of an unsafe berth. And CARCO’s reading of the insurance clause (as relieving CARCO of oil-spill liability) does not square with its reading of the safe-berth clause (as imposing such liability when CARCO fails to exercise due diligence). Finally, CARCO offers an alternative interpretation of the safe-berth clause that focuses on the vessel master’s right instead of the charterer’s duty. This alternative interpretation proceeds from the subclause specifying that the selected berth be one that the vessel may “proceed thereto, lie at, and depart therefrom always safely afloat, any lighterage [i.e., transfer of goods between vessels] being at the expense, risk and peril of the Charterer.” Id., at 8a. On CARCO’s reading, that subclause means that the vessel master has a right to refuse entry into a berth that the master perceives to be unsafe, and the charterer must pay any expenses resulting from the refusal. We have, to be sure, recognized that similarly worded safe-berth clauses may implicitly denote a vessel master’s right to refuse entry and the charterer’s resultant obligation to bear the costs of that refusal. See Mencke v. Cargo of Java Sugar, 187 U.S. 248 (1902); The Gazelle and Cargo, 128 U.S. 474 (1888). But that a charterer may be liable for expenses when a vessel master justifiably refuses to enter an unsafe berth in no way abates the scope of the charterer’s liability when a vessel in fact enters an unsafe berth. And a tacit recognition of a vessel master’s right of refusal does not overwrite the safe-berth clause’s express prescription of a warranty of safety. The dissent, too, offers an alternative interpretation. It claims that if the safe-berth clause binds the charterer to a warranty of safety, the clause must bind the vessel master to effectively the same warranty—due to the clause’s statement that “ ‘[t]he vessel shall load and discharge at [a] safe place or wharf.’ ” Post, at 6 (quoting Addendum to Brief for Petitioners 8a). Because that would “creat[e] contradictory warranties of safety,” the dissent continues, the safe-berth clause must not bind the charterer to a warranty of safety (or, apparently, impose an obligation on the charterer at all). Post, at 7. This conclusion does not follow because the conflict diagnosed by the dissent does not exist. The safe-berth clause says that “[t]he vessel shall load and discharge at any safe place or wharf, . . . which shall be designated and procured by the Charterer.” Addendum to Brief for Petitioners 8a. Plainly, that means that the “safe place or wharf . . . shall be designated and procured by the Charterer.” Ibid. The vessel master’s duty is only to “load and discharge” at the chosen safe berth. Ibid. (Not, as the dissent urges, at any safe berth the vessel master so desires regardless of the charterer’s contractually required selection. Post, at 6, n. 4.) On its face, the vessel master’s duty creates no tension with the charterer’s duty. And it strains common sense to insist (as the dissent does) that the vessel master implicitly has a separate, dueling obligation regarding the safety of the berth, when the clause explicitly assigns that responsibility to the charterer. Post, at 6–7. Perhaps the dissent says it best: We must “reject [this] interpretation that . . . ‘se[ts] up . . . two clauses in conflict with one another.’ ” Post, at 6 (quoting Mastrobuono v. Shearson Lehman Hutton, Inc., 514 U.S. 52, 64 (1995)). We instead take the safe-berth clause at face value. It requires the charterer to select a safe berth, and that requirement here amounts to a warranty of safety. B CARCO’s remaining arguments point to authorities that have purportedly construed safe-berth clauses to contain limitations on liability. These arguments find no foothold in the language of the charter party at issue here. And none is otherwise convincing. CARCO asserts, for instance, that a leading admiralty treatise has urged that safe-berth clauses ought not be interpreted as establishing a warranty. See G. Gilmore & C. Black, Law of Admiralty §4–4, p. 205 (2d ed. 1975) (Gilmore & Black). Gilmore and Black’s position, however, stemmed from their belief that vessel masters or vessel owners are generally better positioned than charterers to bear the liability of an unsafe berth. See ibid. (reasoning that charterers “may know nothing of the safety of ports and berths, and [are] much less certain to be insured against” liability for losses stemming from an unsafe berth).[8] Gilmore and Black also acknowledged that, as of 1975, many courts had not interpreted safe-berth clauses in the manner that they proposed. See id., at 204, and n. 34a, 206, and n. 36. Whatever Gilmore and Black sought to prevail upon courts to adopt as a prescriptive matter does not alter the plain meaning of the safe-berth clause here. CARCO next contends that in Atkins v. Disintegrating Co., 18 Wall. 272 (1874), this Court acknowledged that safe-berth clauses do not embody a warranty of safety. That greatly overreads Atkins. In that case, this Court affirmed a District Court’s ruling that, although the berth selected by the charterer was not safe, the vessel master had “waived” the protection of the safe-berth clause. Atkins v. Fibre Disintegrating Co., 2 F. Cas. 78, 79 (EDNY 1868); see Atkins, 18 Wall., at 299. No one posits that the District Court’s waiver holding has any significance in this case. CARCO, however, points to language in the District Court’s opinion observing that the “safe” berth referenced in the charter party “impl[ied one] which th[e] vessel could enter and depart from without legal restraint, and without incurring more than the ordinary perils of the seas.” Atkins, 2 F. Cas., at 79. But the District Court’s remark—that a berth may be safe even if certain perils lurk within—did not bear on its finding that the berth in question was unsafe or its holding that the vessel master had “waived” the protection of the safe-berth clause. When this Court approved of the District Court’s “views” and “conclusions,” Atkins, 18 Wall., at 299, it did not adopt as controlling precedent—for all safe-berth clauses going forward—an observation that was not controlling even for the District Court. Also misplaced is CARCO’s reliance on Orduna S. A., 913 F.2d 1149. True, the Fifth Circuit there held that a similarly unqualified safe-berth clause imposed a duty of due diligence. Id., at 1157. But in so holding, the court did not purport to interpret the language of the safe-berth clause at issue in that case. Id., at 1156–1157. Instead, it looked principally to tort law and policy considerations. See, e.g., id., at 1156 (“requiring negligence as a predicate for the charterer’s liability does not increase the risk that the vessel will be exposed to an unsafe berth”); id., at 1157 (“no legitimate legal or social policy is furthered by making the charterer warrant the safety of the berth it selects”). Neither tort principles nor policy objectives, however, override the safe-berth clause’s unambiguous meaning. More consistent with traditional contract analysis is the Second Circuit’s long line of decisions interpreting the language of unqualified safe-berth clauses to embody an express warranty of safety. See, e.g., Paragon Oil Co., 310 F. 2d, at 172–173 (“the express terms of [the] contract” established a “warranty” obliging the charterer “to furnish, not only a place which he believes to be safe, but a place where the chartered vessel can discharge ‘always afloat’ ” (some internal quotation marks omitted)); Park S. S. Co. v. Cities Serv. Oil Co., 188 F.2d 804, 805–806 (CA2 1951) (“the natural meaning of ‘safe place’ is a place entirely safe, not an area only part of which is safe,” and “the charter party was an express assurance that the berth was safe”); Cities Serv. Transp. Co. v. Gulf Refining Co., 79 F.2d 521 (CA2 1935) (per curiam) (the “charter party was itself an express assurance . . . that at the berth ‘indicated’ the ship would be able to lie ‘always afloat’ ”). Those decisions, which focused on the controlling contract language, all point in the same direction: When the language of a safe-berth clause obliges a charterer to select a safe berth without qualifying the charterer’s duty or the assurance of safety that language establishes a warranty. That aligns with our decision today.[9] III We conclude that the language of the safe-berth clause here unambiguously establishes a warranty of safety, and that CARCO has identified “no reason to contravene the clause’s obvious meaning.” Kirby, 543 U. S., at 31–32. We emphasize, however, that our decision today “does no more than provide a legal backdrop against which future [charter parties] will be negotiated.” Id., at 36. Charterers remain free to contract around unqualified language that would otherwise establish a warranty of safety, by expressly limiting the extent of their obligations or liability. * * * For the foregoing reasons, we conclude that the plain language of the safe-berth clause establishes a warranty of safety and therefore affirm the judgment of the Third Circuit. It is so ordered. Notes 1 An allision is “[t]he contact of a vessel with a stationary object such as an anchored vessel or a pier.” Black’s Law Dictionary 94 (11th ed. 2019). 2 The parties agree that the safe-berth clause also encompasses what is often referred to as a “safe-port clause.” The safe-port clause here provides that “[t]he vessel . . . shall, with all convenient dispatch, proceed as ordered to Loading Port(s) named . . . , or so near thereunto as she may safely get (always afloat), . . . and being so loaded shall forthwith proceed, . . . direct to the Discharging Port[s], or so near thereunto as she may safely get (always afloat), and deliver said cargo.” Addendum to Brief for Petitioners 4a. The parties do not dispute that the two clauses should be read in conjunction. 3 The central pillar of the dissent is that the safe-berth clause merely bestows upon the charterer “the right to ‘designat[e]’ the place of discharge,” and thus apparently creates no duty to select a safe berth (much less a warranty of safety). Post, at 2 (opinion of Thomas, J.) (quoting Addendum to Brief for Petitioners 8a; emphasis added); see also post, at 3 (“the charterer has a right of selection”). That sidesteps the safe-berth clause’s plain terms, which prescribe that the charterer “shall . . . designat[e] and procur[e]” a “safe place or wharf.” Addendum to Brief for Petitioners 8a (emphasis added). As we have said before, “the word ‘shall’ usually connotes a requirement.” Kingdomware Technologies, Inc. v. United States, 579 U. S. ___, ___ (2016) (slip op., at 9); see also, e.g., Lexecon Inc. v. Milberg Weiss Bershad Hynes & Lerach, 523 U.S. 26, 35 (1998). The text thus forecloses the dissent’s permissive view that the charterer merely has an elective “right” to select a berth of discharge but no duty to do so. And even CARCO disclaims that atextual position. See Brief for Petitioners 21. 4 Because the materiality of the charterer’s assurance of safety is plain on the face of the charter party, the specific materiality issue here raises no question of fact for a jury to resolve. That is not to say that the materiality of a statement in a charter party is always a question of law. Nor does the materiality analysis here bear on wholly different materiality inquiries. For not all questions of materiality are alike: Sometimes materiality is a question of law. See, e.g., 30 Williston §75:30, at 108 (whether an alteration of a contract is material). Other times, it involves factual determinations uniquely suited for a jury. See, e.g., TSC Industries, Inc. v. Northway, Inc., 426 U.S. 438, 450 (1976) (whether a company’s misstatements to the public are material for securities-fraud purposes). The dissent’s insistence that materiality is a question of fact “ ‘in other contexts’ ”—such as securities fraud—thus is inapposite. Post, at 8 (quoting United States v. Gaudin, 515 U.S. 506, 512 (1995)). 5 It also bears mention that many other industry form charter par-ties—not selected by CARCO and Star Tankers—explicitly limit the liability that may flow from a charterer’s selection of a berth. See, e.g., 2EJ. Force & L. Lambert, Benedict on Admiralty, ch. XXVII, §27–567 (rev. 7th ed. 2019) (INTERTANKVOY form specifies that “[c]harterers shall exercise due diligence to ascertain that any places to which they order the vessel are safe for the vessel and that she will lie there always afloat”). 6 After all, language that limits liability is necessary to overcome the default rule of strict liability for contractual breach. Supra, at 8–9. That stands in contrast to the established principle that charter parties can, at least in the circumstances here, create warranties without invoking express warranty language. Supra, at 7–8. The dissent overlooks this distinction when it claims that the absence of express warranty language in the safe-berth clause and the presence of it elsewhere in the charter party imply that no warranty may be found here. Post, at 4. 7 At oral argument, CARCO urged that the abandoned anchor was not only “a peril of the sea” but also “an abnormal occurrence.” Tr. of Oral Arg. 28–29. CARCO’s “abnormal occurrence” argument appears to rest on a recent decision by the Supreme Court of the United Kingdom interpreting a safe-berth clause not to impose liability if an “abnormal occurrence” rendered the selected berth unsafe. See Gard Marine & Energy Ltd. v. China Nat. Chartering Co., [2017] UKSC 35 (The Ocean Victory). In its opening brief to this Court, however, CARCO did not cite The Ocean Victory or argue that the abandoned anchor here constituted an “abnormal occurrence.” 8 The dissent’s claim that Gilmore and Black looked to “ ‘the very words of the usual clauses,’ ” post, at 3, n. 1 (quoting Gilmore & Black §4–4, at 204), relies on a discussion not of the charterer’s obligation under the safe-berth clause but of the vessel master’s lack of such obligation, Gilmore & Black §4–4, at 204–205. At most, Gilmore and Black “suggested” that interpreting safe-berth clauses to relieve vessel masters of any obligation to enter an unsafe berth “might easily be read to contradict” any “affirmative liability” on the part of the charterer “in case of mishap.” See id., §4–4, at 205. But that supposition is at odds with the language of the safe-berth clause here, which (as even CARCO acknowledges) plainly contemplates at least some liability for the charterer’s designation of an unsafe berth. Supra, at 6–7, and n. 3, 8. And as explained, a vessel master’s ability to refuse entry into an unsafe berth does not logically or textually diminish a charterer’s liability when the vessel master in fact enters an unsafe berth selected by the charterer. Supra, at 11. 9 The parties also dispute whether the prevailing industry usage of safe-berth clauses supports reading the safe-berth clause here as a warranty or as a promise of due diligence. Because the express language of the safe-berth clause is susceptible to only one meaning, we need not address these arguments.

589.US.2019_18-1171

Entertainment Studios Network (ESN), an African-American-owned television-network operator, sought to have cable television conglomerate Comcast Corporation carry its channels. Comcast refused, citing lack of programming demand, bandwidth constraints, and a preference for programming not offered by ESN. ESN and the National Association of African American-Owned Media (collectively, ESN) sued, alleging that Comcast’s behavior violated 42 U. S. C. §1981, which guarantees “[a]ll persons . . . the same right . . . to make and enforce contracts . . . as is enjoyed by white citizens.” The District Court dismissed the complaint for failing plausibly to show that, but for racial animus, Comcast would have contracted with ESN. The Ninth Circuit reversed, holding that ESN needed only to plead facts plausibly showing that race played “some role” in the defendant’s decisionmaking process and that, under this standard, ESN had pleaded a viable claim. Held: A §1981 plaintiff bears the burden of showing that the plaintiff’s race was a but-for cause of its injury, and that burden remains constant over the life of the lawsuit. Pp. 3–13. (a) To prevail, a tort plaintiff typically must prove but-for causation. See University of Tex. Southwestern Medical Center v. Nassar, 570 U.S. 338, 347. Normally, too, the essential elements of a claim remain constant throughout the lawsuit. See, e.g., Lujan v. Defenders of Wildlife, 504 U.S. 555, 561. ESN suggests that §1981 creates an exception to one or both of these general principles, either because a §1981 plaintiff only bears the burden of showing that race was a “motivating factor” in the defendant’s challenged decision or because, even when but-for causation applies at trial, a plausible “motivating factor” showing is all that is necessary to overcome a motion to dismiss at the pleading stage. Pp. 3–12. (1) Several clues, taken collectively, make clear that §1981 follows the usual rules. The statute’s text suggests but-for causation: An ordinary English speaker would not say that a plaintiff did not enjoy the “same right” to make contracts “as is enjoyed by white citizens” if race was not a but-for cause affecting the plaintiff’s ability to contract. Nor does the text suggest that the test should be different in the face of a motion to dismiss. The larger structure and history of the Civil Rights Act of 1866 provide further clues. When enacted, §1981 did not provide a private enforcement mechanism for violations. That right was judicially created, see Johnson v. Railway Express Agency, Inc., 421 U.S. 454, 459, but even in that era, the Court usually insisted that the legal elements of implied causes of action be at least as demanding as those found in analogous statutory causes of action. That rule supplies useful guidance here, where a neighboring section of the 1866 Act uses the terms “on account of” and “by reason of,” §2, 14Stat. 27—phrases often held to indicate but-for causation—and gives no hint that a different rule might apply at different times in the life of a lawsuit. Another provision provides that in cases not provided for by the Act, the common law shall govern, §3, ibid., which in 1866, usually treated a showing of but-for causation as a prerequisite to a tort suit. This Court’s precedents confirm what the statute’s language and history indicate. See, e.g., Johnson, 421 U. S., at 459–460; Buchanan v. Warley, 245 U.S. 60, 78–79. Pp. 4–8. (2) ESN urges applying the “motivating factor” causation test in Title VII of the Civil Rights Act of 1964 to §1981 cases. But this Court has already twice rejected such efforts in other contexts, see, e.g., Gross v. FBL Financial Services, Inc., 557 U.S. 167, and there is no reason to think it would fit any better here. Moreover, when that test was added to Title VII in the Civil Rights Act of 1991, Congress also amended §1981 without mentioning “motivating factors.” Even if ESN is correct that those amendments clarified that §1981 addresses not just contractual outcomes but the whole contracting process, its claim that a process-oriented right necessarily pairs with a motivating factor causal standard is mistaken. The burden-shifting framework of McDonnell Douglas Corp. v. Green, 411 U.S. 792, also supplies no support for the innovations ESN seeks. Pp. 8–12. (b) The court of appeals should determine in the first instance how the operative amended complaint in this case fares under the proper standard. P. 13. 743 Fed. Appx. 106, vacated and remanded. Gorsuch, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Breyer, Alito, Sotomayor, Kagan, and Kavanaugh, JJ., joined, and in which Ginsburg, J., joined except for the footnote. Ginsburg, J., filed an opinion concurring in part and concurring in the judgment.

Few legal principles are better established than the rule requiring a plaintiff to establish causation. In the law of torts, this usually means a plaintiff must first plead and then prove that its injury would not have occurred “but for” the defendant’s unlawful conduct. The plaintiffs before us suggest that 42 U. S. C. §1981 departs from this traditional arrangement. But looking to this particular statute’s text and history, we see no evidence of an exception. I This case began after negotiations between two media companies failed. African-American entrepreneur Byron Allen owns Entertainment Studios Network (ESN), the operator of seven television networks—Justice Central.TV, Comedy.TV, ES.TV, Pets.TV, Recipe.TV, MyDestination.TV, and Cars.TV. For years, ESN sought to have Comcast, one of the nation’s largest cable television conglomerates, carry its channels. But Comcast refused, citing lack of demand for ESN’s programming, bandwidth constraints, and its preference for news and sports programming that ESN didn’t offer. With bargaining at an impasse, ESN sued. Seeking billions in damages, the company alleged that Comcast systematically disfavored “100% African American-owned media companies.” ESN didn’t dispute that, during negotiations, Comcast had offered legitimate business reasons for refusing to carry its channels. But, ESN contended, these reasons were merely pretextual. To help obscure its true discriminatory intentions and win favor with the Federal Communications Commission, ESN asserted, Comcast paid civil rights groups to advocate publicly on its behalf. As relevant here, ESN alleged that Comcast’s behavior violated 42 U. S. C. §1981(a), which guarantees, among other things, “[a]ll persons . . . the same right . . . to make and enforce contracts . . . as is enjoyed by white citizens.” Much motions practice followed. Comcast sought to dismiss ESN’s complaint, and eventually the district court agreed, holding that ESN’s pleading failed to state a claim as a matter of law. The district court twice allowed ESN a chance to remedy its complaint’s deficiencies by identifying additional facts to support its case. But each time, the court concluded, ESN’s efforts fell short of plausibly showing that, but for racial animus, Comcast would have contracted with ESN. After three rounds of pleadings, motions, and dismissals, the district court decided that further amendments would prove futile and entered a final judgment for Comcast. The Ninth Circuit reversed. As that court saw it, the district court used the wrong causation standard when assessing ESN’s pleadings. A §1981 plaintiff doesn’t have to point to facts plausibly showing that racial animus was a “but for” cause of the defendant’s conduct. Instead, the Ninth Circuit held, a plaintiff must only plead facts plausibly showing that race played “some role” in the defendant’s decisionmaking process. 743 Fed. Appx. 106, 107 (2018); see also National Assn. of African American-Owned Media v. Charter Communications, Inc., 915 F.3d 617, 626 (CA9 2019) (describing the test as whether “discriminatory intent play[ ed] any role”). And under this more forgiving causation standard, the court continued, ESN had pleaded a viable claim. Other circuits dispute the Ninth Circuit’s understanding of §1981. Like the district court in this case, for example, the Seventh Circuit has held that “to be actionable, racial prejudice must be a but-for cause . . . of the refusal to transact.” Bachman v. St. Monica’s Congregation, 902 F.2d 1259, 1262–1263 (1990). To resolve the disagreement among the circuits over §1981’s causation requirement, we agreed to hear this case. 587 U. S. ___ (2019). II It is “textbook tort law” that a plaintiff seeking redress for a defendant’s legal wrong typically must prove but-for causation. University of Tex. Southwestern Medical Center v. Nassar, 570 U.S. 338, 347 (2013) (citing W. Keeton, D. Dobbs, R. Keeton, & D. Owen, Prosser and Keeton on Law of Torts 265 (5th ed. 1984)). Under this standard, a plaintiff must demonstrate that, but for the defendant’s unlawful conduct, its alleged injury would not have occurred. This ancient and simple “but for” common law causation test, we have held, supplies the “default” or “background” rule against which Congress is normally presumed to have legislated when creating its own new causes of action. 570 U. S., at 346–347 (citing Los Angeles Dept. of Water and Power v. Manhart, 435 U.S. 702, 711 (1978)). That includes when it comes to federal antidiscrimination laws like §1981. See 570 U. S., at 346–347 (Title VII retaliation); Gross v. FBL Financial Services, Inc., 557 U.S. 167, 176–177 (2009) (Age Discrimination in Employment Act of 1967). Normally, too, the essential elements of a claim remain constant through the life of a lawsuit. What a plaintiff must do to satisfy those elements may increase as a case progresses from complaint to trial, but the legal elements themselves do not change. So, to determine what the plaintiff must plausibly allege at the outset of a lawsuit, we usually ask what the plaintiff must prove in the trial at its end. See, e.g., Lujan v. Defenders of Wildlife, 504 U.S. 555, 561 (1992); Dura Pharmaceuticals, Inc. v. Broudo, 544 U.S. 336, 346–347 (2005); Ashcroft v. Iqbal, 556 U.S. 662, 678–679 (2009). ESN doesn’t seriously dispute these general principles. Instead, it suggests §1981 creates an exception to one or both of them. At times, ESN seems to argue that a §1981 plaintiff only bears the burden of showing that race was a “motivating factor” in the defendant’s challenged decision, not a but-for cause of its injury. At others, ESN appears to concede that a §1981 plaintiff does have to prove but-for causation at trial, but contends the rules should be different at the pleading stage. According to this version of ESN’s argument, a plaintiff should be able to overcome at least a motion to dismiss if it can allege facts plausibly showing that race was a “motivating factor” in the defendant’s decision. ESN admits this arrangement would allow some claims to proceed past the pleading stage that are destined to fail later as a matter of law. Still, the company insists, that is what the statute demands. A We don’t doubt that most rules bear their exceptions. But, taken collectively, clues from the statute’s text, its history, and our precedent persuade us that §1981 follows the general rule. Here, a plaintiff bears the burden of showing that race was a but-for cause of its injury. And, while the materials the plaintiff can rely on to show causation may change as a lawsuit progresses from filing to judgment, the burden itself remains constant. Congress passed the Civil Rights Act of 1866 in the aftermath of the Civil War to vindicate the rights of former slaves. Section 1 of that statute included the language found codified today in §1981(a), promising that “[a]ll persons . . . shall have the same right . . . to make and enforce contracts, to sue, be parties, [and] give evidence . . . as is enjoyed by white citizens.” 42 U. S. C. §1981; Civil Rights Act of 1866, 14Stat. 27. While the statute’s text does not expressly discuss causation, it is suggestive. The guarantee that each person is entitled to the “same right . . . as is enjoyed by white citizens” directs our attention to the counterfactual—what would have happened if the plaintiff had been white? This focus fits naturally with the ordinary rule that a plaintiff must prove but-for causation. If the defendant would have responded the same way to the plaintiff even if he had been white, an ordinary speaker of English would say that the plaintiff received the “same” legally protected right as a white person. Conversely, if the defendant would have responded differently but for the plaintiff ’s race, it follows that the plaintiff has not received the same right as a white person. Nor does anything in the statute signal that this test should change its stripes (only) in the face of a motion to dismiss. The larger structure and history of the Civil Rights Act of 1866 provide further clues. Nothing in the Act specifically authorizes private lawsuits to enforce the right to contract. Instead, this Court created a judicially implied private right of action, definitively doing so for the first time in 1975. See Johnson v. Railway Express Agency, Inc., 421 U.S. 454, 459 (1975); see also Jett v. Dallas Independent School Dist., 491 U.S. 701, 720 (1989). That was during a period when the Court often “assumed it to be a proper judicial function to provide such remedies as are necessary to make effective a statute’s purpose.” Ziglar v. Abbasi, 582 U. S. ___, ___ (2017) (slip op., at 8) (internal quotation marks omitted). With the passage of time, of course, we have come to appreciate that, “[l]ike substantive federal law itself, private rights of action to enforce federal law must be created by Congress” and “[r]aising up causes of action where a statute has not created them may be a proper function for common-law courts, but not for federal tribunals.” Alexander v. Sandoval, 532 U.S. 275, 286–287 (2001) (internal quotation marks omitted). Yet, even in the era when this Court routinely implied causes of action, it usually insisted on legal elements at least as demanding as those Congress specified for analogous causes of action actually found in the statutory text. See, e.g., Blue Chip Stamps v. Manor Drug Stores, 421 U.S. 723, 736 (1975). That rule supplies useful guidance here. Though Congress did not adopt a private enforcement mechanism for violations of §1981, it did establish criminal sanctions in a neighboring section. That provision permitted the prosecution of anyone who “depriv[es]” a person of “any right” protected by the substantive provisions of the Civil Rights Act of 1866 “on account of ” that person’s prior “condition of slavery” or “by reason of ” that person’s “color or race.” §2, 14Stat. 27. To prove a violation, then, the government had to show that the defendant’s challenged actions were taken “ ‘on account of ’ ” or “ ‘by reason of ’ ” race—terms we have often held indicate a but-for causation requirement. Gross, 557 U. S., at 176–177. Nor did anything in the statute hint that a different and more forgiving rule might apply at one particular stage in the litigation. In light of the causation standard Congress specified for the cause of action it expressly endorsed, it would be more than a little incongruous for us to employ the laxer rules ESN proposes for this Court’s judicially implied cause of action. Other provisions of the 1866 statute offer further guidance. Not only do we generally presume that Congress legislates against the backdrop of the common law. Nassar, 570 U. S., at 347. The Civil Rights Act of 1866 made this background presumption explicit, providing that “in all cases where [the laws of the United States] are not adapted to the object [of carrying the statute into effect] the common law . . . shall . . . govern said courts in the trial and disposition of such cause.” §3, 14Stat. 27. And, while there were exceptions, the common law in 1866 often treated a showing of but-for causation as a prerequisite to a tort suit. See, e.g., Hayes v. Michigan Central R. Co., 111 U.S. 228, 241 (1884); Smith, Legal Cause in Actions of Tort, 25 Harv. L. Rev. 103, 108–109 (1911); White, The Emergence and Doctrinal Development of Tort Law, 1870–1930, 11 U. St. Thomas L. J. 463, 464–465 (2014); 1 F. Hilliard, Law of Torts 78–79 (1866); 1 T. Sedgwick, Measure of Damages 199 (9th ed. 1912). Nor did this prerequisite normally wait long to make its appearance; if anything, pleadings standards back then were generally even stricter than they are in federal practice today. See generally, e.g., Lugar, Common Law Pleading Modified versus the Federal Rules, 52 W. Va. L. Rev. 137 (1950). This Court’s precedents confirm all that the statute’s language and history indicate. When it first inferred a private cause of action under §1981, this Court described it as “afford[ing] a federal remedy against discrimination . . . on the basis of race,” language (again) strongly suggestive of a but-for causation standard. Johnson, 421 U. S., at 459–460 (emphasis added). Later, in General Building Contractors Assn., Inc. v. Pennsylvania, 458 U.S. 375 (1982), the Court explained that §1981 was “designed to eradicate blatant deprivations of civil rights,” such as where “a private offeror refuse[d] to extend to [an African-American], . . . because he is [an African-American], the same opportunity to enter into contracts as he extends to white offerees.” Id., at 388 (emphasis deleted; internal quotation marks omitted). Once more, the Court spoke of §1981 using language—because of—often associated with but-for causation. Nassar, 570 U. S., at 350. Nor did anything in these decisions even gesture toward the possibility that this rule of causation sometimes might be overlooked or modified in the early stages of a case. This Court’s treatment of a neighboring provision, §1982, supplies a final telling piece of evidence. Because §1982 was also first enacted as part of the Civil Rights Act of 1866 and uses nearly identical language as §1981, the Court’s “precedents have . . . construed §§1981 and 1982 similarly.” CBOCS West, Inc. v. Humphries, 553 U.S. 442, 447 (2008). Section 1982 guarantees all citizens “the same right . . . as is enjoyed by white citizens . . . to inherit, purchase, lease, sell, hold, and convey real and personal property.” And this Court has repeatedly held that a claim arises under §1982 when a citizen is not allowed “to acquire property . . . because of color.” Buchanan v. Warley, 245 U.S. 60, 78–79 (1917) (emphasis added); see also Jones v. Alfred H. Mayer Co., 392 U.S. 409, 419 (1968); Runyon v. McCrary, 427 U.S. 160, 170–171 (1976). If a §1982 plaintiff must show the defendant’s challenged conduct was “because of ” race, it is unclear how we might demand less from a §1981 plaintiff. Certainly ESN offers no compelling reason to read two such similar statutes so differently. B What does ESN offer in reply? The company asks us to draw on, and then innovate with, the “motivating factor” causation test found in Title VII of the Civil Rights Act of 1964. But a critical examination of Title VII’s history reveals more than a few reasons to be wary of any invitation to import its motivating factor test into §1981. This Court first adopted Title VII’s motivating factor test in Price Waterhouse v. Hopkins, 490 U.S. 228 (1989). There, a plurality and two Justices concurring in the judgment held that a Title VII plaintiff doesn’t have to prove but-for causation; instead, it’s enough to show that discrimination was a motivating factor in the defendant’s decision. Id., at 249–250 (plurality opinion); see also id., at 258–259 (White, J., concurring in judgment); id., at 268–269 (O’Connor, J., concurring in judgment). Once a plaintiff meets this lesser standard, the plurality continued, the defendant may defeat liability by establishing that it would have made the same decision even if it had not taken the plaintiff ’s race (or other protected trait) into account. In essence, Price Waterhouse took the burden of proving but-for causation from the plaintiff and handed it to the defendant as an affirmative defense. Id., at 246. But this arrangement didn’t last long. Congress soon displaced Price Waterhouse in favor of its own version of the motivating factor test. In the Civil Rights Act of 1991, Congress provided that a Title VII plaintiff who shows that discrimination was even a motivating factor in the defendant’s challenged employment decision is entitled to declaratory and injunctive relief. §107, 105Stat. 1075. A defendant may still invoke lack of but-for causation as an affirmative defense, but only to stave off damages and reinstatement, not liability in general. 42 U. S. C. §§2000e–2(m), 2000e–5(g)(2)(B); see also Desert Palace, Inc. v. Costa, 539 U.S. 90, 94–95 (2003). While this is all well and good for understanding Title VII, it’s hard to see what any of it might tell us about §1981. Title VII was enacted in 1964; this Court recognized its motivating factor test in 1989; and Congress replaced that rule with its own version two years later. Meanwhile, §1981 dates back to 1866 and has never said a word about motivating factors. So we have two statutes with two distinct histories, and not a shred of evidence that Congress meant them to incorporate the same causation standard. Worse yet, ESN’s fallback position—that we should borrow the motivating factor concept only at the pleadings stage—is foreign even to Title VII practice. To accept ESN’s invitation to consult, tinker with, and then engraft a test from a modern statute onto an old one would thus require more than a little judicial adventurism, and look a good deal more like amending a law than interpreting one. What’s more, it’s not as if Congress forgot about §1981 when it adopted the Civil Rights Act of 1991. At the same time that it added the motivating factor test to Title VII, Congress also amended §1981. See Civil Rights Act of 1991, §101, 105Stat. 1072 (adding new subsections (b) and (c) to §1981). But nowhere in its amendments to §1981 did Congress so much as whisper about motivating factors. And where, as here, Congress has simultaneously chosen to amend one statute in one way and a second statute in another way, we normally assume the differences in language imply differences in meaning. Gross, 557 U. S., at 174–175; see also Russello v. United States, 464 U.S. 16, 23 (1983). Still, ESN tries to salvage something from the 1991 law. It reminds us that one of the amendments to §1981 defined the term “make and enforce contracts” to include “making, performance, modification, and termination of contracts, and the enjoyment of all benefits, privileges, terms, and conditions of the contractual relationship.” 42 U. S. C. §1981(b). In all this, ESN asks us to home in on one word, “making.” By using this particular word, ESN says, Congress clarified that §1981(a) guarantees not only the right to equivalent contractual outcomes (a contract with the same final terms), but also the right to an equivalent contracting process (no extra hurdles on the road to securing that contract). And, ESN continues, if the statute addresses the whole contracting process, not just its outcome, a motivating factor causation test fits more logically than the traditional but-for test. Comcast and the government disagree. As they see it, the Civil Rights Act of 1866 unambiguously protected only outcomes—the right to contract, sue, be a party, and give evidence. When Congress sought to define some of these terms in 1991, it merely repeated one word from the original 1866 Act (make) in a different form (making). No reasonable reader, Comcast and the government contend, would think that the addition of the present participle form of a verb already in the statute carries such a radically different meaning and so extends §1981 liability in the new directions ESN suggests. And, we are told, the statute’s original and continuing focus on contractual outcomes (not processes) is more consistent with the traditional but-for test of causation. This debate, we think, misses the point. Of course, Congress could write an employment discrimination statute to protect only outcomes or to provide broader protection. But, for our purposes today, none of this matters. The difficulty with ESN’s argument lies in its mistaken premise that a process-oriented right necessarily pairs with a motivating factor causal standard. The inverse argument—that an outcome-oriented right implies a but-for causation standard—is just as flawed. Either causal standard could conceivably apply regardless of the legal right §1981 protects. We need not and do not take any position on whether §1981 as amended protects only outcomes or protects processes too, a question not passed on below or raised in the petition for certiorari. Our point is simply that a §1981 plaintiff first must show that he was deprived of the protected right and then establish causation—and that these two steps are analytically distinct.[1] Unable to latch onto either Price Waterhouse or the Civil Rights Act of 1991, ESN is left to cast about for some other hook to support its arguments about §1981’s operation. In a final effort, it asks us to consider the burden-shifting framework of McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 804 (1973). Like the motivating factor test, McDonnell Douglas is a product of Title VII practice. Under its terms, once a plaintiff establishes a prima facie case of race discrimination through indirect proof, the defendant bears the burden of producing a race-neutral explanation for its action, after which the plaintiff may challenge that explanation as pretextual. Texas Dept. of Community Affairs v. Burdine, 450 U.S. 248, 257–258 (1981). This burden shifting, ESN contends, is comparable to the regime it proposes for §1981. It is nothing of the kind. Whether or not McDonnell Douglas has some useful role to play in §1981 cases, it does not mention the motivating factor test, let alone endorse its use only at the pleadings stage. Nor can this come as a surprise: This Court didn’t introduce the motivating factor test into Title VII practice until years after McDonnell Douglas. For its part, McDonnell Douglas sought only to supply a tool for assessing claims, typically at summary judgment, when the plaintiff relies on indirect proof of discrimination. See 411 U. S., at 802–805; see also Furnco Constr. Corp. v. Waters, 438 U.S. 567, 577 (1978); Malamud, The Last Minuet: Disparate Treatment After Hicks, 93 Mich. L. Rev. 2229, 2259 (1995). Because McDonnell Douglas arose in a context where but-for causation was the undisputed test, it did not address causation standards. So nothing in the opinion involves ESN’s preferred standard. Under McDonnell Douglas’s terms, too, only the burden of production ever shifts to the defendant, never the burden of persuasion. See Burdine, 450 U. S., at 254–255; Postal Service Bd. of Governors v. Aikens, 460 U.S. 711, 715–716 (1983). So McDonnell Douglas can provide no basis for allowing a complaint to survive a motion to dismiss when it fails to allege essential elements of a plaintiff’s claim. III All the traditional tools of statutory interpretation persuade us that §1981 follows the usual rules, not any exception. To prevail, a plaintiff must initially plead and ultimately prove that, but for race, it would not have suffered the loss of a legally protected right. We do not, however, pass on whether ESN’s operative amended complaint “contain[s] sufficient factual matter, accepted as true, to ‘state a claim to relief that is plausible on its face’ ” under the but-for causation standard. Iqbal, 556 U. S., at 678–679. The Ninth Circuit has yet to consider that question because it assessed ESN’s pleadings under a different and mistaken test. To allow that court the chance to determine the sufficiency of ESN’s pleadings under the correct legal rule in the first instance, we vacate the judgment of the court of appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 The concurrence proceeds to offer a view on the nature of the right, while correctly noting that the Court reserves the question for another day. We reserve the question because “we are a court of review, not of first view,” Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7 (2005), and do not normally strain to address issues that are less than fully briefed and that the district and appellate courts have had no opportunity to consider. Such restraint is particularly appropriate here, where addressing the issue is entirely unnecessary to our resolution of the case.

590.US.2019_18-260

The Clean Water Act forbids “any addition” of any pollutant from “any point source” to “navigable waters” without an appropriate permit from the Environmental Protection Agency (EPA). §§ 301(a), 502(12), 86Stat. 844, 886. The Act defines “pollutant” broadly, §502(6); defines a “point source” as “ ‘any discernible, confined and discrete conveyance . . . from which pollutants are or may be discharged,’ ” including, e.g., any “ ‘container,’ ” “ ‘pipe, ditch, channel, tunnel, conduit,’ ” or “ ‘well,’ ” §502(14); and defines the term “discharge of a pollutant” as “ ‘any addition of any pollutant to navigable waters [including navigable streams, rivers, the ocean, or coastal waters] from any point source,’ ” §502(12). It then uses those terms in making “unlawful” “ ‘the discharge of any pollutant by any person’ ” without an appropriate permit. §301. Petitioner County of Maui’s wastewater reclamation facility collects sewage from the surrounding area, partially treats it, and each day pumps around 4 million gallons of treated water into the ground through four wells. This effluent then travels about a half mile, through groundwater, to the Pacific Ocean. Respondent environmental groups brought a citizens’ Clean Water Act suit, alleging that Maui was “discharg[ing]” a “pollutant” to “navigable waters” without the required permit. The District Court found that the discharge from Maui’s wells into the nearby groundwater was “functionally one into navigable water,” 24 F. Supp. 3d 980, 998, and granted summary judgment to the environmental groups. The Ninth Circuit affirmed, stating that a permit is required when “pollutants are fairly traceable from the point source to a navigable water.” 886 F.3d 737, 749. Held: The statutory provisions at issue require a permit when there is a direct discharge from a point source into navigable waters or when there is the functional equivalent of a direct discharge. Pp. 4–18. (a) Statutory context limits the reach of the phrase “from any point source” to a range of circumstances narrower than that which the Ninth Circuit’s “fairly traceable” interpretation suggests. At the same time, it is significantly broader than the total exclusion of all discharges through groundwater, as urged by Maui and by the Solicitor General as amicus curiae. Pp. 4–5. (b) The Ninth Circuit’s “fairly traceable” limitation could allow EPA to assert permitting authority over the release of pollutants that reach navigable waters many years after their release. But Congress did not intend to provide EPA with such broad authority. First, to interpret “from” so broadly might require a permit in unexpected circumstances, such as, e.g., the 100-year migration of pollutants through 250 miles of groundwater to a river. Second, the statute’s structure indicates that, as to groundwater pollution and nonpoint source pollution, Congress left substantial responsibility and autonomy to the States and did not give EPA authority that could seriously interfere with this state responsibility. Third, the Act’s legislative history strongly supports the conclusion that the permitting provision does not extend so far. Finally, longstanding regulatory practice shows that EPA has successfully applied the permitting provision to pollution discharges from point sources that reached navigable waters through groundwater using a narrower interpretation than that of the Ninth Circuit. Pp. 5–10. (c) Maui, the Government, and the two dissents argue for interpretations that, in light of the statute’s language, structure, and purposes, are also too extreme. Pp. 10–15. (1) Maui and the Solicitor General argue that the statute’s permitting requirement does not apply if a pollutant, having emerged from a “point source,” must travel through any amount of groundwater before reaching navigable waters. That narrow interpretation would risk serious interference with EPA’s ability to regulate point source discharges, and Congress would not have intended to create such a large and obvious loophole in one of the Clean Water Act’s key regulatory innovations. P. 10. (2) Reading “from” in the phrase “from any point source” together with “conveyance” in the point source definition “any . . . conveyance,” Maui argues that the meaning of “from any point source” is not about where the pollution originated, but about how it got there. Thus, Maui claims, a permit is required only if a point source ultimately delivers the pollutant to navigable waters. By contrast, if a pollutant travels through groundwater, then the groundwater is the conveyance and no permit is required. But Maui’s definition of “from” as connoting a means does not fit in context. Coupling “from” with “to” is strong evidence that Congress was referring to a destination (“navigable waters”) and an origin (“any point source”). That Maui’s reading would create a serious loophole in the permitting regime also indicates that it is unreasonable. Pp. 10–11. (3) The Solicitor General argues that the proper interpretation of the statute is the one reflected in EPA’s recent Interpretive Statement, namely, that “all releases of pollutants to groundwater” are excluded from the scope of the permitting program, “even where pollutants are conveyed to jurisdictional surface waters via groundwater.” 84 Fed. Reg. 16810, 16811. That reading, which would open a loophole allowing easy evasion of the statutory provision’s basic purposes, is neither persuasive nor reasonable. EPA is correct that Congress did not require a permit for all discharges to groundwater, and it did authorize study and funding related to groundwater pollution. But the most that the study and funding provisions show is that Congress thought that the problem of pollution in groundwater would primarily be addressed by the States or perhaps by other federal statutes. EPA’s new interpretation is also difficult to reconcile with the statute’s reference to “any addition” of a pollutant to navigable waters; with the statute’s inclusion of “wells” in the “point source” definition, since wells would ordinarily discharge pollutants through groundwater; and with statutory provisions that allow EPA to delegate its permitting authority to a State only if the State, inter alia, provides “ ‘adequate authority’ ” to “ ‘control the disposal of pollutants into wells,’ ” §402(b). Pp. 11–13. (4) Perhaps, as the dissents suggest, the statute’s language could be narrowed by reading the statute to refer only to the pollutant’s immediate origin, but there is no linguistic basis for this limitation. Pp. 13–15. (d) The statute’s words reflect Congress’ basic aim to provide federal regulation of identifiable sources of pollutants entering navigable waters without undermining the States’ longstanding regulatory authority over land and groundwater. The reading of the statute that best captures Congress’ meaning, reflected in the statute’s words, structure, and purposes, is that a permit is required when there is a discharge from a point source directly into navigable waters or when there is the functional equivalent of a direct discharge. Many factors may be relevant to determining whether a particular discharge is the functional equivalent of one directly into navigable waters. Time and distance will be the most important factors in most cases, but other relevant factors may include, e.g., the nature of the material through which the pollutant travels and the extent to which the pollutant is diluted or chemically changed as it travels. Courts will provide additional guidance through decisions in individual cases. The underlying statutory objectives can also provide guidance, and EPA can provide administrative guidance. Although this interpretation does not present as clear a line as the other interpretations proffered, the EPA has applied the permitting provision to some discharges through groundwater for over 30 years, with no evidence of inadministrability or an unmanageable expansion in the statute’s scope. Pp. 15–18. 886 F.3d 737, vacated and remanded. Breyer, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Sotomayor, Kagan, and Kavanaugh, JJ., joined. Kav- anaugh, J., filed a concurring opinion. Thomas, J., filed a dissenting opinion, in which Gorsuch, J., joined. Alito, J., filed a dissenting opinion.

The Clean Water Act forbids the “addition” of any pollutant from a “point source” to “navigable waters” without the appropriate permit from the Environmental Protection Agency (EPA). Federal Water Pollution Control Act, §§301(a), 502(12)(A), as amended by the Federal Water Pollution Control Act Amendments of 1972 (Clean Water Act) §2, 86Stat. 844, 886, 33 U. S. C. §§1311(a), 1362(12)(A). The question presented here is whether the Act “requires a permit when pollutants originate from a point source but are conveyed to navigable waters by a nonpoint source,” here, “groundwater.” Pet. for Cert. i. Suppose, for example, that a sewage treatment plant discharges polluted water into the ground where it mixes with groundwater, which, in turn, flows into a navigable river, or perhaps the ocean. Must the plant’s owner seek an EPA permit before emitting the pollutant? We conclude that the statutory provisions at issue require a permit if the addition of the pollutants through groundwater is the functional equivalent of a direct discharge from the point source into navigable waters. I A Congress’ purpose as reflected in the language of the Clean Water Act is to “ ‘restore and maintain the . . . integrity of the Nation’s waters,’ ” §101(a), 86Stat. 816. Prior to the Act, Federal and State Governments regulated water pollution in large part by setting water quality standards. See EPA v. California ex rel. State Water Resources Control Bd., 426 U.S. 200, 202–203 (1976). The Act restructures federal regulation by insisting that a person wishing to discharge any pollution into navigable waters first obtain EPA’s permission to do so. See id., at 203–205; Milwaukee v. Illinois, 451 U.S. 304, 310–311 (1981). The Act’s provisions use specific definitional language to achieve this result. First, the Act defines “pollutant” broadly, including in its definition, for example, any solid waste, incinerator residue, “ ‘heat,’ ” “ ‘discarded equipment,’ ” or sand (among many other things). §502(6), 86Stat. 886. Second, the Act defines a “point source” as “ ‘any discernible, confined and discrete conveyance . . . from which pollutants are or may be discharged,’ ” including, for example, any “ ‘container,’ ” “ ‘pipe, ditch, channel, tunnel, conduit,’ ” or “ ‘well.’ ” §502(14), id., at 887. Third, it defines the term “discharge of a pollutant” as “ ‘any addition of any pollutant to navigable waters [including navigable streams, rivers, the ocean, or coastal waters] from any point source.’ ” §502(12), id., at 886. The Act then sets forth a statutory provision that, using these terms, broadly states that (with certain exceptions) “ ‘the discharge of any pollutant by any person’ ” without an appropriate permit “ ‘shall be unlawful.’ ” §301, id., at 844. The question here, as we have said, is whether, or how, this statutory language applies to a pollutant that reaches navigable waters only after it leaves a “point source” and then travels through groundwater before reaching navigable waters. In such an instance, has there been a “discharge of a pollutant,” that is, has there been “any addition of any pollutant to navigable waters from any point source? ” B The petitioner, the County of Maui, operates a wastewater reclamation facility on the island of Maui, Hawaii. The facility collects sewage from the surrounding area, partially treats it, and pumps the treated water through four wells hundreds of feet underground. This effluent, amounting to about 4 million gallons each day, then travels a further half mile or so, through groundwater, to the ocean. In 2012, several environmental groups, the respondents here, brought this citizens’ Clean Water Act lawsuit against Maui. See §505(a), id., at 888. They claimed that Maui was “discharg[ing]” a “pollutant” to “navigable waters,” namely, the Pacific Ocean, without the permit required by the Clean Water Act. The District Court, relying in part upon a detailed study of the discharges, found that a considerable amount of effluent from the wells ended up in the ocean (a navigable water). It wrote that, because the “path to the ocean is clearly ascertainable,” the discharge from Maui’s wells into the nearby groundwater was “functionally one into navigable water.” 24 F. Supp. 3d 980, 998 (Haw. 2014). And it granted summary judgment in favor of the environmental groups. See id., at 1005. The Ninth Circuit affirmed the District Court, but it described the relevant statutory standard somewhat differently. The appeals court wrote that a permit is required when “the pollutants are fairly traceable from the point source to a navigable water such that the discharge is the functional equivalent of a discharge into the navigable water.” 886 F.3d 737, 749 (2018) (emphasis added). The court left “for another day the task of determining when, if ever, the connection between a point source and a navigable water is too tenuous to support liability . . . .” Ibid. Maui petitioned for certiorari. In light of the differences in the standards adopted by the different Courts of Appeals, we granted the petition. Compare, e.g., 886 F. 3d, at 749 (“fairly traceable”), with Upstate Forever v. Kinder Morgan Energy Partners, L. P., 887 F.3d 637, 651 (CA4 2018) (“direct hydrological connection”), and Kentucky Waterways Alliance v. Kentucky Util. Co., 905 F.3d 925, 932–938 (CA6 2018) (discharges through groundwater are excluded from the Act’s permitting requirements). II The linguistic question here concerns the statutory word “from.” Is pollution that reaches navigable waters only through groundwater pollution that is “from” a point source, as the statute uses the word? The word “from” is broad in scope, but context often imposes limitations. “Finland,” for example, is often not the right kind of answer to the question, “Where have you come from?” even if long ago you were born there. The parties here disagree dramatically about the scope of the word “from” in the present context. The environmental groups, the respondents, basically adopt the Ninth Circuit’s view—that the permitting requirement applies so long as the pollutant is “fairly traceable” to a point source even if it traveled long and far (through groundwater) before it reached navigable waters. They add that the release from the point source must be “a proximate cause of the addition of pollutants to navigable waters.” Brief for Respondents 20. Maui, on the other hand, argues that the statute creates a “bright-line test.” Brief for Petitioner 27–28. A point source or series of point sources must be “the means of delivering pollutants to navigable waters.” Id., at 28. They add that, if “at least one nonpoint source (e.g., unconfined rainwater runoff or groundwater)” lies “between the point source and the navigable water,” then the permit requirement “does not apply.” Id., at 54. A pollutant is “from” a point source only if a point source is the last “conveyance” that conducted the pollutant to navigable waters. The Solicitor General, as amicus curiae, supports Maui, at least in respect to groundwater. Reiterating the position taken in a recent EPA “Interpretive Statement,” see 84 Fed. Reg. 16810 (2019), he argues that, given the Act’s structure and history, “a release of pollutants to groundwater is not subject to” the Act’s permitting requirement “even if the pollutants subsequently migrate to jurisdictional surface waters,” such as the ocean. Brief for United States as Amicus Curiae 12 (capitalization omitted). We agree that statutory context limits the reach of the statutory phrase “from any point source” to a range of circumstances narrower than that which the Ninth Circuit’s interpretation suggests. At the same time, it is significantly broader than the total exclusion of all discharges through groundwater described by Maui and the Solicitor General. III Virtually all water, polluted or not, eventually makes its way to navigable water. This is just as true for groundwater. See generally 2 Van Nostrand’s Scientific Encyclopedia 2600 (10th ed. 2008) (defining “Hydrology”). Given the power of modern science, the Ninth Circuit’s limitation, “fairly traceable,” may well allow EPA to assert permitting authority over the release of pollutants that reach navigable waters many years after their release (say, from a well or pipe or compost heap) and in highly diluted forms. See, e.g., Brief for Aquatic Scientists et al. as Amici Curiae 13–28. The respondents suggest that the standard can be narrowed by adding a “proximate cause” requirement. That is, to fall within the permitting provision, the discharge from a point source must “proximately cause” the pollutants’ eventual addition to navigable waters. But the term “proximate cause” derives from general tort law, and it takes on its specific content based primarily on “policy” considerations. See CSX Transp., Inc. v. McBride, 564 U.S. 685, 701 (2011) (plurality opinion). In the context of water pollution, we do not see how it significantly narrows the statute beyond the words “fairly traceable” themselves. Our view is that Congress did not intend the point source-permitting requirement to provide EPA with such broad authority as the Ninth Circuit’s narrow focus on traceability would allow. First, to interpret the word “from” in this literal way would require a permit in surprising, even bizarre, circumstances, such as for pollutants carried to navigable waters on a bird’s feathers, or, to mention more mundane instances, the 100-year migration of pollutants through 250 miles of groundwater to a river. Second, and perhaps most important, the structure of the statute indicates that, as to groundwater pollution and nonpoint source pollution, Congress intended to leave substantial responsibility and autonomy to the States. See, e.g., §101(b), 86Stat. 816 (stating Congress’ purpose in this regard). Much water pollution does not come from a readily identifiable source. See 3 Van Nostrand’s Scientific Encyclopedia, at 5801 (defining “Water Pollution”). Rainwater, for example, can carry pollutants (say, as might otherwise collect on a roadway); it can pollute groundwater, and pollution collected by unchanneled rainwater runoff is not ordinarily considered point source pollution. Over many decades, and with federal encouragement, the States have developed methods of regulating nonpoint source pollution through water quality standards, and otherwise. See, e.g., Nonpoint Source Program, Annual Report (California) 6 (2016–2017) (discussing state timberland management programs to address addition of sediment-pollutants to navigable waters); id., at 10–11 (discussing regulations of vineyards to control water pollution); id. at 17–19 (discussing livestock grazing management, including utilization ratios and time restrictions); Nonpoint Source Management Program, Annual Report (Maine) 8–10 (2018) (discussing installation of livestock fencing and planting of vegetation to reduce nonpoint source pollution); Oklahoma’s Nonpoint Source Management Program, Annual Report 5, 14 (2017) (discussing program to encourage voluntary no-till farming to reduce sediment pollution). The Act envisions EPA’s role in managing nonpoint source pollution and groundwater pollution as limited to studying the issue, sharing information with and collecting information from the States, and issuing monetary grants. See §§105, 208, 86Stat. 825, 839; see also Water Quality Act of 1987, §316, 101Stat. 52 (establishing Nonpoint Source Management Programs). Although the Act grants EPA specific authority to regulate certain point source pollution (it can also delegate some of this authority to the States acting under EPA supervision, see §402(b), 86Stat. 880), these permitting provisions refer to “point sources” and “navigable waters,” and say nothing at all about nonpoint source regulation or groundwater regulation. We must doubt that Congress intended to give EPA the authority to apply the word “from” in a way that could interfere as seriously with States’ traditional regulatory authority—authority the Act preserves and promotes—as the Ninth Circuit’s “fairly traceable” test would. Third, those who look to legislative history to help interpret a statute will find that this Act’s history strongly supports our conclusion that the permitting provision does not extend so far. Fifty years ago, when Congress was considering the bills that became the Clean Water Act, William Ruckelshaus, the first EPA Administrator, asked Congress to grant EPA authority over “ground waters” to “assure that we have control over the water table . . . so we can . . . maintai[n] a control over all the sources of pollution, be they discharged directly into any stream or through the ground water table.” Water Pollution Control Legislation–1971 (Proposed Amendments to Existing Legislation): Hearings before the House Committee on Public Works, 92d Cong., 1st Sess., 230 (1971). Representative Les Aspin similarly pointed out that there were “conspicuou[s ]” references to groundwater in all sections of the bill except the permitting section at issue here. Water Pollution Control Legislation–1971: Hearings before the House Committee on Public Works on H. R. 11896 and H. R. 11895, 92d Cong., 1st Sess., 727 (1972). The Senate Committee on Public Works “recognize[d] the essential link between ground and surface waters.” S. Rep. No. 92–414, p. 73 (1971). But Congress did not accept these requests for general EPA authority over groundwater. It rejected Representative Aspin’s amendment that would have extended the permitting provision to groundwater. Instead, Congress provided a set of more specific groundwater-related measures such as those requiring States to maintain “affirmative controls over the injection or placement in wells” of “any pollutants that may affect ground water.” Ibid. These specific state-related programs were, in the words of the Senate Public Works Committee, “designed to protect ground waters and eliminate the use of deep well disposal as an uncontrolled alternative to toxic and pollution control.” Ibid. The upshot is that Congress was fully aware of the need to address groundwater pollution, but it satisfied that need through a variety of state-specific controls. Congress left general groundwater regulatory authority to the States; its failure to include groundwater in the general EPA permitting provision was deliberate. Finally, longstanding regulatory practice undermines the Ninth Circuit’s broad interpretation of the statute. EPA itself for many years has applied the permitting provision to pollution discharges from point sources that reached navigable waters only after traveling through groundwater. See, e.g., United States Steel Corp. v. Train, 556 F.2d 822, 832 (CA7 1977) (permit for “deep waste-injection well” on the shore of navigable waters). But, in doing so, EPA followed a narrower interpretation than that of the Ninth Circuit. See, e.g., In re Bethlehem Steel Corp., 2 E. A. D. 715, 718 (EAB 1989) (Act’s permitting requirement applies only to injection wells “that inject into ground water with a physically and temporally direct hydrologic connection to surface water”). EPA has opposed applying the Act’s permitting requirements to discharges that reach groundwater only after lengthy periods. See McClellan Ecological Seepage Situation (MESS) v. Cheney, 763 F. Supp. 431, 437 (ED Cal. 1989) (United States argued that permitting provisions do not apply when it would take “literally dozens, and perhaps hundreds, of years for any pollutants” to reach navigable waters); Greater Yellowstone Coalition v. Larson, 641 F. Supp. 2d 1120, 1139 (Idaho 2009) (same in respect to instances where it would take “between 60 and 420 years” for pollutants to travel “one to four miles” through groundwater before reaching navigable waters). Indeed, in this very case (prior to its recent Interpretive Statement, see infra, at 12–13), EPA asked the Ninth Circuit to apply a more limited “direct hydrological connection” test. See Brief for United States as Amicus Curiae in No. 15–17447 (CA9), pp. 13–20. The Ninth Circuit did not accept this suggestion. We do not defer here to EPA’s interpretation of the statute embodied in this practice. Indeed, EPA itself has changed its mind about the meaning of the statutory provision. See infra, at 12–14. But this history, by showing that a comparatively narrow view of the statute is administratively workable, offers some additional support for the view that Congress did not intend as broad a delegation of regulatory authority as the Ninth Circuit test would allow. As we have said, the specific meaning of the word “from” necessarily draws its meaning from context. The apparent breadth of the Ninth Circuit’s “fairly traceable” approach is inconsistent with the context we have just described. IV A Maui and the Solicitor General argue that the statute’s permitting requirement does not apply if a pollutant, having emerged from a “point source,” must travel through any amount of groundwater before reaching navigable waters. That interpretation is too narrow, for it would risk serious interference with EPA’s ability to regulate ordinary point source discharges. Consider a pipe that spews pollution directly into coastal waters. There is an “addition of ” a “pollutant to navigable waters from [a] point source.” Hence, a permit is required. But Maui and the Government read the permitting requirement not to apply if there is any amount of groundwater between the end of the pipe and the edge of the navigable water. See Tr. of Oral Arg. 5–6, 24–25. If that is the correct interpretation of the statute, then why could not the pipe’s owner, seeking to avoid the permit requirement, simply move the pipe back, perhaps only a few yards, so that the pollution must travel through at least some groundwater before reaching the sea? Cf. Brief for State of Maryland et al. as Amici Curiae 9, n. 4. We do not see how Congress could have intended to create such a large and obvious loophole in one of the key regulatory innovations of the Clean Water Act. Cf. California ex rel. State Water Resources Control Bd., 426 U. S., at 202–204 (basic purpose of Clean Water Act is to regulate pollution at its source); The Emily, 9 Wheat. 381, 390 (1824) (rejecting an interpretation that would facilitate “evasion of the law”). B Maui argues that the statute’s language requires its reading. That language requires a permit for a “discharge.” A “discharge” is “any addition” of a pollutant to navigable waters “from any point source.” And a “point source” is “any discernible, confined and discrete conveyance” (such as a pipe, ditch, well, etc.). Reading “from” and “conveyance” together, Maui argues that the statutory meaning of “from any point source” is not about where the pollution originated, but about how it got there. Under what Maui calls the means-of-delivery test, a permit is required only if a point source itself ultimately delivers the pollutant to navigable waters. Under this view, if the pollutant must travel through groundwater to reach navigable waters, then it is the groundwater, not the pipe, that is the conveyance. Congress sometimes adopts less common meanings of common words, but this esoteric definition of “from,” as connoting a means, does not remotely fit in this context. The statute couples the word “from” with the word “to”—strong evidence that Congress was referring to a destination (“navigable waters”) and an origin (“any point source”). Further underscoring that Congress intended this every day meaning is that the object of “from” is a “point source”—a source, again, connoting an origin. That Maui’s proffered interpretation would also create a serious loophole in the permitting regime also indicates it is an unreasonable one. C The Solicitor General agrees that, as a general matter, the permitting requirement applies to at least some additions of pollutants to navigable waters that come indirectly from point sources. See Brief for United States as Amicus Curiae 33–35. But the Solicitor General argues that the proper interpretation of the statute is the one reflected in EPA’s recent Interpretive Statement. After receiving more than 50,000 comments from the public, and after the Ninth Circuit released its opinion in this case, EPA wrote that “the best, if not the only, reading” of the statutory provisions is that “all releases of pollutants to groundwater” are excluded from the scope of the permitting program, “even where pollutants are conveyed to jurisdictional surface waters via groundwater.” 84 Fed. Reg. 16810, 16811. Neither the Solicitor General nor any party has asked us to give what the Court has referred to as Chevron deference to EPA’s interpretation of the statute. See Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 844 (1984). Even so, we often pay particular attention to an agency’s views in light of the agency’s expertise in a given area, its knowledge gained through practical experience, and its familiarity with the interpretive demands of administrative need. See United States v. Mead Corp., 533 U.S. 218, 234–235 (2001); Skidmore v. Swift & Co., 323 U.S. 134, 139–140 (1944). But here, as we have explained, to follow EPA’s reading would open a loophole allowing easy evasion of the statutory provision’s basic purposes. Such an interpretation is neither persuasive nor reasonable. EPA correctly points out that Congress did not require a permit for all discharges to groundwater; rather, Congress authorized study and funding related to groundwater pollution. See Brief for United States as Amicus Curiae 15–19. But there is quite a gap between “not all” and “none.” The statutory text itself alludes to no exception for discharges through groundwater. These separate provisions for study and funding that EPA points to would be a “surprisingly indirect route” to convey “an important and easily expressed message”—that the permit requirement simply does not apply if the pollutants travel through groundwater. Landgraf v. USI Film Products, 511 U.S. 244, 262 (1994). In truth, the most these provisions show is that Congress thought that the problem of groundwater pollution, as distinct from navigable water pollution, would primarily be addressed by the States or perhaps by other federal statutes. EPA’s new interpretation is also difficult to reconcile with the statute’s reference to “any addition” of a pollutant to navigable waters. Cf. Milwaukee, 451 U. S., at 318 (“Every point source discharge is prohibited unless covered by a permit” (footnote omitted)). It is difficult to reconcile EPA’s interpretation with the statute’s inclusion of “wells” in the definition of “point source,” for wells most ordinarily would discharge pollutants through groundwater. And it is difficult to reconcile EPA’s interpretation with the statutory provisions that allow EPA to delegate its permitting authority to a State only if the State (among other things) provides “ ‘adequate authority’ ” to “ ‘control the disposal of pollutants into wells.’ ” §402(b), 86Stat. 881. What need would there be for such a proviso if the federal permitting program the State replaces did not include such discharges (from wells through groundwater) in the first place? In short, EPA’s oblique argument about the statute’s references to groundwater cannot overcome the statute’s structure, its purposes, or the text of the provisions that actually govern. D Perhaps, as the two dissents suggest, the language could be narrowed to similar effect by reading the statute to refer only to the pollutant’s immediate origin. See post, at 2–3 (opinion of Thomas, J.); post, at 8 (opinion of Alito, J.). But there is no linguistic basis here to so limit the statute in that way. Again, whether that is the correct reading turns on context. Justice Thomas insists that in the case of a discharge through groundwater, the pollutants are added “from the groundwater.” Post, at 2. Indeed, but that does not mean they are not also “from the point source.” Ibid. When John comes to the hotel, John might have come from the train station, from Baltimore, from Europe, from any two of those three places, or from all three. A sign that asks all persons who arrive from Baltimore to speak to the desk clerk includes those who took a taxi from the train station. There is nothing unnatural about such a construction. As the plurality correctly noted in Rapanos v. United States, 547 U.S. 715 (2006), the statute here does not say “directly” from or “immediately” from. Id., at 743 (opinion of Scalia, J.). Indeed, the expansive language of the provision—any addition from any point source—strongly suggests its scope is not so limited. Justice Alito appears to believe that there are only two possible ways to read “from”: as referring either to the immediate source, or else to the original source. Post, at 5, 8. Because he agrees that the statute cannot reasonably be read always to reach the original source, he concludes the statute must refer only to the immediate origin. But as the foregoing example illustrates, context may indicate that “from” includes an intermediate stop—Baltimore, not Europe or the train station. Justice Thomas relies on the word “addition,” but we fail to see how that word limits the statute to discharges directly to navigable waters. Ordinary language abounds in counter examples: A recipe might instruct to “add the drippings from the meat to the gravy”; that instruction does not become incomprehensible, or even peculiar, simply because the drippings will have first collected in a pan or on a cutting board. And while it would be an unusual phrasing (as statutory phrasings often are), we do not see how the recipe’s meaning would transform if it instead said to “add the drippings to the gravy from the meat.” To take another example: If Timmy is told to “add water to the bath from the well” he will know just what it means—even though he will have to use a bucket to complete the task. And although Justice Thomas resists the inevitable implications of his reading of the statute, post, at 5–6, that reading would create the same loopholes as those offered by the petitioner and the Government, and more. It would necessarily exclude a pipe that drains onto the beach next to navigable waters, even if the pollutants then flow to those waters. It also seems to exclude a pipe that hangs out over the water and adds pollutants to the air, through which the pollutants fall to navigable waters. The absurdity of such an interpretation is obvious enough. We therefore reject this reading as well: Like Maui’s and the Government’s, it is inconsistent with the statutory text and simultaneously creates a massive loophole in the permitting scheme that Congress established. E For the reasons set forth in Part III and in this Part, we conclude that, in light of the statute’s language, structure, and purposes, the interpretations offered by the parties, the Government, and the dissents are too extreme. V Over the years, courts and EPA have tried to find general language that will reflect a middle ground between these extremes. The statute’s words reflect Congress’ basic aim to provide federal regulation of identifiable sources of pollutants entering navigable waters without undermining the States’ longstanding regulatory authority over land and groundwater. We hold that the statute requires a permit when there is a direct discharge from a point source into navigable waters or when there is the functional equivalent of a direct discharge. We think this phrase best captures, in broad terms, those circumstances in which Congress intended to require a federal permit. That is, an addition falls within the statutory requirement that it be “from any point source” when a point source directly deposits pollutants into navigable waters, or when the discharge reaches the same result through roughly similar means. Time and distance are obviously important. Where a pipe ends a few feet from navigable waters and the pipe emits pollutants that travel those few feet through groundwater (or over the beach), the permitting requirement clearly applies. If the pipe ends 50 miles from navigable waters and the pipe emits pollutants that travel with groundwater, mix with much other material, and end up in navigable waters only many years later, the permitting requirements likely do not apply. The object in a given scenario will be to advance, in a manner consistent with the statute’s language, the statutory purposes that Congress sought to achieve. As we have said (repeatedly), the word “from” seeks a “point source” origin, and context imposes natural limits as to when a point source can properly be considered the origin of pollution that travels through groundwater. That context includes the need, reflected in the statute, to preserve state regulation of groundwater and other nonpoint sources of pollution. Whether pollutants that arrive at navigable waters after traveling through groundwater are “from” a point source depends upon how similar to (or different from) the particular discharge is to a direct discharge. The difficulty with this approach, we recognize, is that it does not, on its own, clearly explain how to deal with middle instances. But there are too many potentially relevant factors applicable to factually different cases for this Court now to use more specific language. Consider, for example, just some of the factors that may prove relevant (depending upon the circumstances of a particular case): (1) transit time, (2) distance traveled, (3) the nature of the material through which the pollutant travels, (4) the extent to which the pollutant is diluted or chemically changed as it travels, (5) the amount of pollutant entering the navigable waters relative to the amount of the pollutant that leaves the point source, (6) the manner by or area in which the pollutant enters the navigable waters, (7) the degree to which the pollution (at that point) has maintained its specific identity. Time and distance will be the most important factors in most cases, but not necessarily every case. At the same time, courts can provide guidance through decisions in individual cases. The Circuits have tried to do so, often using general language somewhat similar to the language we have used. And the traditional common-law method, making decisions that provide examples that in turn lead to ever more refined principles, is sometimes useful, even in an era of statutes. The underlying statutory objectives also provide guidance. Decisions should not create serious risks either of undermining state regulation of groundwater or of creating loopholes that undermine the statute’s basic federal regulatory objectives. EPA, too, can provide administrative guidance (within statutory boundaries) in numerous ways, including through, for example, grants of individual permits, promulgation of general permits, or the development of general rules. Indeed, over the years, EPA and the States have often considered the Act’s application to discharges through groundwater. Both Maui and the Government object that to subject discharges to navigable waters through groundwater to the statute’s permitting requirements, as our interpretation will sometimes do, would vastly expand the scope of the statute, perhaps requiring permits for each of the 650,000 wells like petitioner’s or for each of the over 20 million septic systems used in many Americans’ homes. Brief for Petitioner 44–48; Brief for United States as Amicus Curiae 24–25. Cf. Utility Air Regulatory Group v. EPA, 573 U.S. 302, 324 (2014). But EPA has applied the permitting provision to some (but not to all) discharges through groundwater for over 30 years. See supra, at 8–9. In that time we have seen no evidence of unmanageable expansion. EPA and the States also have tools to mitigate those harms, should they arise, by (for example) developing general permits for recurring situations or by issuing permits based on best practices where appropriate. See, e.g., 40 CFR §122.44(k) (2019). Judges, too, can mitigate any hardship or injustice when they apply the statute’s penalty provision. That provision vests courts with broad discretion to set a penalty that takes account of many factors, including “any good-faith efforts to comply” with the Act, the “seriousness of the violation,” the “economic impact of the penalty on the violator,” and “such other matters as justice may require.” See 33 U. S. C. §1319(d). We expect that district judges will exercise their discretion mindful, as we are, of the complexities inherent to the context of indirect discharges through groundwater, so as to calibrate the Act’s penalties when, for example, a party could reasonably have thought that a permit was not required. In sum, we recognize that a more absolute position, such as the means-of-delivery test or that of the Government or that of the Ninth Circuit, may be easier to administer. But, as we have said, those positions have consequences that are inconsistent with major congressional objectives, as revealed by the statute’s language, structure, and purposes. We consequently understand the permitting requirement, §301, as applicable to a discharge (from a point source) of pollutants that reach navigable waters after traveling through groundwater if that discharge is the functional equivalent of a direct discharge from the point source into navigable waters. VI Because the Ninth Circuit applied a different standard, we vacate its judgment and remand the case for further proceedings consistent with this opinion. It is so ordered.

591.US.2019_18-587

In 2012, the Department of Homeland Security (DHS) issued a memorandum announcing an immigration relief program known as Deferred Action for Childhood Arrivals (DACA), which allows certain unauthorized aliens who arrived in the United States as children to apply for a two-year forbearance of removal. Those granted such relief become eligible for work authorization and various federal benefits. Some 700,000 aliens have availed themselves of this opportunity. Two years later, DHS expanded DACA eligibility and created a related program known as Deferred Action for Parents of Americans and Lawful Permanent Residents (DAPA). If implemented, that program would have made 4.3 million parents of U. S. citizens or lawful permanent residents eligible for the same forbearance from removal, work eligibility, and other benefits as DACA recipients. Texas, joined by 25 other States, secured a nationwide preliminary injunction barring implementation of both the DACA expansion and DAPA. The Fifth Circuit upheld the injunction, concluding that the program violated the Immigration and Nationality Act (INA), which carefully defines eligibility for benefits. This Court affirmed by an equally divided vote, and the litigation then continued in the District Court. In June 2017, following a change in Presidential administrations, DHS rescinded the DAPA Memorandum, citing, among other reasons, the ongoing suit by Texas and new policy priorities. That September, the Attorney General advised Acting Secretary of Homeland Security Elaine C. Duke that DACA shared DAPA’s legal flaws and should also be rescinded. The next day, Duke acted on that advice. Taking into consideration the Fifth Circuit and Supreme Court rulings and the Attorney General’s letter, Duke decided to terminate the program. She explained that DHS would no longer accept new applications, but that existing DACA recipients whose benefits were set to expire within six months could apply for a two-year renewal. For all other DACA recipients, previously issued grants of relief would expire on their own terms, with no prospect for renewal. Several groups of plaintiffs challenged Duke’s decision to rescind DACA, claiming that it was arbitrary and capricious in violation of the Administrative Procedure Act (APA) and infringed the equal protection guarantee of the Fifth Amendment’s Due Process Clause. District Courts in California (Regents, No. 18–587), New York (Batalla Vidal, No. 18–589), and the District of Columbia (NAACP, No. 18–588) all ruled for the plaintiffs. Each court rejected the Government’s arguments that the claims were unreviewable under the APA and that the INA deprived the courts of jurisdiction. In Regents and Batalla Vidal, the District Courts further held that the equal protection claims were adequately alleged, and they entered coextensive nationwide preliminary injunctions based on the conclusion that the plaintiffs were likely to succeed on their APA claims. The District Court in NAACP took a different approach. It deferred ruling on the equal protection challenge but granted partial summary judgment to the plaintiffs on their APA claim, finding that the rescission was inadequately explained. The court then stayed its order for 90 days to permit DHS to reissue a memorandum rescinding DACA, this time with a fuller explanation of the conclusion that DACA was unlawful. Two months later, Duke’s successor, Secretary Kirstjen M. Nielsen, responded to the court’s order. She declined to disturb or replace Duke’s rescission decision and instead explained why she thought her predecessor’s decision was sound. In addition to reiterating the illegality conclusion, she offered several new justifications for the rescission. The Government moved for the District Court to reconsider in light of this additional explanation, but the court concluded that the new reasoning failed to elaborate meaningfully on the illegality rationale. The Government appealed the various District Court decisions to the Second, Ninth, and D. C. Circuits, respectively. While those appeals were pending, the Government filed three petitions for certiorari before judgment. Following the Ninth Circuit affirmance in Regents, this Court granted certiorari. Held: The judgment in No. 18–587 is vacated in part and reversed in part; the judgment in No. 18–588 is affirmed; the February 13, 2018 order in No. 18–589 is vacated, the November 9, 2017 order is affirmed in part, and the March 29, 2018 order is reversed in part; and all of the cases are remanded. No. 18–587, 908 F.3d 476, vacated in part and reversed in part; No. 18–588, affirmed; and No. 18–589, February 13, 2018 order vacated, November 9, 2017 order affirmed in part, and March 29, 2018 order reversed in part; all cases remanded. The Chief Justice delivered the opinion of the Court, except as to Part IV, concluding: 1. DHS’s rescission decision is reviewable under the APA and is within this Court’s jurisdiction. Pp. 9–13. (a) The APA’s “basic presumption of judicial review” of agency action, Abbott Laboratories v. Gardner, 387 U.S. 136, 140, can be rebutted by showing that the “agency action is committed to agency discretion by law,” 5 U. S. C. §701(a)(2). In Heckler v. Chaney, the Court held that this narrow exception includes an agency’s decision not to institute an enforcement action. 470 U.S. 821, 831–832. The Government contends that DACA is a general non-enforcement policy equivalent to the individual non-enforcement decision in Chaney. But the DACA Memorandum did not merely decline to institute enforcement proceedings; it created a program for conferring affirmative immigration relief. Therefore, unlike the non-enforcement decision in Chaney, DACA’s creation—and its rescission—is an “action [that] provides a focus for judicial review.” Id., at 832. In addition, by virtue of receiving deferred action, 700,000 DACA recipients may request work authorization and are eligible for Social Security and Medicare. Access to such benefits is an interest “courts often are called upon to protect.” Ibid. DACA’s rescission is thus subject to review under the APA. Pp. 9–12. (b) The two jurisdictional provisions of the INA invoked by the Government do not apply. Title 8 U. S. C. §1252(b)(9), which bars review of claims arising from “action[s]” or “proceeding[s] brought to remove an alien,” is inapplicable where, as here, the parties do not challenge any removal proceedings. And the rescission is not a decision “to commence proceedings, adjudicate cases, or execute removal orders” within the meaning of §1252(g). Pp. 12–13. 2. DHS’s decision to rescind DACA was arbitrary and capricious under the APA. Pp. 13–26. (a) In assessing the rescission, the Government urges the Court to consider not just the contemporaneous explanation offered by Acting Secretary Duke but also the additional reasons supplied by Secretary Nielsen nine months later. Judicial review of agency action, however, is limited to “the grounds that the agency invoked when it took the action.” Michigan v. EPA, 576 U.S. 743, 758. If those grounds are inadequate, a court may remand for the agency to offer “a fuller explanation of the agency’s reasoning at the time of the agency action,” Pension Benefit Guaranty Corporation v. LTV Corp., 496 U.S. 633, 654 (emphasis added), or to “deal with the problem afresh” by taking new agency action, SEC v. Chenery Corp., 332 U.S. 194, 201. Because Secretary Nielsen chose not to take new action, she was limited to elaborating on the agency’s original reasons. But her reasoning bears little relationship to that of her predecessor and consists primarily of impermissible “post hoc rationalization.” Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 420. The rule requiring a new decision before considering new reasons is not merely a formality. It serves important administrative law values by promoting agency accountability to the public, instilling confidence that the reasons given are not simply convenient litigating positions, and facilitating orderly review. Each of these values would be markedly undermined if this Court allowed DHS to rely on reasons offered nine months after the rescission and after three different courts had identified flaws in the original explanation. Pp. 13–17. (b) Acting Secretary Duke’s rescission memorandum failed to consider important aspects of the problem before the agency. Although Duke was bound by the Attorney General’s determination that DACA is illegal, see 8 U. S. C. §1103(a)(1), deciding how best to address that determination involved important policy choices reserved for DHS. Acting Secretary Duke plainly exercised such discretionary authority in winding down the program, but she did not appreciate the full scope of her discretion. The Attorney General concluded that the legal defects in DACA mirrored those that the courts had recognized in DAPA. The Fifth Circuit, the highest court to offer a reasoned opinion on DAPA’s legality, found that DAPA violated the INA because it extended eligibility for benefits to a class of unauthorized aliens. But the defining feature of DAPA (and DACA) is DHS’s decision to defer removal, and the Fifth Circuit carefully distinguished that forbearance component from the associated benefits eligibility. Eliminating benefits eligibility while continuing forbearance thus remained squarely within Duke’s discretion. Yet, rather than addressing forbearance in her decision, Duke treated the Attorney General’s conclusion regarding the illegality of benefits as sufficient to rescind both benefits and forbearance, without explanation. That reasoning repeated the error in Motor Vehicle Manufacturers Association of the United States, Inc. v. State Farm— treating a rationale that applied to only part of a policy as sufficient to rescind the entire policy. 463 U.S. 29, 51. While DHS was not required to “consider all policy alternatives,” ibid., deferred action was “within the ambit of the existing” policy, ibid.; indeed, it was the centerpiece of the policy. In failing to consider the option to retain deferred action, Duke “failed to supply the requisite ‘reasoned analysis.’ ” Id., at 57. That omission alone renders Duke’s decision arbitrary and capricious, but it was not the only defect. Duke also failed to address whether there was “legitimate reliance” on the DACA Memorandum. Smiley v. Citibank (South Dakota), N. A., 517 U.S. 735, 742. Certain features of the DACA policy may affect the strength of any reliance interests, but those features are for the agency to consider in the first instance. DHS has flexibility in addressing any reliance interests and could have considered various accommodations. While the agency was not required to pursue these accommodations, it was required to assess the existence and strength of any reliance interests, and weigh them against competing policy concerns. Its failure to do so was arbitrary and capricious. Pp. 17–26. The Chief Justice, joined by Justice Ginsburg, Justice Breyer, and Justice Kagan, concluded in Part IV that respondents’ claims fail to establish a plausible inference that the rescission was motivated by animus in violation of the equal protection guarantee of the Fifth Amendment. Pp. 27–29. Roberts, C. J., delivered the opinion of the Court, except as to Part IV. Ginsburg, Breyer, and Kagan, JJ., joined that opinion in full, and Sotomayor, J., joined as to all but Part IV. Sotomayor, J., filed an opinion concurring in part, concurring in the judgment in part, and dissenting in part. Thomas, J., filed an opinion concurring in the judgment in part and dissenting in part, in which Alito and Gorsuch, JJ., joined. Alito, J., and Kavanaugh, J., filed opinions concurring in the judgment in part and dissenting in part. Notes 1 Together with No. 18–588, Trump, President of the United States, et al. v. National Association for the Advancement of Colored People et al., on certiorari before judgment to the United States Court of Appeals for the District of Columbia Circuit, and No. 18–589, Wolf, Acting Secretary of Homeland Security, et al. v. Batalla Vidal et al., on certiorari before judgment to the United States Court of Appeals for the Second Circuit.

. Alternatively, the agency can “deal with the problem afresh” by taking new agency action. SEC v. Chenery Corp., 332 U.S. 194, 201 (1947) (Chenery II). An agency taking this route is not limited to its prior reasons but must comply with the procedural requirements for new agency action. The District Court’s remand thus presented DHS with a choice: rest on the Duke Memorandum while elaborating on its prior reasoning, or issue a new rescission bolstered by new reasons absent from the Duke Memorandum. Secretary Nielsen took the first path. Rather than making a new decision, she “decline[d] to disturb the Duke memorandum’s rescission” and instead “provide[d] further explanation” for that action. App. to Pet. for Cert. 121a. Indeed, the Government’s subsequent request for reconsideration described the Nielsen Memorandum as “additional explanation for [Duke’s] decision” and asked the District Court to “leave in place [Duke’s] September 5, 2017 decision to rescind the DACA policy.” Motion to Revise Order in No. 17–cv–1907 etc. (D DC), pp. 2, 19. Contrary to the position of the Government before this Court, and of Justice Kavanaugh in dissent, post, at 4 (opinion concurring in judgment in part and dissenting in part), the Nielsen Memorandum was by its own terms not a new rule implementing a new policy. Because Secretary Nielsen chose to elaborate on the reasons for the initial rescission rather than take new administrative action, she was limited to the agency’s original reasons, and her explanation “must be viewed critically” to ensure that the rescission is not upheld on the basis of impermissible “post hoc rationalization.” Overton Park, 401 U. S., at 420. But despite purporting to explain the Duke Memorandum, Secretary Nielsen’s reasoning bears little relationship to that of her predecessor. Acting Secretary Duke rested the rescission on the conclusion that DACA is unlawful. Period. See App. to Pet. for Cert. 117a. By contrast, Secretary Nielsen’s new memorandum offered three “separate and independently sufficient reasons” for the rescission, id., at 122a, only the first of which is the conclusion that DACA is illegal. Her second reason is that DACA is, at minimum, legally questionable and should be terminated to maintain public confidence in the rule of law and avoid burdensome litigation. No such justification can be found in the Duke Memorandum. Legal uncertainty is, of course, related to illegality. But the two justifications are meaningfully distinct, especially in this context. While an agency might, for one reason or another, choose to do nothing in the face of uncertainty, illegality presumably requires remedial action of some sort. The policy reasons that Secretary Nielsen cites as a third basis for the rescission are also nowhere to be found in the Duke Memorandum. That document makes no mention of a preference for legislative fixes, the superiority of case-by-case decisionmaking, the importance of sending a message of robust enforcement, or any other policy consideration. Nor are these points included in the legal analysis from the Fifth Circuit and the Attorney General. They can be viewed only as impermissible post hoc rationalizations and thus are not properly before us. The Government, echoed by Justice Kavanaugh, protests that requiring a new decision before considering Nielsen’s new justifications would be “an idle and useless formality.” NLRB v. Wyman-Gordon Co., 394 U.S. 759, 766, n. 6 (1969) (plurality opinion). See also post, at 5. Procedural requirements can often seem such. But here the rule serves important values of administrative law. Requiring a new decision before considering new reasons promotes “agency accountability,” Bowen v. American Hospital Assn., 476 U.S. 610, 643 (1986), by ensuring that parties and the public can respond fully and in a timely manner to an agency’s exercise of authority. Considering only contemporaneous explanations for agency action also instills confidence that the reasons given are not simply “convenient litigating position[s].” Christopher v. SmithKline Beecham Corp., 567 U.S. 142, 155 (2012) (internal quotation marks omitted). Permitting agencies to invoke belated justifications, on the other hand, can upset “the orderly functioning of the process of review,” SEC v. Chenery Corp., 318 U.S. 80, 94 (1943), forcing both litigants and courts to chase a moving target. Each of these values would be markedly undermined were we to allow DHS to rely on reasons offered nine months after Duke announced the rescission and after three different courts had identified flaws in the original explanation. Justice Kavanaugh asserts that this “foundational principle of administrative law,” Michigan, 576 U. S., at 758, actually limits only what lawyers may argue, not what agencies may do. Post, at 5. While it is true that the Court has often rejected justifications belatedly advanced by advocates, we refer to this as a prohibition on post hoc rationalizations, not advocate rationalizations, because the problem is the timing, not the speaker. The functional reasons for requiring contemporaneous explanations apply with equal force regardless whether post hoc justifications are raised in court by those appearing on behalf of the agency or by agency officials themselves. See American Textile Mfrs. Institute, Inc. v. Donovan, 452 U.S. 490, 539 (1981) (“[T]he post hoc rationalizations of the agency . . . cannot serve as a sufficient predicate for agency action.”); Overton Park, 401 U. S., at 419 (rejecting “litigation affidavits” from agency officials as “merely ‘post hoc’ rationalizations”).[3] Justice Holmes famously wrote that “[m]en must turn square corners when they deal with the Government.” Rock Island, A. & L. R. Co. v. United States, 254 U.S. 141, 143 (1920). But it is also true, particularly when so much is at stake, that “the Government should turn square corners in dealing with the people.” St. Regis Paper Co. v. United States, 368 U.S. 208, 229 (1961) (Black, J., dissenting). The basic rule here is clear: An agency must defend its actions based on the reasons it gave when it acted. This is not the case for cutting corners to allow DHS to rely upon reasons absent from its original decision. B We turn, finally, to whether DHS’s decision to rescind DACA was arbitrary and capricious. As noted earlier, Acting Secretary Duke’s justification for the rescission was succinct: “Taking into consideration” the Fifth Circuit’s conclusion that DAPA was unlawful because it conferred benefits in violation of the INA, and the Attorney General’s conclusion that DACA was unlawful for the same reason, she concluded—without elaboration—that the “DACA program should be terminated.” App. to Pet. for Cert. 117a.[4] Respondents maintain that this explanation is deficient for three reasons. Their first and second arguments work in tandem, claiming that the Duke Memorandum does not adequately explain the conclusion that DACA is unlawful, and that this conclusion is, in any event, wrong. While those arguments carried the day in the lower courts, in our view they overlook an important constraint on Acting Secretary Duke’s decisionmaking authority—she was bound by the Attorney General’s legal determination. The same statutory provision that establishes the Secretary of Homeland Security’s authority to administer and enforce immigration laws limits that authority, specifying that, with respect to “all questions of law,” the determinations of the Attorney General “shall be controlling.” 8 U. S. C. §1103(a)(1). Respondents are aware of this constraint. Indeed they emphasized the point in the reviewability sections of their briefs. But in their merits arguments, respondents never addressed whether or how this unique statutory provision might affect our review. They did not discuss whether Duke was required to explain a legal conclusion that was not hers to make. Nor did they discuss whether the current suits challenging Duke’s rescission decision, which everyone agrees was within her legal authority under the INA, are proper vehicles for attacking the Attorney General’s legal conclusion. Because of these gaps in respondents’ briefing, we do not evaluate the claims challenging the explanation and correctness of the illegality conclusion. Instead we focus our attention on respondents’ third argument—that Acting Secretary Duke “failed to consider . . . important aspect[s] of the problem” before her. Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U.S. 29, 43 (1983). Whether DACA is illegal is, of course, a legal determination, and therefore a question for the Attorney General. But deciding how best to address a finding of illegality moving forward can involve important policy choices, especially when the finding concerns a program with the breadth of DACA. Those policy choices are for DHS. Acting Secretary Duke plainly exercised such discretionary authority in winding down the program. See App. to Pet. for Cert. 117a–118a (listing the Acting Secretary’s decisions on eight transition issues). Among other things, she specified that those DACA recipients whose benefits were set to expire within six months were eligible for two-year renewals. Ibid. But Duke did not appear to appreciate the full scope of her discretion, which picked up where the Attorney General’s legal reasoning left off. The Attorney General concluded that “the DACA policy has the same legal . . . defects that the courts recognized as to DAPA.” App. 878. So, to understand those defects, we look to the Fifth Circuit, the highest court to offer a reasoned opinion on the legality of DAPA. That court described the “core” issue before it as the “Secretary’s decision” to grant “eligibility for benefits”—including work authorization, Social Security, and Medicare—to unauthorized aliens on “a class-wide basis.” Texas, 809 F. 3d, at 170; see id., at 148, 184. The Fifth Circuit’s focus on these benefits was central to every stage of its analysis. See id., at 155 (standing); id., at 163 (zone of interest); id., at 164 (applicability of §1252(g)); id., at 166 (reviewability); id., at 176–177 (notice and comment); id., at 184 (substantive APA). And the Court ultimately held that DAPA was “manifestly contrary to the INA” precisely because it “would make 4.3 million otherwise removable aliens” eligible for work authorization and public benefits. Id., at 181–182 (internal quotation marks omitted).[5] But there is more to DAPA (and DACA) than such benefits. The defining feature of deferred action is the decision to defer removal (and to notify the affected alien of that decision). See App. to Pet. for Cert. 99a. And the Fifth Circuit was careful to distinguish that forbearance component from eligibility for benefits. As it explained, the “challenged portion of DAPA’s deferred-action program” was the decision to make DAPA recipients eligible for benefits. See Texas, 809 F. 3d, at 168, and n. 108. The other “[p]art of DAPA,” the court noted, “involve[d] the Secretary’s decision—at least temporarily—not to enforce the immigration laws as to a class of what he deem[ed] to be low-priority illegal aliens.” Id., at 166. Borrowing from this Court’s prior description of deferred action, the Fifth Circuit observed that “the states do not challenge the Secretary’s decision to ‘decline to institute proceedings, terminate proceedings, or decline to execute a final order of deportation.’ ” Id., at 168 (quoting Reno, 525 U. S., at 484). And the Fifth Circuit underscored that nothing in its decision or the preliminary injunction “requires the Secretary to remove any alien or to alter” the Secretary’s class-based “enforcement priorities.” Texas, 809 F. 3d, at 166, 169. In other words, the Secretary’s forbearance authority was unimpaired. Acting Secretary Duke recognized that the Fifth Circuit’s holding addressed the benefits associated with DAPA. In her memorandum she explained that the Fifth Circuit concluded that DAPA “conflicted with the discretion authorized by Congress” because the INA “ ‘flatly does not permit the reclassification of millions of illegal aliens as lawfully present and thereby make them newly eligible for a host of federal and state benefits, including work authorization.’ ” App. to Pet. for Cert. 114a (quoting Texas, 809 F. 3d, at 184). Duke did not characterize the opinion as one about forbearance. In short, the Attorney General neither addressed the forbearance policy at the heart of DACA nor compelled DHS to abandon that policy. Thus, removing benefits eligibility while continuing forbearance remained squarely within the discretion of Acting Secretary Duke, who was responsible for “[e]stablishing national immigration enforcement policies and priorities.” 116Stat. 2178, 6 U. S. C. §202(5). But Duke’s memo offers no reason for terminating forbearance. She instead treated the Attorney General’s conclusion regarding the illegality of benefits as sufficient to rescind both benefits and forbearance, without explanation. That reasoning repeated the error we identified in one of our leading modern administrative law cases, Motor Vehicle Manufacturers Association of the United States, Inc. v. State Farm Mutual Automobile Insurance Co. There, the National Highway Traffic Safety Administration (NHTSA) promulgated a requirement that motor vehicles produced after 1982 be equipped with one of two passive restraints: airbags or automatic seatbelts. 463 U. S., at 37–38, 46. Four years later, before the requirement went into effect, NHTSA concluded that automatic seatbelts, the restraint of choice for most manufacturers, would not provide effective protection. Based on that premise, NHTSA rescinded the passive restraint requirement in full. Id., at 38. We concluded that the total rescission was arbitrary and capricious. As we explained, NHTSA’s justification supported only “disallow[ing] compliance by means of ” automatic seatbelts. Id., at 47. It did “not cast doubt” on the “efficacy of airbag technology” or upon “the need for a passive restraint standard.” Ibid. Given NHTSA’s prior judgment that “airbags are an effective and cost-beneficial lifesaving technology,” we held that “the mandatory passive restraint rule [could] not be abandoned without any consideration whatsoever of an airbags-only requirement.” Id., at 51. While the factual setting is different here, the error is the same. Even if it is illegal for DHS to extend work authorization and other benefits to DACA recipients, that conclusion supported only “disallow[ing]” benefits. Id., at 47. It did “not cast doubt” on the legality of forbearance or upon DHS’s original reasons for extending forbearance to childhood arrivals. Ibid. Thus, given DHS’s earlier judgment that forbearance is “especially justified” for “productive young people” who were brought here as children and “know only this country as home,” App. to Pet. for Cert. 98a–99a, the DACA Memorandum could not be rescinded in full “without any consideration whatsoever” of a forbearance-only policy, State Farm, 463 U. S., at 51.[6] The Government acknowledges that “[d]eferred action coupled with the associated benefits are the two legs upon which the DACA policy stands.” Reply Brief 21. It insists, however, that “DHS was not required to consider whether DACA’s illegality could be addressed by separating” the two. Ibid. According to the Government, “It was not arbitrary and capricious for DHS to view deferred action and its collateral benefits as importantly linked.” Ibid. Perhaps. But that response misses the point. The fact that there may be a valid reason not to separate deferred action from benefits does not establish that DHS considered that option or that such consideration was unnecessary. The lead dissent acknowledges that forbearance and benefits are legally distinct and can be decoupled. Post, at 21–22, n. 14 (opinion of Thomas, J). It contends, however, that we should not “dissect” agency action “piece by piece.” Post, at 21. The dissent instead rests on the Attorney General’s legal determination—which considered only benefits—“to supply the ‘reasoned analysis’ ” to support rescission of both benefits and forbearance. Post, at 22 (quoting State Farm, 463 U. S., at 42). But State Farm teaches that when an agency rescinds a prior policy its reasoned analysis must consider the “alternative[s]” that are “within the ambit of the existing [policy].” Id., at 51. Here forbearance was not simply “within the ambit of the existing [policy],” it was the centerpiece of the policy: DACA, after all, stands for “Deferred Action for Childhood Arrivals.” App. to Pet. for Cert. 111a (emphasis added). But the rescission memorandum contains no discussion of forbearance or the option of retaining forbearance without benefits. Duke “entirely failed to consider [that] important aspect of the problem.” State Farm, 463 U. S., at 43. That omission alone renders Acting Secretary Duke’s decision arbitrary and capricious. But it is not the only defect. Duke also failed to address whether there was “legitimate reliance” on the DACA Memorandum. Smiley v. Citibank (South Dakota), N. A., 517 U.S. 735, 742 (1996). When an agency changes course, as DHS did here, it must “be cognizant that longstanding policies may have ‘engendered serious reliance interests that must be taken into account.’ ” Encino Motorcars, LLC v. Navarro, 579 U. S. ___, ___ (2016) (slip op., at 9) (quoting Fox Television, 556 U. S., at 515). “It would be arbitrary and capricious to ignore such matters.” Id., at 515. Yet that is what the Duke Memorandum did. For its part, the Government does not contend that Duke considered potential reliance interests; it counters that she did not need to. In the Government’s view, shared by the lead dissent, DACA recipients have no “legally cognizable reliance interests” because the DACA Memorandum stated that the program “conferred no substantive rights” and provided benefits only in two-year increments. Reply Brief 16–17; App. to Pet. for Cert. 125a. See also post, at 23–24 (opinion of Thomas, J). But neither the Government nor the lead dissent cites any legal authority establishing that such features automatically preclude reliance interests, and we are not aware of any. These disclaimers are surely pertinent in considering the strength of any reliance interests, but that consideration must be undertaken by the agency in the first instance, subject to normal APA review. There was no such consideration in the Duke Memorandum. Respondents and their amici assert that there was much for DHS to consider. They stress that, since 2012, DACA recipients have “enrolled in degree programs, embarked on careers, started businesses, purchased homes, and even married and had children, all in reliance” on the DACA program. Brief for Respondent Regents of Univ. of California et al. in No. 18–587, p. 41 (Brief for Regents). The consequences of the rescission, respondents emphasize, would “radiate outward” to DACA recipients’ families, including their 200,000 U. S.-citizen children, to the schools where DACA recipients study and teach, and to the employers who have invested time and money in training them. See id., at 41–42; Brief for Respondent State of New York et al. in No. 18–589, p. 42 (Brief for New York). See also Brief for 143 Businesses as Amici Curiae 17 (estimating that hiring and training replacements would cost employers $6.3 billion). In addition, excluding DACA recipients from the lawful labor force may, they tell us, result in the loss of $215 billion in economic activity and an associated $60 billion in federal tax revenue over the next ten years. Brief for Regents 6. Meanwhile, States and local governments could lose $1.25 billion in tax revenue each year. Ibid. These are certainly noteworthy concerns, but they are not necessarily dispositive. To the Government and lead dissent’s point, DHS could respond that reliance on forbearance and benefits was unjustified in light of the express limitations in the DACA Memorandum. Or it might conclude that reliance interests in benefits that it views as unlawful are entitled to no or diminished weight. And, even if DHS ultimately concludes that the reliance interests rank as serious, they are but one factor to consider. DHS may determine, in the particular context before it, that other interests and policy concerns outweigh any reliance interests. Making that difficult decision was the agency’s job, but the agency failed to do it. DHS has considerable flexibility in carrying out its responsibility. The wind-down here is a good example of the kind of options available. Acting Secretary Duke authorized DHS to process two-year renewals for those DACA recipients whose benefits were set to expire within six months. But Duke’s consideration was solely for the purpose of assisting the agency in dealing with “administrative complexities.” App. to Pet. for Cert. 116a–118a. She should have considered whether she had similar flexibility in addressing any reliance interests of DACA recipients. The lead dissent contends that accommodating such interests would be “another exercise of unlawful power,” post, at 23 (opinion of Thomas, J.), but the Government does not make that argument and DHS has already extended benefits for purposes other than reliance, following consultation with the Office of the Attorney General. App. to Pet. for Cert. 116a. Had Duke considered reliance interests, she might, for example, have considered a broader renewal period based on the need for DACA recipients to reorder their affairs. Alternatively, Duke might have considered more accommodating termination dates for recipients caught in the middle of a time-bounded commitment, to allow them to, say, graduate from their course of study, complete their military service, or finish a medical treatment regimen. Or she might have instructed immigration officials to give salient weight to any reliance interests engendered by DACA when exercising individualized enforcement discretion. To be clear, DHS was not required to do any of this or to “consider all policy alternatives in reaching [its] decision.” State Farm, 463 U. S., at 51. Agencies are not compelled to explore “every alternative device and thought conceivable by the mind of man.” Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc., 435 U.S. 519, 551 (1978). But, because DHS was “not writing on a blank slate,” post, at 22, n. 14 (opinion of Thomas, J.), it was required to assess whether there were reliance interests, determine whether they were significant, and weigh any such interests against competing policy concerns. The lead dissent sees all the foregoing differently. In its view, DACA is illegal, so any actions under DACA are themselves illegal. Such actions, it argues, must cease immediately and the APA should not be construed to impede that result. See post, at 19–23 (opinion of Thomas, J.). The dissent is correct that DACA was rescinded because of the Attorney General’s illegality determination. See ante, at 20. But nothing about that determination foreclosed or even addressed the options of retaining forbearance or accommodating particular reliance interests. Acting Secretary Duke should have considered those matters but did not. That failure was arbitrary and capricious in violation of the APA. IV Lastly, we turn to respondents’ claim that the rescis- sion violates the equal protection guarantee of the Fifth Amendment. The parties dispute the proper framing of this claim. The Government contends that the allegation that the Executive, motivated by animus, ended a program that disproportionately benefits certain ethnic groups is a selective enforcement claim. Such a claim, the Government asserts, is barred by our decision in Reno v. American-Arab Anti-Discrimination Committee. See 525 U. S., at 488 (holding that “an alien unlawfully in this country has no constitutional right to assert selective enforcement as a defense against his deportation”). Respondents counter that their claim falls outside the scope of that precedent because they are not challenging individual enforcement proceedings. We need not resolve this debate because, even if the claim is cognizable, the allegations here are insufficient. To plead animus, a plaintiff must raise a plausible inference that an “invidious discriminatory purpose was a motivating factor” in the relevant decision. Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 266 (1977). Possible evidence includes disparate impact on a particular group, “[d]epartures from the normal procedural sequence,” and “contemporary statements by members of the decisionmaking body.” Id., at 266–268. Tracking these factors, respondents allege that animus is evidenced by (1) the disparate impact of the rescission on Latinos from Mexico, who represent 78% of DACA recipients; (2) the unusual history behind the rescission; and (3) pre- and post-election statements by President Trump. Brief for New York 54–55. None of these points, either singly or in concert, establishes a plausible equal protection claim. First, because Latinos make up a large share of the unauthorized alien population, one would expect them to make up an outsized share of recipients of any cross-cutting immigration relief program. See B. Baker, DHS, Office of Immigration Statistics, Population Estimates, Illegal Alien Population Residing in the United States: January 2015, Table 2 (Dec. 2018), https://www.dhs.gov/sites/default/files/publications/ 18_1214_PLCY_pops-est-report.pdf. Were this fact sufficient to state a claim, virtually any generally applicable immigration policy could be challenged on equal protection grounds. Second, there is nothing irregular about the history leading up to the September 2017 rescission. The lower courts concluded that “DACA received reaffirmation by [DHS] as recently as three months before the rescission,” 908 F. 3d, at 519 (quoting 298 F. Supp. 3d, at 1315), referring to the June 2017 DAPA rescission memo, which stated that DACA would “remain in effect,” App. 870. But this reasoning confuses abstention with reaffirmation. The DAPA memo did not address the merits of the DACA policy or its legality. Thus, when the Attorney General later determined that DACA shared DAPA’s legal defects, DHS’s decision to reevaluate DACA was not a “strange about-face.” 908 F. 3d, at 519. It was a natural response to a newly identified problem. Finally, the cited statements are unilluminating. The relevant actors were most directly Acting Secretary Duke and the Attorney General. As the Batalla Vidal court acknowledged, respondents did not “identif[y] statements by [either] that would give rise to an inference of discriminatory motive.” 291 F. Supp. 3d, at 278. Instead, respondents contend that President Trump made critical statements about Latinos that evince discriminatory intent. But, even as interpreted by respondents, these statements—remote in time and made in unrelated contexts—do not qualify as “contemporary statements” probative of the decision at issue. Arlington Heights, 429 U. S., at 268. Thus, like respondents’ other points, the statements fail to raise a plausible inference that the rescission was motivated by animus. * * * We do not decide whether DACA or its rescission are sound policies. “The wisdom” of those decisions “is none of our concern.” Chenery II, 332 U. S., at 207. We address only whether the agency complied with the procedural requirement that it provide a reasoned explanation for its action. Here the agency failed to consider the conspicuous issues of whether to retain forbearance and what if anything to do about the hardship to DACA recipients. That dual failure raises doubts about whether the agency appreciated the scope of its discretion or exercised that discretion in a reasonable manner. The appropriate recourse is therefore to remand to DHS so that it may consider the problem anew. The judgment in NAACP, No. 18–588, is affirmed.[7] The judgment in Regents, No. 18–587, is vacated in part and reversed in part. And in Batalla Vidal, No. 18–589, the February 13, 2018 order granting respondents’ motion for a preliminary injunction is vacated, the November 9, 2017 order partially denying the Government’s motion to dismiss is affirmed in part, and the March 29, 2018 order partially denying the balance of the Government’s motion to dismiss is reversed in part. All three cases are remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 Plaintiffs also raised notice and comment claims, which uniformly failed below, and assorted due process challenges, some of which survived motions to dismiss. Those claims are not before us. 2 In a related challenge not at issue here, the District Court for the District of Maryland granted partial summary judgment in favor of the Government. Casa de Maryland v. United States Dept. of Homeland Security, 284 F. Supp. 3d 758 (2018). After the Government filed petitions for certiorari in the instant cases, the Fourth Circuit reversed that decision and vacated Acting Secretary Duke’s rescission as arbitrary and capricious. Casa de Maryland v. United States Dept. of Homeland Security, 924 F.3d 684 (2019), cert. pending, No. 18–1469. The Fourth Circuit has since stayed its mandate. 3 Justice Kavanaugh further argues that the contemporaneous explanation requirement applies only to agency adjudications, not rulemakings. Post, at 5–6 (opinion concurring in judgment in part and dissenting in part). But he cites no authority limiting this basic principle—which the Court regularly articulates in the context of rulemakings—to adjudications. The Government does not even raise this unheralded argument. 4 The Government contends that Acting Secretary Duke also focused on litigation risk. Although the background section of her memo references a letter from the Texas Attorney General threatening to challenge DACA, the memo never asserts that the rescission was intended to avert litigation. And, given the Attorney General’s conclusion that the policy was unlawful—and thus presumably could not be maintained or defended in its current form—it is difficult to see how the risk of litigation carried any independent weight. 5 As the Fifth Circuit noted, DAPA recipients were eligible for Social Security and Medicare benefits because they had been designated “lawfully present.” Texas, 809 F. 3d, at 168. Lawful presence is a statutory prerequisite for receipt of certain benefits. See id., at 148 (citing 8 U. S. C. §1611). It is not the same as forbearance nor does it flow inexorably from forbearance. Thus, while deferred action recipients have been designated lawfully present for purposes of Social Security and Medicare eligibility, see 8 CFR §1.3; 42 CFR §417.422(h), agencies can also exclude them from this designation, see 45 CFR §152.2(8) (2019) (specifying that DACA recipients are not considered lawfully present for purposes of coverage under the Affordable Care Act). 6 The three-page memorandum that established DACA is devoted entirely to forbearance, save for one sentence directing USCIS to “determine whether [DACA recipients] qualify for work authorization.” App. to Pet. for Cert. 101a. The benefits associated with DACA flow from a separate regulation. See 8 CFR §1.3(a)(4)(vi); see also 42 CFR §417.422(h) (cross-referencing 8 CFR §1.3). Thus, DHS could have addressed the Attorney General’s determination that such benefits were impermissible under the INA by amending 8 CFR §1.3 to exclude DACA recipients from those benefits without rescinding the DACA Memorandum and the forbearance policy it established. But Duke’s rescission memo shows no cognizance of this possibility. 7 Our affirmance of the NAACP order vacating the rescission makes it unnecessary to examine the propriety of the nationwide scope of the injunctions issued by the District Courts in Regents and Batalla Vidal.

591.US.2019_19-161

The Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA) provides for the expedited removal of certain “applicants” seeking admission into the United States, whether at a designated port of entry or elsewhere. 8 U. S. C. §1225(a)(1). An applicant may avoid expedited removal by demonstrating to an asylum officer a “credible fear of persecution,” defined as “a significant possibility . . . that the alien could establish eligibility for asylum.” §1225(b)(1)(B)(v). An applicant who makes this showing is entitled to “full consideration” of an asylum claim in a standard removal hearing. 8 CFR §208.30(f). An asylum officer’s rejection of a credible-fear claim is reviewed by a supervisor and may then be appealed to an immigration judge. §§208.30(e)(8), 1003.42(c), (d)(1). But IIRIRA limits the review that a federal court may conduct on a petition for a writ of habeas corpus. 8 U. S. C. §1252(e)(2). In particular, courts may not review “the determination” that an applicant lacks a credible fear of persecution. §1252(a)(2)(A)(iii). Respondent Vijayakumar Thuraissigiam is a Sri Lankan national who was stopped just 25 yards after crossing the southern border without inspection or an entry document. He was detained for expedited removal. An asylum officer rejected his credible-fear claim, a supervising officer agreed, and an Immigration Judge affirmed. Respondent then filed a federal habeas petition, asserting for the first time a fear of persecution based on his Tamil ethnicity and political views and requesting a new opportunity to apply for asylum. The District Court dismissed the petition, but the Ninth Circuit reversed, holding that, as applied here, §1252(e)(2) violates the Suspension Clause and the Due Process Clause. Held: 1. As applied here, §1252(e)(2) does not violate the Suspension Clause. Pp. 11–33. (a) The Suspension Clause provides that “[t]he Privilege of the Writ of Habeas Corpus shall not be suspended, unless when in Cases of Rebellion or Invasion the public Safety may require it.” Art. I, §9, cl. 2. This Court has held that, at a minimum, the Clause “protects the writ as it existed in 1789,” when the Constitution was adopted. INS v. St. Cyr, 533 U.S. 289, 301. Habeas has traditionally provided a means to seek release from unlawful detention. Respondent does not seek release from custody, but an additional opportunity to obtain asylum. His claims therefore fall outside the scope of the writ as it existed when the Constitution was adopted. Pp. 11–15. (b) Respondent contends that three bodies of case law support his argument that the Suspension Clause guarantees a broader habeas right, but none do. Pp. 15–33. (1) Respondent first points to British and American cases decided before or around the Constitution’s adoption. All those cases show is that habeas was used to seek release from detention in a variety of circumstances. Respondent argues that some cases show aliens using habeas to remain in a country. But the relief ordered in those cases was simply release; an alien petitioner’s ability to remain in the country was due to immigration law, or lack thereof. The relief that a habeas court may order and the collateral consequences of that relief are two entirely different things. Pp. 15–23. (2) Although respondent claims to rely on the writ as it existed in 1789, his argument focuses on this Court’s decisions during the “finality era,” which takes its name from a feature of the Immigration Act of 1891 making certain immigration decisions “final.” In Nishimura Ekiu v. United States, 142 U.S. 651, the Court interpreted the Act to preclude judicial review only of questions of fact. Federal courts otherwise retained authority under the Habeas Corpus Act of 1867 to determine whether an alien was detained in violation of federal law. Thus, when aliens sought habeas relief during the finality era, the Court exercised habeas jurisdiction that was conferred by the habeas statute, not because it was required by the Suspension Clause—which the Court did not mention. Pp. 23–32. (3) The Court’s more recent decisions in Boumediene v. Bush, 553 U.S. 723, and St. Cyr, 533 U.S. 289, also do not support respondent’s argument. Boumediene was not about immigration at all, and St. Cyr reaffirmed that the common-law habeas writ provided a vehicle to challenge detention and could be invoked by aliens already in the country who were held in custody pending deportation. It did not approve respondent’s very different attempted use of the writ. Pp. 32–33. 2. As applied here, §1252(e)(2) does not violate the Due Process Clause. More than a century of precedent establishes that, for aliens seeking initial entry, “the decisions of executive or administrative officers, acting within powers expressly conferred by Congress, are due process of law.” Nishimura Ekiu, 142 U. S., at 660. Respondent argues that this rule does not apply to him because he succeeded in making it 25 yards into U. S. territory. But the rule would be meaningless if it became inoperative as soon as an arriving alien set foot on U. S. soil. An alien who is detained shortly after unlawful entry cannot be said to have “effected an entry.” Zadvydas v. Davis, 533 U.S. 678, 693. An alien in respondent’s position, therefore, has only those rights regarding admission that Congress has provided by statute. In respondent’s case, Congress provided the right to a “determin[ation]” whether he had “a significant possibility” of “establish[ing] eligibility for asylum,” and he was given that right. §§1225(b)(1)(B)(ii), (v). Pp. 34–36. 917 F.3d 1097, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a concurring opinion. Breyer, J., filed an opinion concurring in the judgment, in which Ginsburg, J., joined. Sotomayor, J., filed a dissenting opinion, in which Kagan, J., joined.

Every year, hundreds of thousands of aliens are apprehended at or near the border attempting to enter this country illegally. Many ask for asylum, claiming that they would be persecuted if returned to their home countries. Some of these claims are valid, and by granting asylum, the United States lives up to its ideals and its treaty obligations. Most asylum claims, however, ultimately fail, and some are fraudulent. In 1996, when Congress enacted the Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA), 110Stat. 3009–546, it crafted a system for weeding out patently meritless claims and expeditiously removing the aliens making such claims from the country. It was Congress’s judgment that detaining all asylum seekers until the full-blown removal process is completed would place an unacceptable burden on our immigration system and that releasing them would present an undue risk that they would fail to appear for removal proceedings. This case concerns the constitutionality of the system Congress devised. Among other things, IIRIRA placed restrictions on the ability of asylum seekers to obtain review under the federal habeas statute, but the United States Court of Appeals for the Ninth Circuit held that these restrictions are unconstitutional. According to the Ninth Circuit, they unconstitutionally suspend the writ of habeas corpus and violate asylum seekers’ right to due process. We now review that decision and reverse. Respondent’s Suspension Clause argument fails because it would extend the writ of habeas corpus far beyond its scope “when the Constitution was drafted and ratified.” Boumediene v. Bush, 553 U.S. 723, 746 (2008). Indeed, respondent’s use of the writ would have been unrecognizable at that time. Habeas has traditionally been a means to secure release from unlawful detention, but respondent invokes the writ to achieve an entirely different end, namely, to obtain additional administrative review of his asylum claim and ultimately to obtain authorization to stay in this country. Respondent’s due process argument fares no better. While aliens who have established connections in this country have due process rights in deportation proceedings, the Court long ago held that Congress is entitled to set the conditions for an alien’s lawful entry into this country and that, as a result, an alien at the threshold of initial entry cannot claim any greater rights under the Due Process Clause. See Nishimura Ekiu v. United States, 142 U.S. 651, 660 (1892). Respondent attempted to enter the country illegally and was apprehended just 25 yards from the border. He therefore has no entitlement to procedural rights other than those afforded by statute. In short, under our precedents, neither the Suspension Clause nor the Due Process Clause of the Fifth Amendment requires any further review of respondent’s claims, and IIRIRA’s limitations on habeas review are constitutional as applied. I A We begin by briefly outlining the provisions of immigration law that are pertinent to this case. Under those provisions, several classes of aliens are “inadmissible” and therefore “removable.” 8 U. S. C. §§1182, 1229a(e)(2)(A). These include aliens who lack a valid entry document “at the time of application for admission.” §1182(a)(7)(A)(i)(I). An alien who arrives at a “port of entry,” i.e., a place where an alien may lawfully enter, must apply for admission. An alien like respondent who is caught trying to enter at some other spot is treated the same way. §§1225(a)(1), (3). If an alien is inadmissible, the alien may be removed. The usual removal process involves an evidentiary hearing before an immigration judge, and at that hearing an alien may attempt to show that he or she should not be removed. Among other things, an alien may apply for asylum on the ground that he or she would be persecuted if returned to his or her home country. §1229a(b)(4); 8 CFR §1240.11(c) (2020). If that claim is rejected and the alien is ordered removed, the alien can appeal the removal order to the Board of Immigration Appeals and, if that appeal is unsuccessful, the alien is generally entitled to review in a federal court of appeals. 8 U. S. C. §§1229a(c)(5), 1252(a). As of the first quarter of this fiscal year, there were 1,066,563 pending removal proceedings. See Executive Office for Immigration Review (EOIR), Adjudication Statistics: Pending Cases (Jan. 2020). The average civil appeal takes approximately one year.[1] During the time when removal is being litigated, the alien will either be detained, at considerable expense, or allowed to reside in this country, with the attendant risk that he or she may not later be found. §1226(a). Congress addressed these problems by providing more expedited procedures for certain “applicants for admission.” For these purposes, “[a]n alien present in the United States who has not been admitted or who arrives in the United States (whether or not at a designated port of arrival . . . )” is deemed “an applicant for admission.” §1225(a)(1).[2] An applicant is subject to expedited removal if, as relevant here, the applicant (1) is inadmissible because he or she lacks a valid entry document; (2) has not “been physically present in the United States continuously for the 2-year period immediately prior to the date of the determination of inadmissibility”; and (3) is among those whom the Secretary of Homeland Security has designated for expedited removal. §§1225(b)(1)(A)(i), (iii)(I)–(II).[3] Once “an immigration officer determines” that a designated applicant “is inadmissible,” “the officer [must] order the alien removed from the United States without further hearing or review.” §1225(b)(1)(A)(i). Applicants can avoid expedited removal by claiming asylum. If an applicant “indicates either an intention to apply for asylum” or “a fear of persecution,” the immigration officer “shall refer the alien for an interview by an asylum officer.” §§1225(b)(1)(A)(i)–(ii). The point of this screening interview is to determine whether the applicant has a “credible fear of persecution.” §1225(b)(1)(B)(v). The applicant need not show that he or she is in fact eligible for asylum—a “credible fear” equates to only a “significant possibility” that the alien would be eligible. Ibid. Thus, while eligibility ultimately requires a “well-founded fear of persecution on account of,” among other things, “race” or “political opinion,” §§1101(a)(42)(A), 1158(b)(1)(A), all that an alien must show to avoid expedited removal is a “credible fear.”[4] If the asylum officer finds an applicant’s asserted fear to be credible,[5] the applicant will receive “full consideration” of his asylum claim in a standard removal hearing. 8 CFR §208.30(f ); see 8 U. S. C. §1225(b)(1)(B)(ii). If the asylum officer finds that the applicant does not have a credible fear, a supervisor will review the asylum officer’s determination. 8 CFR §208.30(e)(8). If the supervisor agrees with it, the applicant may appeal to an immigration judge, who can take further evidence and “shall make a de novo determination.” §§1003.42(c), (d)(1); see 8 U. S. C. §1225(b)(1)(B)(iii)(III). An alien subject to expedited removal thus has an opportunity at three levels to obtain an asylum hearing, and the applicant will obtain one unless the asylum officer, a supervisor, and an immigration judge all find that the applicant has not asserted a credible fear. Over the last five years, nearly 77% of screenings have resulted in a finding of credible fear.[6] And nearly half the remainder (11% of the total number of screenings) were closed for administrative reasons, including the alien’s withdrawal of the claim.[7] As a practical matter, then, the great majority of asylum seekers who fall within the category subject to expedited removal do not receive expedited removal and are instead afforded the same procedural rights as other aliens. Whether an applicant who raises an asylum claim receives full or only expedited review, the applicant is not entitled to immediate release. Applicants “shall be detained pending a final determination of credible fear of persecution and, if found not to have such a fear, until removed.” §1225(b)(1)(B)(iii)(IV). Applicants who are found to have a credible fear may also be detained pending further consideration of their asylum applications. §1225(b)(1)(B)(ii); see Jennings v. Rodriguez, 583 U. S. ___, ___, ___ (2018) (slip op., at 3, 13).[8] B The IIRIRA provision at issue in this case, §1252(e)(2), limits the review that an alien in expedited removal may obtain via a petition for a writ of habeas corpus. That provision allows habeas review of three matters: first, “whether the petitioner is an alien”; second, “whether the petitioner was ordered removed”; and third, whether the petitioner has already been granted entry as a lawful permanent resident, refugee, or asylee. §§1252(e)(2)(A)–(C). If the petitioner has such a status, or if a removal order has not “in fact” been “issued,” §1252(e)(5), the court may order a removal hearing, §1252(e)(4)(B). A major objective of IIRIRA was to “protec[t] the Executive’s discretion” from undue interference by the courts; indeed, “that can fairly be said to be the theme of the legislation.” Reno v. American-Arab Anti-Discrimination Comm., 525 U.S. 471, 486 (1999) (AAADC). In accordance with that aim, §1252(e)(5) provides that “[t]here shall be no review of whether the alien is actually inadmissible or entitled to any relief from removal.” And “[n]otwithstanding” any other “habeas corpus provision”—including 28 U. S. C. §2241—“no court shall have jurisdiction to review” any other “individual determination” or “claim arising from or relating to the implementation or operation of an order of [expedited] removal.” §1252(a)(2)(A)(i). In particular, courts may not review “the determination” that an alien lacks a credible fear of persecution. §1252(a)(2)(A)(iii); see also §§1252(a)(2)(A)(ii), (iv) (other specific limitations). Even without the added step of judicial review, the credible-fear process and abuses of it can increase the burdens currently “overwhelming our immigration system.” 84 Fed. Reg. 33841 (2019).[9] The past decade has seen a 1,883% increase in credible-fear claims, and in 2018 alone, there were 99,035 claims. See id., at 33838 (data for fiscal years 2008 to 2018). The majority have proved to be meritless. Many applicants found to have a credible fear—about 50% over the same 10-year period—did not pursue asylum. See EOIR, Adjudication Statistics: Rates of Asylum Filings in Cases Originating With a Credible Fear Claim (Nov. 2018); see also 84 Fed. Reg. 33841 (noting that many instead abscond). In 2019, a grant of asylum followed a finding of credible fear just 15% of the time. See EOIR, Asylum Decision Rates in Cases Originating With a Credible Fear Claim (Oct. 2019). Fraudulent asylum claims can also be difficult to detect,[10] especially in a screening process that is designed to be expedited and that is currently handling almost 100,000 claims per year. The question presented thus has significant consequences for the immigration system. If courts must review credible-fear claims that in the eyes of immigration officials and an immigration judge do not meet the low bar for such claims, expedited removal would augment the burdens on that system. Once a fear is asserted, the process would no longer be expedited. C Respondent Vijayakumar Thuraissigiam, a Sri Lankan national, crossed the southern border without inspection or an entry document at around 11 p.m. one night in January 2017. App. 38. A Border Patrol agent stopped him within 25 yards of the border, and the Department detained him for expedited removal. Id., at 37–39, 106; see §§1182(a)(7)(A)(i)(I), 1225(b)(1)(A)(ii), and (b)(1)(B)(iii)(IV). He claimed a fear of returning to Sri Lanka because a group of men had once abducted and severely beaten him, but he said that he did not know who the men were, why they had assaulted him, or whether Sri Lankan authorities would protect him in the future. Id., at 80. He also affirmed that he did not fear persecution based on his race, political opinions, or other protected characteristics. Id., at 76–77; see §1101(a)(42)(A). The asylum officer credited respondent’s account of the assault but determined that he lacked a “credible” fear of persecution, as defined by §1225(b)(1)(B)(v), because he had offered no evidence that could have made him eligible for asylum (or other removal relief ). Id., at 83, 87, 89; see §1158(b)(1)(A). The supervising officer agreed and signed the removal order. Id., at 54, 107. After hearing further testimony from respondent, an Immigration Judge affirmed on de novo review and returned the case to the Department for removal. Id., at 97. Respondent then filed a federal habeas petition. Asserting for the first time a fear of persecution based on his Tamil ethnicity and political views, id., at 12–13, he argued that he “should have passed the credible fear stage,” id., at 30. But, he alleged, the immigration officials deprived him of “a meaningful opportunity to establish his claims” and violated credible-fear procedures by failing to probe past his denial of the facts necessary for asylum. Id., at 27, 32. Allegedly they also failed to apply the “correct standard” to his claims—the “significant possibility” standard—despite its repeated appearance in the records of their decisions. Id., at 30; see id., at 53, 84–89, 97. Respondent requested “a writ of habeas corpus, an injunction, or a writ of mandamus directing [the Department] to provide [him] a new opportunity to apply for asylum and other applicable forms of relief.” Id., at 33. His petition made no mention of release from custody. The District Court dismissed the petition, holding that §§1252(a)(2) and (e)(2) and clear Ninth Circuit case law foreclosed review of the negative credible-fear determination that resulted in respondent’s expedited removal order. 287 F. Supp. 3d 1077, 1081 (SD Cal. 2018). The court also rejected respondent’s argument “that the jurisdictional limitations of §1252(e) violate the Suspension Clause,” again relying on Circuit precedent. Id., at 1082–1083. The Ninth Circuit reversed. It found that our Suspension Clause precedent demands “reference to the writ as it stood in 1789.” 917 F.3d 1097, 1111 (2019). But without citing any pre-1789 case about the scope of the writ, the court held that §1252(e)(2) violates the Suspension Clause. See id., at 1113–1119. The court added that respondent “has procedural due process rights,” specifically the right “ ‘to expedited removal proceedings that conformed to the dictates of due process.’ ” Id., at 1111, n. 15 (quoting United States v. Raya-Vaca, 771 F.3d 1195, 1203 (CA9 2014)). Although the decision applied only to respondent, petitioners across the Circuit have used it to obtain review outside the scope of §1252(e)(2), and petitioners elsewhere have attempted to follow suit.[11] The Ninth Circuit’s decision invalidated the application of an important provision of federal law and conflicted with a decision from another Circuit, see Castro v. United States Dept. of Homeland Security, 835 F.3d 422 (CA3 2016). We granted certiorari, 589 U. S. ___ (2019). II A The Suspension Clause provides that “[t]he Privilege of the Writ of Habeas Corpus shall not be suspended, unless when in Cases of Rebellion or Invasion the public Safety may require it.” U. S. Const., Art. I, §9, cl. 2. In INS v. St. Cyr, 533 U.S. 289 (2001), we wrote that the Clause, at a minimum, “protects the writ as it existed in 1789,” when the Constitution was adopted. Id., at 301 (internal quotation marks omitted). And in this case, respondent agrees that “there is no reason” to consider whether the Clause extends any further. Brief for Respondent 26, n. 12. We therefore proceed on that basis.[12] B This principle dooms respondent’s Suspension Clause argument, because neither respondent nor his amici have shown that the writ of habeas corpus was understood at the time of the adoption of the Constitution to permit a petitioner to claim the right to enter or remain in a country or to obtain administrative review potentially leading to that result. The writ simply provided a means of contesting the lawfulness of restraint and securing release. In 1768, Blackstone’s Commentaries—usually a “satisfactory exposition of the common law of England,” Schick v. United States, 195 U.S. 65, 69 (1904)—made this clear. Blackstone wrote that habeas was a means to “remov[e] the injury of unjust and illegal confinement.” 3 W. Blackstone, Commentaries on the Laws of England 137 (emphasis deleted). Justice Story described the “common law” writ the same way. See 3 Commentaries on the Constitution of the United States §1333, p. 206 (1833). Habeas, he explained, “is the appropriate remedy to ascertain . . . whether any person is rightfully in confinement or not.” Ibid. We have often made the same point. See, e.g., Preiser v. Rodriguez, 411 U.S. 475, 484 (1973) (“It is clear . . . from the common-law history of the writ . . . that the essence of habeas corpus is an attack by a person in custody upon the legality of that custody, and that the traditional function of the writ is to secure release from illegal custody”); Wilkinson v. Dotson, 544 U.S. 74, 79 (2005) (similar); Munaf v. Geren, 553 U.S. 674, 693 (2008) (similar). In this case, however, respondent did not ask to be released.[13] Instead, he sought entirely different relief: vacatur of his “removal order” and “an order directing [the Department] to provide him with a new . . . opportunity to apply for asylum and other relief from removal.” App. 14 (habeas petition). See also id., at 31 (“a fair procedure to apply for asylum, withholding of removal, and CAT relief”); id., at 14 (“a new, meaningful opportunity to apply for asylum and other relief from removal”). Such relief might fit an injunction or writ of mandamus—which tellingly, his petition also requested, id., at 33—but that relief falls outside the scope of the common-law habeas writ. Although the historic role of habeas is to secure release from custody, the Ninth Circuit did not suggest that release, at least in the traditional sense of the term,[14] was required. Instead, what it found to be necessary was a “meaningful opportunity” for review of the procedures used in determining that respondent did not have a credible fear of persecution. 917 F. 3d, at 1117. Thus, even according to the Ninth Circuit, respondent’s petition did not call for traditional habeas relief. Not only did respondent fail to seek release, he does not dispute that confinement during the pendency of expedited asylum review, and even during the additional proceedings he seeks, is lawful. Nor could he. It is not disputed that he was apprehended in the very act of attempting to enter this country; that he is inadmissible because he lacks an entry document, see §§1182(a)(7)(A), 1225(b)(1)(A)(i); and that, under these circumstances, his case qualifies for the expedited review process, including “[m]andatory detention” during his credible-fear review, §§1225(b)(1)(B)(ii), (iii)(IV). Moreover, simply releasing him would not provide the right to stay in the country that his petition ultimately seeks. Without a change in status, he would remain subject to arrest, detention, and removal. §§1226(a), 1229a(e)(2). While respondent does not claim an entitlement to release, the Government is happy to release him—provided the release occurs in the cabin of a plane bound for Sri Lanka. That would be the equivalent of the habeas relief Justice Story ordered in a case while riding circuit. He issued a writ requiring the release of a foreign sailor who jumped ship in Boston, but he provided for the sailor to be released into the custody of the master of his ship. Ex parte D’Olivera, 7 F. Cas. 853, 854 (No. 3,967) (CC Mass. 1813). Respondent does not want anything like that. His claim is more reminiscent of the one we rejected in Munaf. In that case, American citizens held in U. S. custody in Iraq filed habeas petitions in an effort to block their transfer to Iraqi authorities for criminal prosecution. See 553 U. S., at 692. Rejecting this use of habeas, we noted that “[h]abeas is at its core a remedy for unlawful executive detention” and that what these individuals wanted was not “simple release” but an order requiring them to be brought to this country. Id., at 693, 697. Claims so far outside the “core” of habeas may not be pursued through habeas. See, e.g., Skinner v. Switzer, 562 U.S. 521, 535, n. 13 (2011). Like the habeas petitioners in Munaf, respondent does not want “simple release” but, ultimately, the opportunity to remain lawfully in the United States. That he seeks to stay in this country, while the habeas petitioners in Munaf asked to be brought here from Iraq, see post, at 19–20 (opinion of Sotomayor, J.), is immaterial. In this case as in Munaf, the relief requested falls outside the scope of the writ as it was understood when the Constitution was adopted. See Castro, 835 F. 3d, at 450–451 (Hardiman, J., concurring dubitante) (“Petitioners here seek to alter their status in the United States in the hope of avoiding release to their homelands. That prayer for relief . . . dooms the merits of their Suspension Clause argument” (emphasis deleted)). III Disputing this conclusion, respondent argues that the Suspension Clause guarantees a broader habeas right. To substantiate this claim, he points to three bodies of case law: British and American cases decided prior to or around the time of the adoption of the Constitution, decisions of this Court during the so-called “finality era” (running from the late 19th century to the mid-20th century), and two of our more recent cases. None of these sources support his argument. A Respondent and amici supporting his position have done considerable research into the use of habeas before and around the time of the adoption of the Constitution,[15] but they have not unearthed evidence that habeas was then used to obtain anything like what is sought here, namely, authorization for an alien to remain in a country other than his own or to obtain administrative or judicial review leading to that result. All that their research (and the dissent’s) shows is that habeas was used to seek release from detention in a variety of circumstances. In fact, respondent and his amici do not argue that their cases show anything more. See Brief for Respondent 27 (arguing that habeas was “available” at the founding “to test all forms of physical restraint”); Brief for Scholars of the Law of Habeas Corpus as Amici Curiae 11 (the “historical record . . . demonstrates that the touchstone for access to the writ” was “whether the petitioner challenges control of his person”). Because respondent seeks to use habeas to obtain something far different from simple release, his cause is not aided by the many release cases that he and his amici have found. Thus, for present purposes, it is immaterial that habeas was used to seek release from confinement that was imposed for, among other things, contempt of court (see Bushell’s Case, Vaugh. 135, 124 Eng. Rep. 1006 (C. P. 1670)), debt (see Hollingshead’s Case, 1 Salk. 351, 91 Eng. Rep. 307 (K. B. 1702); Rex v. Nathan, 2 Str. 880, 93 Eng. Rep. 914 (K. B. 1724)), medical malpractice (see Dr. Groenvelt’s Case, 1 Raym. Ld. 213, 91 Eng. Rep. 1038 (K. B. 1702)), failing to pay an assessment for sewers (see Hetley v. Boyer, Cro. Jac. 336, 79 Eng. Rep. 287 (K. B. 1613)), failure to lend the King money (see Darnel’s Case, 3 How. St. Tr. 1 (K. B. 1627)), carrying an authorized “dagg,” i.e., handgun (see Gardener’s Case, Cro. Eliz. 821, 78 Eng. Rep. 1048 (K. B. 1600)), “impressment” into military service or involuntary servitude (see St. Cyr, 533 U. S., at 302), or refusing to pay a colonial tax (see Oldham & Wishnie 496). Nor does it matter that common-law courts sometimes ordered or considered ordering release in circumstances that would be beyond the reach of any habeas statute ever enacted by Congress, such as release from private custody. See, e.g., Rex v. Delaval, 3 Burr. 1434, 1435–1437, 97 Eng. Rep. 913, 914 (K. B. 1763) (release of young woman from “indentures of apprenticeship”); Rex v. Clarkson, 1 Str. 444, 93 Eng. Rep. 625 (K. B. 1722) (release from boarding school); Lister’s Case, 8 Mod. 22, 88 Eng. Rep. 17 (K. B. 1721) (release of wife from estranged husband’s restraint). What matters is that all these cases are about release from restraint. Accord, Preiser, 411 U. S., at 484–485, and nn. 3–5.[16] Respondent and his amici note that habeas petitioners were sometimes released on the condition that they conform to certain requirements. See Brief for Respondent 30; Legal Historians Brief 18. For example, they cite a case in which a man was released on condition that he treat his wife well and support her, and another in which a man was released on condition that he issue an apology. Ibid. But what respondent sought in this case is nothing like that. Respondent does not seek an order releasing him on the condition that he do or refrain from doing something. What he wants—further review of his asylum claim—is not a condition with which he must comply. Equally irrelevant is the practice, discussed in the dissent, of allowing the executive to justify or cure a defect in detention before requiring release. See post, at 16–18. Respondent does not seek this sort of conditional release either, because the legality of his detention is not in question. Respondent contends that two cases show that habeas could be used to secure the right of a non-citizen to remain in a foreign country, but neither proves his point. His first case, involving a Scot named Murray, is one for which no official report is available for us to review.[17] We could hardly base our decision here on such a decision.[18] His second case, Somerset v. Stewart, Lofft. 1, 98 Eng. Rep. 499 (K. B. 1772), is celebrated but does not aid respondent. James Somerset was a slave who was “detain[ed]” on a ship bound for Jamaica, and Lord Mansfield famously ordered his release on the ground that his detention as a slave was unlawful in England. Id., at 19, 98 Eng. Rep., at 510. This relief, release from custody, fell within the historic core of habeas, and Lord Mansfield did not order anything else. It may well be that a collateral consequence of Somerset’s release was that he was allowed to remain in England, but if that is so, it was due not to the writ issued by Lord Mansfield, but to English law regarding entitlement to reside in the country. At the time, England had nothing like modern immigration restrictions. As late as 1816, the word “deportation” apparently “was not to be found in any English dictionary.” The Use of the Crown’s Power of Deportation Under the Aliens Act, 1793–1826, in J. Dinwiddy, Radicalism and Reform in Britain, 1780–1850, p. 150, n. 4 (1992); see also, e.g., Craies, The Right of Aliens To Enter British Territory, 6 L. Q. Rev. 27, 35 (1890) (“England was a complete asylum to the foreigner who did not offend against its laws”); Haycraft, Alien Legislation and the Prerogative of the Crown, 13 L. Q. Rev. 165, 180 (1897) (“There do not appear to have been any transactions in Parliament or in the [Crown’s] Privy Council directly affecting [deportation] from the time of Elizabeth [I] to that of George III”).[19] For a similar reason, respondent cannot find support in early 19th-century American cases in which deserting foreign sailors used habeas to obtain their release from the custody of American officials. In none of the cases involving deserters that have been called to our attention did the court order anything more than simple release from custody. As noted, Justice Story ordered a sailor’s release into the custody of his ship’s master. See Ex parte D’Olivera, 7 F. Cas., at 854. Other decisions, while ordering the release of detained foreign deserters because no statute authorized detention, chafed at having to order even release. See Case of the Deserters from the British Frigate L’Africaine, 3 Am. L. J. & Misc. Repertory 132, 135–136 (Md. 1810) (reporting judge’s statement “that he never would interfere to prevent” the British consul himself from detaining British deserters); Case of Hippolyte Dumas, 2 Am. L. J. & Misc. Repertory 86, 87 (Pa. 1809) (noting “inconvenience” that U. S. law did not discourage desertion of foreign sailors); Commonwealth v. Holloway, 1 Serg. & Rawle 392, 396 (Pa. 1815) (opinion of Tilghman, C. J.) (same); id., at 397 (opinion of Yeates, J.) (same). These cases thus do not contemplate the quite different relief that respondent asks us to sanction here. In these cases, as in Somerset, it may be that the released petitioners were able to remain in the United States as a collateral consequence of release, but if so, that was due not to the writs ordering their release, but to U. S. immigration law or the lack thereof. These decisions came at a time when an “open door to the immigrant was the . . . federal policy.” Harisiades v. Shaughnessy, 342 U.S. 580, 588, n. 15 (1952); see also St. Cyr, 533 U. S., at 305 (first immigration regulation enacted in 1875). So release may have had the side effect of enabling these individuals to remain in this country, but that is beside the point. The relief that a habeas court may order and the collateral consequences of that relief are two entirely different things. Ordering an individual’s release from custody may have the side effect of enabling that person to pursue all sorts of opportunities that the law allows. For example, release may enable a qualified surgeon to operate on a patient; a licensed architect may have the opportunity to design a bridge; and a qualified pilot may be able to fly a passenger jet. But a writ of habeas could not be used to compel an applicant to be afforded those opportunities or as a means to obtain a license as a surgeon, architect, or pilot. Similarly, while the release of an alien may give the alien the opportunity to remain in the country if the immigration laws permit, we have no evidence that the writ as it was known in 1789 could be used to require that aliens be permitted to remain in a country other than their own, or as a means to seek that permission. Respondent’s final examples involve international extradition, but these cases are no more pertinent than those already discussed. For one thing, they post-date the founding era. England was not a party to any extradition treaty in 1789, and this country’s first extradition treaty was the Jay Treaty of 1794. See 1 J. Moore, Extradition and Interstate Rendition §§7, 78, pp. 10, 89 (1891). In any event, extradition cases, similar to the deserter cases, illustrate nothing more than the use of habeas to secure release from custody when not in compliance with the extradition statute and relevant treaties. As noted by a scholar on whose work respondent relies, these cases “examine[d] the lawfulness of magistrates’ decisions permitting the executive to detain aliens.” Neuman, Habeas Corpus, Executive Detention, and the Removal of Aliens, 98 Colum. L. Rev. 961, 1003 (1998). In these cases, as in all the others noted above, habeas was used “simply” to seek release from allegedly unlawful detention. Benson v. McMahon, 127 U.S. 457, 463 (1888). See also, e.g., In re Stupp, 23 F. Cas. 296, 303 (No. 13,563) (CC SDNY 1875).[20] Despite pages of rhetoric, the dissent is unable to cite a single pre-1789 habeas case in which a court ordered relief that was anything like what respondent seeks here. The dissent instead contends that “the Suspension Clause inquiry does not require a close (much less precise) factual match with historical habeas precedent,” post, at 11, and then discusses cases that are not even close to this one. The dissent reveals the true nature of its argument by suggesting that there are “inherent difficulties [in] a strict originalist approach in the habeas context because of, among other things, the dearth of reasoned habeas decisions at the founding.” Ibid. But respondent does not ask us to hold that the Suspension Clause guarantees the writ as it might have evolved since the adoption of the Constitution. On the contrary, as noted at the outset of this discussion, he rests his argument on “the writ as it existed in 1789.” Brief for Respondent 26, n. 12. What the dissent merely implies, one concurring opinion states expressly, arguing that the scope of the writ guaranteed by the Suspension Clause “may change ‘depending upon the circumstances’ ” and thus may allow certain aliens to seek relief other than release. Post, at 3 (Breyer, J., concurring in judgment) (quoting Boumediene, 553 U. S., at 779). But that is not respondent’s argument, and as a general rule “we rely on the parties to frame the issues for decision and assign to courts the role of neutral arbiter of matters the parties present.” United States v. Sineneng-Smith, 590 U. S. ___, ___ (2020) (slip op., at 3) (internal quotation marks omitted). In any event, the concurrence’s snippets of quotations from Boumediene are taken entirely out of context. They relate to the question whether the statutory review procedures for Guantanamo detainees seeking release from custody provided an adequate substitute for a habeas petition seeking release. See infra, at 32–33. They do not suggest that any habeas writ guaranteed by the Suspension Clause permits a petitioner to obtain relief that goes far beyond the “core” of habeas as “a remedy for unlawful executive detention.” Munaf, 553 U. S., at 693.[21] B We now proceed to consider the second body of case law on which respondent relies, decisions of this Court during the “finality era,” which takes its name from a feature of the Immigration Act of 1891 making certain immigration decisions “final.” Although respondent claims that his argument is supported by “the writ as it existed in 1789,” Brief for Respondent 26, n. 12, his argument focuses mainly on this body of case law, which began a century later. These cases, he claims, held that “the Suspension Clause mandates a minimum level of judicial review to ensure that the Executive complies with the law in effectuating removal.” Id., at 11–12. The Ninth Circuit also relied heavily on these cases and interpreted them to “suggest that the Suspension Clause requires review of legal and mixed questions of law and fact related to removal orders.” 917 F. 3d, at 1117. This interpretation of the “finality era” cases is badly mistaken. Those decisions were based not on the Suspension Clause but on the habeas statute and the immigration laws then in force. The habeas statute in effect during this time was broad in scope. It authorized the federal courts to review whether a person was being held in custody in violation of any federal law, including immigration laws. Thus, when aliens claimed that they were detained in violation of immigration statutes, the federal courts considered whether immigration authorities had complied with those laws. This, of course, required that the immigration laws be interpreted, and at the start of the finality era, this Court interpreted the 1891 Act’s finality provision to block review of only questions of fact. Accordingly, when writs of habeas corpus were sought by aliens who were detained on the ground that they were not entitled to enter this country, the Court considered whether, given the facts found by the immigration authorities, the detention was consistent with applicable federal law. But the Court exercised that review because it was authorized to do so by statute. The decisions did not hold that this review was required by the Suspension Clause. In this country, the habeas authority of federal courts has been addressed by statute from the very beginning. The Judiciary Act of 1789, §14, 1Stat. 82, gave the federal courts the power to issue writs of habeas corpus under specified circumstances, but after the Civil War, Congress enacted a much broader statute. That law, the Habeas Corpus Act of 1867, provided that “the several courts of the United States . . . shall have power to grant writs of habeas corpus in all cases where any person may be restrained of his or her liberty in violation of the constitution, or of any treaty or law of the United States.” Judiciary Act of Feb. 5, 1867, §1, 14Stat. 385. The Act was “of the most comprehensive character,” bringing “within the habeas corpus jurisdiction of every court and of every judge every possible case of privation of liberty contrary” to federal law. Ex parte McCardle, 6 Wall. 318, 325–326 (1868). This jurisdiction was “impossible to widen.” Id., at 326; see Fay v. Noia, 372 U.S. 391, 415 (1963) (noting the Act’s “expansive language” and “imperative tone”). The 1867 statute, unlike the current federal habeas statute, was not subject to restrictions on the issuance of writs in immigration matters, and in United States v. Jung Ah Lung, 124 U.S. 621 (1888), the Court held that an alien in immigration custody could seek a writ under that statute. Id., at 626. This provided the statutory basis for the writs sought in the finality era cases. The Immigration Act of 1891, enacted during one of the country’s great waves of immigration, required the exclusion of certain categories of aliens and established procedures for determining whether aliens fell within one of those categories. The Act required the exclusion of “idiots, insane persons, paupers or persons likely to become a public charge,” persons with infectious diseases, persons with convictions for certain crimes, some individuals whose passage had been paid for by a third party, and certain laborers. Act of Mar. 3, 1891, ch. 551, §1, 26Stat. 1084. Inspection officers were authorized to board arriving vessels and inspect any aliens on board. §8, id., at 1085. And, in the provision of central importance here, the Act provided that “[a]ll decisions made by the inspection officers or their assistants touching the right of any alien to land, when adverse to such right, shall be final unless appeal be taken to the superintendent of immigration, whose action shall be subject to review by the Secretary of the Treasury.” Ibid. Later immigration Acts, which remained in effect until 1952,[22] contained similar provisions. See Act of 1894, 28Stat. 390; Immigration Act of 1907, §25, 34Stat. 907; Immigration Act of 1917, §17, 39Stat. 887. The first of the finality era cases, Nishimura Ekiu v. United States, 142 U.S. 651 (1892), required the Court to address the effect of the 1891 Act’s finality provision in a habeas case. Nishimura Ekiu is the cornerstone of respondent’s argument regarding the finality era cases, so the opinion in that case demands close attention. The case involved an alien who was detained upon arrival based on the immigration inspector’s finding that she was liable to become a public charge. Seeking to be released, the alien applied to the Circuit Court for a writ of habeas corpus and argued that the 1891 Act, if construed to give immigration authorities the “exclusive authority to determine” her right to enter, would violate her constitutional right to the writ of habeas corpus and her right to due process. Id., at 656 (statement of the case). The Circuit Court refused to issue the writ, holding that the determination of the inspector of immigration was not subject to review, and the alien then appealed. This Court upheld the denial of the writ. The Court interpreted the 1891 Act to preclude judicial review only with respect to questions of fact. Id., at 660. And after interpreting the 1891 Act in this way, the Court found that “the act of 1891 is constitutional.” Id., at 664. The Court’s narrow interpretation of the 1891 Act’s finality provision meant that the federal courts otherwise retained the full authority granted by the Habeas Corpus Act of 1867 to determine whether an alien was detained in violation of federal law. Turning to that question, the Court held that the only procedural rights of an alien seeking to enter the country are those conferred by statute. “As to such persons,” the Court explained, “the decisions of executive or administrative officers, acting within powers expressly conferred by Congress, are due process of law.” Id., at 660. The Court therefore considered whether the procedures set out in the 1891 Act had been followed, and finding no violation, affirmed the denial of the writ. Id., at 661–664. What is critical for present purposes is that the Court did not hold that the Suspension Clause imposed any limitations on the authority of Congress to restrict the issuance of writs of habeas corpus in immigration matters. Respondent interprets Nishimura Ekiu differently. See Brief for Respondent 13–15. As he reads the decision, the Court interpreted the 1891 Act to preclude review of all questions related to an alien’s entitlement to enter the country. Any other interpretation, he contends, would fly in the face of the statutory terms. But, he maintains, the Court held that this limitation violated the Suspension Clause except with respect to questions of fact, and it was for this reason that the Court considered whether the procedures specified by the 1891 Act were followed. In other words, he reads Nishimura Ekiu as holding that the 1891 Act’s finality provision was unconstitutional in most of its applications (i.e., to all questions other than questions of fact). This interpretation is wrong. The opinion in Nishimura Ekiu states unequivocally that “the act of 1891 is constitutional,” id., at 664, not that it is constitutional only in part. And if there is any ambiguity in the opinion regarding the Court’s interpretation of the finality provision, the later decision in Gegiow v. Uhl, 239 U.S. 3 (1915), left no doubt. What Nishimura Ekiu meant, Gegiow explained, was that the immigration authorities’ factual findings were conclusive (as Gegiow put it, “[t]he conclusiveness of the decisions of immigration officers . . . is conclusiveness upon matters of fact”) and therefore, the Court was “not forbidden by the statute to consider” in a habeas proceeding “whether the reasons” for removing an alien “agree with the requirements of the act.” 239 U. S., at 9. In light of this interpretation, the Nishimura Ekiu Court had no occasion to decide whether the Suspension Clause would have tolerated a broader limitation, and there is not so much as a hint in the opinion that the Court considered this question. Indeed, the opinion never even mentions the Suspension Clause, and it is utterly implausible that the Court would hold sub silentio that Congress had violated that provision. Holding that an Act of Congress unconstitutionally suspends the writ of habeas corpus is momentous. See Boumediene, 553 U. S., at 773 (noting “the care Congress has taken throughout our Nation’s history” to avoid suspension). The Justices on the Court at the beginning of the finality era had seen historic occasions when the writ was suspended—during the Civil War by President Lincoln and then by Congress, and later during Reconstruction by President Grant. See Hamdi v. Rumsfeld, 542 U.S. 507, 563 (2004) (Scalia, J., dissenting) (discussing these events). The suspension of habeas during this era played a prominent role in our constitutional history. See Ex parte Merryman, 17 F. Cas. 144, 151–152 (No. 9,487) (CC Md. 1861) (Taney, C. J.); Ex parte Milligan, 4 Wall. 2, 116, 131 (1866). (Two of the Justices at the beginning of the finality era were on the Court when Ex parte Milligan was decided.) The Justices knew a suspension of the writ when they saw one, and it is impossible to believe that the Nishimura Ekiu Court identified another occasion when Congress had suspended the writ and based its decision on the Suspension Clause without even mentioning that provision. The dissent’s interpretation of Nishimura Ekiu is different from respondent’s. According to the dissent, Nishimura Ekiu interpreted the 1891 Act as it did based on the doctrine of constitutional avoidance. See post, at 22. This reading has no support in the Court’s opinion, which never mentions the Suspension Clause or the avoidance doctrine and never explains why the Clause would allow Congress to preclude review of factual findings but nothing more. But even if there were some basis for this interpretation, it would not benefit respondent, and that is undoubtedly why he has not made the argument. IIRIRA unequivocally bars habeas review of respondent’s claims, see §1252(e)(2), and he does not argue that it can be read any other way. The avoidance doctrine “has no application in the absence of ambiguity.” Warger v. Shauers, 574 U.S. 40, 50 (2014) (internal quotation marks and ellipsis omitted). Thus, if Nishimura Ekiu’s interpretation were based on constitutional avoidance, it would still not answer the interpretive question here. When we look to later finality era cases, any suggestion of a Suspension Clause foundation becomes even less plausible. None of those decisions mention the Suspension Clause or even hint that they are based on that provision, and these omissions are telling. On notable occasions during that time, the writ was suspended—in the Philippines in 1906[23] and Hawaii in 1941.[24] During World War II, the Court held that “enemy aliens” could utilize habeas “unless there was suspension of the writ.” In re Yamashita, 327 U.S. 1, 9 (1946). And the Court invoked the Suspension Clause in holding that the Executive lacked authority to intern a Japanese-American citizen. See Ex parte Endo, 323 U.S. 283, 297–299 (1944). If the Justices during that time had thought that the Suspension Clause provided the authority they were exercising in the many cases involving habeas petitions by aliens detained prior to entry, it is hard to believe that this important fact would have escaped mention. Respondent suggests that Nishimura Ekiu cannot have interpreted the 1891 Act’s finality provision to apply only to factual questions because the statutory text categorically bars all review. The important question here, however, is what the Court did in Nishimura Ekiu, not whether its interpretation was correct, and in any event, there was a reasonable basis for the Court’s interpretation. The determinations that the immigration officials were required to make under the 1891 Act were overwhelmingly factual in nature. The determination in Nishimura’s case—that she was likely to become a public charge—seems to have been a pure question of fact, and the other grounds for exclusion under the Act involved questions that were either solely or at least primarily factual in nature. If we were now called upon to determine the meaning of a provision like the finality provision in the 1891 Act, our precedents would provide the basis for an argument in favor of the interpretation that the Nishimura Ekiu Court reached. The presumption in favor of judicial review, see, e.g., Guerrero-Lasprilla v. Barr, 589 U. S. ___, ___ (2020) (slip op., at 6); Nasrallah v. Barr, 590 U. S. ___, ___–___ (2020) (slip op., at 7–9), could be invoked. So could the rule that “[i]mplications from statutory text or legislative history are not sufficient to repeal habeas jurisdiction.” St. Cyr, 533 U. S., at 299; accord, Ex parte Yerger, 8 Wall. 85, 105 (1869). Thus, respondent’s interpretation of the decision in Nishimura Ekiu is wrong, and the same is true of his understanding of the later finality era cases. Rather than relying on the Suspension Clause, those cases simply involved the exercise of the authority conferred by the habeas statute then in effect. This was true of Nishimura Ekiu, Gegiow, and every other finality era case that respondent cites in support of his Suspension Clause argument. See, e.g., Gonzales v. Williams, 192 U.S. 1 (1904); Yee Won v. White, 256 U.S. 399 (1921); Tod v. Waldman, 266 U.S. 113 (1924); United States ex rel. Polymeris v. Trudell, 284 U.S. 279 (1932); United States ex rel. Johnson v. Shaughnessy, 336 U.S. 806 (1949); United States ex rel. Knauff v. Shaughnessy, 338 U.S. 537 (1950); Shaughnessy v. United States ex rel. Mezei, 345 U.S. 206 (1953); United States ex rel. Accardi v. Shaughnessy, 347 U.S. 260 (1954). Some finality era cases presented pure questions of law, while others involved the application of a legal test to particular facts. At least one involved an alien who had entered illegally. See id., at 262. But none was based on the Suspension Clause. No majority opinion even mentioned the Suspension Clause.[25] Indeed, any mention of the Constitution was rare—and unhelpful to respondent’s arguments here.[26] And in all the cited cases concerning aliens detained at entry, unlike the case now before us, what was sought—and the only relief considered—was release. Indeed, in an early finality era case, the Court took pains to note that it did not “express any opinion” on whether an alien was entitled to enter. Lem Moon Sing v. United States, 158 U.S. 538, 549 (1895). Like the dissent, respondent makes much of certain statements in Heikkila v. Barber, 345 U.S. 229 (1953), which he interprets to substantiate his interpretation of Nishimura Ekiu and the subsequent entry cases discussed above. But he takes these statements out of context and reads far too much into them. Heikkila was not a habeas case, and the question before the Court was whether a deportation order was reviewable under the Administrative Procedure Act (APA). The Court held that the order was not subject to APA review because the Immigration Act of 1917 foreclosed “judicial review”—as opposed to review in habeas. 345 U. S., at 234–235. Nothing in Heikkila suggested that the 1891 Act had been found to be partly unconstitutional, and Heikkila certainly did not address the scope of the writ of habeas corpus in 1789. In sum, the Court exercised habeas jurisdiction in the finality era cases because the habeas statute conferred that authority, not because it was required by the Suspension Clause. As a result, these cases cannot support respondent’s argument that the writ of habeas corpus as it was understood when the Constitution was adopted would have allowed him to claim the right to administrative and judicial review while still in custody. C We come, finally, to the more recent cases on which respondent relies. The most recent, Boumediene, is not about immigration at all. It held that suspected foreign terrorists could challenge their detention at the naval base in Guantanamo Bay, Cuba. They had been “apprehended on the battlefield in Afghanistan” and elsewhere, not while crossing the border. 553 U. S., at 734. They sought only to be released from Guantanamo, not to enter this country. See, e.g., Brief for Petitioner Al Odah et al. in Al Odah v. United States, decided with Boumediene v. Bush, O. T. 2007, No. 06–1196, p. 39 (arguing that “habeas contemplates but one remedy,” “release”). And nothing in the Court’s discussion of the Suspension Clause suggested that they could have used habeas as a means of gaining entry. Rather, the Court reaffirmed that release is the habeas remedy though not the “exclusive” result of every writ, given that it is often “appropriate” to allow the executive to cure defects in a detention. 553 U. S., at 779. Respondent’s other recent case is St. Cyr, in which the Court’s pertinent holding rejected the argument that certain provisions of IIRIRA and the Antiterrorism and Effective Death Penalty Act of 1996 that did not refer expressly to habeas should nevertheless be interpreted as stripping the authority conferred by the habeas statute. In refusing to adopt that interpretation, the Court enlisted a quartet of interpretive canons: “the strong presumption in favor of judicial review of administrative action,” “the longstanding rule requiring a clear statement of congressional intent to repeal habeas jurisdiction,” the rule that a “clear indication” of congressional intent is expected when a proposed interpretation would push “the outer limits of Congress’ power,” and the canon of constitutional avoidance. 533 U. S., at 298–300. In connection with this final canon, the Court observed: “Because of [the Suspension] Clause, some ‘judicial intervention in deportation cases’ is unquestionably ‘required by the Constitution.’ ” Id., at 300 (quoting Heikkila, 345 U. S., at 235). Respondent pounces on this statement, but like the Heikkila statement on which it relies, it does nothing for him. The writ of habeas corpus as it existed at common law provided a vehicle to challenge all manner of detention by government officials, and the Court had held long before that the writ could be invoked by aliens already in the country who were held in custody pending deportation. St. Cyr reaffirmed these propositions, and this statement in St. Cyr does not signify approval of respondent’s very different attempted use of the writ, which the Court did not consider.[27] IV In addition to his Suspension Clause argument, respondent contends that IIRIRA violates his right to due process by precluding judicial review of his allegedly flawed credible-fear proceeding. Brief for Respondent 38–45. The Ninth Circuit agreed, holding that respondent “had a constitutional right to expedited removal proceedings that conformed to the dictates of due process.” 917 F. 3d, at 1111, n. 15 (internal quotation marks omitted). And the Ninth Circuit acknowledged, ibid., that this holding conflicted with the Third Circuit’s decision upholding §1252(e)(2) on the ground that applicants for admission lack due process rights regarding their applications, see Castro, 835 F. 3d, at 445–446. Since due process provided an independent ground for the decision below and since respondent urges us to affirm on this ground, it is hard to understand the dissent’s argument that the due process issue was not “seriously in dispute below” or that it is somehow improper for us to decide the issue. Post, at 34. Nor is the dissent correct in defending the Ninth Circuit’s holding. That holding is contrary to more than a century of precedent. In 1892, the Court wrote that as to “foreigners who have never been naturalized, nor acquired any domicil or residence within the United States, nor even been admitted into the country pursuant to law,” “the decisions of executive or administrative officers, acting within powers expressly conferred by Congress, are due process of law.” Nishimura Ekiu, 142 U. S., at 660. Since then, the Court has often reiterated this important rule. See, e.g., Knauff, 338 U. S., at 544 (“Whatever the procedure authorized by Congress is, it is due process as far as an alien denied entry is concerned”); Mezei, 345 U. S., at 212 (same); Landon v. Plasencia, 459 U.S. 21, 32 (1982) (“This Court has long held that an alien seeking initial admission to the United States requests a privilege and has no constitutional rights regarding his application, for the power to admit or exclude aliens is a sovereign prerogative”). Respondent argues that this rule does not apply to him because he was not taken into custody the instant he attempted to enter the country (as would have been the case had he arrived at a lawful port of entry). Because he succeeded in making it 25 yards into U. S. territory before he was caught, he claims the right to be treated more favorably. The Ninth Circuit agreed with this argument. We reject it. It disregards the reason for our century-old rule regarding the due process rights of an alien seeking initial entry. That rule rests on fundamental propositions: “[T]he power to admit or exclude aliens is a sovereign prerogative,” id., at 32; the Constitution gives “the political department of the government” plenary authority to decide which aliens to admit, Nishimura Ekiu, 142 U. S., at 659; and a concomitant of that power is the power to set the procedures to be followed in determining whether an alien should be admitted, see Knauff, 338 U. S., at 544. This rule would be meaningless if it became inoperative as soon as an arriving alien set foot on U. S. soil. When an alien arrives at a port of entry—for example, an international airport—the alien is on U. S. soil, but the alien is not considered to have entered the country for the purposes of this rule. On the contrary, aliens who arrive at ports of entry—even those paroled elsewhere in the country for years pending removal—are “treated” for due process purposes “as if stopped at the border.” Mezei, 345 U. S., at 215; see Leng May Ma v. Barber, 357 U.S. 185, 188–190 (1958); Kaplan v. Tod, 267 U.S. 228, 230–231 (1925). The same must be true of an alien like respondent. As previously noted, an alien who tries to enter the country illegally is treated as an “applicant for admission,” §1225(a)(1), and an alien who is detained shortly after unlawful entry cannot be said to have “effected an entry,” Zadvydas v. Davis, 533 U.S. 678, 693 (2001). Like an alien detained after arriving at a port of entry, an alien like respondent is “on the threshold.” Mezei, 345 U. S., at 212. The rule advocated by respondent and adopted by the Ninth Circuit would undermine the “sovereign prerogative” of governing admission to this country and create a perverse incentive to enter at an unlawful rather than a lawful location. Plasencia, 459 U. S., at 32. For these reasons, an alien in respondent’s position has only those rights regarding admission that Congress has provided by statute. In respondent’s case, Congress provided the right to a “determin[ation]” whether he had “a significant possibility” of “establish[ing] eligibility for asylum,” and he was given that right. §§1225(b)(1)(B)(ii), (v). Because the Due Process Clause provides nothing more, it does not require review of that determination or how it was made. As applied here, therefore, §1252(e)(2) does not violate due process.[28] * * * Because the Ninth Circuit erred in holding that §1252(e)(2) violates the Suspension Clause and the Due Process Clause, we reverse the judgment and remand the case with directions that the application for habeas corpus be dismissed. It is so ordered. Notes 1 See Administrative Office of the U. S. Courts, Federal Judicial Caseload Statistics, U. S. Courts of Appeals—Median Time Intervals in Months for Civil and Criminal Appeals Terminated on the Merits (2019) (Table B–4A) (time calculated for non-prisoner appeals from the filing of a notice of appeal to the last opinion or final order). 2 When respondent entered the country, aliens were treated as applicants for admission if they were “encountered within 14 days of entry without inspection and within 100 air miles of any U. S. international land border.” 69 Fed. Reg. 48879 (2004). 3 This authority once belonged to the Attorney General, who is still named in the statute. See 6 U. S. C. §251(2) (transferring authority over “[t]he detention and removal program” to the Department). 4 A grant of asylum enables an alien to enter the country, but even if an applicant qualifies, an actual grant of asylum is discretionary. §1158(b)(1)(A). 5 The asylum officer also considers an alien’s potential eligibility for withholding of removal under §1231(b)(3) or relief under the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (CAT). 8 CFR §§208.30(e)(2)–(3). Respondent’s habeas petition alleges that “he can show a significan[t] possibility that he could establish eligibility for asylum, withholding of removal, and CAT claims.” App. 31–32. But he says in his petition that he left Sri Lanka “to seek asylum in the United States.” Id., at 15. He discusses the criteria only for asylum. Id., at 21; see also Brief for Respondent 4. And he now alleges that he was improperly “denied asylum.” Id., at 5. Moreover, the gravamen of his petition is that he faces persecution in Sri Lanka “because of ” his Tamil ethnicity and political opinions. App. 13. To obtain withholding or CAT relief on that basis, he would need to show “a greater likelihood of persecution or torture at home than is necessary for asylum.” Moncrieffe v. Holder, 569 U.S. 184, 187, n. 1 (2013). And he would not avoid removal, only removal to Sri Lanka. 8 U. S. C. §1231(b)(3)(A); 8 CFR §208.16(f ). We therefore read his petition as it is plainly intended: to seek another opportunity to apply for asylum. 6 See GAO, Immigration: Actions Needed To Strengthen USCIS’s Oversight and Data Quality of Credible and Reasonable Fear Screenings 13–15, and fig. 2 (GAO–20–250, Feb. 2020). 7 See id., at 16, n. b. 8 The Department may grant temporary parole “for urgent humanitarian reasons or significant public benefit.” 8 U. S. C. §1182(d)(5)(A); see also 8 CFR §§212.5(b), 235.3(b)(2)(iii), and (4)(ii). 9 References to the factual material in this regulation are not endorsements of the regulation itself. And like the immigration officials in this case, we do not question the basis for respondent’s asserted fear. See infra, at 9. But we note the Department’s view that credible-fear claims can be asserted “in the hope of a lengthy asylum process that will enable [the claimants] to remain in the United States for years . . . despite their statutory ineligibility for relief ” and that an influx of meritless claims can delay the adjudication of meritorious ones; strain detention capacity and degrade detention conditions; cause the release of many inadmissible aliens into States and localities that must shoulder the resulting costs; divert Department resources from protecting the border; and aggravate “the humanitarian crisis created by human smugglers.” 84 Fed. Reg. 33831; see also, e.g., Violent Crime Control and Law Enforcement Act of 1994, §130010(a)(3)(C), 108Stat. 2030 (legislative finding of “a drain on limited resources resulting from the high cost of processing frivolous asylum claims”); Arizona v. United States, 567 U.S. 387, 397–398 (2012); Homeland Security Advisory Council, Final Emergency Interim Report 1, 7–8 (Apr. 16, 2019); Letter from K. Nielsen, Secretary of Homeland Security, to Members of Congress 1–2 (Mar. 28, 2019); GAO, Asylum: Additional Actions Needed To Assess and Address Fraud Risks 24 (GAO–16–50, Dec. 2015) (GAO Fraud Report); Congressional Budget Office, The Impact of Unauthorized Immigrants on the Budgets of State and Local Governments 8–9 (Dec. 2007); Brief for State of Arizona et al. as Amici Curiae 9–12. 10 See, e.g., GAO Fraud Report 32–33 (discussing Operation Fiction Writer, a criminal investigation of attorneys and application preparers who counseled asylum seekers to lie about religious persecution and forced abortions); Asylum Fraud: Abusing America’s Compassion? Hearing before the Subcommittee on Immigration and Border Security of the House Committee on the Judiciary, 113th Cong., 2d Sess. (2014) (testimony of Louis D. Crocetti, Jr.) (describing study in which 58% of randomly selected asylum applications exhibited indicators of possible fraud and 12% were determined to be fraudulent). 11 See, e.g., Mnatsakanyan v. United States Dept. of Homeland Security, 2020 WL 1245371, *5 (SD Cal., Mar. 16, 2020) (“Given the identical claims here as in Thuraissigiam, the Court concludes it has jurisdiction over Petitioner’s habeas petition under the Suspension Clause”); Kaur v. Barr, 2019 WL 4974425, *3 (D Ariz., Oct. 8, 2019) (granting stay of removal in light of the decision below); Rodrigues v. McAleenan, 2020 WL 363041, *2, *6 (ND Tex., Jan. 22, 2020) (declining to follow the decision below). 12 The original meaning of the Suspension Clause is the subject of controversy. In INS v. St. Cyr, 533 U.S. 289 (2001), the majority and dissent debated whether the Clause independently guarantees the availability of the writ or simply restricts the temporary withholding of its operation. Compare id., at 300, with id., at 336–341 (Scalia, J., dissenting). See also Ex parte Bollman, 4 Cranch 75, 95 (1807). We do not revisit that question. Nor do we consider whether the scope of the writ as it existed in 1789 defines the boundary of the constitutional protection to which the St. Cyr Court referred, since the writ has never encompassed respondent’s claims. We also do not reconsider whether the common law allowed the issuance of a writ on behalf of an alien who lacked any allegiance to the country. Compare Boumediene v. Bush, 553 U.S. 723, 746–747 (2008) (forming “no certain conclusions”), with Brief for Criminal Justice Legal Foundation as Amicus Curiae 5–13. See also Hamburger, Beyond Protection, 109 Colum. L. Rev. 1823, 1847 (2009); P. Halliday, Habeas Corpus: From England to Empire 204 (2010) (Halliday). 13 In his brief, respondent states that “he requests an entirely ordinary habeas remedy: conditional release pending a lawful adjudication. J. A. 33.” Brief for Respondent 29. Citing the same page, the dissent argues that respondent “asked the district court to ‘[i]ssue a writ of habeas corpus’ without further limitation on the kind of relief that might entail.” Post, at 7 (opinion of Sotomayor, J.) (quoting App. 33). However, neither on the cited page nor at any other place in the habeas petition is release, conditional or otherwise, even mentioned. And in any event, as we discuss infra, at 15–21, the critical point is that what he sought in the habeas petition and still seeks—a writ “directing [the Department] to provide [him] a new opportunity to apply for asylum,” App. 33—is not a form of relief that was available in habeas at the time of the adoption of the Constitution. 14 Although the Ninth Circuit never mentioned release, its opinion might be read to suggest that gaining a right to remain in this country would constitute a release from the “restraint” of exclusion. See 917 F.3d 1097, 1117 (2019). No evidence has been called to our attention that the writ was understood in 1789 to apply to any comparable form of restraint. 15 Respondent and his amici rely primarily on British cases decided before the adoption of the Constitution. “There is widespread agreement that the common-law writ of habeas corpus was in operation in all thirteen of the British colonies that rebelled in 1776,” but “almost no reported decisio[n] from the period.” Oldham & Wishnie, The Historical Scope of Habeas Corpus and INS v. St. Cyr, 16 Geo. Immigration L. J. 485, 496 (2002) (Oldham & Wishnie) (internal quotation marks omitted). 16 Respondent’s amici also point out that, during the English Civil War, Parliament created a national religion and a “bewildering array of committees” to manage the war. Brief for Legal Historians as Amici Curiae 10 (Legal Historians Brief ) (internal quotation marks omitted). They argue that “[h]abeas corpus was readily available to test the legality of their actions.” Ibid. But according to their source, the challenged actions were “imprisonment orders,” including imprisonment of clergymen who refused to conform. Halliday 163–164. 17 Respondent cites a secondary source, which in turn cites to the National Archives in London. See Brief for Respondent 27 (citing Halliday 236). 18 Whether the founding generation understood habeas relief more broadly than described by Blackstone, Justice Story, and our prior cases, see supra, at 12, cannot be settled by a single case or even a few obscure and possibly aberrant cases. And in any event, what is said here about Murray’s case provides little support for respondent’s position. In 1677, we are told, Murray was imprisoned in England so that he could be “ ‘sent into Scotland’ ” for a criminal trial, but the King’s Bench twice issued a writ of habeas corpus requiring his release. Brief for Respondent 27 (quoting Halliday 236). Putting aside the “delicate” relationship between England and Scotland at the time, Boumediene, 553 U. S., at 749, issuance of a writ to secure the release of a person held in pretrial custody is far afield from what respondent wants here. 19 This regime lasted until after 1789, when the Aliens Act of 1793 authorized justices of the peace to imprison “without bail or mainprize” (i.e., bond) any alien found without a passport, who could then be “sen[t] out of th[e] realm.” An Act for Regulating Immigration into Great Britain, 33 Geo. III, ch. 4, §§11, 29. 20 Amici supporting respondent make an additional argument. They contend that “[i]n eighteenth century practice, the authority of English judges to review habeas petitions was not constrained by past decisions” and that these judges felt free to innovate in order to ensure that justice was done. Legal Historians Brief 5–6. But the role of federal courts under our Constitution is very different from that of those English judges. The English judges “were considered agents of the Crown, designed to assist the King in the exercise of his power.” Boumediene, 553 U. S., at 740. The court with primary habeas jurisdiction, after all, was called the King’s Bench, on which the King “was theoretically always present.” Halliday & White, The Suspension Clause: English Text, Imperial Contexts, and American Implications, 94 Va. L. Rev. 575, 594, 598, and n. 49 (2008). Habeas was an exercise of the King’s prerogative “to have an account . . . why the liberty of any of his subjects is restrained.” 2 J. Story, Commentaries on the Constitution of the United States §1335, p. 207 (1833); accord, Legal Historians Brief 5–7. In our federal courts, by contrast, the scope of habeas has been tightly regulated by statute, from the Judiciary Act of 1789 to the present day, and precedent is as binding in a habeas case as in any other. See, e.g., Jenkins v. Hutton, 582 U. S. ___, ___ (2017) (per curiam) (slip op., at 4). 21 This concurrence imagines three horrible possibilities that it fears could come to pass unless we interpret the Suspension Clause to protect the right to some undefined category of relief beyond release from custody. See post, at 2 (opinion of Breyer, J.). But its interpretation is neither necessary nor obviously sufficient to prevent the possibilities it fears. First, if a citizen were detained for deportation, today’s opinion would not prevent the citizen from petitioning for release. Second, if respondent’s “procedural” claims do not merit habeas review, as the concurrence concludes, post, at 8, it is not clear why habeas should help the concurrence’s hypothetical alien whose credible-fear claim was rejected based on forged evidence. Both respondent and this hypothetical alien assert procedural irregularities. Does the availability of habeas review depend on a judge’s view of the severity of the irregularity asserted? Finally, there is the hypothetical alien denied asylum on the ground that Judaism is not a religion. Such a decision would of course be ridiculous, but why it would not raise a question of “brute fac[t]” that falls outside the concurrence’s interpretation of the Suspension Clause, post, at 5, is again not clear. Whatever may be said about the concurrence’s hypotheticals, it is possible to imagine all sorts of abuses not even remotely related to unauthorized executive detention that could be imposed on people in this country if the Constitution allowed Congress to deprive the courts of any jurisdiction to entertain claims regarding such abuses. If that were to happen, it would no doubt be argued that constitutional provisions other than the Suspension Clause guaranteed judicial review. We have no occasion to consider such arguments here. 22 See Shaughnessy v. Pedreiro, 349 U.S. 48, 51–52 (1955) (interpreting 1952 Immigration and Nationality Act, 66Stat. 163, to provide for review of deportation orders). 23 While the Philippines was a Territory, its government suspended habeas to deal with “ ‘certain organized bands’ ” of rebels. Fisher v. Baker, 203 U.S. 174, 179–181 (1906) (quoting resolution). 24 The Governor of Hawaii suspended habeas, with President Roosevelt’s approval, after the attack on Pearl Harbor. See Duncan v. Kahanamoku, 327 U.S. 304, 307–308, 324 (1946). 25 In a concurrence in United States ex rel. Turner v. Williams, 194 U.S. 279 (1904), Justice Brewer stated without elaboration and without citing any authority that the Suspension Clause prohibits Congress from “oust[ing] the courts from the duty of inquiry respecting both law and facts” in habeas cases. Id., at 295. No other Justice joined that opinion. 26 In Fong Yue Ting v. United States, 149 U.S. 698, 713 (1893), and many other cases, the Court noted that the Constitution gives Congress plenary power to set requirements for admission. 27 The Government notes other distinctions between St. Cyr and this case, including that the alien in St. Cyr raised a pure question of law, while respondent raises at best a mixed question of law and fact. We have no need to consider these distinctions. 28 Although respondent, during his interviews with immigration officials, does not appear to have provided any information tying the assault he suffered at the hands of those who arrived at his home in a van to persecution on the basis of ethnicity or political opinion, his counseled petition offers details about “white va[n]” attacks against Tamils in Sri Lanka. App. 25–26 (internal quotation marks omitted). As now portrayed, his assault resembles those incidents. Department officials and immigration judges may reopen cases or reconsider decisions, see 8 CFR §§103.5(a)(1), (5), and 1003.23(b)(1), and the Executive always has discretion not to remove, see AAADC, 525 U. S., at 483–484.

591.US.2019_18-1195

The Montana Legislature established a program that grants tax credits to those who donate to organizations that award scholarships for private school tuition. To reconcile the program with a provision of the Montana Constitution that bars government aid to any school “controlled in whole or in part by any church, sect, or denomination,” Art. X, §6(1), the Montana Department of Revenue promulgated “Rule 1,” which prohibited families from using the scholarships at religious schools. Three mothers who were blocked by Rule 1 from using scholarship funds for their children’s tuition at Stillwater Christian School sued the Department in state court, alleging that the Rule discriminated on the basis of their religious views and the religious nature of the school they had chosen. The trial court enjoined Rule 1. Reversing, the Montana Supreme Court held that the program, unmodified by Rule 1, aided religious schools in violation of the Montana Constitution’s no-aid provision. The Court further held that the violation required invalidating the entire program. Held: The application of the no-aid provision discriminated against religious schools and the families whose children attend or hope to attend them in violation of the Free Exercise Clause of the Federal Constitution. Pp. 6–22. (a) The Free Exercise Clause “protects religious observers against unequal treatment” and against “laws that impose special disabilities on the basis of religious status.” Trinity Lutheran Church of Columbia, Inc. v. Comer, 582 U. S. ___, ___. In Trinity Lutheran, this Court held that disqualifying otherwise eligible recipients from a public benefit “solely because of their religious character” imposes “a penalty on the free exercise of religion that triggers the most exacting scrutiny.” Id., at ___. Here, the application of Montana’s no-aid provision excludes religious schools from public benefits solely because of religious status. As a result, strict scrutiny applies. Pp. 6–12. (b) Contrary to the Department’s contention, this case is not governed by Locke v. Davey, 540 U.S. 712. The plaintiff in Locke was denied a scholarship “because of what he proposed to do—use the funds to prepare for the ministry,” an essentially religious endeavor. Trinity Lutheran, 582 U. S., at ___. By contrast, Montana’s no-aid provision does not zero in on any essentially religious course of instruction but rather bars aid to a religious school “simply because of what it is”—a religious school. Id., at ___. Locke also invoked a “historic and substantial” state interest in not funding the training of clergy, 540 U. S., at 725, but no comparable tradition supports Montana’s decision to disqualify religious schools from government aid. Pp. 12–16. (c) The proposed alternative approach involving a flexible case-by-case analysis is inconsistent with Trinity Lutheran. The protections of the Free Exercise Clause do not depend on a varying case-by-case analysis regarding whether discrimination against religious adherents would serve ill-defined interests. Pp. 16–18. (d) To satisfy strict scrutiny, government action “must advance ‘interests of the highest order’ and must be narrowly tailored in pursuit of those interests.” Church of Lukumi Babalu Aye, Inc. v. Hialeah, 508 U.S. 520, 546. Montana’s interest in creating greater separation of church and State than the Federal Constitution requires “cannot qualify as compelling” in the face of the infringement of free exercise here. Trinity Lutheran, 582 U. S., at ___. The Department’s argument that the no-aid provision actually promotes religious freedom is unavailing because an infringement of First Amendment rights cannot be justified by a State’s alternative view that the infringement advances religious liberty. The Department’s argument is especially unconvincing because the infringement here broadly burdens not only religious schools but also the families whose children attend them. The Department suggests that the no-aid provision safeguards public education by ensuring that government support is not diverted to private schools, but that interest does not justify a no-aid provision that requires only religious private schools to bear its weight. Pp. 18–20. (e) Because the Free Exercise Clause barred the application of the no-aid provision here, the Montana Supreme Court had no authority to invalidate the program on the basis of that provision. The Department argues that the invalidation of the entire program prevented a free exercise violation, but the Department overlooks the Montana Supreme Court’s threshold error of federal law. Had the Montana Supreme Court recognized that the application of the no-aid provision was barred by the Free Exercise Clause, the Court would have had no basis for invalidating the program. The Court was obligated to disregard the no-aid provision and decide this case consistent with the Federal Constitution. Pp. 20–22. 393 Mont. 446, 435 P.3d 603, reversed and remanded. Roberts, C. J., delivered the opinion of the Court, in which Thomas, Alito, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a concurring opinion, in which Gorsuch, J., joined. Alito, J., and Gorsuch, J., filed concurring opinions. Ginsburg, J., filed a dissenting opinion, in which Kagan, J., joined. Breyer, J., filed a dissenting opinion, in which Kagan, J., joined as to Part I. Sotomayor, J., filed a dissenting opinion.

The Montana Legislature established a program to provide tuition assistance to parents who send their children to private schools. The program grants a tax credit to anyone who donates to certain organizations that in turn award scholarships to selected students attending such schools. When petitioners sought to use the scholarships at a religious school, the Montana Supreme Court struck down the program. The Court relied on the “no-aid” provision of the State Constitution, which prohibits any aid to a school controlled by a “church, sect, or denomination.” The question presented is whether the Free Exercise Clause of the United States Constitution barred that application of the no-aid provision. I A In 2015, the Montana Legislature sought “to provide parental and student choice in education” by enacting a scholarship program for students attending private schools. 2015 Mont. Laws p. 2168, §7. The program grants a tax credit of up to $150 to any taxpayer who donates to a participating “student scholarship organization.” Mont. Code Ann. §§15–30–3103(1), –3111(1) (2019). The scholarship organizations then use the donations to award scholarships to children for tuition at a private school. §§15–30–3102(7)(a), –3103(1)(c).[1] So far only one scholarship organization, Big Sky Scholarships, has participated in the program. Big Sky focuses on providing scholarships to families who face financial hardship or have children with disabilities. Scholarship organizations like Big Sky must, among other requirements, maintain an application process for awarding the scholarships; use at least 90% of all donations on scholarship awards; and comply with state reporting and monitoring requirements. §§15–30–3103(1), –3105(1), –3113(1). A family whose child is awarded a scholarship under the program may use it at any “qualified education provider”—that is, any private school that meets certain accreditation, testing, and safety requirements. See §15–30–3102(7). Virtually every private school in Montana qualifies. Upon receiving a scholarship, the family designates its school of choice, and the scholarship organization sends the scholarship funds directly to the school. §15–30–3104(1). Neither the scholarship organization nor its donors can restrict awards to particular types of schools. See §§15–30–3103(1)(b), –3111(1). The Montana Legislature allotted $3 million annually to fund the tax credits, beginning in 2016. §15–30–3111(5)(a). If the annual allotment is exhausted, it increases by 10% the following year. Ibid. The program is slated to expire in 2023. 2015 Mont. Laws p. 2186, §33. The Montana Legislature also directed that the program be administered in accordance with Article X, section 6, of the Montana Constitution, which contains a “no-aid” provision barring government aid to sectarian schools. See Mont. Code Ann. §15–30–3101. In full, that provision states: “Aid prohibited to sectarian schools. . . . The legislature, counties, cities, towns, school districts, and public corporations shall not make any direct or indirect appropriation or payment from any public fund or monies, or any grant of lands or other property for any sectarian purpose or to aid any church, school, academy, seminary, college, university, or other literary or scientific institution, controlled in whole or in part by any church, sect, or denomination.” Mont. Const., Art. X, §6(1). Shortly after the scholarship program was created, the Montana Department of Revenue promulgated “Rule 1,” over the objection of the Montana Attorney General. That administrative rule prohibited families from using scholarships at religious schools. Mont. Admin. Rule §42.4.802(1)(a) (2015). It did so by changing the definition of “qualified education provider” to exclude any school “owned or controlled in whole or in part by any church, religious sect, or denomination.” Ibid. The Department explained that the Rule was needed to reconcile the scholarship program with the no-aid provision of the Montana Constitution. The Montana Attorney General disagreed. In a letter to the Department, he advised that the Montana Constitution did not require excluding religious schools from the program, and if it did, it would “very likely” violate the United States Constitution by discriminating against the schools and their students. See Complaint in No. DV–15–1152A (Dist. Ct. Flathead Cty.), Exh. 3, pp. 2, 5–6. The Attorney General is not representing the Department in this case. B This suit was brought by three mothers whose children attend Stillwater Christian School in northwestern Montana. Stillwater is a private Christian school that meets the statutory criteria for “qualified education providers.” It serves students in prekindergarten through 12th grade, and petitioners chose the school in large part because it “teaches the same Christian values that [they] teach at home.” App. to Pet. for Cert. 152; see id., at 138, 167. The child of one petitioner has already received scholarships from Big Sky, and the other petitioners’ children are eligible for scholarships and planned to apply. While in effect, however, Rule 1 blocked petitioners from using scholarship funds for tuition at Stillwater. To overcome that obstacle, petitioners sued the Department of Revenue in Montana state court. Petitioners claimed that Rule 1 conflicted with the statute that created the scholarship program and could not be justified on the ground that it was compelled by the Montana Constitution’s no-aid provision. Petitioners further alleged that the Rule discriminated on the basis of their religious views and the religious nature of the school they had chosen for their children. The trial court enjoined Rule 1, holding that it was based on a mistake of law. The court explained that the Rule was not required by the no-aid provision, because that provision prohibits only “appropriations” that aid religious schools, “not tax credits.” Id., at 94. The injunctive relief freed Big Sky to award scholarships to students regardless of whether they attended a religious or secular school. For the school year beginning in fall 2017, Big Sky received 59 applications and ultimately awarded 44 scholarships of $500 each. The next year, Big Sky received 90 applications and awarded 54 scholarships of $500 each. Several families, most with incomes of $30,000 or less, used the scholarships to send their children to Stillwater Christian. In December 2018, the Montana Supreme Court reversed the trial court. 393 Mont. 446, 435 P.3d 603. The Court first addressed the scholarship program unmodified by Rule 1, holding that the program aided religious schools in violation of the no-aid provision of the Montana Constitution. In the Court’s view, the no-aid provision “broadly and strictly prohibits aid to sectarian schools.” Id., at 459, 435 P. 3d, at 609. The scholarship program provided such aid by using tax credits to “subsidize tuition payments” at private schools that are “religiously affiliated” or “controlled in whole or in part by churches.” Id., at 464–467, 435 P. 3d, at 612–613. In that way, the scholarship program flouted the State Constitution’s “guarantee to all Montanans that their government will not use state funds to aid religious schools.” Id., at 467, 435 P. 3d, at 614. The Montana Supreme Court went on to hold that the violation of the no-aid provision required invalidating the entire scholarship program. The Court explained that the program provided “no mechanism” for preventing aid from flowing to religious schools, and therefore the scholarship program could not “under any circumstance” be construed as consistent with the no-aid provision. Id., at 466–468, 435 P. 3d, at 613–614. As a result, the tax credit is no longer available to support scholarships at either religious or secular private schools. The Montana Supreme Court acknowledged that “an overly-broad” application of the no-aid provision “could implicate free exercise concerns” and that “there may be a case” where “prohibiting the aid would violate the Free Exercise Clause.” Id., at 468, 435 P. 3d, at 614. But, the Court concluded, “this is not one of those cases.” Ibid. Finally, the Court agreed with petitioners that the Department had exceeded its authority in promulgating Rule 1. The Court explained that the statute creating the scholarship program had broadly defined qualifying schools to include all private schools, including religious ones, and the Department lacked authority to “transform” that definition with an administrative rule. Id., at 468–469, 435 P. 3d, at 614–615. Several Justices wrote separately. All agreed that Rule 1 was invalid, but they expressed differing views on whether the scholarship program was consistent with the Montana and United States Constitutions. Justice Gustafson’s concurrence argued that the program violated not only Montana’s no-aid provision but also the Federal Establishment and Free Exercise Clauses. Id., at 475–479, 435 P. 3d, at 619–621. Justice Sandefur echoed the majority’s conclusion that applying the no-aid provision was consistent with the Free Exercise Clause, and he dismissed the “modern jurisprudence” of that Clause as “unnecessarily complicate[d]” due to “increasingly value-driven hairsplitting and overstretching.” Id., at 482–484, 435 P. 3d, at 623–624. Two Justices dissented. Justice Rice would have held that the scholarship program was permissible under the no-aid provision. He criticized the majority for invalidating the program “sua sponte,” contending that no party had challenged it under the State Constitution. Id., at 495, 435 P. 3d, at 631. Justice Baker also would have upheld the program. In her view, the no-aid provision did not bar the use of scholarships at religious schools, and free exercise concerns could arise under the Federal Constitution if it did. Id., at 493–494, 435 P. 3d, at 630. We granted certiorari. 588 U. S. ___ (2019). II A The Religion Clauses of the First Amendment provide that “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof.” We have recognized a “ ‘play in the joints’ between what the Establishment Clause permits and the Free Exercise Clause compels.” Trinity Lutheran Church of Columbia, Inc. v. Comer, 582 U. S. ___, ___ (2017) (slip op., at 6) (quoting Locke v. Davey, 540 U.S. 712, 718 (2004)). Here, the parties do not dispute that the scholarship program is permissible under the Establishment Clause. Nor could they. We have repeatedly held that the Establishment Clause is not offended when religious observers and organizations benefit from neutral government programs. See, e.g., Locke, 540 U. S., at 719; Rosenberger v. Rector and Visitors of Univ. of Va., 515 U.S. 819, 839 (1995). See also Trinity Lutheran, 582 U. S., at ___ (slip op., at 6) (noting the parties’ agreement that the Establishment Clause was not violated by including churches in a playground resurfacing program). Any Establishment Clause objection to the scholarship program here is particularly unavailing because the government support makes its way to religious schools only as a result of Montanans independently choosing to spend their scholarships at such schools. See Locke, 540 U. S., at 719; Zelman v. Simmons-Harris, 536 U.S. 639, 649–653 (2002). The Montana Supreme Court, however, held as a matter of state law that even such indirect government support qualified as “aid” prohibited under the Montana Constitution. The question for this Court is whether the Free Exercise Clause precluded the Montana Supreme Court from applying Montana’s no-aid provision to bar religious schools from the scholarship program. For purposes of answering that question, we accept the Montana Supreme Court’s interpretation of state law—including its determination that the scholarship program provided impermissible “aid” within the meaning of the Montana Constitution—and we assess whether excluding religious schools and affected families from that program was consistent with the Federal Constitution.[2] The Free Exercise Clause, which applies to the States under the Fourteenth Amendment, “protects religious observers against unequal treatment” and against “laws that impose special disabilities on the basis of religious status.” Trinity Lutheran, 582 U. S., at ___, ___ (slip op., at 6, 9) (internal quotation marks and alterations omitted); see Cantwell v. Connecticut, 310 U.S. 296, 303 (1940). Those “basic principle[s ]” have long guided this Court. Trinity Lutheran, 582 U. S., at ___–___ (slip op., at 6–9). See, e.g., Everson v. Board of Ed. of Ewing, 330 U.S. 1, 16 (1947) (a State “cannot exclude individual Catholics, Lutherans, Mohammedans, Baptists, Jews, Methodists, Non-believers, Presbyterians, or the members of any other faith, because of their faith, or lack of it, from receiving the benefits of public welfare legislation”); Lyng v. Northwest Indian Cemetery Protective Assn., 485 U.S. 439, 449 (1988) (the Free Exercise Clause protects against laws that “penalize religious activity by denying any person an equal share of the rights, benefits, and privileges enjoyed by other citizens”). Most recently, Trinity Lutheran distilled these and other decisions to the same effect into the “unremarkable” conclusion that disqualifying otherwise eligible recipients from a public benefit “solely because of their religious character” imposes “a penalty on the free exercise of religion that triggers the most exacting scrutiny.” 582 U. S., at ___–___ (slip op., at 9–10). In Trinity Lutheran, Missouri provided grants to help nonprofit organizations pay for playground resurfacing, but a state policy disqualified any organization “owned or controlled by a church, sect, or other religious entity.” Id., at ___ (slip op., at 2). Because of that policy, an otherwise eligible church-owned preschool was denied a grant to resurface its playground. Missouri’s policy discriminated against the Church “simply because of what it is—a church,” and so the policy was subject to the “strictest scrutiny,” which it failed. Id., at ___–___ (slip op., at 11–15). We acknowledged that the State had not “criminalized” the way in which the Church worshipped or “told the Church that it cannot subscribe to a certain view of the Gospel.” Id., at ___ (slip op., at 11). But the State’s discriminatory policy was “odious to our Constitution all the same.” Id., at ___ (slip op., at 15). Here too Montana’s no-aid provision bars religious schools from public benefits solely because of the religious character of the schools. The provision also bars parents who wish to send their children to a religious school from those same benefits, again solely because of the religious character of the school. This is apparent from the plain text. The provision bars aid to any school “controlled in whole or in part by any church, sect, or denomination.” Mont. Const., Art. X, §6(1). The provision’s title—“Aid prohibited to sectarian schools”—confirms that the provision singles out schools based on their religious character. Ibid. And the Montana Supreme Court explained that the provision forbids aid to any school that is “sectarian,” “religiously affiliated,” or “controlled in whole or in part by churches.” 393 Mont., at 464–467, 435 P. 3d, at 612–613. The provision plainly excludes schools from government aid solely because of religious status. See Trinity Lutheran, 582 U. S., at ___–___ (slip op., at 9–10). The Department counters that Trinity Lutheran does not govern here because the no-aid provision applies not because of the religious character of the recipients, but because of how the funds would be used—for “religious education.” Brief for Respondents 38. In Trinity Lutheran, a majority of the Court concluded that the Missouri policy violated the Free Exercise Clause because it discriminated on the basis of religious status. A plurality declined to address discrimination with respect to “religious uses of funding or other forms of discrimination.” 582 U. S., at ___, n. 3 (slip op., at 14, n. 3). The plurality saw no need to consider such concerns because Missouri had expressly discriminated “based on religious identity,” ibid., which was enough to invalidate the state policy without addressing how government funds were used. This case also turns expressly on religious status and not religious use. The Montana Supreme Court applied the no-aid provision solely by reference to religious status. The Court repeatedly explained that the no-aid provision bars aid to “schools controlled in whole or in part by churches,” “sectarian schools,” and “religiously-affiliated schools.” 393 Mont., at 463–467, 435 P. 3d, at 611–613. Applying this provision to the scholarship program, the Montana Supreme Court noted that most of the private schools that would benefit from the program were “religiously affiliated” and “controlled by churches,” and the Court ultimately concluded that the scholarship program ran afoul of the Montana Constitution by aiding “schools controlled by churches.” Id., at 466–467, 435 P. 3d, at 613–614. The Montana Constitution discriminates based on religious status just like the Missouri policy in Trinity Lutheran, which excluded organizations “owned or controlled by a church, sect, or other religious entity.” 582 U. S., at ___ (slip op., at 2). The Department points to some language in the decision below indicating that the no-aid provision has the goal or effect of ensuring that government aid does not end up being used for “sectarian education” or “religious education.” 393 Mont., at 460, 466–467, 435 P. 3d, at 609, 613–614. The Department also contrasts what it characterizes as the “completely non-religious” benefit of playground resurfacing in Trinity Lutheran with the unrestricted tuition aid at issue here. Tr. of Oral Arg. 31. General school aid, the Department stresses, could be used for religious ends by some recipients, particularly schools that believe faith should “permeate[ ]” everything they do. Brief for Respondents 39 (quoting State ex rel. Chambers v. School Dist. No. 10, 155 Mont. 422, 438, 472 P.2d 1013, 1021 (1970)). See also post, at 8, 13 (Breyer, J., dissenting). Regardless, those considerations were not the Montana Supreme Court’s basis for applying the no-aid provision to exclude religious schools; that hinged solely on religious status. Status-based discrimination remains status based even if one of its goals or effects is preventing religious organizations from putting aid to religious uses. Undeterred by Trinity Lutheran, the Montana Supreme Court applied the no-aid provision to hold that religious schools could not benefit from the scholarship program. 393 Mont., at 464–468, 435 P. 3d, at 612–614. So applied, the provision “impose[s] special disabilities on the basis of religious status” and “condition[s] the availability of benefits upon a recipient’s willingness to surrender [its] religiously impelled status.” Trinity Lutheran, 582 U. S., at ___–___ (slip op., at 9–10) (quoting Church of Lukumi Babalu Aye, Inc. v. Hialeah, 508 U.S. 520, 533 (1993), and McDaniel v. Paty, 435 U.S. 618, 626 (1978) (plurality opinion) (alterations omitted)). To be eligible for government aid under the Montana Constitution, a school must divorce itself from any religious control or affiliation. Placing such a condition on benefits or privileges “inevitably deters or discourages the exercise of First Amendment rights.” Trinity Lutheran, 582 U. S., at ___ (slip op., at 11) (quoting Sherbert v. Verner, 374 U.S. 398, 405 (1963) (alterations omitted)). The Free Exercise Clause protects against even “indirect coercion,” and a State “punishe[s] the free exercise of religion” by disqualifying the religious from government aid as Montana did here. Trinity Lutheran, 582 U. S., at ___–___ (slip op., at 10–11) (internal quotation marks omitted). Such status-based discrimination is subject to “the strictest scrutiny.” Id., at ___ (slip op., at 11). None of this is meant to suggest that we agree with the Department, Brief for Respondents 36–40, that some lesser degree of scrutiny applies to discrimination against religious uses of government aid. See Lukumi, 508 U. S., at 546 (striking down law designed to ban religious practice involving alleged animal cruelty, explaining that a law “target[ing] religious conduct for distinctive treatment or advanc[ing] legitimate governmental interests only against conduct with a religious motivation will survive strict scrutiny only in rare cases”). Some Members of the Court, moreover, have questioned whether there is a meaningful distinction between discrimination based on use or conduct and that based on status. See Trinity Lutheran, 582 U. S., at ___–___ (slip op., at 1–2) (Gorsuch, J., joined by Thomas, J., concurring in part) (citing, e.g., Lukumi, 508 U.S. 520, and Thomas v. Review Bd. of Ind. Employment Security Div., 450 U.S. 707 (1981)). We acknowledge the point but need not examine it here. It is enough in this case to conclude that strict scrutiny applies under Trinity Lutheran because Montana’s no-aid provision discriminates based on religious status. B Seeking to avoid Trinity Lutheran, the Department contends that this case is instead governed by Locke v. Davey, 540 U.S. 712 (2004). See also post, at 5 (Breyer, J., dissenting); post, at 9 (Sotomayor, J., dissenting). Locke also involved a scholarship program. The State of Washington provided scholarships paid out of the State’s general fund to help students pursuing postsecondary education. The scholarships could be used at accredited religious and nonreligious schools alike, but Washington prohibited students from using the scholarships to pursue devotional theology degrees, which prepared students for a calling as clergy. This prohibition prevented Davey from using his scholarship to obtain a degree that would have enabled him to become a pastor. We held that Washington had not violated the Free Exercise Clause. Locke differs from this case in two critical ways. First, Locke explained that Washington had “merely chosen not to fund a distinct category of instruction”: the “essentially religious endeavor” of training a minister “to lead a congregation.” Id., at 721. Thus, Davey “was denied a scholarship because of what he proposed to do—use the funds to prepare for the ministry.” Trinity Lutheran, 582 U. S., at ___ (slip op., at 12). Apart from that narrow restriction, Washington’s program allowed scholarships to be used at “pervasively religious schools” that incorporated religious instruction throughout their classes. Locke, 540 U. S., at 724–725. By contrast, Montana’s Constitution does not zero in on any particular “essentially religious” course of instruction at a religious school. Rather, as we have explained, the no-aid provision bars all aid to a religious school “simply because of what it is,” putting the school to a choice between being religious or receiving government benefits. Trinity Lutheran, 582 U. S., at ___ (slip op., at 12). At the same time, the provision puts families to a choice between sending their children to a religious school or receiving such benefits. Second, Locke invoked a “historic and substantial” state interest in not funding the training of clergy, 540 U. S., at 725, explaining that “opposition to . . . funding ‘to support church leaders’ lay at the historic core of the Religion Clauses,” Trinity Lutheran, 582 U. S., at ___ (slip op., at 13) (quoting Locke, 540 U. S., at 722). As evidence of that tradition, the Court in Locke emphasized that the propriety of state-supported clergy was a central subject of founding-era debates, and that most state constitutions from that era prohibited the expenditure of tax dollars to support the clergy. See id., at 722–723. But no comparable “historic and substantial” tradition supports Montana’s decision to disqualify religious schools from government aid. In the founding era and the early 19th century, governments provided financial support to private schools, including denominational ones. “Far from prohibiting such support, the early state constitutions and statutes actively encouraged this policy.” L. Jorgenson, The State and the Non-Public School, 1825–1925, p. 4 (1987); e.g., R. Gabel, Public Funds for Church and Private Schools 210, 217–218, 221, 241–243 (1937); C. Kaestle, Pillars of the Republic: Common Schools and American Society, 1760–1860, pp. 166–167 (1983). Local governments provided grants to private schools, including religious ones, for the education of the poor. M. McConnell, et al., Religion and the Constitution 318–319 (4th ed. 2016). Even States with bans on government-supported clergy, such as New Jersey, Pennsylvania, and Georgia, provided various forms of aid to religious schools. See Kaestle, supra, at 166–167; Gabel, supra, at 215–218, 241–245, 372–374; cf. Locke, 540 U. S., at 723. Early federal aid (often land grants) went to religious schools. McConnell, supra, at 319. Congress provided support to denominational schools in the District of Columbia until 1848, ibid., and Congress paid churches to run schools for American Indians through the end of the 19th century, see Quick Bear v. Leupp, 210 U.S. 50, 78 (1908); Gabel, supra, at 521–523. After the Civil War, Congress spent large sums on education for emancipated freedmen, often by supporting denominational schools in the South through the Freedmen’s Bureau. McConnell, supra, at 323.[3] The Department argues that a tradition against state support for religious schools arose in the second half of the 19th century, as more than 30 States—including Montana—adopted no-aid provisions. See Brief for Respondents 40–42 and App. D. Such a development, of course, cannot by itself establish an early American tradition. Justice Sotomayor questions our reliance on aid provided during the same era by the Freedmen’s Bureau, post, at 10 (dissenting opinion), but we see no inconsistency in recognizing that such evidence may reinforce an early practice but cannot create one. In addition, many of the no-aid provisions belong to a more checkered tradition shared with the Blaine Amendment of the 1870s. That proposal—which Congress nearly passed—would have added to the Federal Constitution a provision similar to the state no-aid provisions, prohibiting States from aiding “sectarian” schools. See Mitchell v. Helms, 530 U.S. 793, 828 (2000) (plurality opinion). “[I]t was an open secret that ‘sectarian’ was code for ‘Catholic.’ ” Ibid.; see Jorgenson, supra, at 70. The Blaine Amendment was “born of bigotry” and “arose at a time of pervasive hostility to the Catholic Church and to Catholics in general”; many of its state counterparts have a similarly “shameful pedigree.” Mitchell, 530 U. S., at 828–829 (plurality opinion); see Jorgenson, supra, at 69–70, 216; Jeffries & Ryan, A Political History of the Establishment Clause, 100 Mich. L. Rev. 279, 301–305 (2001). The no-aid provisions of the 19th century hardly evince a tradition that should inform our understanding of the Free Exercise Clause. The Department argues that several States have rejected referendums to overturn or limit their no-aid provisions, and that Montana even re-adopted its own in the 1970s, for reasons unrelated to anti-Catholic bigotry. See Brief for Respondents 20, 42. But, on the other side of the ledger, many States today—including those with no-aid provisions—provide support to religious schools through vouchers, scholarships, tax credits, and other measures. See Brief for Oklahoma et al. as Amici Curiae 29–31, 33–35; Brief for Petitioners 5. According to petitioners, 20 of 37 States with no-aid provisions allow religious options in publicly funded scholarship programs, and almost all allow religious options in tax credit programs. Reply Brief 22, n. 9. All to say, we agree with the Department that the historical record is “complex.” Brief for Respondents 41. And it is true that governments over time have taken a variety of approaches to religious schools. But it is clear that there is no “historic and substantial” tradition against aiding such schools comparable to the tradition against state-supported clergy invoked by Locke. C Two dissenters would chart new courses. Justice Sotomayor would grant the government “some room” to “single . . . out” religious entities “for exclusion,” based on what she views as “the interests embodied in the Religion Clauses.” Post, at 8, 9 (quoting Trinity Lutheran, 582 U. S., at ___, ___ (Sotomayor, J., dissenting) (slip op., at 8, 9)). Justice Breyer, building on his solo opinion in Trinity Lutheran, would adopt a “flexible, context-specific approach” that “may well vary” from case to case. Post, at 14, 16; see Trinity Lutheran, 582 U. S., at ___ (Breyer, J., concurring in judgment). As best we can tell, courts applying this approach would contemplate the particular benefit and restriction at issue and discern their relationship to religion and society, taking into account “context and consequences measured in light of [the] purposes” of the Religion Clauses. Post, at 16–17, 19 (quoting Van Orden v. Perry, 545 U.S. 677, 700 (2005) (Breyer, J., concurring in judgment)). What is clear is that Justice Breyer would afford much freer rein to judges than our current regime, arguing that “there is ‘no test-related substitute for the exercise of legal judgment.’ ” Post, at 19 (quoting Van Orden, 545 U. S., at 700 (opinion of Breyer, J.)). The simplest response is that these dissents follow from prior separate writings, not from the Court’s decision in Trinity Lutheran or the decades of precedent on which it relied. These precedents have “repeatedly confirmed” the straightforward rule that we apply today: When otherwise eligible recipients are disqualified from a public benefit “solely because of their religious character,” we must apply strict scrutiny. Trinity Lutheran, 582 U. S., at ___–___ (slip op., at 6–10). This rule against express religious discrimination is no “doctrinal innovation.” Post, at 13 (opinion of Breyer, J.). Far from it. As Trinity Lutheran explained, the rule is “unremarkable in light of our prior decisions.” 582 U. S., at ___ (slip op., at 10). For innovation, one must look to the dissents. Their “room[y]” or “flexible” approaches to discrimination against religious organizations and observers would mark a significant departure from our free exercise precedents. The protections of the Free Exercise Clause do not depend on a “judgment-by-judgment analysis” regarding whether discrimination against religious adherents would somehow serve ill-defined interests. Cf. Medellín v. Texas, 552 U.S. 491, 514 (2008). D Because the Montana Supreme Court applied the no-aid provision to discriminate against schools and parents based on the religious character of the school, the “strictest scrutiny” is required. Supra, at 9, 12 (quoting Trinity Lutheran, 582 U. S., at ___ (slip op., at 11)). That “stringent standard,” id., at ___ (slip op., at 14), is not “watered down but really means what it says,” Lukumi, 508 U. S., at 546 (internal quotation marks and alterations omitted). To satisfy it, government action “must advance ‘interests of the highest order’ and must be narrowly tailored in pursuit of those interests.” Ibid. (quoting McDaniel, 435 U. S., at 628). The Montana Supreme Court asserted that the no-aid provision serves Montana’s interest in separating church and State “more fiercely” than the Federal Constitution. 393 Mont., at 467, 435 P. 3d, at 614. But “that interest cannot qualify as compelling” in the face of the infringement of free exercise here. Trinity Lutheran, 582 U. S., at ___ (slip op., at 14). A State’s interest “in achieving greater separation of church and State than is already ensured under the Establishment Clause . . . is limited by the Free Exercise Clause.” Ibid. (quoting Widmar v. Vincent, 454 U.S. 263, 276 (1981)). The Department, for its part, asserts that the no-aid provision actually promotes religious freedom. In the Department’s view, the no-aid provision protects the religious liberty of taxpayers by ensuring that their taxes are not directed to religious organizations, and it safeguards the freedom of religious organizations by keeping the government out of their operations. See Brief for Respondents 17–23. An infringement of First Amendment rights, however, cannot be justified by a State’s alternative view that the infringement advances religious liberty. Our federal system prizes state experimentation, but not “state experimentation in the suppression of free speech,” and the same goes for the free exercise of religion. Boy Scouts of America v. Dale, 530 U.S. 640, 660 (2000). Furthermore, we do not see how the no-aid provision promotes religious freedom. As noted, this Court has repeatedly upheld government programs that spend taxpayer funds on equal aid to religious observers and organizations, particularly when the link between government and religion is attenuated by private choices. A school, concerned about government involvement with its religious activities, might reasonably decide for itself not to participate in a government program. But we doubt that the school’s liberty is enhanced by eliminating any option to participate in the first place. The Department’s argument is especially unconvincing because the infringement of religious liberty here broadly affects both religious schools and adherents. Montana’s no-aid provision imposes a categorical ban—“broadly and strictly” prohibiting “any type of aid” to religious schools. 393 Mont., at 462–463, 435 P. 3d, at 611. This prohibition is far more sweeping than the policy in Trinity Lutheran, which barred churches from one narrow program for playground resurfacing—causing “in all likelihood” only “a few extra scraped knees.” 582 U. S., at ___ (slip op., at 15). And the prohibition before us today burdens not only religious schools but also the families whose children attend or hope to attend them. Drawing on “enduring American tradition,” we have long recognized the rights of parents to direct “the religious upbringing” of their children. Wisconsin v. Yoder, 406 U.S. 205, 213–214, 232 (1972). Many parents exercise that right by sending their children to religious schools, a choice protected by the Constitution. See Pierce v. Society of Sisters, 268 U.S. 510, 534–535 (1925). But the no-aid provision penalizes that decision by cutting families off from otherwise available benefits if they choose a religious private school rather than a secular one, and for no other reason. The Department also suggests that the no-aid provision advances Montana’s interests in public education. According to the Department, the no-aid provision safeguards the public school system by ensuring that government support is not diverted to private schools. See Brief for Respondents 19, 25. But, under that framing, the no-aid provision is fatally underinclusive because its “proffered objectives are not pursued with respect to analogous nonreligious conduct.” Lukumi, 508 U. S., at 546. On the Department’s view, an interest in public education is undermined by diverting government support to any private school, yet the no-aid provision bars aid only to religious ones. A law does not advance “an interest of the highest order when it leaves appreciable damage to that supposedly vital interest unprohibited.” Id., at 547 (internal quotation marks and alterations omitted). Montana’s interest in public education cannot justify a no-aid provision that requires only religious private schools to “bear [its] weight.” Ibid. A State need not subsidize private education. But once a State decides to do so, it cannot disqualify some private schools solely because they are religious. III The Department argues that, at the end of the day, there is no free exercise violation here because the Montana Supreme Court ultimately eliminated the scholarship program altogether. According to the Department, now that there is no program, religious schools and adherents cannot complain that they are excluded from any generally available benefit. Two dissenters agree. Justice Ginsburg reports that the State of Montana simply chose to “put all private school parents in the same boat” by invalidating the scholarship program, post, at 5–6, and Justice Sotomayor describes the decision below as resting on state law grounds having nothing to do with the federal Free Exercise Clause, see post, at 1, 6. The descriptions are not accurate. The Montana Legislature created the scholarship program; the Legislature never chose to end it, for policy or other reasons. The program was eliminated by a court, and not based on some innocuous principle of state law. Rather, the Montana Supreme Court invalidated the program pursuant to a state law provision that expressly discriminates on the basis of religious status. The Court applied that provision to hold that religious schools were barred from participating in the program. Then, seeing no other “mechanism” to make absolutely sure that religious schools received no aid, the court chose to invalidate the entire program. 393 Mont., at 466–468, 435 P. 3d, at 613–614. The final step in this line of reasoning eliminated the program, to the detriment of religious and non-religious schools alike. But the Court’s error of federal law occurred at the beginning. When the Court was called upon to apply a state law no-aid provision to exclude religious schools from the program, it was obligated by the Federal Constitution to reject the invitation. Had the Court recognized that this was, indeed, “one of those cases” in which application of the no-aid provision “would violate the Free Exercise Clause,” id., at 468, 435 P. 3d, at 614, the Court would not have proceeded to find a violation of that provision. And, in the absence of such a state law violation, the Court would have had no basis for terminating the program. Because the elimination of the program flowed directly from the Montana Supreme Court’s failure to follow the dictates of federal law, it cannot be defended as a neutral policy decision, or as resting on adequate and independent state law grounds.[4] The Supremacy Clause provides that “the Judges in every State shall be bound” by the Federal Constitution, “any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.” Art. VI, cl. 2. “[T]his Clause creates a rule of decision” directing state courts that they “must not give effect to state laws that conflict with federal law[ ].” Armstrong v. Exceptional Child Center, Inc., 575 U.S. 320, 324 (2015). Given the conflict between the Free Exercise Clause and the application of the no-aid provision here, the Montana Supreme Court should have “disregard[ed]” the no-aid provision and decided this case “conformably to the [C]onstitution” of the United States. Marbury v. Madison, 1 Cranch 137, 178 (1803). That “supreme law of the land” condemns discrimination against religious schools and the families whose children attend them. Id., at 180. They are “member[s] of the community too,” and their exclusion from the scholarship program here is “odious to our Constitution” and “cannot stand.” Trinity Lutheran, 582 U. S., at ___, ___ (slip op., at 11, 15).[5] * * * The judgment of the Montana Supreme Court is reversed, and the case is remanded for further proceedings not inconsistent with this opinion. It is so ordered. Notes 1 The Legislature provided the same tax credit to taxpayers who donate to public schools for the purpose of supporting innovative educational programs or curing technology deficiencies at such schools. See Mont. Code Ann. §15–30–3110 (2019). 2 Justice Sotomayor argues that the Montana Supreme Court “expressly declined to reach any federal issue.” Post, at 6 (dissenting opinion). Not so. As noted, supra, at 5, the Montana Supreme Court recognized that certain applications of the no-aid provision could “violate the Free Exercise Clause.” 393 Mont. 446, 468, 435 P.3d 603, 614 (2018). But the Court expressly concluded that “this is not one of those cases.” Ibid. 3 Justice Breyer sees “no meaningful difference” between concerns animating bans on support for clergy and bans on support for religious schools. Post, at 8–10. But evidently early American governments did. See supra, at 14. Justice Breyer contests particular examples but acknowledges that some bans on clergy support did not bar certain “sponsorship” of religious schools. Post, at 10. And, central to the issue here, he certainly does not identify a consistent early tradition, of the sort invoked in Locke, against support for religious schools. Virginia’s opposition to establishing university theology professorships and chartering theological seminaries, see ibid., do not fit the bill. Buckley, After Disestablishment: Thomas Jefferson’s Wall of Separation in Antebellum Virginia, 61 J. So. Hist. 445, 452–453 (1995). Justice Breyer also invokes Madison’s objections to the Virginia Assessment Bill, post, at 8–9, but Madison objected in part because the Bill provided special support to certain churches and clergy, thereby “violat[ing] equality by subjecting some to peculiar burdens.” Memorial and Remonstrance Against Religious Assessments, Art. 4, reprinted in Everson, 330 U. S., at 66 (appendix to dissenting opinion of Rutledge, J.); see V. Muñoz, God and the Founders: Madison, Washington, and Jefferson 21–22, 27 (2009). It is far from clear that the same objections extend to programs that provide equal support to all private primary and secondary schools. If anything, excluding religious schools from such programs would appear to impose the “peculiar burdens” feared by Madison. 4 Justice Sotomayor worries that, in light of our decision, the Montana Supreme Court must “order the State to recreate” a scholarship program that “no longer exists.” Post, at 6 (dissenting opinion). But it was the Montana Supreme Court that eliminated the program, in the decision below, which remains under review. Our reversal of that decision simply restores the status quo established by the Montana Legislature before the Court’s error of federal law. We do not consider any alterations the Legislature may choose to make in the future. 5 In light of this holding, we do not address petitioners’ claims that the no-aid provision, as applied, violates the Equal Protection Clause or the Establishment Clause.

590.US.2019_18-1334

In 2016, in response to a fiscal crisis in Puerto Rico, Congress invoked its Article IV power to “make all needful Rules and Regulations respecting the Territory . . . belonging to the United States,” §3, cl. 2, to enact the Puerto Rico Oversight, Management, and Economic Stability Act (PROMESA). PROMESA created a Financial Oversight and Management Board, whose seven voting members are to be appointed by the President without the Senate’s advice and consent. Congress authorized the Board to file for bankruptcy on behalf of Puerto Rico or its instrumentalities, to supervise and modify Puerto Rico’s laws and budget, and to gather evidence and conduct investigations in support of these efforts. After President Obama selected the Board’s members, the Board filed bankruptcy petitions on behalf of the Commonwealth and five of its entities. Both court and Board had decided a number of matters when several creditors moved to dismiss the proceedings on the ground that the Board members’ selection violated the Constitution’s Appointments Clause, which says that the President “shall nominate, and by and with the Advice and Consent of the Senate, shall appoint . . . all . . . Officers of the United States . . . .” Art. II, §2, cl. 2. The court denied the motions, but the First Circuit reversed. It held that the Board members’ selection violated the Appointments Clause but also concluded that any Board actions taken prior to its decision were valid under the “de facto officer” doctrine. Held: 1. The Appointments Clause constrains the appointments power as to all officers of the United States, even those who exercise power in or in relation to Puerto Rico. The Constitution’s structure provides strong reason to believe that this is so. The Appointments Clause reflects an allocation of responsibility, between President and Senate, in cases involving appointment to high federal office. Concerned about possible manipulation of appointments, the Founders both concentrated the appointment power and distributed it, ensuring that primary responsibility for important nominations would fall on the President while also ensuring that the Senate’s advice and consent power would provide a check on that power. Other, similar structural constraints in the Constitution apply to all exercises of federal power, including those related to Article IV entities. Cf., e.g., Metropolitan Washington Airports Authority v. Citizens for Abatement of Aircraft Noise, Inc., 501 U.S. 252, 270–271 (MWAA). The objectives advanced by the Appointments Clause counsel strongly in favor of applying that Clause to all officers of the United States, even those with powers and duties related to Puerto Rico. Indeed, the Clause’s text firmly indicates that it applies to the appointment of all “Officers of the United States.” And history confirms this reading. Congress’ longstanding practice of requiring the Senate’s advice and consent for territorial Governors with important federal duties supports the inference that Congress expected the Appointments Clause to apply to at least some officials with supervisory authority over the Territories. Pp. 5–9. 2. The Appointments Clause does not restrict the appointment or selection of the Board members. Pp. 9–21. (a) The Appointments Clause does not restrict the appointment of local officers that Congress vests with primarily local duties. The Clause’s language suggests a distinction between federal officers—who exercise power of the National Government—and nonfederal officers—who exercise power of some other government. Pursuant to Article I, §8, cl. 17, and Article IV, §3, Congress has long legislated for entities that are not States—the District of Columbia and the Territories. In so doing, Congress has both made local law directly and also created local government structures, staffed by local officials, who themselves have made and enforced local law. This suggests that when Congress creates local offices using these two unique powers, the officers exercise power of the local government, not the Federal Government. Historical practice indicates that a federal law’s creation of an office does not automatically make its holder an officer of the United States. Congress has for more than two centuries created local offices for the Territories and District of Columbia that are filled through election or local executive appointment. And the history of Puerto Rico—whose public officials with important local responsibilities have been selected in ways that the Appointments Clause does not describe—is consistent with the history of other entities that fall within Article IV’s scope and with the history of the District of Columbia. This historical practice indicates that when an officer of one of these local governments has primarily local duties, he is not an officer of the United States within the meaning of the Appointments Clause. Pp. 9–14. (b) The Board members here have primarily local powers and duties. PROMESA says that the Board is “an entity within the territorial government” that “shall not be considered a department, agency, establishment, or instrumentality of the Federal Government,” §101(c), 130Stat. 553, and Congress gave the Board a structure, duties, and related powers that are consistent with this statement. The Board’s broad investigatory powers—administering oaths, issuing subpoenas, taking evidence, and demanding data from governments and creditors alike—are backed by Puerto Rican, not federal, law. Its powers to oversee the development of Puerto Rico’s fiscal and budgetary plans are also quintessentially local. And in exercising its power to initiate bankruptcy proceedings, the Board acts on behalf of, and in the interests of, Puerto Rico. Pp. 14–17. (c) Buckley v. Valeo, 424 U.S. 1, Freytag v. Commissioner, 501 U.S. 868, and Lucia v. SEC, 585 U. S. ___, do not provide the relevant legal test here, for each considered an Appointments Clause problem concerning the importance or significance of duties that were indisputably federal or national in nature. Nor do Lebron v. National Railroad Passenger Corporation, 513 U.S. 374, or MWAA, 501 U.S. 252, help. Lebron considered whether Amtrak was a governmental or a private entity, but the fact that the Board is a Government entity does not answer the “primarily local versus primarily federal” question. And the MWAA Court expressly declined to address Appointments Clause questions. However, the Court’s analysis in O’Donoghue v. United States, 289 U.S. 516, and Palmore v. United States, 411 U.S. 389, does provide a rough analogy. In O’Donoghue, the Court found that Article III’s tenure and salary protections applied to judges of the District of Columbia courts because those courts exercised the judicial power of the United States. But the Court reached the seemingly opposite conclusion in Palmore, a case decided after Congress had altered the nature of the District of Columbia local courts so that its judges adjudicated primarily local issues. Pp. 17–21. 3. Given the conclusion reached here, there is no need to consider whether to overrule the “Insular Cases” and their progeny, see, e.g., Downes v. Bidwell, 182 U.S. 244, 287, to consider the application of the de facto officer doctrine, see Ryder v. United States, 515 U.S. 177, or to decide questions about the application of the Federal Relations Act and Public Law 600. Pp. 21–22. 915 F.3d 838, reversed and remanded. Breyer, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Alito, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., and Sotomayor, J., filed opinions concurring in the judgment. Notes 1 Together with No. 18–1475, Aurelius Investment, LLC, et al. v. Commonwealth of Puerto Rico et al., No. 18–1496, Official Committee of Unsecured Creditors of All Title III Debtors Other Than COFINA v. Aurelius Investment, LLC, et al., No. 18–1514, United States v. Aurelius Investment, LLC, et al., and No. 18–1521, Unión de Trabajadores de la Industria Elećtrica y Riego, Inc. v. Financial Oversight and Management Board for Puerto Rico et al., also on certiorari to the same court.

The Constitution’s Appointments Clause says that the President “shall nominate, and by and with the Advice and Consent of the Senate, shall appoint Ambassadors, other public Ministers and Consuls, Judges of the supreme Court, and all other Officers of the United States . . . .” Art. II, §2, cl. 2 (emphasis added). In 2016, Congress enacted the Puerto Rico Oversight, Management, and Economic Stability Act (PROMESA). 130Stat. 549, 48 U. S. C. §2101 et seq. That Act created a Financial Oversight and Management Board, and it provided, as relevant here, that the President could appoint its seven members without “the advice and consent of the Senate,” i.e., without Senate confirmation. The question before us is whether this method of appointment violates the Constitution’s Senate confirmation requirement. In our view, the Appointments Clause governs the appointments of all officers of the United States, including those located in Puerto Rico. Yet two provisions of the Constitution empower Congress to create local offices for the District of Columbia and for Puerto Rico and the Territories. See Art. I, §8, cl. 17; Art. IV, §3, cl. 2. And the Clause’s term “Officers of the United States” has never been understood to cover those whose powers and duties are primarily local in nature and derive from these two constitutional provisions. The Board’s statutory responsibilities consist of primarily local duties, namely, representing Puerto Rico in bankruptcy proceedings and supervising aspects of Puerto Rico’s fiscal and budgetary policies. We therefore find that the Board members are not “Officers of the United States.” For that reason, the Appointments Clause does not dictate how the Board’s members must be selected. I A In 2006, tax advantages that had previously led major businesses to invest in Puerto Rico expired. See Small Business Job Protection Act of 1996, §1601, 110Stat. 1827. Many industries left the island. Emigration increased. And the public debt of Puerto Rico’s government and its instrumentalities soared, rising from $39.2 billion in 2005 to $71 billion in 2016. See Dept. of Treasury, Puerto Rico’s Economic and Fiscal Crisis 1, 3, https://www.treasury.gov/ connect/blog/Documents/Puerto_Ricos_fiscal_challenges.pdf; GAO, U. S. Territories: Public Debt Outlook 12 (GAO–18–160, 2017). Puerto Rico found that it could not service that debt. Yet Puerto Rico could not easily restructure it. The Federal Bankruptcy Code’s municipality-related Chapter 9 did not apply to Puerto Rico (or to the District of Columbia). See 11 U. S. C. §§109(c), 101(52). But at the same time, federal bankruptcy law invalidated Puerto Rico’s own local “debt-restructuring” statutes. Puerto Rico v. Franklin Cal. Tax-Free Trust, 579 U. S. ___ (2016). In 2016, in response to Puerto Rico’s fiscal crisis, Congress enacted PROMESA. 130Stat. 549, 48 U. S. C. §2101 et seq. PROMESA allows Puerto Rico and its entities to file for federal bankruptcy protection. See §§301, 302, 130Stat. 577, 579; cf. 11 U. S. C. §901 (related to bankruptcies of local governments). The filing and subsequent proceedings are to take place in the United States District Court for the District of Puerto Rico, before a federal judge selected by the Chief Justice of the United States. PROMESA §§307–308, 130Stat. 582. PROMESA also created the Financial Oversight and Management Board—with seven members appointed by the President and with the Governor serving as an ex officio member. §§101(b), (e), id., at 553, 554–555. PROMESA gives the Board authority to file for bankruptcy on behalf of Puerto Rico or its instrumentalities. §304(a), id., at 579. The Board can supervise and modify Puerto Rico’s laws (and budget) to “achieve fiscal responsibility and access to the capital markets.” §201(b), id., at 564; see §§201–207, id., at 563–575. And it can gather evidence and conduct investigations in support of these efforts. §104, id., at 558–561. As we have just said, PROMESA gives the President of the United States the power to appoint the Board’s seven members without Senate confirmation, so long as he selects six from lists prepared by congressional leaders. §101(e)(2)(A), id., at 554–555. B On August 31, 2016, President Obama selected the Board’s seven members in the manner just described. The Board established offices in Puerto Rico and New York, and soon filed bankruptcy petitions on behalf of the Commonwealth and (eventually) five Commonwealth entities. Title III Petition in No. 17–BK–3283 (PR); see Order Pursuant to PROMESA Section 304(g), No. 17–BK–3283 (PR, Oct. 9, 2019), Doc. 8829 (consolidating petitions filed on behalf of the Commonwealth of Puerto Rico, the Puerto Rico Sales Tax Financing Corporation, the Puerto Rico Highways and Transportation Authority, the Employees Retirement System of the Government of the Commonwealth of Puerto Rico, the Puerto Rico Electric Power Authority, and the Puerto Rico Public Buildings Authority). And the Chief Justice then selected a federal judge to serve as bankruptcy judge for Puerto Rico. Designation of Presiding District Judge, No. 17–BK–3283 (PR, May 5, 2017), Doc. 4. After both court and Board had decided a number of matters, several creditors moved to dismiss all proceedings on the ground that the Board members’ selection violated the Appointments Clause. The court denied the motions. See In re Financial Oversight and Management Bd. of Puerto Rico, 318 F. Supp. 3d 537, 556–557 (PR 2018). The creditors appealed to the United States Court of Appeals for the First Circuit. That court reversed. It held that the selection of the Board’s members violated the Appointments Clause. 915 F.3d 838, 861 (2019). But it concluded that those Board actions taken prior to its decision remained valid under the “de facto officer” doctrine. Id., at 862–863; see, e.g., McDowell v. United States, 159 U.S. 596, 601 (1895) (judicial decisions could not later be attacked on ground that an unlawfully sitting judge presided); Ball v. United States, 140 U.S. 118, 128–129 (1891) (same). The Board, the United States, and various creditors then filed petitions for certiorari in this Court, some arguing that the appointments were constitutionally valid, others that the de facto officer doctrine did not apply. Compare Pets. for Cert. in Nos. 18–1334, 18–1496, 18–1514 with Pets. for Cert. in Nos. 18–1475, 18–1521. In light of the importance of the questions, we granted certiorari in all the petitions and consolidated them for argument. 588 U. S. ___ (2019). II Congress created the Board pursuant to its power under Article IV of the Constitution to “make all needful Rules and Regulations respecting the Territory . . . belonging to the United States.” §3, cl. 2; see PROMESA §101(b)(2), 130Stat. 553. Some have argued in these cases that the Appointments Clause simply does not apply in the context of Puerto Rico. But, like the Court of Appeals, we believe the Appointments Clause restricts the appointment of all officers of the United States, including those who carry out their powers and duties in or in relation to Puerto Rico. The Constitution’s structure provides strong reason to believe that is so. The Constitution separates the three basic powers of Government—legislative, executive, and judicial—with each branch serving different functions. But the Constitution requires cooperation among the three branches in specified areas. Thus, to become law, proposed legislation requires the agreement of both Congress and the President (or, a supermajority in Congress). See INS v. Chadha, 462 U.S. 919, 955 (1983) (noting that the Constitution prescribes only four specific actions that Congress can take without bicameralism and presentment). At the same time, legislation must be consistent with constitutional constraints, and we usually look to the Judiciary as the ultimate interpreter of those constraints. The Appointments Clause reflects a similar allocation of responsibility, between President and Senate, in cases involving appointment to high federal office. That Clause reflects the Founders’ reaction to “one of [their] generation’s greatest grievances against [pre-Revolutionary] executive power,” the manipulation of appointments. Freytag v. Commissioner, 501 U.S. 868, 883 (1991); see also The Federalist No. 76, p. 455 (C. Rossiter ed. 1961) (A. Hamilton) (the Appointments Clause helps to preserve democratic accountability). The Founders addressed their concerns with the appointment power by both concentrating it and distributing it. On the one hand, they ensured that primary responsibility for nominations would fall on the President, whom they deemed “less vulnerable to interest-group pressure and personal favoritism” than a collective body. Edmond v. United States, 520 U.S. 651, 659 (1997). See also The Federalist No. 76, at 455 (“The sole and undivided responsibility of one man will naturally beget a livelier sense of duty and a more exact regard to reputation”). On the other hand, they ensured that the Senate’s advice and consent power would provide “an excellent check upon a spirit of favoritism in the President and a guard against the appointment of unfit characters.” NLRB v. SW General, Inc., 580 U. S. ___, ___ (2017) (slip op., at 2) (internal quotation marks omitted). By “limiting the appointment power” in this fashion, the Clause helps to “ensure that those who wielded [the appointments power] were accountable to political force and the will of the people.” Freytag, supra, at 884; see also Edmond, 520 U. S., at 659. “The blame of a bad nomination would fall upon the president singly and absolutely,” while “[t]he censure of rejecting a good one would lie entirely at the door of the senate.” Id., at 660 (internal quotation marks omitted). These other structural constraints, designed in part to ensure political accountability, apply to all exercises of federal power, including those related to Article IV entities. Cf., e.g., Metropolitan Washington Airports Authority v. Citizens for Abatement of Aircraft Noise, Inc., 501 U.S. 252, 270–271 (1991) (MWAA) (separation-of-powers principles apply when Congress acts under its Article IV power to legislate “respecting . . . other Property”). See also, e.g., Act of Aug. 7, 1789, ch. 8, 1Stat. 50 (the First Congress using bicameralism and presentment to make rules and regulations for the Northwest Territory). The objectives advanced by the Appointments Clause counsel strongly in favor of that Clause applying to the appointment of all “Officers of the United States.” Why should it be different when such an officer’s duties relate to Puerto Rico or other Article IV entities? Indeed, the Appointments Clause has no Article IV exception. The Clause says in part that the President “shall nominate, and by and with the Advice and Consent of the Senate, shall appoint Ambassadors, other public Ministers and Consuls, Judges of the supreme Court, and all other Officers of the United States, whose Appointments . . . shall be established by Law . . . .” Art. II, §2, cl. 2. That text firmly indicates that it applies to the appointment of all “Officers of the United States.” And history confirms this reading. Before the writing of the Constitution, Congress had enacted an ordinance that allowed Congress to appoint officers to govern the Northwest Territory. As soon as the Constitution became law, the First Congress “adapt[ed]” that ordinance “to the present Constitution of the United States,” Act of Aug. 7, 1789, 1Stat. 51, in large part by providing for an appointment process consistent with the constraints of the Appointments Clause. In particular, it provided for a Presidential-appointment, Senate-confirmation process for high-level territorial appointees who assumed federal, as well as local, duties. See id., at 52, n. (a); §1, id., at 53 (appointment by President, and confirmation by Senate, of Governor, secretary, and members of the upper house); Act of Sept. 11, 1789, ch. 13, §1, 1Stat. 68 (Governor “discharg[ed]” the federal “duties of superintendent of Indian affairs”). Later Congresses took a similar approach to later territorial Governors with federal duties. See Act of June 6, 1900, §10, 31Stat. 325 (appointment of Governor of Territory of Alaska by President with confirmation by Senate); §2, id., at 322 (federal duties of Alaska territorial Governor include entering into contracts in name of the United States and granting reprieves for federal offenses); Act of Mar. 2, 1819, §§3, 10, 3Stat. 494, 495 (similar for Governor of Arkansas). We do not mean to suggest that every time Congress chooses to require advice and consent procedures it does so because they are constitutionally required. At times, Congress may wish to require Senate confirmation for policy reasons. Even so, Congress’ practice of requiring advice and consent for these Governors with important federal duties supports the inference that Congress expected the Appointments Clause to apply to at least some officials with supervisory authority over the Territories. Given the Constitution’s structure, this history, roughly analogous case law, and the absence of any conflicting authority, we conclude that the Appointments Clause constrains the appointments power as to all “Officers of the United States,” even when those officers exercise power in or related to Puerto Rico. III A The more difficult question before us is whether the Board members are officers of the United States such that the Appointments Clause requires Senate confirmation. If they are not officers of the United States, but instead are some other type of officer, the Appointments Clause says nothing about them. (No one suggests that they are “Ambassadors,” “other public Ministers and Consuls,” or “Judges of the supreme Court.”) And as we shall see, the answer to this question turns on whether the Board members have primarily local powers and duties. The language at issue does not offer us much guidance for understanding the key term “of the United States.” The text suggests a distinction between federal officers—officers exercising power of the National Government—and nonfederal officers—officers exercising power of some other government. The Constitution envisions a federalist structure, with the National Government exercising limited federal power and other, local governments—usually state governments—exercising more expansive power. But the Constitution recognizes that for certain localities, there will be no state government capable of exercising local power. Thus, two provisions of the Constitution, Article I, §8, cl. 17, and Article IV, §3, cl. 2, give Congress the power to legislate for those localities in ways “that would exceed its powers, or at least would be very unusual” in other contexts. Palmore v. United States, 411 U.S. 389, 398 (1973). Using these powers, Congress has long legislated for entities that are not States—the District of Columbia and the Territories. See District of Columbia v. John R. Thompson Co., 346 U.S. 100, 104–106 (1953). And, in doing so, Congress has both made local law directly and also created structures of local government, staffed by local officials, who themselves have made and enforced local law. Compare, e.g., Act of Mar. 2, 1962, §401, 76Stat. 17 (changing D. C. liquor tax from $1.25 per gallon to $1.50 per gallon), with District of Columbia Self-Government and Governmental Reorganization Act, 87Stat. 774 (giving local D. C. government primary legislative control over local matters). This structure suggests that when Congress creates local offices using these two unique powers, the officers exercise power of the local government, not the Federal Government. Cf. American Ins. Co. v. 356 Bales of Cotton, 1 Pet. 511, 546 (1828) (Marshall, C. J.) (territorial courts may exercise the judicial power of the Territories without the life tenure and salary protections mandated by Article III for federal judges); Cincinnati Soap Co. v. United States, 301 U.S. 308, 323 (1937) (territorial legislators may exercise the legislative power of the Territories without violating the nondelegation doctrine). History confirms what the Constitution’s text and structure suggest. See NLRB v. Noel Canning, 573 U.S. 513, 524 (2014) (relying on history and structure in interpreting the Recess Appointments Clause). See also McCulloch v. Maryland, 4 Wheat. 316, 401 (1819) (emphasizing the utility of historical practice in interpreting constitutional provisions). Longstanding practice indicates that a federal law’s creation of an office in this context does not automatically make its holder an “Officer of the United States.” Rather, Congress has often used these two provisions to create local offices filled in ways other than those specified in the Appointments Clause. When the First Congress legislated for the Northwest Territories, for example, it created a House of Representatives for the Territory with members selected by election. It also created an upper house of the territorial legislature, whose members were appointed by the President (without Senate confirmation) from lists provided by the elected, lower house. And it created magistrates appointed by the Governor. See Act of Aug. 7, 1789, 1Stat. 51, n. (a). The practice of creating by federal law local offices for the Territories and District of Columbia that are filled through election or local executive appointment has continued unabated for more than two centuries. See, e.g., ibid. (Northwest Territories local offices filled by election); Act of Apr. 7, 1798, §3, 1Stat. 550 (Mississippi, same); Act of May 7, 1800, §2, 2Stat. 59 (Indiana, same); Act of May 15, 1820, §3, 3Stat. 584 (District of Columbia, same); Act of Apr. 30, 1900, §13, 31Stat. 144 (Hawaii, same); Act of Aug. 24, 1912, §4, 37Stat. 513 (Alaska, same); Act of Aug. 23, 1968, §4, 82Stat. 837 (Virgin Islands, same); Act of Sept. 11, 1968, Pub. L. 90–497, §1, 82Stat. 842 (Guam, same); Act of May 4, 1812, §3, 2Stat. 723 (D. C. mayor appoints “all offices”); Act of June 4, 1812, §2, 2Stat. 744 (Missouri Governor, similar); Act of Mar. 2, 1819, §3, 3Stat. 494 (Arkansas, similar); Act of June 6, 1900, §2, 31Stat. 322 (Alaska, similar); Act of Sept. 11, 1968, §1, 82Stat. 843 (Guam, similar). Like Justice Thomas, post, at 6 (opinion concurring in judgment), we think the practice of the First Congress is strong evidence of the original meaning of the Constitution. We find this subsequent history similarly illuminates the text’s meaning. Puerto Rico’s history is no different. It reveals a longstanding practice of selecting public officials with important local responsibilities in ways that the Appointments Clause does not describe. In 1898, at the end of the Spanish-American War, the United States took responsibility for determining the civil rights of Puerto Ricans as well as Puerto Rico’s political status. Treaty of Paris, Art. 9, Dec. 10, 1898, 30Stat. 1759. In 1900, the Foraker Act provided for Presidential appointment (with Senate confirmation) of Puerto Rico’s Governor, the heads of six departments, the legislature’s upper house, and the justices of its high court. Organic Act of 1900, §§ 17, 18, 33, 31Stat. 81, 84. But it also provided for the selection, through popular election, of a lower legislative house with the power (subject to upper house concurrence) to “alter, amend, modify, and repeal any and all laws . . . of every character.” §§27, 32, id., at 82, 84. There is no indication that anyone thought members of the lower house, wielding important local responsibilities, were “Officers of the United States.” Congress replaced the Foraker Act with the Jones Act in 1917. Organic Act of Puerto Rico, ch. 145, 39Stat. 951. Under the Jones Act the Puerto Rican Senate was elected and consequently no longer satisfied the Appointments Clause criteria. See §26, id., at 958. Similarly, the Governor of Puerto Rico nominated four cabinet members, confirmed by the Senate of Puerto Rico. §13, id., at 955–956. The elected legislature retained “all local legislative powers,” including the power to appropriate funds. §§ 25, 34, 37, id., at 958, 962, 964. Congress amended the Jones Act in 1947 to provide for an elected Governor of Puerto Rico, and granted that Governor the power to appoint all cabinet officials. See Act of Aug. 5, 1947, ch. 490, §§ 1, 3, 61Stat. 770, 771. The President retained the power to appoint (with Federal Senate confirmation) judges, an auditor, and the new office of Coordinator of Federal Agencies, who was to supervise federal functions in Puerto Rico and recommend to higher federal officials ways to improve the quality of federal services. §6, id., at 772. In 1950, Congress enacted Public Law 600, “in the nature of a compact” with Puerto Rico and subject to approval by the voters of Puerto Rico. Act of July 3, 1950, ch. 446, §§1, 2, 64Stat. 319. The Act adopted the Jones Act, as amended, as the Puerto Rican Federal Relations Act, and provided for the Jones Act’s substantial (but not complete) repeal upon the effective adoption of a contemplated Puerto Rican constitution. §§4, 5, id., at 319–320. Among the provisions of the Jones Act that Public Law 600 retained were several related to Puerto Rico’s public debt. Congress retained, for example, the triple-tax-exempt nature of Puerto Rican bonds. Jones Act, §3, 39Stat. 953. It also retained a (later repealed) cap on the amount of public debt Puerto Rico or its subdivisions could accumulate. Ibid. In a public referendum, the citizens of Puerto Rico approved Public Law 600—including the limits on debt in §3 of the Federal Relations Act—and then began the constitution-making process. Pub. L. 600, §§2, 3, 64 Stat. 319; see Act of July 3, 1952, 66Stat. 327; A. Fernós-Isern, Original Intent in the Constitution of Puerto Rico 13 (2d ed. 2002). Puerto Rico’s popularly ratified Constitution, which Congress accepted with a few fairly minor changes, does not involve the President or the Senate in the appointment process for local officials. That Constitution provides for the election of Puerto Rico’s Governor and legislators. Art. III, §1; Art. IV, §1. And it provides for gubernatorial appointment (and Puerto Rican Senate confirmation) of cabinet officers. Art. IV, §5. The upshot is that Puerto Rico’s history reflects long-standing use of various methods for selecting officials with primarily local responsibilities. This history is consistent with the history of other entities that fall within the scope of Article IV and with the history of the District of Columbia. See supra, at 10–11. And it comports with our precedents, which have long acknowledged that Congress may structure local governments under Article IV and Article I in ways that do not precisely mirror the constitutional blueprint for the National Government. See, e.g., Benner v. Porter, 9 How. 235, 242 (1850). Cf. Glidden Co. v. Zdanok, 370 U.S. 530, 546 (1962) (plurality opinion) (recognizing that local governments created by Congress could, like governments of the States, “dispense with protections deemed inherent in a separation of governmental powers”). Sometimes Congress has specified the use of methods that would satisfy the Appointments Clause, other times it has specified methods that would not satisfy the Appointments Clause, including elections and appointment by local officials. Officials with primarily local duties have often fallen into the latter categories. We know of no case endorsing an Appointments Clause based challenge to such selection methods. Indeed, to read Appointments Clause constraints as binding Puerto Rican officials with primarily local duties would work havoc with Puerto Rico’s (federally ratified) democratic methods for selecting many of its officials. We thus conclude that while the Appointments Clause does restrict the appointment of “Officers of the United States” with duties in or related to the District of Columbia or an Article IV entity, it does not restrict the appointment of local officers that Congress vests with primarily local duties under Article IV, §3, or Article I, §8, cl. 17. B The question remains whether the Board members have primarily local powers and duties. We note that the Clause qualifies the phrase “Officers of the United States” with the words “whose Appointments . . . shall be established by Law.” And we also note that PROMESA says that the Board is “an entity within the territorial government” and “shall not be considered a department, agency, establishment, or instrumentality of the Federal Government.” §101(c), 130Stat. 553. But the most these words show is that Congress did not intend to make the Board members “Officers of the United States.” It does not prove that, insofar as the Constitution is concerned, they succeeded. But we think they have. Congress did not simply state that the Board is part of the local Puerto Rican government. Rather, Congress also gave the Board a structure, a set of duties, and related powers all of which are consistent with this statement. The government of Puerto Rico pays the Board’s expenses, including the salaries of its employees (the members serve without pay). §107, id., at 562; see §101(g), id., at 556. The Board possesses investigatory powers. It can hold hearings. §104(a), id., at 558. It can issue subpoenas, subject to Puerto Rico’s limits on personal jurisdiction and enforceable under Puerto Rico’s laws. §104(f ), id., at 559. And it can enforce those subpoenas in (and only in) Puerto Rico’s courts. §§104(f )(2), 106(a), id., at 559, 562. From its own offices in or outside of Puerto Rico, the Board works with the elected government of Puerto Rico to develop a fiscal plan that provides “a method to achieve fiscal responsibility and access to the capital markets.” §201(b), id., at 564. If it finds it necessary, the Board can develop its own budget for Puerto Rico which is “deemed . . . approved” and becomes the operative budget. §202(e)(3), id., at 568. It can ensure compliance with the plan and budget by reviewing the Puerto Rico government’s laws and spending and by “direct[ing]” corrections or taking “such [other] actions as it considers necessary,” including preventing a law from taking effect. §§203(d), 204(a), id., at 569, 571. The Board controls the issuance of new debt for Puerto Rico. §207, id., at 575. The Board also may initiate bankruptcy proceedings for Puerto Rico or its instrumentalities. §304(a), id., at 579. It may take any related “action necessary on behalf of,” and it serves as “the representative of,” Puerto Rico or its instrumentalities. §315, id., at 584. These proceedings take place in the U. S. District Court for Puerto Rico. §307, id., at 582. To repeat: The Board has broad investigatory powers: It can administer oaths, issue subpoenas, take evidence and demand data from governments and creditors alike. But these powers are backed by Puerto Rican, not federal, law: Subpoenas are governed by Puerto Rico’s personal jurisdiction statute; false testimony is punishable under the law of Puerto Rico; the Board must seek enforcement of its subpoenas by filing in the courts of Puerto Rico. See §104, id., at 558–561. These powers are primarily local in nature. The Board also oversees the development of Puerto Rico’s fiscal and budgetary plans. It receives and evaluates proposals from the elected Governor and legislature. It can create a budget “deemed” to be that of Puerto Rico. It can intervene when budgetary constraints are violated. And it has authority over the issuance of new debt. §§201–207, id., at 563–575. These powers, too, are quintessentially local. Each concerns the finances of the Commonwealth, not of the United States. The Board members in this respect discharge duties ordinarily held by local officials. Last, the Board has the power to initiate bankruptcy proceedings. But in doing so, it acts not on behalf of the United States, but on behalf of, and in the interests of, Puerto Rico. The proceedings take place in federal court; but the same is true of all persons or entities who seek bankruptcy protection. The Board here acts as a local government that might take precisely the same actions. See, e.g., 11 U. S. C. §§109(c), 921 (related to bankruptcies of local governments). Some Board actions, of course, may have nationwide consequences. But the same can be said of many actions taken by many Governors or other local officials. Taking actions with nationwide consequences does not automatically transform a local official into an “Officer of the United States.” The challengers rely most heavily on the nationwide effects of the bankruptcy proceedings. E.g., Brief for Aurelius et al. 31; Brief for Petitioner Unión de Trabajadores de la Industria Eléctrica y Riego, Inc. (UTIER) 49. But the same might be said of any major municipal, or even corporate, bankruptcy. E.g., In re Detroit, 504 B.R. 97 (Bkrtcy. Ct. ED Mich. 2013) (restructuring $18 billion in municipal debt). In short, the Board possesses considerable power—including the authority to substitute its own judgment for the considered judgment of the Governor and other elected officials. But this power primarily concerns local matters. Congress’ law thus substitutes a different process for determining certain local policies (related to local fiscal responsibility) in respect to local matters. And that is the critical point for current purposes. The local nature of the legislation’s expressed purposes, the representation of local interests in bankruptcy proceedings, the focus of the Board’s powers upon local expenditures, the local logistical support, the reliance on local laws in aid of the Board’s procedural powers—all these features when taken together and judged in the light of Puerto Rico’s history (and that of the Territories and the District of Columbia)—make clear that the Board’s members have primarily local duties, such that their selection is not subject to the constraints of the Appointments Clause. IV The Court of Appeals, pointing to three of this Court’s cases, reached the opposite conclusion. See Buckley v. Valeo, 424 U.S. 1 (1976) (per curiam), Freytag v. Commissioner, 501 U.S. 868, and Lucia v. SEC, 585 U. S. ___ (2018). It pointed out that the Court, in those cases, discussed the term “Officer of the United States,” and it concluded that, for Appointments Clause purposes, an appointee is such an “officer” if “(1) the appointee occupies a ‘continuing’ position established by federal law; (2) the appointee ‘exercis[es] significant authority’; and (3) the significant authority is exercised ‘pursuant to the laws of the United States.’ ” 915 F. 3d, at 856. The Court of Appeals concluded that the Board members satisfied this test. See id., at 856–857. We do not believe these three cases set forth the critical legal test relevant here, however, and we do not apply any test they might enunciate. Each of the cases considered an Appointments Clause problem concerning the importance or significance of duties that were indisputably federal or national in nature. In Buckley, the question was whether members of the Federal Election Commission—appointees carrying out federal-election related duties—were “officers” for Appointments Clause purposes. In Freytag, the Court asked the same question about special federal trial judges serving on federal tax courts. And in Lucia the Court asked the same question about federal administrative law judges carrying out Securities and Exchange Commission duties. Here, PROMESA, a federal law, creates the Board and its duties, and no one doubts their significance. But we cannot stop there. To do so would ignore the history we have discussed—history stretching back to the founding. See supra, at 10–13. And failing to take account of the nature of an appointee’s federally created duties, i.e., whether they are primarily local versus primarily federal, would threaten interference with democratic (or local appointment) selection methods in numerous Article IV Territories and perhaps the District of Columbia as well. See, e.g., 48 U. S. C. §1422 (providing for an elected Governor of Guam); §1591 (same for Virgin Islands); District of Columbia Self-Government Act, §421, 87Stat. 789 (same for D. C. Mayor); §422(2), 87Stat. 790 (describing D. C. Mayor’s appointment powers); 48 U. S. C. §1422c (same for Guam’s Governor); §1597(c) (same for Virgin Islands). There is no reason to understand the Appointments Clause—which, at least in part, seeks to advance democratic accountability and broaden appointments-related responsibility, see supra, at 6–7—as making it significantly more difficult for local residents of such areas to share responsibility for the implementation of (statutorily created) primarily local duties. Neither the text nor the history of the Clause commands such a result. Neither do Lebron v. National Railroad Passenger Corporation, 513 U.S. 374 (1995), or MWAA, 501 U.S. 252, help those challenging the Board’s constitutional legitimacy. Lebron considered whether, for First Amendment purposes, Amtrak was a governmental or a private entity. 513 U. S., at 379. All here agree that the Board is a Government entity, but that fact does not answer the “primarily local versus primarily federal” question. In MWAA, the Court held that separation-of-powers principles forbid Members of Congress to become members of a board that controls federally owned airports. 501 U. S., at 275–276 (relying on Bowsher v. Synar, 478 U.S. 714, 726 (1986), and INS v. Chadha, 462 U.S. 919, 952 (1983)). The Court expressly declined to answer any question related to the Appointments Clause. 501 U. S., at 277, n. 23. While we have found no case from this Court directly on point, we believe that the Court’s analysis in O’Donoghue v. United States, 289 U.S. 516 (1933), and especially Palmore v. United States, 411 U.S. 389, provides a rough analogy. In O’Donoghue, the Court considered whether Article III’s tenure and salary protections applied to judges of the courts in the District of Columbia. The Court held that they did. Those courts, it believed, were “ ‘courts of the United States’ ” and “recipients of the judicial power of the United States.” 289 U. S., at 546, 548. The judges’ salaries consequently could not be reduced. Id., at 551. In Palmore, however, the Court reached what might seem the precisely opposite conclusion. A criminal defendant, invoking O’Donoghue, argued that the D. C. Superior Court Judge could not constitutionally preside over the case because the judge lacked Article III’s tenure protection, namely, life tenure. Palmore, supra, at 390. But the Court rejected the defendant’s argument. Why? How did it explain O’Donoghue? The difference, said the Court, lies in the fact that, in the meantime, Congress had changed the nature of the District of Columbia court. Palmore, supra, at 406–407; see District of Columbia Court Reform and Criminal Procedure Act of 1970, 84Stat. 473. Congress changed what had been a unified court system where judges adjudicated both local and federal issues into separate court systems, in one of which judges adjudicated primarily local issues. §111, id., at 475. Courts in that category had criminal jurisdiction over only those cases brought “ ‘under any law applicable exclusively to the District of Columbia.’ ” Id., at 486. Its judges served for 15-year terms. Id., at 491. This Court, in Palmore, considered a local judge presiding over a local court. Congress had created that court in the exercise of its Article I power to “exercise exclusive Legislation in all Cases whatsoever” over the District of Columbia. See Art I, §8, cl. 17. The “focus” of these courts was “primarily upon . . . matters of strictly local concern.” 411 U. S., at 407. Hence, the nature of those courts was a “far cry” from that of the courts at issue in O’Donoghue. Palmore, 411 U. S., at 406. The Court added that Congress had created non-Article III courts under its Article IV powers. It wrote that Congress could also create non-Article III courts under its Article I powers. Id., at 403, 410. And it held that judges serving on those non-Article III courts lacked Article III protections. Id., at 410. Palmore concerned Article I of the Constitution, not Article IV. And it concerned “the judicial Power of the United States,” not “Officers of the United States.” But it provides a rough analogy. It holds that Article III protections do not apply to an Article I court “focus[ed],” unlike the courts at issue in O’Donoghue, primarily on local matters. Here, Congress expressly invoked a constitutional provision allowing it to make local debt-related law (Article IV); it expressly located the Board within the local government of Puerto Rico; it clearly indicated that it intended the Board’s members to be local officials; and it gave them primarily local powers, duties, and responsibilities. In his concurring opinion, Justice Thomas criticizes the inquiry we set out—whether an officer’s duties are primarily local or primarily federal—as too “amorphous,” post, at 10. But we think this is the test established by the Constitution’s text, as illuminated by historical practice. And we cannot see how Congress could avoid the strictures of the Appointments Clause by adding to a federal officer’s other obligations a large number of local duties. Indeed, we think that our test, tied as it is to both the text and the history of the Appointments Clause, is more rigorous than the bare inquiry into the “nature” of the officer’s authority that Justice Thomas proposes, and we believe it is more faithful to the Clause’s original meaning. Ibid. V We conclude, for the reasons stated, that the Constitution’s Appointments Clause applies to the appointment of officers of the United States with powers and duties in and in relation to Puerto Rico, but that the congressionally mandated process for selecting members of the Financial Oversight and Management Board for Puerto Rico does not violate that Clause. Given this conclusion, we need not consider the request by some of the parties that we overrule the much-criticized “Insular Cases” and their progeny. See, e.g., Downes v. Bidwell, 182 U.S. 244, 287 (1901) (opinion of Brown, J.); Balzac v. Porto Rico, 258 U.S. 298, 309 (1922); Reid v. Covert, 354 U.S. 1, 14 (1957) (plurality opinion) (indicating that the Insular Cases should not be further extended); see also Brief for Official Committee of Unsecured Creditors of All Title III Debtors (Other than COFINA) 20–25 (arguing that the Insular Cases support reversal on the Appointments Clause issue); Brief for UTIER 64–66 (encouraging us to overrule the Insular Cases); Brief for Virgin Islands Bar Association as Amicus Curiae 13–18 (same); Cabranes, Citizenship and the American Empire, 127 U. Pa. L. Rev. 391, 436–442 (1978) (criticizing the Insular Cases); Littlefield, The Insular Cases, 15 Harv. L. Rev. 169 (1901) (same). Those cases did not reach this issue, and whatever their continued validity we will not extend them in these cases. See Reid, supra, at 14. Neither, since we hold the appointment method valid, need we consider the application of the de facto officer doctrine. See Ryder v. United States, 515 U.S. 177 (1995) (discussing the doctrine); see also, e.g., Brief for Aurelius et al. 48–69 (arguing the doctrine does not apply in this context); Brief for UTIER 69–85 (same); Reply Brief for United States 26–47 (insisting to the contrary); Brief for Cross-Respondent COFINA Senior Bondholders’ Coalition 14–46 (same). Finally, as Justice Sotomayor recognizes, post, at 8 (opinion concurring in judgment), we need not, and therefore do not, decide questions concerning the application of the Federal Relations Act and Public Law 600. No party has argued that those Acts bear any significant relation to the answer to the Appointments Clause question now before us. For these reasons, we reverse the judgment of the Court of Appeals and remand the cases for further proceedings consistent with this opinion. It is so ordered.

590.US.2019_18-1048

ThyssenKrupp Stainless USA, LLC, entered into three contracts with F. L. Industries, Inc., for the construction of cold rolling mills at ThyssenKrupp’s steel manufacturing plant in Alabama. Each contract contained a clause requiring arbitration of any contract dispute. F. L. Industries then entered into a subcontractor agreement with petitioner (GE Energy) for the provision of nine motors to power the cold rolling mills. After the motors for the cold rolling mills allegedly failed, Outokumpu Stainless USA, LLC (which acquired ownership of the plant), and its insurers sued GE Energy in Alabama state court. GE Energy removed the case to federal court under 9 U. S. C. §205. It then moved to dismiss and compel arbitration, relying on the arbitration clauses in the F. L. Industries and ThyssenKrupp contracts. The District Court granted the motion, concluding that both Outokumpu and GE Energy were parties to the agreement. The Eleventh Circuit reversed. It concluded that the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York Convention or Convention) allows enforcement of an arbitration agreement only by the parties that actually signed the agreement and that GE Energy was a nonsignatory. It also held that allowing GE Energy to rely on state-law equitable estoppel doctrines to enforce the arbitration agreement would conflict with the Convention’s signatory requirement. Held: The New York Convention does not conflict with domestic equitable estoppel doctrines that permit the enforcement of arbitration agreements by nonsignatories. Pp. 3–12. (a) Chapter 1 of the Federal Arbitration Act (FAA) does not “alter background principles of state contract law regarding the scope of agreements (including the question of who is bound by them).” Arthur Andersen LLP v. Carlisle, 556 U.S. 624, 630. The “ ‘traditional principles’ of state law” that apply under Chapter 1 include doctrines, like equitable estoppel, authorizing contract enforcement by a nonsignatory. Id., at 631–632. The New York Convention is a multilateral treaty addressing international arbitration. One Article of the Convention addresses arbitration agreements—Article II—and one provision of Article II addresses the enforcement of those agreements—Article II(3). Article II(3) provides that courts of a contracting state “shall . . . refer the parties to arbitration” when the parties to an action entered into a written agreement to arbitrate and one of the parties requests such a referral. Chapter 2 of the FAA grants federal courts jurisdiction over actions governed by the Convention. As relevant here, Chapter 2 provides that “Chapter 1 applies to actions and proceedings brought under this chapter to the extent that [Chapter 1] is not in conflict with this chapter or the Convention.” 9 U. S. C. §208. Pp. 3–6. (b) The application of familiar tools of treaty interpretation establishes that the state-law equitable estoppel doctrines permitted under Chapter 1 do not “conflict with . . . the Convention.” §208. Pp. 6–11. (1) The text of the New York Convention does not address whether nonsignatories may enforce arbitration agreements under domestic doctrines such as equitable estoppel. The Convention is simply silent on the issue of nonsignatory enforcement. This silence is dispositive because nothing in the Convention’s text could be read to conflict with the application of domestic equitable estoppel doctrines. Article II(3)—the only provision in the Convention addressing the enforcement of arbitration agreements—contains no exclusionary language; it does not state that arbitration agreements shall be enforced only in the identified circumstances. Given that the Convention was drafted against the backdrop of domestic law, it would be unnatural to read Article II(3) to displace domestic doctrines in the absence of such language. This interpretation is especially appropriate because Article II contemplates using domestic doctrines to fill gaps in the Convention. Pp. 6–7. (2) This interpretation is confirmed by the Convention’s negotiation and drafting history as well as “ ‘the postratification understanding’ of signatory nations,” Medellín v. Texas, 552 U.S. 491, 507. Cherry-picked generalizations from the negotiating and drafting history cannot be used to create a rule that finds no support in the treaty’s text. Here, to the extent that the Convention’s drafting history sheds any light on the treaty’s meaning, it shows only that the drafters sought to impose baseline requirements on contracting states so that signatories would “not be permitted to decline enforcement of such agreements on the basis of parochial views of their desirability or in a manner that would diminish the mutually binding nature of the agreements.” Scherk v. Alberto-Culver Co., 417 U.S. 506, 520, n. 15. The postratification understanding of other contracting states—as evidenced by the “[d]ecisions of the courts of other Convention signatories,” El Al Israel Airlines, Ltd. v. Tsui Yuan Tseng, 525 U.S. 155, 175, and the “postratification conduct” of contracting state governments, Zicherman v. Korean Air Lines Co., 516 U.S. 217, 227—may also serve as an aid to this Court’s interpretation. Here, numerous sources indicate that the New York Convention does not prohibit the application of domestic law addressing the enforcement of arbitration agreements. These sources, however, are from decades after the fi-nalization of the New York Convention’s text in 1958. This diminishes their value as evidence of the original understanding of the treaty’s meaning. Finally, because the Court’s textual analysis and the Executive’s interpretation of the Convention align here, there is no need to determine whether the Executive’s understanding is entitled to “weight” or “deference.” Cf. Edelman v. Lynchburg College, 535 U.S. 106, 114–115, n. 8. Pp. 7–11. (c) The Court of Appeals may address on remand whether GE Energy can enforce the arbitration clauses under equitable estoppel principles and which body of law governs that determination. Pp. 11–12. 902 F.3d 1316, reversed and remanded. Thomas, J., delivered the opinion for a unanimous Court. Sotomayor, J., filed a concurring opinion.

The question in this case is whether the Convention on the Recognition and Enforcement of Foreign Arbitral Awards, June 10, 1958, 21 U. S. T. 2517, T. I. A. S. No. 6997, conflicts with domestic equitable estoppel doctrines that permit the enforcement of arbitration agreements by nonsignatories. We hold that it does not. I In 2007, ThyssenKrupp Stainless USA, LLC, entered into three contracts with F. L. Industries, Inc., for the construction of cold rolling mills at ThyssenKrupp’s steel manufacturing plant in Alabama. Each of the contracts contained an identical arbitration clause. The clause provided that “[a]ll disputes arising between both parties in connection with or in the performances of the Contract . . . shall be submitted to arbitration for settlement.” App. 171. After executing these agreements, F. L. Industries, Inc., entered into a subcontractor agreement with petitioner GE Energy Power Conversion France SAS, Corp. (GE Energy), then known as Converteam SAS. Under that agreement, GE Energy agreed to design, manufacture, and supply motors for the cold rolling mills. Between 2011 and 2012, GE Energy delivered nine motors to the Alabama plant for installation. Soon thereafter, respondent Outokumpu Stainless USA, LLC, acquired ownership of the plant from ThyssenKrupp. According to Outokumpu, GE Energy’s motors failed by the summer of 2015, resulting in substantial damages. In 2016, Outokumpu and its insurers filed suit against GE Energy in Alabama state court. GE Energy removed the case to federal court under 9 U. S. C. §205, which authorizes the removal of an action from state to federal court if the action “relates to an arbitration agreement . . . falling under the Convention [on the Recognition and Enforcement of Foreign Arbitral Awards].” GE Energy then moved to dismiss and compel arbitration, relying on the arbitration clauses in the contracts between F. L. Industries, Inc., and ThyssenKrupp. The District Court granted GE Energy’s motion to dismiss and compel arbitration with Outokumpu and Sompo Japan Insurance Company of America. Outokumpu Stainless USA LLC v. Converteam SAS, 2017 WL 401951 (SD Ala., Jan. 30, 2017).[1] The court held that GE Energy qualified as a party under the arbitration clauses because the contracts defined the terms “Seller” and “Parties” to include subcontractors. Id., at *4. Because the court concluded that both Outokumpu and GE Energy were parties to the agreements, it declined to address GE Energy’s argument that the agreement was enforceable under equitable estoppel. Id., at *1, n. 1. The Eleventh Circuit reversed the District Court’s order compelling arbitration. Outokumpu Stainless USA, LLC v. Converteam SAS, 902 F.3d 1316 (2018). The court interpreted the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York Convention or Convention) to include a “requirement that the parties actually sign an agreement to arbitrate their disputes in order to compel arbitration.” Id., at 1326 (emphasis in original). The court concluded that this requirement was not satisfied because “GE Energy is undeniably not a signatory to the Contracts.” Ibid. It then held that GE Energy could not rely on state-law equitable estoppel doctrines to enforce the arbitration agreement as a nonsignatory because, in the court’s view, equitable estoppel conflicts with the Convention’s signatory requirement. Id., at 1326–1327. Given a conflict between the Courts of Appeals on this question,[2] we granted certiorari. 588 U. S. ___ (2019). II A Chapter 1 of the Federal Arbitration Act (FAA) permits courts to apply state-law doctrines related to the enforcement of arbitration agreements. Section 2 of that chapter provides that an arbitration agreement in writing “shall be . . . enforceable, save upon such grounds as exist at law or in equity for the revocation of any contract.” 9 U. S. C. §2. As we have explained, this provision requires federal courts to “place [arbitration] agreements ‘ “upon the same footing as other contracts.” ’ ” Volt Information Sciences, Inc. v. Board of Trustees of Leland Stanford Junior Univ., 489 U.S. 468, 474 (1989) (quoting Scherk v. Alberto-Culver Co., 417 U.S. 506, 511 (1974)). But it does not “alter background principles of state contract law regarding the scope of agreements (including the question of who is bound by them).” Arthur Andersen LLP v. Carlisle, 556 U.S. 624, 630 (2009). The “traditional principles of state law” that apply under Chapter 1 include doctrines that authorize the enforcement of a contract by a nonsignatory. Id., at 631 (internal quotation marks omitted). For example, we have recognized that arbitration agreements may be enforced by nonsignatories through “ ‘assumption, piercing the corporate veil, alter ego, incorporation by reference, third-party beneficiary theories, waiver and estoppel.’ ” Ibid. (quoting 21 R. Lord, Williston on Contracts §57:19, p. 183 (4th ed. 2001)). This case implicates domestic equitable estoppel doctrines. Generally, in the arbitration context, “equitable estoppel allows a nonsignatory to a written agreement containing an arbitration clause to compel arbitration where a signatory to the written agreement must rely on the terms of that agreement in asserting its claims against the nonsignatory.” Id., at 200 (2017). In Arthur Andersen, we recognized that Chapter 1 of the FAA permits a nonsignatory to rely on state-law equitable estoppel doctrines to enforce an arbitration agreement. 556 U. S., at 631–632. B The New York Convention is a multilateral treaty that addresses international arbitration. 21 U. S. T. 2517, T. I. A. S. No. 6997. It focuses almost entirely on arbitral awards. Article I(1) describes the Convention as applying only to “the recognition and enforcement of arbitral awards.” Id., at 2519. Articles III, IV, and V contain recognition and enforcement obligations related to arbitral awards for contracting states and for parties seeking the enforcement of arbitral awards. Id., at 2519–2520. Article VI addresses when an award can be set aside or suspended. Id., at 2520. And Article VII(1) states that the “Convention shall not . . . deprive any interested party of any right he may have to avail himself of an arbitral award in the manner and to the extent allowed by the law or the treaties of the country where such award is sought to be relied upon.” Id., at 2520–2521. Only one article of the Convention addresses arbitration agreements—Article II. That article contains only three provisions, each one sentence long. Article II(1) requires “[e]ach Contracting State [to] recognize an agreement in writing under which the parties undertake to submit to arbitration all or any differences which have arisen or which may arise between them in respect of a defined legal relationship, whether contractual or not, concerning a subject matter capable of settlement by arbitration.” Id., at 2519. Article II(2) provides that “[t]he term ‘agreement in writing’ shall include an arbitral clause in a contract or an arbitration agreement, signed by the parties or contained in an exchange of letters or telegrams.” Ibid. Finally, Article II(3) states that “[t]he court of a Contracting State, when seized of an action in a matter in respect of which the parties have made an agreement within the meaning of this article, shall, at the request of one of the parties, refer the parties to arbitration, unless it finds that the said agreement is null and void, inoperative or incapable of being performed.” Ibid. C In 1970, the United States acceded to the New York Convention, and Congress enacted implementing legislation in Chapter 2 of the FAA. See 84Stat. 692, 9 U. S. C. §§201–208. Chapter 2 grants federal courts jurisdiction over actions governed by the Convention, §203; establishes venue for such actions, §204; authorizes removal from state court, §205; and empowers courts to compel arbitration, §206. Chapter 2 also states that “Chapter 1 applies to actions and proceedings brought under this chapter to the extent that [Chapter 1] is not in conflict with this chapter or the Convention.” §208. III We must determine whether the equitable estoppel doctrines permitted under Chapter 1 of the FAA, see supra, at 3–4, “conflict with . . . the Convention.” §208. Applying familiar tools of treaty interpretation, we conclude that they do not conflict. A “The interpretation of a treaty, like the interpretation of a statute, begins with its text.” Medellín v. Texas, 552 U.S. 491, 506 (2008). The text of the New York Convention does not address whether nonsignatories may enforce arbitration agreements under domestic doctrines such as equitable estoppel. The Convention is simply silent on the issue of nonsignatory enforcement, and in general, “a matter not covered is to be treated as not covered”—a principle “so obvious that it seems absurd to recite it,” A. Scalia & B. Garner, Reading Law: The Interpretation of Legal Texts 93 (2012). This silence is dispositive here because nothing in the text of the Convention could be read to otherwise prohibit the application of domestic equitable estoppel doctrines. Only one Article of the Convention addresses arbitration agreements—Article II—and only one provision of Article II addresses the enforcement of those agreements—Article II(3). The text of Article II(3) states that courts of a contracting state “shall . . . refer the parties to arbitration” when the parties to an action entered into a written agreement to arbitrate and one of the parties requests referral to arbitration. The provision, however, does not restrict contracting states from applying domestic law to refer parties to arbitration in other circumstances. That is, Article II(3) provides that arbitration agreements must be enforced in certain circumstances, but it does not prevent the application of domestic laws that are more generous in enforcing arbitration agreements. Article II(3) contains no exclusionary language; it does not state that arbitration agreements shall be enforced only in the identified circumstances. Given that the Convention was drafted against the backdrop of domestic law, it would be unnatural to read Article II(3) to displace domestic doctrines in the absence of exclusionary language. Cf. Marx v. General Revenue Corp., 568 U.S. 371, 380–384 (2013). This interpretation is especially appropriate in the context of Article II. Far from displacing domestic law, the provisions of Article II contemplate the use of domestic doctrines to fill gaps in the Convention. For example, Article II(1) refers to disputes “capable of settlement by arbitration,” but it does not identify what disputes are arbitrable, leaving that matter to domestic law. Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 639, n. 21 (1985). Similarly, Article II(3) states that it does not apply to agreements that are “null and void, inoperative or incapable of being performed,” but it fails to define those terms. Again, the Convention requires courts to rely on domestic law to fill the gaps; it does not set out a comprehensive regime that displaces domestic law. In sum, the only provision of the Convention that addresses the enforcement of arbitration agreements is Article II(3). We do not read the nonexclusive language of that provision to set a ceiling that tacitly precludes the use of domestic law to enforce arbitration agreements. Thus, nothing in the text of the Convention “conflict[s] with” the application of domestic equitable estoppel doctrines permitted under Chapter 1 of the FAA. 9 U. S. C. §208. B “Because a treaty ratified by the United States is ‘an agreement among sovereign powers,’ we have also considered as ‘aids to its interpretation’ the negotiation and drafting history of the treaty as well as ‘the postratification understanding’ of signatory nations.” Medellín, 552 U. S., at 507 (quoting Zicherman v. Korean Air Lines Co., 516 U.S. 217, 226 (1996)). These aids confirm our interpretation of the Convention’s text. 1 Our precedents have looked to the “negotiating and drafting history” of a treaty as an aid in determining the shared understanding of the treaty. Id., at 226. Invoking this interpretive aid, Outokumpu argues that the Convention’s drafting history establishes a “rule of consent” that “displace[s] varying local laws.” Brief for Respondents 27. We are unpersuaded. For one, nothing in the text of the Convention imposes a “rule of consent” that displaces domestic law—let alone a rule that allows some domestic-law doctrines and not others, as Outokumpu proposes. The only time the Convention uses the word “consent” is in Article X(3), which addresses ratification and accession procedures. Moreover, the statements relied on by Outokumpu do not address the specific question whether the Convention prohibits the application of domestic law that would allow nonsignatories to compel arbitration. Cherry-picked “generalization[s]” from the negotiating and drafting history cannot be used to create a rule that finds no support in the treaty’s text. Zicherman, 516 U. S., at 227. To the extent the drafting history sheds any light on the meaning of the Convention, it shows only that the drafters sought to impose baseline requirements on contracting states. As this Court has recognized, “[i]n their discussion of [Article II], the delegates to the Convention voiced frequent concern that courts of signatory countries . . . should not be permitted to decline enforcement of such agreements on the basis of parochial views of their desirability or in a manner that would diminish the mutually binding nature of the agreements.” Scherk, 417 U. S., at 520, n. 15 (citing G. Haight, Convention on the Recognition and Enforcement of Foreign Arbitral Awards: Summary Analysis of Record of United Nations Conference, May/June 1958, pp. 24–28 (1958)). Nothing in the drafting history suggests that the Convention sought to prevent contracting states from applying domestic law that permits nonsignatories to enforce arbitration agreements in additional circumstances. 2 “[T]he postratification understanding” of other contracting states may also serve as an aid to our interpretation of a treaty’s meaning. Medellín, 552 U. S., at 507 (internal quotation marks omitted). To discern this understanding, we have looked to the “[d]ecisions of the courts of other Convention signatories,” El Al Israel Airlines, Ltd. v. Tsui Yuan Tseng, 525 U.S. 155, 175 (1999), as well as the “postratification conduct” of the governments of contracting states, Zicherman, 516 U. S., at 227. Here, the weight of authority from contracting states indicates that the New York Convention does not prohibit the application of domestic law addressing the enforcement of arbitration agreements. The courts of numerous contracting states permit enforcement of arbitration agreements by entities who did not sign an agreement. See 1 G. Born, International Commercial Arbitration §10.02, pp. 1418–1484 (2d ed. 2014) (compiling cases). The United States identifies at least one contracting state with domestic legislation illustrating a similar understanding. See Brief for United States as Amicus Curiae 28 (discussing Peru’s national legislation). And GE Energy points to a recommendation issued by the United Nations Commission on International Trade Law that, although not directly addressing Article II(3), adopts a nonexclusive interpretation of Article II(1) and (2). Report of the United Nations Commission on International Trade Law on the Work of Its Thirty-Ninth Session, Recommendation Regarding the Interpretation of Article II, Paragraph 2, and Article VII, Paragraph 1, of the Convention on the Recognition and Enforcement of Foreign Arbitral Awards ¶¶1, 2, U. N. Doc. A/61/17, annex II (July 7, 2006) (UN recommendation). These sources, while generally pointing in one direction, are not without their faults. The court decisions, domestic legislation, and UN recommendation relied on by the parties occurred decades after the finalization of the New York Convention’s text in 1958. This diminishes the value of these sources as evidence of the original shared understanding of the treaty’s meaning. Moreover, unlike the actions and decisions of signatory nations, we have not previously relied on UN recommendations to discern the meaning of treaties. See also Yang v. Majestic Blue Fisheries, LLC, 876 F.3d 996, 1000–1001 (CA9 2017) (declining to give weight to the 2006 UN recommendation). But to the extent this evidence is given any weight, it confirms our interpretation of the Convention’s text. 3 Finally, the parties dispute whether the Executive’s interpretation of the New York Convention should affect our analysis. The United States claims that we should apply a “ ‘canon of deference’ ” and give “ ‘ “great weight” ’ ” to an interpretation set forth by the Executive in an amicus brief submitted to the D. C. Circuit in 2014. Brief for United States as Amicus Curiae 30 (quoting Abbott v. Abbott, 560 U.S. 1, 15 (2010)); see also Brief for United States as Amicus Curiae in No. 13–7004 (CADC), pp. 7, 9. GE Energy echoes this request. Outokumpu, on the other hand, argues that the Executive’s noncontemporaneous interpretation sheds no light on the meaning of the treaty, asserting that the Executive expressed the “opposite . . . view at the time of the Convention’s adoption.” Brief for Respondents 33. Outokumpu asserts that this Court has repeatedly rejected executive interpretations that contradict the treaty’s text or the political branches’ previous understanding of a treaty. Id., at 34–35 (citing, e.g., Chan v. Korean Air Lines, Ltd., 490 U.S. 122, 136 (1989) (Brennan, J., concurring in judgment); Perkins v. Elg, 307 U.S. 325, 328, 337–349 (1939)). We have never provided a full explanation of the basis for our practice of giving weight to the Executive’s interpretation of a treaty. Nor have we delineated the limitations of this practice, if any. But we need not resolve these issues today. Our textual analysis aligns with the Executive’s interpretation so there is no need to determine whether the Executive’s understanding is entitled to “weight” or “deference.” Cf. Edelman v. Lynchburg College, 535 U.S. 106, 114–115, n. 8 (2002) (“[T]here is no need to resolve deference issues when there is no need for deference”). IV The Court of Appeals did not analyze whether Article II(3) of the New York Convention conflicts with equitable estoppel. Instead, the court held that Article II(1) and (2) include a “requirement that the parties actually sign an agreement to arbitrate their disputes in order to compel arbitration.” 902 F. 3d, at 1326. But those provisions address the recognition of arbitration agreements, not who is bound by a recognized agreement. Article II(1) simply requires contracting states to “recognize an agreement in writing,” and Article II(2) defines the term “agreement in writing.” Here, the three agreements at issue were both written and signed.[3] Only Article II(3) speaks to who may request referral under those agreements, and it does not prohibit the application of domestic law. See supra, at 6–7. Because the Court of Appeals concluded that the Convention prohibits enforcement by nonsignatories, the court did not determine whether GE Energy could enforce the arbitration clauses under principles of equitable estoppel or which body of law governs that determination. Those questions can be addressed on remand. We hold only that the New York Convention does not conflict with the enforcement of arbitration agreements by nonsignatories under domestic-law equitable estoppel doctrines. * * * For the foregoing reasons, we reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 The District Court later granted GE Energy’s motion to compel arbitration with additional insurers. Outokumpu Stainless USA LLC v. Converteam SAS, 2017 WL 480716 (SD Ala., Feb. 3, 2017). 2 Compare 902 F.3d 1316, 1326 (CA11 2018), and Yang v. Majestic Blue Fisheries, LLC, 876 F.3d 996, 1001–1002 (CA9 2017), with Aggarao v. MOL Ship Mgmt. Co., 675 F.3d 355, 375 (CA4 2012), and Sourcing Unlimited, Inc. v. Asimco Int’l, Inc., 526 F.3d 38, 48 (CA1 2008). 3 We do not address whether Article II(2) requires a signed agreement.

590.US.2019_18-1150

The Copyright Act grants monopoly protection for “original works of authorship.” 17 U. S. C. §102(a). Under the government edicts doctrine, officials empowered to speak with the force of law cannot be the authors of the works they create in the course of their official duties. The State of Georgia has one official code—the Official Code of Georgia Annotated (OCGA). That Code includes the text of every Georgia statute currently in force, as well as a set of non-binding annotations that appear beneath each statutory provision. The annotations typically include summaries of judicial opinions construing each provision, summaries of pertinent opinions of the state attorney general, and a list of related law review articles and other reference materials. The OCGA is assembled by the Code Revision Commission, a state entity composed mostly of legislators, funded through legislative branch appropriations, and staffed by the Office of Legislative Counsel. The annotations in the current OCGA were produced by Matthew Bender & Co., Inc., a division of the LexisNexis Group, pursuant to a work-for-hire agreement with the Commission. Under the agreement, Lexis drafts the annotations under the supervision of the Commission, which specifies what the annotations must include in exacting detail. The agreement also states that any copyright in the OCGA vests in the State of Georgia, acting through the Commission. Respondent Public.Resource.Org (PRO), a nonprofit dedicated to facilitating public access to government records and legal materials, posted the OCGA online and distributed copies to various organizations and Georgia officials. After sending PRO several cease-and-desist letters, the Commission sued PRO for infringing its copyright in the OCGA annotations. PRO counterclaimed, seeking a declaratory judgment that the entire OCGA, including the annotations, fell in the public domain. The District Court sided with the Commission, holding that the annotations were eligible for copyright protection because they had not been enacted into law. The Eleventh Circuit reversed, rejecting the Commission’s copyright assertion under the government edicts doctrine. Held: The OCGA annotations are ineligible for copyright protection. Pp. 5–18. (a) The government edicts doctrine developed from a trio of 19th-century cases. In Wheaton v. Peters, 8 Pet. 591, the Court held that no reporter can have a copyright in the Court’s opinions and that the Justices cannot confer such a right on any reporter. In Banks v. Manchester, 128 U.S. 244, the Court held that judges could not assert copyright in “whatever work they perform in their capacity as judges”—be it “the opinion or decision, the statement of the case and the syllabus or the head note.” Id., at 253. Finally, in Callaghan v. Myers, 128 U.S. 617, the Court reiterated that an official reporter cannot hold a copyright interest in opinions created by judges. But, confronting an issue not addressed in Wheaton or Banks, the Court upheld the reporter’s copyright interest in several explanatory materials that the reporter had created himself because they came from an author who had no authority to speak with the force of law. The animating principle behind the government edicts doctrine is that no one can own the law. The doctrine gives effect to that principle in the copyright context through construction of the statutory term “author.” For purposes of the Copyright Act, judges cannot be the “author[s]” of “whatever work they perform in their capacity” as lawmakers. Banks, 128 U. S., at 253. Because legislators, like judges, have the authority to make law, it follows that they, too, cannot be “authors.” And, as with judges, the doctrine applies to whatever work legislators perform in their capacity as legislators, including explanatory and procedural materials they create in the discharge of their legislative duties. Pp. 5–9. (b) Applying that framework, Georgia’s annotations are not copyrightable. First, the author of the annotations qualifies as a legislator. Under the Copyright Act, the sole “author” of the annotations is the Commission, 17 U. S. C. §201(b), which functions as an arm of the Georgia Legislature in producing the annotations. Second, the Commission creates the annotations in the discharge of its legislative duties. Pp. 9–11. (c) Georgia argues that excluding the OCGA annotations from copyright protection conflicts with the text of the Copyright Act. First, it notes that §101 lists “annotations” among the kinds of works eligible for copyright protection. That provision, however, refers only to “annotations . . . which . . . represent an original work of authorship.” (Emphasis added.) Georgia’s annotations do not fit that description because they are prepared by a legislative body that cannot be deemed the “author” of the works it creates in its official capacity. Second, Georgia draws a negative inference from the fact that the Act excludes from copyright protection works prepared by Federal Government officials, without establishing a similar rule for State officials. §§101, 105. That rule, however, applies to all federal officials, regardless of the nature and scope of their duties. It does not suggest an intent to displace the much narrower government edits doctrine with respect to the States. Moving on from the text, Georgia invokes what it views as the official position of the Copyright Office, as reflected in the Compendium of U. S. Copyright Office Practices. The Compendium, however, is a non-binding administrative manual and is largely consistent with this Court’s position. Georgia also appeals to copyright policy, but such requests should be addressed to Congress, not the courts. Georgia attempts to frame the government edicts doctrine to focus exclusively on whether a particular work has the force of law. But that understanding cannot be squared with precedent—especially Banks. Moreover, Georgia’s conception of the doctrine as distinguishing between different categories of content with different effects has less of a textual footing than the traditional formulation, which focuses on the identity of the author. Georgia’s characterization of the OCGA annotations as non-binding and non-authoritative undersells the practical significance of the annotations to litigants and citizens. And its approach would logically permit States to hide all non-binding judicial and legislative work product—including dissents and legislative history—behind a paywall. Pp. 11–18. 906 F.3d 1229, affirmed. Roberts, C. J., delivered the opinion of the Court, in which Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a dissenting opinion, in which Alito, J., joined, and in which Breyer, J., joined as to all but Part II–A and footnote 6. Ginsburg, J., filed a dissenting opinion, in which Breyer, J., joined.

The Copyright Act grants potent, decades-long monopoly protection for “original works of authorship.” 17 U. S. C. §102(a). The question in this case is whether that protection extends to the annotations contained in Georgia’s official annotated code. We hold that it does not. Over a century ago, we recognized a limitation on copyright protection for certain government work product, rooted in the Copyright Act’s “authorship” requirement. Under what has been dubbed the government edicts doctrine, officials empowered to speak with the force of law cannot be the authors of—and therefore cannot copyright—the works they create in the course of their official duties. We have previously applied that doctrine to hold that non-binding, explanatory legal materials are not copyrightable when created by judges who possess the authority to make and interpret the law. See Banks v. Manchester, 128 U.S. 244 (1888). We now recognize that the same logic applies to non-binding, explanatory legal materials created by a legislative body vested with the authority to make law. Because Georgia’s annotations are authored by an arm of the legislature in the course of its legislative duties, the government edicts doctrine puts them outside the reach of copyright protection. I A The State of Georgia has one official code—the “Official Code of Georgia Annotated,” or OCGA. The first page of each volume of the OCGA boasts the State’s official seal and announces to readers that it is “Published Under Authority of the State.” The OCGA includes the text of every Georgia statute currently in force, as well as various non-binding supplementary materials. At issue in this case is a set of annotations that appear beneath each statutory provision. The annotations generally include summaries of judicial decisions applying a given provision, summaries of any pertinent opinions of the state attorney general, and a list of related law review articles and similar reference materials. In addition, the annotations often include editor’s notes that provide information about the origins of the statutory text, such as whether it derives from a particular judicial decision or resembles an older provision that has been construed by Georgia courts. See, e.g., OCGA §§51–1–1, 53–4–2 (2019). The OCGA is assembled by a state entity called the Code Revision Commission. In 1977, the Georgia Legislature established the Commission to recodify Georgia law for the first time in decades. The Commission was (and remains) tasked with consolidating disparate bills into a single Code for reenactment by the legislature and contracting with a third party to produce the annotations. A majority of the Commission’s 15 members must be members of the Georgia Senate or House of Representatives. The Commission receives funding through appropriations “provided for the legislative branch of state government.” OCGA §28–9–2(c) (2018). And it is staffed by the Office of Legislative Counsel, which is obligated by statute to provide services “for the legislative branch of government.” §§28–4–3(c)(4), 28–9–4. Under the Georgia Constitution, the Commission’s role in compiling the statutory text and accompanying annotations falls “within the sphere of legislative authority.” Harrison Co. v. Code Revision Comm’n, 244 Ga. 325, 330, 260 S.E.2d 30, 34 (1979). Each year, the Commission submits its proposed statutory text and accompanying annotations to the legislature for approval. The legislature then votes to do three things: (1) “enact[ ]” the “statutory portion of the codification of Georgia laws”; (2) “merge[ ]” the statutory portion “with [the] annotations”; and (3) “publish[ ]” the final merged product “by authority of the state” as “the ‘Official Code of Georgia Annotated.’ ” OCGA §1–1–1 (2019); see Code Revision Comm’n v. Public.Resource.Org, Inc., 906 F.3d 1229, 1245, 1255 (CA11 2018); Tr. of Oral Arg. 8. The annotations in the current OCGA were prepared in the first instance by Matthew Bender & Co., Inc., a division of the LexisNexis Group, pursuant to a work-for-hire agreement with the Commission. The agreement between Lexis and the Commission states that any copyright in the OCGA vests exclusively in “the State of Georgia, acting through the Commission.” App. 567. Lexis and its army of researchers perform the lion’s share of the work in drafting the annotations, but the Commission supervises that work and specifies what the annotations must include in exacting detail. See 906 F. 3d, at 1243–1244; App. 269–278, 286–427 (Commission specifications). Under the agreement, Lexis enjoys the exclusive right to publish, distribute, and sell the OCGA. In exchange, Lexis has agreed to limit the price it may charge for the OCGA and to make an unannotated version of the statutory text available to the public online for free. A hard copy of the complete OCGA currently retails for $412.00. B Public.Resource.Org (PRO) is a nonprofit organization that aims to facilitate public access to government records and legal materials. Without permission, PRO posted a digital version of the OCGA on various websites, where it could be downloaded by the public without charge. PRO also distributed copies of the OCGA to various organizations and Georgia officials. In response, the Commission sent PRO several cease-and-desist letters asserting that PRO’s actions constituted unlawful copyright infringement. When PRO refused to halt its distribution activities, the Commission sued PRO on behalf of the Georgia Legislature and the State of Georgia for copyright infringement. The Commission limited its assertion of copyright to the annotations described above; it did not claim copyright in the statutory text or numbering. PRO counterclaimed, seeking a declaratory judgment that the entire OCGA, including the annotations, fell in the public domain. The District Court sided with the Commission. The Court acknowledged that the annotations in the OCGA presented “an unusual case because most official codes are not annotated and most annotated codes are not official.” Code Revision Comm’n v. Public.Resource.Org, Inc., 244 F. Supp. 3d 1350, 1356 (ND Ga. 2017). But, ultimately, the Court concluded that the annotations were eligible for copyright protection because they were “not enacted into law” and lacked “the force of law.” Ibid. In light of that conclusion, the Court granted partial summary judgment to the Commission and entered a permanent injunction requiring PRO to cease its distribution activities and to remove the digital copies of the OCGA from the internet. The Eleventh Circuit reversed. 906 F.3d 1229. The Court began by reviewing the three 19th-century cases in which we articulated the government edicts doctrine. See Wheaton v. Peters, 8 Pet. 591 (1834); Banks v. Manchester, 128 U.S. 244 (1888); Callaghan v. Myers, 128 U.S. 617 (1888). The Court understood those cases to establish a “rule” based on an interpretation of the statutory term “author” that “works created by courts in the performance of their official duties did not belong to the judges” but instead fell “in the public domain.” 906 F. 3d, at 1239. In the Court’s view, that rule “derive[s] from first principles about the nature of law in our democracy.” Ibid. In a democracy, the Court reasoned, “the People” are “the constructive authors” of the law, and judges and legislators are merely “draftsmen . . . exercising delegated authority.” Ibid. The Court therefore deemed the “ultimate inquiry” to be whether a work is “attributable to the constructive authorship of the People.” Id., at 1242. The Court identified three factors to guide that inquiry: “the identity of the public official who created the work; the nature of the work; and the process by which the work was produced.” Id., at 1254. The Court found that each of those factors cut in favor of treating the OCGA annotations as government edicts authored by the People. It therefore rejected the Commission’s assertion of copyright, vacated the injunction against PRO, and directed that judgment be entered for PRO. We granted certiorari. 588 U. S. ___ (2019). II We hold that the annotations in Georgia’s Official Code are ineligible for copyright protection, though for reasons distinct from those relied on by the Court of Appeals. A careful examination of our government edicts precedents reveals a straightforward rule based on the identity of the author. Under the government edicts doctrine, judges—and, we now confirm, legislators—may not be considered the “authors” of the works they produce in the course of their official duties as judges and legislators. That rule applies regardless of whether a given material carries the force of law. And it applies to the annotations here because they are authored by an arm of the legislature in the course of its official duties. A We begin with precedent. The government edicts doctrine traces back to a trio of cases decided in the 19th century. In this Court’s first copyright case, Wheaton v. Peters, 8 Pet. 591 (1834), the Court’s third Reporter of Decisions, Wheaton, sued the fourth, Peters, unsuccessfully asserting a copyright interest in the Justices’ opinions. Id., at 617 (argument). In Wheaton’s view, the opinions “must have belonged to some one” because “they were new, original,” and much more “elaborate” than law or custom required. Id., at 615. Wheaton argued that the Justices were the authors and had assigned their ownership interests to him through a tacit “gift.” Id., at 614. The Court unanimously rejected that argument, concluding that “no reporter has or can have any copyright in the written opinions delivered by this court” and that “the judges thereof cannot confer on any reporter any such right.” Id., at 668 (opinion). That conclusion apparently seemed too obvious to adorn with further explanation, but the Court provided one a half century later in Banks v. Manchester, 128 U.S. 244 (1888). That case concerned whether Wheaton’s state-court counterpart, the official reporter of the Ohio Supreme Court, held a copyright in the judges’ opinions and several non-binding explanatory materials prepared by the judges. Id., at 249–251. The Court concluded that he did not, explaining that “the judge who, in his judicial capacity, prepares the opinion or decision, the statement of the case and the syllabus or head note” cannot “be regarded as their author or their proprietor, in the sense of [the Copyright Act].” Id., at 253. Pursuant to “a judicial consensus” dating back to Wheaton, judges could not assert copyright in “whatever work they perform in their capacity as judges.” Banks, 128 U. S, at 253 (emphasis in original). Rather, “[t]he whole work done by the judges constitutes the authentic exposition and interpretation of the law, which, binding every citizen, is free for publication to all.” Ibid. (citing Nash v. Lathrop, 142 Mass. 29, 6 N.E. 559 (1886)). In a companion case decided later that Term, Callaghan v. Myers, 128 U.S. 617 (1888), the Court identified an important limiting principle. As in Wheaton and Banks, the Court rejected the claim that an official reporter held a copyright interest in the judges’ opinions. But, resolving an issue not addressed in Wheaton and Banks, the Court upheld the reporter’s copyright interest in several explanatory materials that the reporter had created himself: headnotes, syllabi, tables of contents, and the like. Callaghan, 128 U. S., at 645, 647. Although these works mirrored the judge-made materials rejected in Banks, they came from an author who had no authority to speak with the force of law. Because the reporter was not a judge, he was free to “obtain[ ] a copyright” for the materials that were “the result of his [own] intellectual labor.” 128 U. S., at 647. These cases establish a straightforward rule: Because judges are vested with the authority to make and interpret the law, they cannot be the “author” of the works they prepare “in the discharge of their judicial duties.” Banks, 128 U. S., at 253. This rule applies both to binding works (such as opinions) and to non-binding works (such as headnotes and syllabi). Ibid. It does not apply, however, to works created by government officials (or private parties) who lack the authority to make or interpret the law, such as court reporters. Compare ibid. with Callaghan, 128 U. S., at 647. The animating principle behind this rule is that no one can own the law. “Every citizen is presumed to know the law,” and “it needs no argument to show . . . that all should have free access” to its contents. Nash, 142 Mass., at 35, 6 N. E., at 560 (cited by Banks, 128 U. S., at 253–254). Our cases give effect to that principle in the copyright context through construction of the statutory term “author.” Id., at 253.[1] Rather than attempting to catalog the materials that constitute “the law,” the doctrine bars the officials responsible for creating the law from being considered the “author[s]” of “whatever work they perform in their capacity” as lawmakers. Ibid. (emphasis added). Because these officials are generally empowered to make and interpret law, their “whole work” is deemed part of the “authentic exposition and interpretation of the law” and must be “free for publication to all.” Ibid. If judges, acting as judges, cannot be “authors” because of their authority to make and interpret the law, it follows that legislators, acting as legislators, cannot be either. Courts have thus long understood the government edicts doctrine to apply to legislative materials. See, e.g., Nash, 142 Mass., at 35, 6 N. E., at 560 (judicial opinions and statutes stand “on substantially the same footing” for purposes of the government edicts doctrine); Howell v. Miller, 91 F. 129, 130–131, 137–138 (CA6 1898) (Harlan, J., Circuit Justice, joined by then-Circuit Judge Taft) (analyzing statutes and supplementary materials under Banks and Callaghan and concluding that the materials were copyrightable because they were prepared by a private compiler). Moreover, just as the doctrine applies to “whatever work [judges] perform in their capacity as judges,” Banks, 128 U. S., at 253, it applies to whatever work legislators perform in their capacity as legislators. That of course includes final legislation, but it also includes explanatory and procedural materials legislators create in the discharge of their legislative duties. In the same way that judges cannot be the authors of their headnotes and syllabi, legislators cannot be the authors of (for example) their floor statements, committee reports, and proposed bills. These materials are part of the “whole work done by [legislators],” so they must be “free for publication to all.” Ibid. Under our precedents, therefore, copyright does not vest in works that are (1) created by judges and legislators (2) in the course of their judicial and legislative duties. B 1 Applying that framework, Georgia’s annotations are not copyrightable. The first step is to examine whether their purported author qualifies as a legislator. As we have explained, the annotations were prepared in the first instance by a private company (Lexis) pursuant to a work-for-hire agreement with Georgia’s Code Revision Commission. The Copyright Act therefore deems the Commission the sole “author” of the work. 17 U. S. C. §201(b). Although Lexis expends considerable effort preparing the annotations, for purposes of copyright that labor redounds to the Commission as the statutory author. Georgia agrees that the author is the Commission. Brief for Petitioners 25. The Commission is not identical to the Georgia Legislature, but functions as an arm of it for the purpose of producing the annotations. The Commission is created by the legislature, for the legislature, and consists largely of legislators. The Commission receives funding and staff designated by law for the legislative branch. Significantly, the annotations the Commission creates are approved by the legislature before being “merged” with the statutory text and published in the official code alongside that text at the legislature’s direction. OCGA §1–1–1; see 906 F. 3d, at 1245, 1255; Tr. of Oral Arg. 8. If there were any doubt about the link between the Commission and the legislature, the Georgia Supreme Court has dispelled it by holding that, under the Georgia Constitution, “the work of the Commission; i.e., selecting a publisher and contracting for and supervising the codification of the laws enacted by the General Assembly, including court interpretations thereof, is within the sphere of legislative authority.” Harrison Co., 244 Ga., at 330, 260 S. E. 2d, at 34 (emphasis added). That holding is not limited to the Commission’s role in codifying the statutory text. The Commission’s “legislative authority” specifically includes its “codification of . . . court interpretations” of the State’s laws. Ibid. Thus, as a matter of state law, the Commission wields the legislature’s authority when it works with Lexis to produce the annotations. All of this shows that the Commission serves as an extension of the Georgia Legislature in preparing and publishing the annotations. And it helps explain why the Commission brought this suit asserting copyright in the annotations “on behalf of and for the benefit of ” the Georgia Legislature and the State of Georgia. App. 20.[2] 2 The second step is to determine whether the Commission creates the annotations in the “discharge” of its legislative “duties.” Banks, 128 U. S., at 253. It does. Although the annotations are not enacted into law through bicameralism and presentment, the Commission’s preparation of the annotations is under Georgia law an act of “legislative authority,” Harrison Co., 244 Ga., at 330, 260 S. E. 2d, at 34, and the annotations provide commentary and resources that the legislature has deemed relevant to understanding its laws. Georgia and Justice Ginsburg emphasize that the annotations do not purport to provide authoritative explanations of the law and largely summarize other materials, such as judicial decisions and law review articles. See post, at 3–4 (dissenting opinion). But that does not take them outside the exercise of legislative duty by the Commission and legislature. Just as we have held that the “statement of the case and the syllabus or head note” prepared by judges fall within the “work they perform in their capacity as judges,” Banks, 128 U. S., at 253, so too annotations published by legislators alongside the statutory text fall within the work legislators perform in their capacity as legislators. In light of the Commission’s role as an adjunct to the legislature and the fact that the Commission authors the annotations in the course of its legislative responsibilities, the annotations in Georgia’s Official Code fall within the government edicts doctrine and are not copyrightable. III Georgia resists this conclusion on several grounds. At the outset, Georgia advances two arguments for why, in its view, excluding the OCGA annotations from copyright protection conflicts with the text of the Copyright Act. Both are unavailing. First, Georgia notes that §101 of the Act specifically lists “annotations” among the kinds of works eligible for copyright protection. But that provision refers only to “annotations . . . which . . . represent an original work of authorship.” 17 U. S. C. §101 (emphasis added). The whole point of the government edicts doctrine is that judges and legislators cannot serve as authors when they produce works in their official capacity. While the reference to “annotations” in §101 may help explain why supplemental, explanatory materials are copyrightable when prepared by a private party, or a non-lawmaking official like the reporter in Callaghan, it does not speak to whether those same materials are copyrightable when prepared by a judge or a legislator. In the same way that judicial materials are ineligible for protection even though they plainly qualify as “[l]iterary works . . . expressed in words,” ibid., legislative materials are ineligible for protection even if they happen to fit the description of otherwise copyrightable “annotations.” Second, Georgia draws a negative inference from the fact that the Act excludes from copyright protection “work[s] prepared by an officer or employee of the United States Government as part of that person’s official duties” and does not establish a similar rule for the States. §101; see also §105. But the bar on copyright protection for federal works sweeps much more broadly than the government edicts doctrine does. That bar applies to works created by all federal “officer[s] or employee[s],” without regard for the nature of their position or scope of their authority. Whatever policy reasons might justify the Federal Government’s decision to forfeit copyright protection for its own proprietary works, that federal rule does not suggest an intent to displace the much narrower government edicts doctrine with respect to the States. That doctrine does not apply to non-lawmaking officials, leaving States free to assert copyright in the vast majority of expressive works they produce, such as those created by their universities, libraries, tourism offices, and so on. More generally, Georgia suggests that we should resist applying our government edicts precedents to the OCGA annotations because our 19th-century forebears interpreted the statutory term author by reference to “public policy”—an approach that Georgia believes is incongruous with the “modern era” of statutory interpretation. Brief for Petitioners 21 (internal quotation marks omitted). But we are particularly reluctant to disrupt precedents interpreting language that Congress has since reenacted. As we explained last Term in Helsinn Healthcare S. A. v. Teva Pharmaceuticals USA, Inc., 586 U. S. ___ (2019), when Congress “adopt[s] the language used in [an] earlier act,” we presume that Congress “adopted also the construction given by this Court to such language, and made it a part of the enactment.” Id., at ___ (slip op., at 7) (quoting Shapiro v. United States, 335 U.S. 1, 16 (1948)). A century of cases have rooted the government edicts doctrine in the word “author,” and Congress has repeatedly reused that term without abrogating the doctrine. The term now carries this settled meaning, and “critics of our ruling can take their objections across the street, [where] Congress can correct any mistake it sees.” Kimble v. Marvel Entertainment, LLC, 576 U.S. 446, 456 (2015).[3] Moving on from the text, Georgia invokes what it views as the official position of the Copyright Office, as reflected in the Compendium of U. S. Copyright Office Practices (Compendium). But, as Georgia concedes, the Compendium is a non-binding administrative manual that at most merits deference under Skidmore v. Swift & Co., 323 U.S. 134 (1944). That means we must follow it only to the extent it has the “power to persuade.” Id., at 140. Because our precedents answer the question before us, we find any competing guidance in the Compendium unpersuasive. In any event, the Compendium is largely consistent with our decision. Drawing on Banks, it states that, “[a]s a matter of longstanding public policy, the U. S. Copyright Office will not register a government edict that has been issued by any state, local, or territorial government, including legislative enactments, judicial decisions, administrative rulings, public ordinances, or similar types of official legal materials.” Compendium §313.6(C)(2) (rev. 3d ed. 2017) (emphasis added). And, under Banks, what counts as a “similar” material depends on what kind of officer created the material (i.e., a judge) and whether the officer created it in the course of official (i.e., judicial) duties. See Compendium §313.6(C)(2) (quoting Banks, 128 U. S., at 253, for the proposition that copyright cannot vest “in the products of the labor done by judicial officers in the discharge of their judicial duties”). The Compendium goes on to observe that “the Office may register annotations that summarize or comment upon legal materials . . . unless the annotations themselves have the force of law.” Compendium §313.6(C)(2). But that broad statement—true of annotations created by officials such as court reporters that lack the authority to make or interpret the law—does not engage with the critical issue of annotations created by judges or legislators in their official capacities. Because the Compendium does not address that question and otherwise echoes our government edicts precedents, it is of little relevance here. Georgia also appeals to the overall purpose of the Copyright Act to promote the creation and dissemination of creative works. Georgia submits that, without copyright protection, Georgia and many other States will be unable to induce private parties like Lexis to assist in preparing affordable annotated codes for widespread distribution. That appeal to copyright policy, however, is addressed to the wrong forum. As Georgia acknowledges, “[I]t is generally for Congress, not the courts, to decide how best to pursue the Copyright Clause’s objectives.” Eldred v. Ashcroft, 537 U.S. 186, 212 (2003). And that principle requires adherence to precedent when, as here, we have construed the statutory text and “tossed [the ball] into Congress’s court, for acceptance or not as that branch elects.” Kimble, 576 U. S., at 456. Turning to our government edicts precedents, Georgia insists that they can and should be read to focus exclusively on whether a particular work has “the force of law.” Brief for Petitioners 32 (capitalization deleted). Justice Thomas appears to endorse the same view. See post, at 4. But that framing has multiple flaws. Most obviously, it cannot be squared with the reasoning or results of our cases—especially Banks. Banks, following Wheaton and the “judicial consensus” it inspired, denied copyright protection to judicial opinions without excepting concurrences and dissents that carry no legal force. 128 U. S., at 253 (emphasis deleted). As every judge learns the hard way, “comments in [a] dissenting opinion” about legal principles and precedents “are just that: comments in a dissenting opinion.” Railroad Retirement Bd. v. Fritz, 449 U.S. 166, 177, n. 10 (1980). Yet such comments are covered by the government edicts doctrine because they come from an official with authority to make and interpret the law. Indeed, Banks went even further and withheld copyright protection from headnotes and syllabi produced by judges. 128 U. S., at 253. Surely these supplementary materials do not have the force of law, yet they are covered by the doctrine. The simplest explanation is the one Banks provided: These non-binding works are not copyrightable because of who creates them—judges acting in their judicial capacity. See ibid. The same goes for non-binding legislative materials produced by legislative bodies acting in a legislative capacity. There is a broad array of such works ranging from floor statements to proposed bills to committee reports. Under the logic of Georgia’s “force of law” test, States would own such materials and could charge the public for access to them. Furthermore, despite Georgia’s and Justice Thomas’s purported concern for the text of the Copyright Act, their conception of the government edicts doctrine has less of a textual footing than the traditional formulation. The textual basis for the doctrine is the Act’s “authorship” requirement, which unsurprisingly focuses on—the author. Justice Thomas urges us to dig deeper to “the root” of our government edicts precedents. Post, at 5. But, in our view, the text is the root. The Court long ago interpreted the word “author” to exclude officials empowered to speak with the force of law, and Congress has carried that meaning forward in multiple iterations of the Copyright Act. This textual foundation explains why the doctrine distinguishes between some authors (who are empowered to speak with the force of law) and others (who are not). Compare Callaghan, 128 U. S., at 647, with Banks, 128 U. S., at 253. But the Act’s reference to “authorship” provides no basis for Georgia’s rule distinguishing between different categories of content with different effects.[4] Georgia minimizes the OCGA annotations as non-binding and non-authoritative, but that description undersells their practical significance. Imagine a Georgia citizen interested in learning his legal rights and duties. If he reads the economy-class version of the Georgia Code available online, he will see laws requiring political candidates to pay hefty qualification fees (with no indigency exception), criminalizing broad categories of consensual sexual conduct, and exempting certain key evidence in criminal trials from standard evidentiary limitations—with no hint that important aspects of those laws have been held unconstitutional by the Georgia Supreme Court. See OCGA §§21–2–131, 16–6–2, 16–6–18, 16–15–9 (available at www.legis.ga.gov). Meanwhile, first-class readers with access to the annotations will be assured that these laws are, in crucial respects, unenforceable relics that the legislature has not bothered to narrow or repeal. See §§21–2–131, 16–6–2, 16–6–18, 16–15–9 (available at https://store. lexisnexis . com / products / official - code - of - georgia - annotated - skuSKU6647 for $412.00). If everything short of statutes and opinions were copyrightable, then States would be free to offer a whole range of premium legal works for those who can afford the extra benefit. A State could monetize its entire suite of legislative history. With today’s digital tools, States might even launch a subscription or pay-per-law service. There is no need to assume inventive or nefarious behavior for these concerns to become a reality. Unlike other forms of intellectual property, copyright protection is both instant and automatic. It vests as soon as a work is captured in a tangible form, triggering a panoply of exclusive rights that can last over a century. 17 U. S. C. §§102, 106, 302. If Georgia were correct, then unless a State took the affirmative step of transferring its copyrights to the public domain, all of its judges’ and legislators’ non-binding legal works would be copyrighted. And citizens, attorneys, nonprofits, and private research companies would have to cease all copying, distribution, and display of those works or risk severe and potentially criminal penalties. §§501–506. Some affected parties might be willing to roll the dice with a potential fair use defense. But that defense, designed to accommodate First Amendment concerns, is notoriously fact sensitive and often cannot be resolved without a trial. Cf. Harper & Row, Publishers, Inc. v. Nation Enterprises, 471 U.S. 539, 552, 560–561 (1985). The less bold among us would have to think twice before using official legal works that illuminate the law we are all presumed to know and understand. Thankfully, there is a clear path forward that avoids these concerns—the one we are already on. Instead of examining whether given material carries “the force of law,” we ask only whether the author of the work is a judge or a legislator. If so, then whatever work that judge or legislator produces in the course of his judicial or legislative duties is not copyrightable. That is the framework our precedents long ago established, and we adhere to those precedents today. * * * For the foregoing reasons, we affirm the judgment of the Eleventh Circuit. It is so ordered. Notes 1 The Copyright Act of 1790 granted copyright protection to “the author and authors” of qualifying works. Act of May 31, 1790, §1, 1Stat. 124. This author requirement appears in the current Copyright Act at §102(a), which limits protection to “original works of authorship.” 17 U. S. C. §102(a) (emphasis added); see also §201(a) (copyright “vests initially in the author or authors of the work”). 2 Justice Thomas does not dispute that the Commission is an extension of the legislature; he instead faults us for highlighting the multiple features of the Commission that make clear that this is so. See post, at 16 (dissenting opinion). 3 Justice Thomas disputes the applicability of the Helsinn Healthcare presumption because States have asserted copyright in statutory annotations over the years notwithstanding our government edicts precedents. Post, at 11–12. In Justice Thomas’s view, those assertions prove that our precedents could not have provided clear enough guidance for Congress to incorporate. But that inference from state behavior proves too much. The same study cited by Justice Thomas to support a practice of claiming copyright in non-binding annotations also reports that “many states claim copyright interest in their primary law materials,” including statutes and regulations. Dmitrieva, State Ownership of Copyrights in Primary Law Materials, 23 Hastings Com. & Entertainment L. J. 81, 109 (2000) (emphasis added). Justice Thomas concedes that such assertions are plainly foreclosed by our government edicts precedents. Post, at 4. That interested parties have pursued ambitious readings of our precedents does not mean those precedents are incapable of providing meaningful guidance to us or to Congress. 4 Instead of accepting our predecessors’ textual reasoning at face value, Justice Thomas conjures a trinity of alternative “origin[s] and justification[s]” for the government edicts doctrine that the Court might have had in mind. See post, at 5–7. Without committing to one or all of these possibilities, Justice Thomas suggests that each would yield a rule that requires federal courts to pick out the subset of judicial and legislative materials that independently carry the force of law. But a Court motivated by Justice Thomas’s three-fold concerns might just as easily have read them as supporting a rule that prevents the officials responsible for creating binding materials from qualifying as an “author.” Regardless, it is more “[ ]consistent with the judicial role” to apply the reasoning and results the Court voted on and committed to writing than to speculate about what practical considerations our predecessors “may have had . . . in mind,” what history “may [have] suggest[ed],” or what constitutional concerns “may have animated” our government edicts precedents. Ibid.

589.US.2019_18-776

The Immigration and Nationality Act provides for judicial review of a final Government order directing the removal of an alien from this country. 8 U. S. C. §1252(a). Section 1252(a)(2)(C) limits the scope of that review where the removal rests upon the fact that the alien has committed certain crimes. And §1252(a)(2)(D), the Limited Review Provision, says that in such instances courts may consider only “constitutional claims or questions of law.” Petitioners Guerrero-Lasprilla and Ovalles, aliens who lived in the United States, committed drug crimes and were subsequently ordered removed (Guerrero-Lasprilla in 1998 and Ovalles in 2004). Neither filed a motion to reopen his removal proceedings “within 90 days of the date of entry of [the] final administrative order of removal.” §1229a(c)(7)(C)(i). Nonetheless, Guerrero-Lasprilla (in 2016) and Ovalles (in 2017) asked the Board of Immigration Appeals to reopen their removal proceedings, arguing that the 90-day time limit should be equitably tolled. Both petitioners, who had become eligible for discretionary relief due to various judicial and Board decisions years after their removal, rested their claim for equitable tolling on Lugo-Resendez v. Lynch, 831 F.3d 337, in which the Fifth Circuit had held that the 90-day time limit could be equitably tolled. The Board denied both petitioners’ requests, concluding, inter alia, that they had not demonstrated the requisite due diligence. The Fifth Circuit denied their requests for review, holding that, given the Limited Review Provision, it “lack[ed] jurisdiction” to review petitioners’ “factual” due diligence claims. Petitioners contend that whether the Board incorrectly applied the equitable tolling due diligence standard to the undisputed facts of their cases is a “question of law” that the Provision authorizes courts of appeals to consider. Held: Because the Provision’s phrase “questions of law” includes the application of a legal standard to undisputed or established facts, the Fifth Circuit erred in holding that it had no jurisdiction to consider petitioners’ claims of due diligence for equitable tolling purposes. Pp. 3–13. (a) Nothing in the statute’s language precludes the conclusion that Congress used the term “questions of law” to refer to the application of a legal standard to settled facts. Indeed, this Court has at times referred to the question whether a given set of facts meets a particular legal standard as presenting a legal inquiry. See Neitzke v. Williams, 490 U.S. 319, 326 (“Rule 12(b)(6) authorizes a court to dismiss a claim on the basis of a dispositive issue of law”); Mitchell v. Forsyth, 472 U.S. 511, 528, n. 9 (“[T]he appealable issue is a purely legal one: whether the facts alleged . . . support a claim of violation of clearly established law”); cf. Nelson v. Montgomery Ward & Co., 312 U.S. 373, 376 (“The effect of admitted facts is a question of law”). That judicial usage indicates that the statutory term “questions of law” can reasonably encompass questions about whether settled facts satisfy a legal standard. The Court has sometimes referred to such a question as a “mixed question of law and fact.” See, e.g., U. S. Bank N. A. v. Village at Lakeridge, LLC, 583 U. S. ___, ___. And the Court has often used the phrase “mixed questions” in determining the proper standard for appellate review of a district, bankruptcy, or agency decision that applies a legal standard to underlying facts. But these cases present no such question involving the standard of review. And, in any event, nothing in those cases, nor in the language of the statute, suggests that the statutory phrase “questions of law” excludes the application of law to settled facts. Pp. 4–5. (b) A longstanding presumption, the statutory context, and the statute’s history all support the conclusion that the application of law to undisputed or established facts is a “questio[n] of law” within the meaning of §1252(a)(2)(D). Pp. 5–11. (1) A “well-settled” and “strong presumption,” McNary v. Haitian Refugee Center, Inc., 498 U.S. 479, 496, 498, “favor[s] judicial review of administrative action,” Kucana v. Holder, 558 U.S. 233, 251. That presumption, which can only be overcome by “ ‘ “clear and convincing evidence” ’ ” of congressional intent to preclude judicial review, Reno v. Catholic Social Services, Inc., 509 U.S. 43, 64, has consistently been applied to immigration statutes, Kucana, 558 U. S., at 251. And there is no reason to make an exception here. Because the Court can reasonably interpret the statutory term “questions of law” to encompass the application of law to undisputed facts, and given that a contrary interpretation would result in a barrier to meaningful judicial review, the presumption indicates that “questions of law” does indeed include mixed questions. Pp. 6–7. (2) The Limited Review Provision’s immediate statutory context belies the Government and the dissent’s claim that “questions of law” excludes the application of law to settled facts. The Provision is part of §1252, which also contains §1252(b)(9), the “zipper clause.” The zipper clause is meant to “consolidate judicial review of immigration proceedings into one action in the court of appeals.” INS v. St. Cyr, 533 U.S. 289, 313. The zipper clause’s language makes clear that Congress understood the statutory term “questions of law and fact,” to include the application of law to facts. One interpretation of the zipper clause at the very least disproves the Government’s argument that Congress consistently uses a three-part typology, such that “questions of law” cannot include mixed questions. And another interpretation—that “questions of law” in the zipper clause includes mixed questions—directly supports the holding here and would give the term the same meaning in the zipper clause and the Limited Review Provision. Pp. 7–8. (3) The Provision’s statutory history and relevant precedent also support this conclusion. The Provision was enacted in response to INS v. St. Cyr, in which the Court interpreted the predecessor of §1252(a)(2)(C) to permit habeas corpus review in order to avoid the serious constitutional questions that would arise from a contrary interpretation, 533 U. S., at 299–305, 314. In doing so, the Court suggested that the Constitution, at a minimum, protected the writ of habeas corpus “ ‘as it existed in 1789.’ ” Id., at 300–301. The Court then noted the kinds of review that were traditionally available in a habeas proceeding, which included “detentions based on errors of law, including the erroneous application or interpretation of statutes.” Id., at 302 (emphasis added). Congress took up the Court’s invitation to “provide an adequate substitute [for habeas review] through the courts of appeals,” id., at 314, n. 38. It made clear that the limits on judicial review in various §1252 provisions included habeas review, and it consolidated virtually all review of removal orders in one proceeding in the courts of appeals. Congress also added the Limited Review Provision, permitting review of “constitutional claims or questions of law.” Congress did so, the statutory history strongly suggests, because it sought an “adequate substitute” for habeas in view of St. Cyr’s guidance. If “questions of law” in the Provision does not include the misapplication of a legal standard to undisputed facts, then review would not include an element that St. Cyr said was traditionally reviewable in habeas. Lower court precedent citing St. Cyr and legislative history also support this conclusion. Pp. 8–11. (c) The Government’s additional arguments in favor of its contrary reading are unpersuasive. More than that, the Government’s interpretation is itself difficult to reconcile with the Provision’s basic purpose of providing an adequate substitute for habeas review. Pp. 11–13. No. 18–776, 737 Fed. Appx. 230; No. 18–1015, 741 Fed. Appx. 259, vacated and remanded. Breyer, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a dissenting opinion, in which Alito, J., joined as to all but Part II–A–1. Notes 1 Together with No. 18–1015, Ovalles v. Barr, Attorney General, also on certiorari to the same court.

Section 242(a) of the Immigration and Nationality Act, codified as 8 U. S. C. §1252(a), provides for judicial review of a final Government order directing the removal of an alien from this country. See 66Stat. 163, as amended, 8 U. S. C. §1101 et seq. A subdivision of that section limits the scope of that review where the removal rests upon the fact that the alien has committed certain crimes, including aggravated felonies and controlled substance offenses. §1252(a)(2)(C). Another subdivision, §1252(a)(2)(D), which we shall call the Limited Review Provision, says that in such instances courts may consider only “constitutional claims or questions of law.” The question that these two consolidated cases present is whether the phrase “questions of law” in the Provision includes the application of a legal standard to undisputed or established facts. We believe that it does. I The two petitioners before us, Pedro Pablo Guerrero-Lasprilla and Ruben Ovalles, are aliens who lived in the United States. Each committed a drug crime and consequently became removable. App. 33; Record in No. 18–1015, p. 66. In 1998, an Immigration Judge ordered Guerrero-Lasprilla removed. Record in No. 18–776, p. 137. In 2004, the Board of Immigration Appeals ordered Ovalles removed, reversing a decision by an Immigration Judge. App. to Pet. for Cert. in No. 18–1015, pp. 32a–35a. Both removal orders became administratively final, and both petitioners left the country. Several months after their removal orders became final, each petitioner’s window for filing a timely motion to reopen his removal proceedings closed. That is because the Immigration and Nationality Act permits a person one motion to reopen, “a form of procedural relief that asks the Board to change its decision in light of newly discovered evidence or a change in circumstances.” Dada v. Mukasey, 554 U.S. 1, 12, 14 (2008) (internal quotation marks omitted). But the motion must usually be filed “within 90 days of the date of entry of a final administrative order of removal.” 8 U. S. C. §1229a(c)(7)(C)(i). Nonetheless, Guerrero-Lasprilla (in 2016) and Ovalles (in 2017) asked the Board to reopen their removal proceedings. Recognizing that the 90-day time limit had long since passed, both petitioners argued that the time limit should be equitably tolled. Both petitioners, who had become eligible for discretionary relief due to various judicial and Board decisions years after their removal, rested their claim for equitable tolling on Lugo-Resendez v. Lynch, 831 F.3d 337 (CA5 2016). In that case, the Fifth Circuit had held that the 90-day time limit could be “equitably tolled.” Id., at 344. Guerrero-Lasprilla filed his motion to reopen a month after Lugo-Resendez was decided. App. 5. Ovalles filed his motion to reopen eight months after the decision. Id., at 35. The Board denied both petitioners’ requests for equitable tolling, concluding, inter alia, that they had failed to demonstrate the requisite due diligence. App. to Pet. for Cert. in No. 18–1015, at 6a; App. to Pet. for Cert. in No. 18–776, p. 12a. Guerrero-Lasprilla and Ovalles each asked the Fifth Circuit to review the Board’s decision. See 8 U. S. C. §1252(a)(1); 28 U. S. C. §2342; Reyes Mata v. Lynch, 576 U.S. 143, 147 (2015) (“[C]ircuit courts have jurisdiction when an alien appeals from the Board’s denial of a motion to reopen a removal proceeding”). The Fifth Circuit denied their requests for review, concluding in both cases that “whether an alien acted diligently in attempting to reopen removal proceedings for purposes of equitable tolling is a factual question.” Guerrero-Lasprilla v. Sessions, 737 Fed. Appx. 230, 231 (2018) (per curiam); Ovalles v. Sessions, 741 Fed. Appx. 259, 261 (2018) (per curiam). And, given the Limited Review Provision, it “lack[ed] jurisdiction” to review those “factual” claims. 737 Fed. Appx., at 231; 741 Fed. Appx., at 261. Both petitioners claim that the underlying facts were not in dispute, and they asked us to grant certiorari in order to determine whether their claims that the Board incorrectly applied the equitable tolling due diligence standard to the “undisputed” (or established) facts is a “question of law,” which the Limited Review Provision authorizes courts of appeals to consider. We agreed to do so. II The Limited Review Provision provides that, in this kind of immigration case (involving aliens who are removable for having committed certain crimes), a court of appeals may consider only “constitutional claims or questions of law.” 8 U. S. C. §1252(a)(2)(D). The issue before us is, as we have said, whether the statutory phrase “questions of law” includes the application of a legal standard to undisputed or established facts. If so, the Fifth Circuit erred in holding that it “lack[ed] jurisdiction” to consider the petitioners’ claims of due diligence for equitable tolling purposes. We conclude that the phrase “questions of law” does include this type of review, and the Court of Appeals was wrong to hold the contrary. A Consider the statute’s language. Nothing in that language precludes the conclusion that Congress used the term “questions of law” to refer to the application of a legal standard to settled facts. Indeed, we have at times referred to the question whether a given set of facts meets a particular legal standard as presenting a legal inquiry. Do the facts alleged in a complaint, taken as true, state a claim for relief under the applicable legal standard? See Fed. Rule Civ. Proc. 12(b)(6); Neitzke v. Williams, 490 U.S. 319, 326 (1989) (“Rule 12(b)(6) authorizes a court to dismiss a claim on the basis of a dispositive issue of law”). Did a Government official’s alleged conduct violate clearly established law? See Mitchell v. Forsyth, 472 U.S. 511, 528, n. 9 (1985) (“[T]he appealable issue is a purely legal one: whether the facts alleged . . . support a claim of violation of clearly established law”); cf. Nelson v. Montgomery Ward & Co., 312 U.S. 373, 376 (1941) (“The effect of admitted facts is a question of law”). Even the dissent concedes that we have sometimes referred to mixed questions as raising a legal inquiry. See post, at 3–4 (opinion of Thomas, J.). While that judicial usage alone does not tell us what Congress meant by the statutory term “questions of law,” it does indicate that the term can reasonably encompass questions about whether settled facts satisfy a legal standard. We have sometimes referred to such a question, which has both factual and legal elements, as a “mixed question of law and fact.” See, e.g., U. S. Bank N. A. v. Village at Lakeridge, LLC, 583 U. S. ___, ___ (2018) (slip op., at 7) (“[W]hether the historical facts found satisfy the legal test chosen” is a “so-called ‘mixed question’ of law and fact” (citing Pullman-Standard v. Swint, 456 U.S. 273, 289, n. 19 (1982))). And we have often used the phrase “mixed questions” in determining the proper standard for appellate review of a district, bankruptcy, or agency decision that applies a legal standard to underlying facts. The answer to the “proper standard” question may turn on practical considerations, such as whether the question primarily “require[s] courts to expound on the law, particularly by amplifying or elaborating on a broad legal standard” (often calling for review de novo), or rather “immerse[s] courts in case-specific factual issues” (often calling for deferential review). Village at Lakeridge, 583 U. S., at ___ (slip op., at 8). But these cases present no such question involving the standard of review. And, in any event, nothing in those cases forecloses the conclusion that the application of law to settled facts can be encompassed within the statutory phrase “questions of law.” Nor is there anything in the language of the statute that suggests that “questions of law” excludes the application of law to settled facts. B The Government, respondent here, argues to the contrary. Namely, the Government claims that Congress intended to exclude from judicial review all mixed questions. We do not agree. Rather, a longstanding presumption, the statutory context, and the statute’s history all support the conclusion that the application of law to undisputed or established facts is a “questio[n] of law” within the meaning of §1252(a)(2)(D). 1 Consider first “a familiar principle of statutory construction: the presumption favoring judicial review of administrative action.” Kucana v. Holder, 558 U.S. 233, 251 (2010). Under that “well-settled” and “strong presumption,” McNary v. Haitian Refugee Center, Inc., 498 U.S. 479, 496, 498 (1991), when a statutory provision “is reasonably susceptible to divergent interpretation, we adopt the reading that accords with traditional understandings and basic principles: that executive determinations generally are subject to judicial review.” Kucana, 558 U. S., at 251 (quoting Gutierrez de Martinez v. Lamagno, 515 U.S. 417, 434 (1995); internal quotation marks omitted); see McNary, 498 U. S., at 496 (“[G]iven [that] presumption . . . , it is most unlikely that Congress intended to foreclose all forms of meaningful judicial review”). The presumption can only be overcome by “clear and convincing evidence” of congressional intent to preclude judicial review. Reno v. Catholic Social Services, Inc., 509 U.S. 43, 64 (1993) (quoting Abbott Laboratories v. Gardner, 387 U.S. 136, 141 (1967); internal quotation marks omitted); see Cuozzo Speed Technologies, LLC v. Lee, 579 U. S. ___, ___–___ (2016) (slip op., at 9–10). We have “consistently applied” the presumption of reviewability to immigration statutes. Kucana, 558 U. S., at 251. And we see no reason to make an exception here. The dissent’s “doubts” about the presumption, see post, at 6–9, do not undermine our recognition that it is a “well-settled” principle of statutory construction, McNary, 498 U. S., at 496. Notably, even the Government does not dispute the soundness of the presumption or its applicability here. See Brief for Respondent 47–48 (arguing only that the presumption is overcome). As discussed above, we can reasonably interpret the statutory term “questions of law” to encompass the application of law to undisputed facts. See supra, at 4–5. And as we explain further below, infra, at 13, interpreting the Limited Review Provision to exclude mixed questions would effectively foreclose judicial review of the Board’s determinations so long as it announced the correct legal standard. The resulting barrier to meaningful judicial review is thus a strong indication, given the presumption, that “questions of law” does indeed include the application of law to established facts. That is particularly so given that the statutory context and history point to the same result. 2 Consider next the Limited Review Provision’s immediate statutory context. That context belies the Government and the dissent’s claim that “questions of law” refers only to “pure” questions and necessarily excludes the application of law to settled facts. See Brief for Respondent 19–26; post, at 3–6. The Limited Review Provision forms part of §1252, namely, §1252(a)(2)(D). The same statutory section contains a provision, §1252(b)(9), which we have called a “ ‘zipper clause.’ ” INS v. St. Cyr, 533 U.S. 289, 313 (2001). We have explained that Congress intended the zipper clause to “consolidate judicial review of immigration proceedings into one action in the court of appeals.” Ibid. (internal quotation marks omitted). The zipper clause reads in part as follows: “Judicial review of all questions of law and fact, including interpretation and application of constitutional and statutory provisions, arising from any action taken . . . to remove an alien from the United States under this subchapter shall be available only in judicial review of a final order under this section.” §1252(b)(9) (emphasis added). Because it is meant to consolidate judicial review, the zipper clause must encompass mixed questions. Indeed, the clause by its very language includes the “application of [a] statutory provisio[n].” Ibid. The zipper clause accordingly makes clear that Congress understood the statutory term “questions of law and fact” to include the application of law to facts. Reread the zipper clause: It uses the terms “[(1)] questions of law and [(2)] fact, including” the “application of ” statutes, i.e., the application of law to fact. Ibid. (emphasis added). Thus, there are three possibilities: Congress either used (1) “questions of law,” (2) “fact,” or (3) the combination of both terms to encompass mixed questions. Even the Government does not argue that Congress used “questions of fact” alone to cover mixed questions. Congress thus either meant the term “questions of law” alone to include mixed questions, or it used both “questions of law” and questions of “fact” to encompass mixed questions. The latter interpretation at the very least disproves the Government’s argument that Congress consistently uses a three-part typology, referring to mixed questions separately from questions of law or questions of fact (such that “questions of law” cannot include mixed questions). See Brief for Respondent 21; see also post, at 3 (arguing that this Court has often used that three-part typology and thus “questions of law” must exclude mixed questions). And the former interpretation directly supports the conclusion that “questions of law” includes mixed questions. That interpretation gives “questions of law” the same meaning across both provisions. Notably, when Congress enacted the Limited Review Provision, it added language to the end of the zipper clause (following the language quoted above) to clarify that, except as provided elsewhere in §1252, “ ‘no court shall have jurisdiction’ ” to “ ‘review . . . such questions of law or fact.’ ” §106, 119Stat. 311. There is thus every reason to think that Congress used the phrase “questions of law” to have the same meaning in both provisions. 3 Consider also the Limited Review Provision’s statutory history and the relevant precedent. The parties agree that Congress enacted the Limited Review Provision in response to this Court’s decision in St. Cyr. See Brief for Respondent 16, 27–31; Brief for Petitioners 31–33. In that case, the Court evaluated the effect of various allegedly jurisdiction-stripping provisions, including the predecessor to §1252(a)(2)(C). That predecessor (which today is modified by the Limited Review Provision) essentially barred judicial review of removal orders based on an alien’s commission of certain crimes. See St. Cyr, 533 U. S., at 298, 311 (citing §1252(a)(2)(C) (1994 ed., Supp. V)). This Court interpreted that predecessor and the other purportedly jurisdiction-stripping provisions as not barring (i.e., as permitting) review in habeas corpus proceedings, to avoid the serious constitutional questions that would be raised by a contrary interpretation. See St. Cyr, 533 U. S., at 299–305, 314. In doing so, the Court suggested that the Constitution, at a minimum, protected the writ of habeas corpus “ ‘as it existed in 1789.’ ” Id., at 300–301. The Court then noted the kinds of review that were traditionally available in a habeas proceeding, which included “detentions based on errors of law, including the erroneous application or interpretation of statutes.” Id., at 302 (emphasis added). And it supported this view by citing cases from the 18th and early 19th centuries. See id., at 302–303, and nn. 18–23. English cases consistently demonstrate that the “erroneous application . . . of statutes” includes the misapplication of a legal standard to the facts of a particular case. See, e.g., Hollingshead’s Case, 1 Salk. 351, 91 Eng. Rep. 307 (K. B. 1702); King v. Nathan, 2 Str. 880, 93 Eng. Rep. 914 (K. B. 1724); King v. Rudd, 1 Cowp. 331, 334–337, 98 Eng. Rep. 1114, 1116–1117 (K. B. 1775); King v. Pedley, 1 Leach 325, 326, 168 Eng. Rep. 265, 266 (1784). The Court ultimately made clear that “Congress could, without raising any constitutional questions, provide an adequate substitute [for habeas review] through the courts of appeals.” St. Cyr., 533 U. S., at 314, n. 38. Congress took up this suggestion. It made clear that the limits on judicial review in various provisions of §1252 included habeas review, and it consolidated virtually all review of removal orders in one proceeding in the courts of appeals. See §106(a), 119Stat. 310–311 (inserting specific references to 28 U. S. C. §2241 and “ ‘any other habeas corpus provision’ ”). At the same time, Congress added the Limited Review Provision, which permits judicial review of “ ‘constitutional claims or questions of law,’ ” the words directly before us now. 119Stat. 310. This statutory history strongly suggests that Congress added the words before us because it sought an “adequate substitute” for habeas in view of St. Cyr’s guidance. See supra, at 9. If so, then the words “questions of law” in the Limited Review Provision must include the misapplication of a legal standard to undisputed facts, for otherwise review would not include an element that St. Cyr said was traditionally reviewable in habeas. We reach the same conclusion through reference to lower court precedent. After we decided St. Cyr, numerous Courts of Appeals held that habeas review included review of the application of law to undisputed facts. See Cadet v. Bulger, 377 F.3d 1173, 1184 (CA11 2004) (“[W]e hold that the scope of habeas review available in [28 U. S. C.] §2241 petitions by aliens challenging removal orders . . . includes . . . errors of law, including both statutory interpretations and application of law to undisputed facts or adjudicated facts”); Ogbudimkpa v. Ashcroft, 342 F.3d 207, 222 (CA3 2003) (same); Mu-Xing Wang v. Ashcroft, 320 F.3d 130, 143 (CA2 2003) (same); Singh v. Ashcroft, 351 F.3d 435, 441–442 (CA9 2003) (“[O]ther courts have rejected the Government’s argument that only ‘purely legal questions of statutory interpretation’ permit the exercise of habeas jurisdiction. . . . We agree with those rulings”). We normally assume that Congress is “aware of relevant judicial precedent” when it enacts a new statute. Merck & Co. v. Reynolds, 559 U.S. 633, 648 (2010). Thus, we should assume that Congress, aware of this precedent (and wishing to substitute review in the courts of appeals for habeas review), would have intended the phrase “questions of law” to include the application of a legal standard to established or undisputed facts. Those who deem legislative history a useful interpretive tool will find that the congressional history of the Limited Review Provision supports this analysis. The House Conference Report refers to St. Cyr and adds that Congress’ amendments are designed to “provide an ‘adequate and effective’ alternative to habeas corpus” in the courts of appeals. H. R. Conf. Rep. No. 109–72, p. 175 (2005) (citing St. Cyr, 533 U. S., at 314, n. 38). The Report adds that the amendments “would not change the scope of review that criminal aliens currently receive.” H. R. Conf. Rep. No. 109–72, at 175. And as we know, that “scope of review” included review of decisions applying a legal standard to undisputed or established facts. That is what this Court, in St. Cyr, had said was traditionally available in habeas; and it was how courts of appeals then determined the scope of habeas review. Notably, the legislative history indicates that Congress was well aware of the state of the law in the courts of appeals in light of St. Cyr. See H. R. Conf. Rep. No. 109–72, at 174 (discussing issues on which the Courts of Appeals agreed and those on which they had split after St. Cyr). The statutory history and precedent, as well as the legislative history, thus support the conclusion that the statutory term “questions of law” includes the application of a legal standard to established facts. III The Government makes two significant arguments that we have not yet discussed. First, it points out that §1252(a)(2)(C) forbids (subject to the Limited Review Provision) review of a removal order based on an alien’s commission of certain crimes. If the words “questions of law” include “mixed questions,” then for such aliens, the Limited Review Provision excludes only (or primarily) agency fact-finding from review. But if Congress intended no more than that, then why, the Government asks, did it not just say so directly rather than eliminate judicial review and then restore it for “constitutional claims or questions of law?” Brief for Respondent 49–50. One answer to this question is that the Limited Review Provision applies to more of the statute than the immediately preceding subparagraph. See §1252(a)(2)(D) (applying notwithstanding “subparagraph (B) or (C), or in any other provision of this chapter (other than this section)”). Another answer is that Congress did not write the Limited Review Provision on a blank slate. Rather, subparagraph (C) initially forbade judicial review, and Congress then simply wrote another subparagraph reflecting our description in St. Cyr of the review traditionally available in habeas (or a substitute for habeas in the courts of appeals). See supra, at 8–10. That statutory history also illustrates why the dissent errs in relying so significantly on language in subparagraph (C) proscribing judicial review. See post, at 5–6, 9 (referring to the “sweeping” and “broad” language of subparagraph (C)). A broad and sweeping reading of subparagraph (C) was precisely what this Court rejected in St. Cyr, and Congress enacted subparagraph (D) in response to that opinion. Subparagraph (C)—constrained as it is by subparagraph (D)—must thus be read in that context. Second, the Government argues that our interpretation will undercut Congress’ efforts to severely limit and streamline judicial review of an order removing aliens convicted of certain crimes. See Brief for Respondent 29–30; see also post, at 11, n. 5 (noting that the legislative history indicates that Congress intended to streamline removal proceedings by limiting judicial review). The Limited Review Provision, however, will still forbid appeals of factual determinations—an important category in the removal context. And that Provision, taken together with other contemporaneous amendments to §1252, does streamline judicial review relative to the post-St. Cyr regime, by significantly curtailing habeas proceedings in district courts. More than that, the Government’s interpretation is itself difficult to reconcile with the Provision’s basic purpose of providing an adequate substitute for habeas review. That interpretation would forbid review of any Board decision applying a properly stated legal standard, irrespective of how mistaken that application might be. By reciting the standard correctly, the Board would be free to apply it in a manner directly contrary to well-established law. The Government, recognizing the extreme results of its interpretation, suggested at oral argument that the courts of appeals might still be able to review certain “categori[es]” of applications, such as whether someone being in a coma always, sometimes, or never requires equitable tolling. See Tr. of Oral Arg. 38. The Government, however, left the nature and rationale of this approach unclear. The approach does not overcome the problem we have just raised, and seems difficult to reconcile with the language and purposes of the statute. * * * For these reasons, we reverse the Fifth Circuit’s “jurisdictional” decisions, vacate its judgments, and remand these cases for further proceedings consistent with this opinion. It is so ordered.

589.US.2019_17-1678

Respondent, United States Border Patrol Agent Jesus Mesa, Jr., shot and killed Sergio Adrián Hernández Güereca, a 15-year-old Mexican national, in a tragic and disputed cross-border incident. Mesa was standing on U. S. soil when he fired the bullets that struck and killed Hernández, who was on Mexican soil, after having just run back across the border following entry onto U. S. territory. Agent Mesa contends that Hernández was part of an illegal border crossing attempt, while petitioners, Hernández’s parents, claim he was playing a game with his friends that involved running back and forth across the culvert separating El Paso, Texas, from Ciudad Juarez, Mexico. The shooting drew international attention, and the Department of Justice investigated, concluded that Agent Mesa had not violated Customs and Border Patrol policy or training, and declined to bring charges against him. The United States also denied Mexico’s request for Agent Mesa to be extradited to face criminal charges in Mexico. Petitioners sued for damages in U. S. District Court under Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388, alleging that Mesa violated Hernández’s Fourth and Fifth Amendment rights. The District Court dismissed their claims, and the United States Court of Appeals for the Fifth Circuit affirmed. After this Court vacated that decision and remanded for further consideration in light of Ziglar v. Abbasi, 582 U. S. ___, the Fifth Circuit again affirmed, refusing to recognize a Bivens claim for a cross-border shooting. Held: Bivens’ holding does not extend to claims based on a cross-border shooting. Pp. 4–20. (a) In Bivens, the Court implied a Fourth Amendment claim for damages even though no federal statute authorized such a claim. The Court later extended Bivens’ reach to cover claims under the Fifth and Eighth Amendments. See Davis v. Passman, 442 U.S. 228; Carlson v. Green, 446 U.S. 14. But Bivens’ expansion has since become “a ‘disfavored’ judicial activity,” Abbasi, supra, at ___, and the Court has generally expressed doubt about its authority to recognize causes of action not expressly created by Congress, see, e.g., Jesner v. Arab Bank, PLC, 584 U. S. ___, ___. When considering whether to extend Bivens, the Court uses a two-step inquiry that first asks whether the request involves a claim that arises in a “new context” or involves a “new category of defendants.” Correctional Services Corp. v. Malesko, 534 U.S. 61, 68. If so, the Court then asks whether there are any “special factors [that] counse[l] hesitation” about granting the extension. Abbasi, supra, at ___. Pp. 4–8. (b) Petitioners’ Bivens claims arise in a new context. Their claims are based on the same constitutional provisions as claims in cases in which damages remedies were previously recognized, but the context—a cross-border shooting—is significantly “different . . . from previous Bivens cases.” Abbasi, supra, ___. It involves a “risk of disruptive intrusion by the Judiciary into the functioning of other branches.” Abbasi, supra, ___. Pp. 8–9. (c) Multiple, related factors counsel hesitation before extending Bivens remedies into this new context. Pp. 9–19. (1) The expansion of a Bivens remedy that impinges on foreign relations—an arena “so exclusively entrusted to the political branches . . . as to be largely immune from judicial inquiry,” Haig v. Agee, 453 U.S. 280, 292—risks interfering with the Executive Branch’s “lead role in foreign policy,” Medellín v. Texas, 552 U.S. 491, 524. A cross-border shooting affects the interests of two countries and, as happened here, may lead to disagreement. It is not for this Court to arbitrate between the United States and Mexico, which both have legitimate and important interests at stake and have sought to reconcile those interests through diplomacy. Pp. 9–12. (2) Another factor is the risk of undermining border security. The U. S. Customs and Border Protection Agency is responsible for preventing the illegal entry of dangerous persons and goods into the United States, and the conduct of their agents positioned at the border has a clear and strong connection to national security. This Court has not extended Bivens where doing so would interfere with the system of military discipline created by statute and regulation, see, e.g., Chappell v. Wallace, 462 U.S. 296, and a similar consideration is applicable to the framework established by the political branches for addressing cases in which it is alleged that lethal force at the border was unlawfully employed by a border agent. Pp. 12–14. (3) Moreover, Congress has repeatedly declined to authorize the award of damages against federal officials for injury inflicted outside U. S. borders. For example, recovery under 42 U. S. C. §1983 is available only to “citizen[s] of the United States or other person[s] within the jurisdiction thereof.” The Federal Tort Claims Act bars “[a]ny claim arising in a foreign country.” 28 U. S. C. §2680(k). And the Torture Victim Protection Act of 1991, note following 28 U. S. C. §1350, cannot be used by an alien to sue a United States officer. When Congress has provided compensation for injuries suffered by aliens outside the United States, it has done so by empowering Executive Branch officials to make payments under circumstances found to be appropriate. See, e.g., Foreign Claims Act, 10 U. S. C. §2734. Congress’s decision not to allow suit in these contexts further indicates that the Judiciary should not create a cause of action that extends across U. S. borders either. Pp. 14–18. (4) These factors can all be condensed to the concern for respecting the separation of powers. The most important question is whether Congress or the courts should create a damages remedy. Here the answer is Congress. Congress’s failure to act does not compel the Court to step into its shoes. Pp. 19–20. 885 F.3d 811, affirmed. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a concurring opinion, in which Gorsuch, J., joined. Ginsburg, J., filed a dissenting opinion, in which Breyer, Sotomayor, and Kagan, JJ., joined.

We are asked in this case to extend Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388 (1971), and create a damages remedy for a cross-border shooting. As we have made clear in many prior cases, however, the Constitution’s separation of powers requires us to exercise caution before extending Bivens to a new “context,” and a claim based on a cross-border shooting arises in a context that is markedly new. Unlike any previously recognized Bivens claim, a cross-border shooting claim has foreign relations and national security implications. In addition, Congress has been notably hesitant to create claims based on allegedly tortious conduct abroad. Because of the distinctive characteristics of cross-border shooting claims, we refuse to extend Bivens into this new field. I The facts of this tragic case are set forth in our earlier opinion in this matter, Hernández v. Mesa, 582 U. S. ___ (2017) (per curiam). Sergio Adrián Hernández Güereca, a 15-year-old Mexican national, was with a group of friends in a concrete culvert that separates El Paso, Texas, from Ciudad Juarez, Mexico. The border runs through the center of the culvert, which was designed to hold the waters of the Rio Grande River but is now largely dry. Border Patrol Agent Jesus Mesa, Jr., detained one of Hernández’s friends who had run onto the United States’ side of the culvert. After Hernández, who was also on the United States’ side, ran back across the culvert onto Mexican soil, Agent Mesa fired two shots at Hernández; one struck and killed him on the other side of the border. Petitioners and Agent Mesa disagree about what Hernández and his friends were doing at the time of shooting. According to petitioners, they were simply playing a game, running across the culvert, touching the fence on the U. S. side, and then running back across the border. According to Agent Mesa, Hernández and his friends were involved in an illegal border crossing attempt, and they pelted him with rocks.[1] The shooting quickly became an international incident, with the United States and Mexico disagreeing about how the matter should be handled. On the United States’ side, the Department of Justice conducted an investigation. When it finished, the Department, while expressing regret over Hernández’s death, concluded that Agent Mesa had not violated Customs and Border Patrol policy or training, and it declined to bring charges or take other action against him. Mexico was not and is not satisfied with the U. S. investigation. It requested that Agent Mesa be extradited to face criminal charges in a Mexican court, a request that the United States has denied. Petitioners, Hernández’s parents, were also dissatisfied and therefore brought suit for damages in the United States District Court for the Western District of Texas. Among other claims, they sought recovery of damages under Bivens, alleging that Mesa violated Hernández’s Fourth and Fifth Amendment rights. The District Court granted Mesa’s motion to dismiss, and the Court of Appeals for the Fifth Circuit sitting en banc has twice affirmed this dismissal. On the first occasion, the court held that Hernández was not entitled to Fourth Amendment protection because he was “a Mexican citizen who had no ‘significant voluntary connection’ to the United States” and “was on Mexican soil at the time he was shot.” Hernandez v. United States, 785 F.3d 117, 119 (CA5 2015) (per curiam). It further concluded that Mesa was entitled to qualified immunity on petitioners’ Fifth Amendment claim. Id., at 120. After granting review, we vacated the Fifth Circuit’s decision and remanded the case, instructing the court “to consider how the reasoning and analysis” of Ziglar v. Abbasi, 582 U. S. ___ (2017), our most recent explication of Bivens, “[might] bear on this case.” Hernández, 582 U. S., at ___ (slip op., at 5). We found it “appropriate for the Court of Appeals, rather than this Court, to address the Bivens question in the first instance.” Ibid. And with the Bivens issue unresolved, we thought it “imprudent” to resolve the “sensitive” question whether the Fourth Amendment applies to a cross-border shooting. Ibid. In addition, while rejecting the ground on which the Court of Appeals had held that Agent Mesa was entitled to qualified immunity, we declined to decide whether he was entitled to qualified immunity on a different ground or whether petitioners’ claim was cognizable under the Fifth Amendment. Id., at ___–___ (slip op., at 5–6). On remand, the en banc Fifth Circuit evaluated petitioners’ case in light of Abbasi and refused to recognize a Bivens claim for a cross-border shooting. 885 F.3d 811 (CA5 2018). The court reasoned that such an incident presents a “ ‘new context’ ” and that multiple factors—including the incident’s relationship to foreign affairs and national security, the extraterritorial aspect of the case, and Congress’s “repeated refusals” to create a damages remedy for injuries incurred on foreign soil––counseled against an extension of Bivens. 885 F. 3d, at 816–823. We granted certiorari, 587 U. S. ___ (2019), and now affirm. II In Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388, the Court broke new ground by holding that a person claiming to be the victim of an unlawful arrest and search could bring a Fourth Amendment claim for damages against the responsible agents even though no federal statute authorized such a claim. The Court subsequently extended Bivens to cover two additional constitutional claims: in Davis v. Passman, 442 U.S. 228 (1979), a former congressional staffer’s Fifth Amendment claim of dismissal based on sex, and in Carlson v. Green, 446 U.S. 14 (1980), a federal prisoner’s Eighth Amendment claim for failure to provide adequate medical treatment. After those decisions, however, the Court changed course. Bivens, Davis, and Carlson were the products of an era when the Court routinely inferred “causes of action” that were “not explicit” in the text of the provision that was allegedly violated. Abbasi, 582 U. S., at ___ (slip op., at 8). As Abbasi recounted: “During this ‘ancien regime,’ . . . the Court assumed it to be a proper judicial function to ‘provide such remedies as are necessary to make effective’ a statute’s purpose . . . . Thus, as a routine matter with respect to statutes, the Court would imply causes of action not explicit in the statutory text itself.” Ibid. (quoting Alexander v. Sandoval, 532 U.S. 275, 287 (2001); J. I. Case Co. v. Borak, 377 U.S. 426, 433 (1964)). Bivens extended this practice to claims based on the Constitution itself. 582 U. S., at ___ (slip op., at 8); Bivens, 403 U. S., at 402 (Harlan, J., concurring in judgment) (Court can infer availability of damages when, “in its view, damages are necessary to effectuate” the “policy underpinning the substantive provisio[n]”). In later years, we came to appreciate more fully the tension between this practice and the Constitution’s separation of legislative and judicial power. The Constitution grants legislative power to Congress; this Court and the lower federal courts, by contrast, have only “judicial Power.” Art. III, §1. But when a court recognizes an implied claim for damages on the ground that doing so furthers the “purpose” of the law, the court risks arrogating legislative power. No law “ ‘pursues its purposes at all costs.’ ” American Express Co. v. Italian Colors Restaurant, 570 U.S. 228, 234 (2013) (quoting Rodriguez v. United States, 480 U.S. 522, 525–526 (1987) (per curiam)). Instead, lawmaking involves balancing interests and often demands compromise. See Board of Governors, FRS v. Dimension Financial Corp., 474 U.S. 361, 373–374 (1986). Thus, a lawmaking body that enacts a provision that creates a right or prohibits specified conduct may not wish to pursue the provision’s purpose to the extent of authorizing private suits for damages. For this reason, finding that a damages remedy is implied by a provision that makes no reference to that remedy may upset the careful balance of interests struck by the lawmakers. See ibid. This problem does not exist when a common-law court, which exercises a degree of lawmaking authority, fleshes out the remedies available for a common-law tort. Analogizing Bivens to the work of a common-law court, petitioners and some of their amici make much of the fact that common-law claims against federal officers for intentional torts were once available. See, e.g., Brief for Petitioners 10–20. But Erie R. Co. v. Tompkins, 304 U.S. 64, 78 (1938), held that “[t]here is no federal general common law,” and therefore federal courts today cannot fashion new claims in the way that they could before 1938. See Alexander, 532 U. S., at 287 (“ ‘Raising up causes of action where a statute has not created them may be a proper function for common-law courts, but not for federal tribunals’ ”). With the demise of federal general common law, a federal court’s authority to recognize a damages remedy must rest at bottom on a statute enacted by Congress, see id., at 286 (“private rights of action to enforce federal law must be created by Congress”), and no statute expressly creates a Bivens remedy. Justice Harlan’s Bivens concurrence argued that this power is inherent in the grant of federal question jurisdiction, see 403 U. S., at 396 (majority opinion); id., at 405 (opinion of Harlan, J.), but our later cases have demanded a clearer manifestation of congressional intent, see Abbasi, 582 U. S., at ___–___ (slip op., at 10–12). In both statutory and constitutional cases, our watchword is caution. For example, in Jesner v. Arab Bank, PLC, 584 U. S. ___, ___–___ (2018) (slip op., at 18–19) we expressed doubt about our authority to recognize any causes of action not expressly created by Congress. See also Abbasi, 582 U. S., at ___ (slip op., at 9) (“If the statute does not itself so provide, a private cause of action will not be created through judicial mandate”). And we declined to recognize a claim against a foreign corporation under the Alien Tort Statute. Jesner, 584 U. S., at ___ (slip op., at 29). In constitutional cases, we have been at least equally reluctant to create new causes of action. We have recognized that Congress is best positioned to evaluate “whether, and the extent to which, monetary and other liabilities should be imposed upon individual officers and employees of the Federal Government” based on constitutional torts. Abbasi, 582 U. S., at ___ (slip op., at 10). We have stated that expansion of Bivens is “a ‘disfavored’ judicial activity,” 582 U. S., at ___ (slip op., at 11) (quoting Ashcroft v. Iqbal, 556 U.S. 662, 675 (2009)), and have gone so far as to observe that if “the Court’s three Bivens cases [had] been . . . decided today,” it is doubtful that we would have reached the same result, 582 U. S., at ___ (slip op., at 11). And for almost 40 years, we have consistently rebuffed requests to add to the claims allowed under Bivens. See 582 U. S., at ___ (slip op., at 23); Minneci v. Pollard, 565 U.S. 118 (2012); Wilkie v. Robbins, 551 U.S. 537 (2007); Correctional Services Corp. v. Malesko, 534 U.S. 61 (2001); FDIC v. Meyer, 510 U.S. 471 (1994); Schweiker v. Chilicky, 487 U.S. 412 (1988); United States v. Stanley, 483 U.S. 669 (1987); Chappell v. Wallace, 462 U.S. 296 (1983); Bush v. Lucas, 462 U.S. 367 (1983). When asked to extend Bivens, we engage in a two-step inquiry. We first inquire whether the request involves a claim that arises in a “new context” or involves a “new category of defendants.” Malesko, 534 U. S., at 68. And our understanding of a “new context” is broad. We regard a context as “new” if it is “different in a meaningful way from previous Bivens cases decided by this Court.” Abbasi, 582 U. S., at ___ (slip op., at 16). When we find that a claim arises in a new context, we proceed to the second step and ask whether there are any “ ‘ “special factors [that] counse[l] hesitation” ’ ” about granting the extension. Id., at ___ (slip op., at 12) (quoting Carlson, 446 U. S., at 18, in turn quoting Bivens, 403 U. S., at 396). If there are––that is, if we have reason to pause before applying Bivens in a new context or to a new class of defendants—we reject the request. We have not attempted to “create an exhaustive list” of factors that may provide a reason not to extend Bivens, but we have explained that “central to [this] analysis” are “separation-of-powers principles.” Abbasi, 582 U. S., at ___ (slip op., at 12). We thus consider the risk of interfering with the authority of the other branches, and we ask whether “there are sound reasons to think Congress might doubt the efficacy or necessity of a damages remedy,” id., at ___ (slip op., at 13), and “whether the Judiciary is well suited, absent congressional action or instruction, to consider and weigh the costs and benefits of allowing a damages action to proceed,” id., at ___ (slip op., at 12). III A The Bivens claims in this case assuredly arise in a new context. Petitioners contend that their Fourth and Fifth Amendment claims do not involve a new context because Bivens and Davis involved claims under those same two amendments, but that argument rests on a basic misunderstanding of what our cases mean by a new context. A claim may arise in a new context even if it is based on the same constitutional provision as a claim in a case in which a damages remedy was previously recognized. Compare Carlson, 446 U. S., at 16–18 (allowing Bivens remedy for an Eighth Amendment claim for failure to provide adequate medical treatment), with Malesko, 534 U. S., at 71–74 (declining to create a Bivens remedy in similar circumstances because the suit was against a private prison operator, not federal officials). And once we look beyond the constitutional provisions invoked in Bivens, Davis, and the present case, it is glaringly obvious that petitioners’ claims involve a new context, i.e., one that is meaningfully different. Bivens concerned an allegedly unconstitutional arrest and search carried out in New York City, 403 U. S., at 389; Davis concerned alleged sex discrimination on Capitol Hill, 442 U. S., at 230. There is a world of difference between those claims and petitioners’ cross-border shooting claims, where “the risk of disruptive intrusion by the Judiciary into the functioning of other branches” is significant. Abbasi, 582 U. S., at ___ (slip op., at 16); see Parts III–B and III–C, infra. Because petitioners assert claims that arise in a new context, we must proceed to the next step and ask whether there are factors that counsel hesitation. As we will explain, there are multiple, related factors that raise warning flags. B The first is the potential effect on foreign relations. “The political branches, not the Judiciary, have the responsibility and institutional capacity to weigh foreign-policy concerns.” Jesner, 584 U. S., at ___ (slip op., at 19). Indeed, we have said that “matters relating ‘to the conduct of foreign relations . . . are so exclusively entrusted to the political branches of government as to be largely immune from judicial inquiry or interference.’ ” Haig v. Agee, 453 U.S. 280, 292 (1981) (quoting Harisiades v. Shaughnessy, 342 U.S. 580, 589 (1952)). “Thus, unless Congress specifically has provided otherwise, courts traditionally have been reluctant to intrude upon the authority of the Executive in [these matters].” Department of Navy v. Egan, 484 U.S. 518, 530 (1988). We must therefore be especially wary before allowing a Bivens remedy that impinges on this arena. A cross-border shooting is by definition an international incident; it involves an event that occurs simultaneously in two countries and affects both countries’ interests. Such an incident may lead to a disagreement between those countries, as happened in this case. The United States, through the Executive Branch, which has “ ‘the lead role in foreign policy,’ ” Medellín v. Texas, 552 U.S. 491, 524 (2008) (alteration omitted), has taken the position that this incident should be handled in a particular way—namely, that Agent Mesa should not face charges in the United States nor be extradited to stand trial in Mexico. As noted, the Executive decided not to take action against Agent Mesa because it found that he “did not act inconsistently with [Border Patrol] policy or training regarding use of force.” DOJ Press Release. We presume that Border Patrol policy and training incorporate both the Executive’s understanding of the Fourth Amendment’s prohibition of unreasonable seizures and the Executive’s assessment of circumstances at the border. Thus, the Executive judged Agent Mesa’s conduct by what it regards as reasonable conduct by an agent under the circumstances that Mesa faced at the time of the shooting, and based on the application of those standards, it declined to prosecute. The Executive does not want a Mexican criminal court to judge Agent Mesa’s conduct by whatever standards would be applicable under Mexican law; nor does it want a jury in a Bivens action to apply its own understanding of what constituted reasonable conduct by a Border Patrol agent under the circumstances of this case. Such a jury determination, the Executive claims, would risk the “ ‘ “embarrassment of our government abroad” through “multifarious pronouncements by various departments on one question.” ’ ” Brief for United States as Amicus Curiae 18 (quoting Sanchez-Espinoza v. Reagan, 770 F.2d 202, 209 (CADC 1985) (Scalia, J.)). The Government of Mexico has taken a different view of what should be done. It has requested that Agent Mesa be extradited for criminal prosecution in a Mexican court under Mexican law, and it has supported petitioners’ Bivens suit. In a brief filed in this Court, Mexico suggests that shootings by Border Patrol agents are a persistent problem and argues that the United States has an obligation under international law, specifically Article 6(1) of the International Covenant on Civil and Political Rights, Dec. 19, 1966, S. Treaty Doc. No. 95–20, 999 U. N. T. S. 174, to provide a remedy for the shooting in this case. Brief for Government of United Mexican States as Amicus Curiae 2, 20–22. Mexico states that it “has a responsibility to look after the well-being of its nationals” and that “it is a priority to Mexico to see that the United States provides adequate means to hold the agents accountable and to compensate the victims.” Id., at 3. Both the United States and Mexico have legitimate and important interests that may be affected by the way in which this matter is handled. The United States has an interest in ensuring that agents assigned the difficult and important task of policing the border are held to standards and judged by procedures that satisfy United States law and do not undermine the agents’ effectiveness and morale. Mexico has an interest in exercising sovereignty over its territory and in protecting and obtaining justice for its nationals. It is not our task to arbitrate between them. In the absence of judicial intervention, the United States and Mexico would attempt to reconcile their interests through diplomacy––and that has occurred. The broad issue of violence along the border, the occurrence of cross- border shootings, and this particular matter have been addressed through diplomatic channels. In 2014, Mexico and the United States established a joint Border Violence Prevention Council, and the two countries have addressed cross-border shootings through the United States-Mexico bilateral Human Rights Dialogue.[2] Following the Justice Department investigation in the present case, the United States reaffirmed its commitment to “work with the Mexican government within existing mechanisms and agreements to prevent future incidents.” DOJ Press Release. For these reasons, petitioners’ assertion that their claims have “nothing to do with the substance or conduct of U. S. foreign . . . policy,” Brief for Petitioners 29, is plainly wrong.[3] C Petitioners are similarly incorrect in deprecating the Fifth Circuit’s conclusion that the issue here implicates an element of national security. One of the ways in which the Executive protects this country is by attempting to control the movement of people and goods across the border, and that is a daunting task. The United States’ border with Mexico extends for 1,900 miles, and every day thousands of persons and a large volume of goods enter this country at ports of entry on the southern border.[4] The lawful passage of people and goods in both directions across the border is beneficial to both countries. Unfortunately, there is also a large volume of illegal cross-border traffic. During the last fiscal year, approximately 850,000 persons were apprehended attempting to enter the United States illegally from Mexico,[5] and large quantities of drugs were smuggled across the border.[6] In addition, powerful criminal organizations operating on both sides of the border present a serious law enforcement problem for both countries.[7] On the United States’ side, the responsibility for attempting to prevent the illegal entry of dangerous persons and goods rests primarily with the U. S. Customs and Border Protection Agency, and one of its main responsibilities is to “detect, respond to, and interdict terrorists, drug smugglers and traffickers, human smugglers and traffickers, and other persons who may undermine the security of the United States.” 6 U. S. C. §211(c)(5). While Border Patrol agents often work miles from the border, some, like Agent Mesa, are stationed right at the border and have the responsibility of attempting to prevent illegal entry. For these reasons, the conduct of agents positioned at the border has a clear and strong connection to national security, as the Fifth Circuit understood. 885 F. 3d, at 819. Petitioners protest that “ ‘shooting people who are just walking down a street in Mexico’ ” does not involve national security, Brief for Petitioners 28, but that misses the point. The question is not whether national security requires such conduct––of course, it does not––but whether the Judiciary should alter the framework established by the political branches for addressing cases in which it is alleged that lethal force was unlawfully employed by an agent at the border. Cf. Abbasi, 582 U. S., at ___ (slip op., at 19) (explaining that “[n]ational-security policy is the prerogative of the Congress and President”). We have declined to extend Bivens where doing so would interfere with the system of military discipline created by statute and regulation, see Chappell, 462 U.S. 296; Stanley, 483 U.S. 669, and a similar consideration is applicable here. Since regulating the conduct of agents at the border unquestionably has national security implications, the risk of undermining border security provides reason to hesitate before extending Bivens into this field. See Abbasi, 582 U. S., at ___ (slip op., at 19) (“Judicial inquiry into the national-security realm raises ‘concerns for the separation of powers’ ” (quoting Christopher v. Harbury, 536 U.S. 403, 417 (2002))). D Our reluctance to take that step is reinforced by our survey of what Congress has done in statutes addressing related matters. We frequently “loo[k] to analogous statutes for guidance on the appropriate boundaries of judge-made causes of action.” Jesner, 584 U. S., at ___ (opinion of Kennedy, J.) (slip op., at 19). When foreign relations are implicated, it “is even more important . . . ‘to look for legislative guidance before exercising innovative authority over substantive law.’ ” Id., at ___ (slip op., at 20) (quoting Sosa v. Alvarez-Machain, 542 U.S. 692, 726 (2004)). Accordingly, it is “telling,” Abbasi, 582 U. S., at ___ (slip op., at 20), that Congress has repeatedly declined to authorize the award of damages for injury inflicted outside our borders. A leading example is 42 U. S. C. §1983, which permits the recovery of damages for constitutional violations by officers acting under color of state law. We have described Bivens as a “more limited” “federal analog” to §1983. Hartman v. Moore, 547 U.S. 250, 254, n. 2 (2006). It is therefore instructive that Congress chose to make §1983 available only to “citizen[s] of the United States or other person[s] within the jurisdiction thereof.” It would be “anomalous to impute . . . a judicially implied cause of action beyond the bounds [Congress has] delineated for [a] comparable express caus[e] of action.” Blue Chip Stamps v. Manor Drug Stores, 421 U.S. 723, 736 (1975). Thus, the limited scope of §1983 weighs against recognition of the Bivens claim at issue here. Section 1983’s express limitation to the claims brought by citizens and persons subject to United States jurisdiction is especially significant, but even if this explicit limitation were lacking, we would presume that §1983 did not apply abroad. See RJR Nabisco, Inc. v. European Community, 579 U. S. ___, ___ (2016) (slip op., at 7) (“Absent clearly expressed congressional intent to the contrary, federal laws will be construed to have only domestic application”). We presume that statutes do not apply extraterritorially to “ensure that the Judiciary does not erroneously adopt an interpretation of U. S. law that carries foreign policy consequences not clearly intended by the political branches.” Kiobel v. Royal Dutch Petroleum Co., 569 U.S. 108, 116 (2013); see also EEOC v. Arabian American Oil Co., 499 U.S. 244, 248 (1991). If this danger provides a reason for caution when Congress has enacted a statute but has not provided expressly whether it applies abroad, we have even greater reason for hesitation in deciding whether to extend a judge-made cause of action beyond our borders. “[T]he danger of unwarranted judicial interference in the conduct of foreign policy is magnified” where “the question is not what Congress has done but instead what courts may do.” Kiobel, 569 U. S., at 116. Where Congress has not spoken at all, the likelihood of impinging on its foreign affairs authority is especially acute. Congress’s treatment of ordinary tort claims against federal officers is also revealing. As petitioners and their amici stress, the traditional way in which civil litigation addressed abusive conduct by federal officers was by subjecting them to liability for common-law torts. See Brief for Petitioners 10–17. For many years, such claims could be raised in state or federal court,[8] and this Court occasionally considered tort suits against federal officers for extraterritorial injuries. See, e.g., Mitchell v. Harmony, 13 How. 115 (1852) (affirming award in trespass suit brought by U. S. citizen against U. S. Army officer who seized personal property in Mexico during the Mexican-American war). After Erie, federal common-law claims were out, but we recognized the continuing viability of state-law tort suits against federal officials as recently as Westfall v. Erwin, 484 U.S. 292 (1988). In response to that decision, Congress passed the so-called Westfall Act, formally the Federal Employees Liability Reform and Tort Compensation Act of 1988, 28 U. S. C. §2679. That Act makes the Federal Tort Claims Act (FTCA) “the exclusive remedy for most claims against Government employees arising out of their official conduct.” Hui v. Castaneda, 559 U.S. 799, 806 (2010).[9] Thus, a person injured by a federal employee may seek recovery directly from the United States under the FTCA, but the FTCA bars “[a]ny claim arising in a foreign country.” §2680(k).[10] The upshot is that claims that would otherwise permit the recovery of damages are barred if the injury occurred abroad. Yet another example is provided by the Torture Victim Protection Act of 1991, note following 28 U. S. C. §1350, which created a cause of action that may be brought by an alien in a U. S. court under the Alien Tort Statute, §1350. Under the Torture Victim Protection Act, a damages action may be brought by or on behalf of a victim of torture or an extrajudicial killing carried out by a person who acted under the authority of a foreign state. Consequently, this provision, which is often employed to seek redress for acts committed abroad,[11] cannot be used to sue a United States officer. See Meshal v. Higgenbotham, 804 F.3d 417, 430 (CADC 2015) (Kavanaugh, J., concurring). These statutes form a pattern that is important for present purposes. When Congress has enacted statutes creating a damages remedy for persons injured by United States Government officers, it has taken care to preclude claims for injuries that occurred abroad. Instead, when Congress has provided compensation for injuries suffered by aliens outside the United States, it has done so by empowering Executive Branch officials to make payments under circumstances found to be appropriate. Thus, the Foreign Claims Act, 10 U. S. C. §2734, first enacted during World War II, ch. 645, 55Stat. 880, allows the Secretary of Defense to appoint claims commissions to settle and pay claims for personal injury and property damage resulting from the noncombat activities of the Armed Forces outside this country. §2734(a). Similarly, §2734a allows the Secretary of Defense and the Secretary of Homeland Security to make payments pursuant to “an international agreement which provides for the settlement or adjudication and cost sharing of claims against the United States” that arise out of “acts or omissions” of the Armed Forces. §2734a(a); see also 22 U. S. C. §2669(b) (State Department may settle and pay certain claims for death, injury, or property loss or damage “for the purpose of promoting and maintaining friendly relations with foreign countries”); §2669–1 (Secretary of State has authority to pay tort claims arising in foreign countries in connection with State Department operations); 21 U. S. C. §904 (Attorney General has authority to pay tort claims arising in connection with the operations of the Drug Enforcement Administration abroad). This pattern of congressional action—refraining from authorizing damages actions for injury inflicted abroad by Government officers, while providing alternative avenues for compensation in some situations—gives us further reason to hesitate about extending Bivens in this case. E In sum, this case features multiple factors that counsel hesitation about extending Bivens, but they can all be condensed to one concern––respect for the separation of powers. See Abbasi, 582 U. S., at ___ (slip op., at 12). “Foreign policy and national security decisions are ‘delicate, complex, and involve large elements of prophecy’ for which ‘the Judiciary has neither aptitude, facilities[,] nor responsibility.’ ” Jesner, 584 U. S., at ___ (Gorsuch, J., concurring part and concurring in judgment) (slip op., at 5) (quoting Chicago & Southern Air Lines, Inc. v. Waterman S. S. Corp., 333 U.S. 103, 111 (1948)). To avoid upsetting the delicate web of international relations, we typically presume that even congressionally crafted causes of action do not apply outside our borders. These concerns are only heightened when judges are asked to fashion constitutional remedies. Congress, which has authority in the field of foreign affairs, has chosen not to create liability in similar statutes, leaving the resolution of extraterritorial claims brought by foreign nationals to executive officials and the diplomatic process. Congress’s decision not to provide a judicial remedy does not compel us to step into its shoes. “The absence of statutory relief for a constitutional violation . . . does not by any means necessarily imply that courts should award money damages against the officers responsible for the violation.” Schweiker, 487 U. S., at 421–422; see also Stanley, 483 U. S., at 683 (“[I]t is irrelevant to a ‘special factors’ analysis whether the laws currently on the books afford [plaintiff] an ‘adequate’ federal remedy for his injuries”).[12] When evaluating whether to extend Bivens, the most important question “is ‘who should decide’ whether to provide for a damages remedy, Congress or the courts?” Abbasi, 582 U. S., at ___ (slip op., at 12) (quoting Bush, 462 U. S., at 380). The correct “answer most often will be Congress.” 582 U. S., at ___ (slip op., at 12). That is undoubtedly the answer here. * * * The judgment of the United States Court of Appeals for the Fifth Circuit is affirmed. It is so ordered. Notes 1 See App. to Pet. for Cert. 198–199; Dept. of Justice, Federal Officials Close Investigation Into the Death of Sergio Hernandez-Guereca(Apr. 27, 2012), https://www.justice.gov/opa/pr/federal-officials-close-investigation-death-sergio-hernandez-guereca (hereinafter DOJ Press Release). 2 See Dept. of Homeland Security, Written Testimony for HouseComm. on Oversight and Govt. Reform Hearing (Sept. 9, 2015), https: / / www .dhs.gov/news/2015/09/09/written-testimony-dhs-southern-border-and-approaches-campaign-joint-task-force-west (discussing creation of Border Violence Prevention Council); Dept. of Homeland Security, Border Violence Prevention Council Fact Sheet, https://www.dhs.gov/sites/default/files/publications/bvpc-fact-sheet.pdf (outlining areas of collaboration); Dept. of State, Joint Statement on the U. S.-Mexico Bilateral High Level Dialogue on Human Rights (Oct. 27, 2016), https://2009-2017.state.gov/r/pa/prs/ps/2016/10/263759.htm (noting discussion of “the use of force at the border”). 3 It is no answer to argue, as Mexico does, that refusing to extend Bivens “is what [would] negatively affect international relations.” Brief for Government of United Mexican States as Amicus Curiae 12. When a third party intervenes and takes sides in a dispute between two countries, one country is likely to be pleased and the other displeased. But no matter which side the third party supports, it will have injected itself into their relations. 4 See Dept. of Transp., Bureau of Transp. Statistics, Border Crossing/Entry Data, https://explore.dot.gov/views/BorderCrossingData/Monthly (detailing the millions of individuals and vehicles that cross theU. S.-Mexico border each month); U. S. Int’l Trade Comm’n, The Year in Trade 2018, p. 190 (USITC Pub. No. 4986, 2019) (explaining that in 2018 the United States imported $346.5 billion of goods from Mexico). 5 Dept. of Homeland Security, U. S. Customs and Border Protection, Southwest Border Migration FY 2019, https://cbp.gov/newsroom/stats/sw-border-migration/fy-2019. 6 Dept. of Homeland Security, U. S. Customs and Border Protection, CBP Enforcement Statistics FY2019, https://cbp.gov/newsroom/stats/cbp-enforcement-statistics-fy2019 (explaining that in FY2019, Border Patrol officers seized 11,682 pounds of cocaine, 266,882 pounds of marijuana, and 14,434 pounds of methamphetamine). 7 Cong. Research Serv., Mexico: Organized Crime and Drug Trafficking Organizations, Summary (2019) (“Mexican drug trafficking organizations . . . pose the greatest crime threat to the United States”); Dept. of Justice, Drug Enforcement Admin., 2018 National Drug Threat Assessment 97 (DEA–DCT–DIR–032–18) (explaining that “Mexican [transnational criminal organizations] . . . maintain the greatest drug trafficking influence in the United States”). 8 State-law claims could be asserted in federal court if the parties’ citizenship was diverse, and federal common-law claims could be raised until Erie R. Co. v. Tompkins, 304 U.S. 64 (1938). 9 The Act also permits claims “brought for a violation of the Constitution.” 28 U. S. C. §2679(b)(2)(A). By enacting this provision, Congress made clear that it was not attempting to abrogate Bivens, but the provi-sion certainly does not suggest, as one of petitioners’ amici contends, thatCongress “intended for a robust enforcement of Bivens remedies.” Brief for Institute for Justice as Amicus Curiae 21. Instead, the provision simply left Bivens where it found it. It is not a license to create a new Bivens remedy in a context we have never before addressed, see Correctional Services Corp. v. Malesko, 534 U.S. 61, 68 (2001). 10 Petitioners contend that Congress excluded claims arising abroad in order to avoid subjecting the United States to liability under foreign law, something that cannot occur under Bivens. Reply Brief 11. But neither the legislative history recounted in Sosa v. Alvarez-Machain, 542 U.S. 692, 707 (2004), nor anything else offered by petitioners shows that this was the only reason for this limitation. And the fact remains that the FTCA does not permit claims for torts committed abroad, a limitation that is consistent with Congress’s general practice of avoiding extraterritorial legislation. See, e.g., Kiobel v. Royal Dutch Petroleum Co., 569 U.S. 108, 115–116 (2013). 11 See, e.g., Samantar v. Yousuf, 560 U.S. 305, 308 (2010) (bringing claim under the Torture Victim Protection Act against the former First Vice President and Minister of Defense of Somalia for alleged torture and extrajudicial killing in Somalia). 12 Indeed, in Abbasi we explained that existence of alternative remedies was merely a further reason not to create Bivens liability. See 582 U. S., at ___ (slip op., at 22) (“[W]hen alternative methods of relief are available, a Bivens remedy is usually not”).

589.US.2019_18-7739

A criminal defendant who wants to “preserve a claim of error” for appellate review must first inform the trial judge “of [1] the action the party wishes the court to take, or [2] the party’s objection to the court’s action and the grounds for that objection.” Fed. Rule Crim. Proc. 51(b). Petitioner Holguin-Hernandez was convicted on drug charges and sentenced to 60 months in prison and five years of supervised release while he was still serving a term of supervised release for an earlier conviction. The Government asked the District Court to impose an additional consecutive prison term of 12 to 18 months for violating the conditions of the earlier term. Petitioner countered that 18 U. S. C. §3553’s sentencing factors either did not support imposing any additional time or supported a sentence of less than 12 months. The court nonetheless imposed a consecutive 12-month term. Petitioner argued on appeal that this sentence was unreasonably long because it was “ ‘greater than necessar[y]’ to accomplish the goals of sentencing,” Kimbrough v. United States, 552 U.S. 85, 101, but the Fifth Circuit held that he had forfeited that argument by failing to object to the reasonableness of the sentence in the District Court. Held: Petitioner’s district-court argument for a specific sentence (nothing or less than 12 months) preserved his claim on appeal that the sentence imposed was unreasonably long. A party who informs the court of the “action” he “wishes the court to take,” Rule 51(b), ordinarily brings to the court’s attention his objection to a contrary decision. That is certainly true where, as here, the defendant advocates for a sentence shorter than the one actually imposed. Judges, having in mind their “overarching duty” under §3553(a) “to ‘impose a sentence sufficient, but not greater than necessary,’ to serve the purposes of sentencing,” would ordinarily understand that a defendant in that circumstance was making the argument that the shorter sentence would be “ ‘sufficient’ ” and a longer sentence “ ‘greater than necessary.’ ” Pepper v. United States, 562 U.S. 476, 493 (quoting §3553(a)). Nothing more is needed to preserve a claim that a longer sentence is unreasonable. Defendants need not also refer to the “reasonableness” of a sentence. Rule 51 abolished the requirement of making formal “exceptions” to a district court’s decision. And, in any event, reasonableness pertains to the standard of “appellate review” of a trial court’s sentencing decision, Gall v. United States, 552 U.S. 38, 46 (emphasis added); it is not the substantive standard that trial courts apply under §3553(a). A defendant who, by advocating for a particular sentence, communicates to the trial judge his view that a longer sentence is “greater than necessary” has thereby informed the court of the legal error at issue in an appellate challenge to the substantive reasonableness of the sentence. Other issues raised by the Government and amicus are not addressed here because they were not considered by the Fifth Circuit. Pp. 4–6. 746 Fed. Appx. 403, vacated and remanded. Breyer, J., delivered the opinion for a unanimous Court. Alito, J., filed a concurring opinion, in which Gorsuch, J., joined.

A criminal defendant who wishes a court of appeals to consider a claim that a ruling of a trial court was in error must first make his objection known to the trial-court judge. The Federal Rules of Criminal Procedure provide two ways of doing so. They say that “[a] party may preserve a claim of error by informing the court . . . of [1] the action the party wishes the court to take, or [2] the party’s objection to the court’s action and the grounds for that objection.” Fed. Rule Crim. Proc. 51(b). Errors “not brought to the court’s attention” in one of these two ways are subject to review only insofar as they are “plain.” Rule 52(b); see United States v. Olano, 507 U.S. 725, 732–736 (1993). In this case, a criminal defendant argued in the District Court that the sentencing factors set forth in 18 U. S. C. §3553(a) did not support imposing any prison time for a supervised-release violation. At the very least, the defendant contended, any term of imprisonment should be less than 12 months long. The judge nevertheless imposed a sentence of 12 months. The question is whether the defendant’s district-court argument for a specific sentence (namely, nothing or less than 12 months) preserved his claim on appeal that the 12-month sentence was unreasonably long. We think that it did. I The petitioner in this case, Gonzalo Holguin-Hernandez, was convicted of drug trafficking and sentenced to 60 months in prison and five years of supervised release. At the time of his conviction, he was also serving a term of supervised release related to an earlier crime. The Government asked the court to find that petitioner had violated the conditions of that earlier term, to revoke it, and to impose an additional consecutive prison term consistent with the pertinent Sentencing Guidelines, namely, 12 to 18 months in prison. See United States Sentencing Commission, Guidelines Manual §§7B1.4(a), 7B1.3(f ) (Nov. 2018). Petitioner’s counsel argued that there “would be no reason under [18 U. S. C. §]3553 that an additional consecutive sentence would get [petitioner’s] attention any better than” the five years in prison the court had already imposed for the current trafficking offense. App. 10. She added that the petitioner understood that, if he offended again, he was “going to serve his life in prison.” Ibid. And she urged the court to impose either “no additional time or certainly less than the [G]uidelines.” Ibid. At the least, she said, the court should “depart” from the Guidelines, imposing a sentence “below” the applicable range “because it is a substantial sentence and to me over represents the role that he played in” the underlying offense. Ibid. The court then imposed a consecutive term of 12 months, a sentence at the bottom of, but not below, the Guidelines range. See id., at 11. The judge indicated that he did not disagree with counsel’s argument, but thought that circumstances justified a greater sentence. He asked counsel if there was “[a]nything further.” Ibid. Counsel said that there was not. See ibid. Petitioner appealed, arguing that the 12-month sentence was unreasonably long in that it was “ ‘greater than necessar[y]’ to accomplish the goals of sentencing.” Kimbrough v. United States, 552 U.S. 85, 101 (2007) (quoting 18 U. S. C. §3553(a)); see also, e.g., Gall v. United States, 552 U.S. 38, 49–50 (2007) (noting the District Court’s obligation to “consider all of the §3553(a) factors to determine” the “appropriate sentence”); 18 U. S. C. §3583(e) (making these factors applicable in substantial part to proceedings to revoke or modify a term of supervised release). The Court of Appeals held that petitioner had forfeited this argument by failing to “object in the district court to the reasonableness of the sentence imposed.” 746 Fed. Appx. 403 (CA5 2018) (per curiam). The court would, of course, consider whether the error petitioner asserted was “plain.” See ibid.; Rule 52(b) (permitting review of a plain error “even though it was not brought to the court’s attention”). But it found no plain error, and so it affirmed. Petitioner sought review in this Court and, in light of differences among the Courts of Appeals, we granted his petition for certiorari. Compare 746 Fed. Appx. 403 with, e.g., United States v. Curry, 461 F.3d 452, 459 (CA4 2006); United States v. Vonner, 516 F.3d 382, 389 (CA6 2008) (en banc); United States v. Castro-Juarez, 425 F.3d 430, 433–434 (CA7 2005); United States v. Sullivan, 327 Fed. Appx. 643, 645 (CA7 2009); United States v. Autery, 555 F.3d 864, 868–871 (CA9 2009); United States v. Torres-Duenas, 461 F.3d 1178, 1183 (CA10 2006); United States v. Gonzalez-Mendez, 545 Fed. Appx. 848, 849, and n. 1 (CA11 2013); United States v. Bras, 483 F.3d 103, 113 (CADC 2007). Because the Government agrees with petitioner that the Fifth Circuit’s approach is inconsistent with the Federal Rules of Criminal Procedure, we appointed K. Winn Allen to defend the judgment below as amicus curiae. He has ably discharged his responsibilities. II Congress has instructed sentencing courts to impose sentences that are “ ‘sufficient, but not greater than necessary, to comply with’ ” (among other things) certain basic objectives, including the need for “just punishment, deterrence, protection of the public, and rehabilitation.” Dean v. United States, 581 U. S. ___, ___ (2017) (slip op., at 4) (quoting 18 U. S. C. §3553(a)(2); emphasis added); see Pepper v. United States, 562 U.S. 476, 491, 493 (2011). If the trial court follows proper procedures and gives adequate consideration to these and the other listed factors, then the question for an appellate court is simply, as here, whether the trial court’s chosen sentence was “reasonable” or whether the judge instead “abused his discretion in determining that the §3553(a) factors supported” the sentence imposed. Gall, 552 U. S., at 56; see United States v. Booker, 543 U.S. 220, 261–262 (2005). By “informing the court” of the “action” he “wishes the court to take,” Fed. Rule Crim. Proc. 51(b), a party ordinarily brings to the court’s attention his objection to a contrary decision. See Rule 52(b). And that is certainly true in cases such as this one, where a criminal defendant advocates for a sentence shorter than the one ultimately imposed. Judges, having in mind their “overarching duty” under §3553(a), would ordinarily understand that a defendant in that circumstance was making the argument (to put it in statutory terms) that the shorter sentence would be “ ‘sufficient’ ” and a longer sentence “ ‘greater than necessary’ ” to achieve the purposes of sentencing. Pepper, 562 U. S., at 493 (quoting §3553(a)). Nothing more is needed to preserve the claim that a longer sentence is unreasonable. We do not agree with the Court of Appeals’ suggestion that defendants are required to refer to the “reasonableness” of a sentence to preserve such claims for appeal. See 746 Fed. Appx. 403; United States v. Peltier, 505 F.3d 389, 391 (CA5 2007). The rulemakers, in promulgating Rule 51, intended to dispense with the need for formal “exceptions” to a trial court’s rulings. Rule 51(a); see also Advisory Committee’s 1944 Notes on Fed. Rule Crim. Proc. 51, 18 U. S. C. App., p. 591. They chose not to require an objecting party to use any particular language or even to wait until the court issues its ruling. Rule 51(b) (a party may “infor[m] the court” of its position either “when the court ruling or order is made or” when it is “sought”). The question is simply whether the claimed error was “brought to the court’s attention.” Rule 52(b). Here, it was. The Court of Appeals properly noted that, to win on appeal, a defendant making such a claim must show that the trial court’s decision was not “reasonable.” Gall, 552 U. S., at 56. But that fact is not relevant to the issue here. Our decisions make plain that reasonableness is the label we have given to “the familiar abuse-of-discretion standard” that “applies to appellate review” of the trial court’s sentencing decision. Id., at 46 (emphasis added); see Kimbrough, 552 U. S., at 90–91; Rita v. United States, 551 U.S. 338, 351 (2007); Booker, 543 U. S., at 261. The substantive standard that Congress has prescribed for trial courts is the “parsimony principle” enshrined in §3553(a). Dean, 581 U. S., at ___ (slip op., at 4); see Pepper, 562 U. S., at 491. A defendant who, by advocating for a particular sentence, communicates to the trial judge his view that a longer sentence is “greater than necessary” has thereby informed the court of the legal error at issue in an appellate challenge to the substantive reasonableness of the sentence. He need not also refer to the standard of review. III The Government and amicus raise other issues. They ask us to decide what is sufficient to preserve a claim that a trial court used improper procedures in arriving at its chosen sentence. And they ask us to decide when a party has properly preserved the right to make particular arguments supporting its claim that a sentence is unreasonably long. We shall not consider these matters, however, for the Court of Appeals has not considered them. See, e.g., Tapia v. United States, 564 U.S. 319, 335 (2011); Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7 (2005). We hold only that the defendant here properly preserved the claim that his 12-month sentence was unreasonably long by advocating for a shorter sentence and thereby arguing, in effect, that this shorter sentence would have proved “sufficient,” while a sentence of 12 months or longer would be “greater than necessary” to “comply with” the statutory purposes of punishment. 18 U. S. C. §3553(a). The judgment of the Court of Appeals is vacated, and the case is remanded for further proceedings consistent with this opinion. It is so ordered.

589.US.2019_18-1116

The Employee Retirement Income Security Act of 1974 (ERISA) requires plaintiffs with “actual knowledge” of an alleged fiduciary breach to file suit within three years of gaining that knowledge, 29 U. S. C. §1113(2), rather than within the 6-year period that would otherwise apply. Respondent Sulyma worked at Intel Corporation from 2010 to 2012 and participated in two Intel retirement plans. In October 2015, he sued petitioners—administrators of those plans—alleging that they had managed the plans imprudently. Petitioners countered that the suit was untimely under §1113(2) because Sulyma filed it more than three years after they had disclosed their investment decisions to him. Al- though Sulyma had visited the website that hosted many of these disclosures many times, he testified that he did not remember reviewing the relevant disclosures and that he had been unaware of the allegedly imprudent investments while working at Intel. The District Court granted summary judgment to petitioners under §1113(2). The Ninth Circuit reversed. That court agreed with petitioners that Sulyma could have known about the investments from the disclosures, but held that his testimony created a dispute as to when he gained “actual knowledge” for purposes of §1113(2). Held: A plaintiff does not necessarily have “actual knowledge” under §1113(2) of the information contained in disclosures that he receives but does not read or cannot recall reading. To meet §1113(2)’s “actual knowledge” requirement, the plaintiff must in fact have become aware of that information. Pp. 5–12. (a) ERISA’s “plain and unambiguous statutory language” must be enforced “according to its terms.” Hardt v. Reliance Standard Life Ins. Co., 560 U.S. 242, 251. Although ERISA does not define the phrase “actual knowledge,” its meaning is plain. Dictionaries confirm that, to have “actual knowledge” of a piece of information, one must in fact be aware of it. Legal dictionaries give “actual knowledge” the same meaning. The law will sometimes impute knowledge—often called “constructive” knowledge—to a person who fails to learn something that a reasonably diligent person would have learned. The addition of “actual” in §1113(2) signals that the plaintiff’s knowledge must be more than hypothetical. Congress has repeatedly drawn the same “linguistic distinction,” Merck & Co. v. Reynolds, 559 U.S. 633, 647, elsewhere in ERISA. When Congress has included both actual and constructive knowledge in ERISA limitations provisions, Congress has done so explicitly. But Congress has never added to §1113(2) the language it has used in those other provisions to encompass both forms of knowledge. Pp. 5–8. (b) Petitioners’ arguments for a broader reading of §1113(2) based on text, context, purpose, and statutory history all founder on Congress’s choice of the word “actual.” Petitioners may well be correct that heeding the plain meaning of §1113(2) substantially diminishes the protection that it provides for ERISA fiduciaries. But if policy considerations suggest that the current scheme should be altered, Congress must be the one to do it. Pp. 8–11. (c) This opinion does not foreclose any of the “usual ways” to prove actual knowledge at any stage in the litigation. Farmer v. Brennan, 511 U.S. 825, 842. Plaintiffs who recall reading particular disclosures will be bound by oath to say so in their depositions. Actual knowledge can also be proved through “inference from circumstantial evidence.” Ibid. And this opinion does not preclude defendants from contending that evidence of “willful blindness” supports a finding of “actual knowledge.” Cf. Global-Tech Appliances, Inc. v. SEB S. A., 563 U.S. 754, 769. Pp. 11–12. 909 F.3d 1069, affirmed. Alito, J., delivered the opinion for a unanimous Court.

The Employee Retirement Income Security Act of 1974 (ERISA) requires plaintiffs with “actual knowledge” of an alleged fiduciary breach to file suit within three years of gaining that knowledge rather than within the 6-year period that would otherwise apply. §413(a)(2)(A), 88Stat. 889, as amended, 29 U. S. C. §1113. The question here is whether a plaintiff necessarily has “actual knowledge” of the information contained in disclosures that he receives but does not read or cannot recall reading. We hold that he does not and therefore affirm. I A Retirement plans governed by ERISA must have at least one named fiduciary, §1102(a)(1), who must manage the plan prudently and solely in the interests of participants and their beneficiaries, §1104(a). Fiduciaries who breach these duties are personally liable to the plan for any resulting losses. §1109(a). ERISA authorizes participants and their beneficiaries, as well as co-fiduciaries and the Secretary of Labor, to sue for that relief. §1132(a)(2). Such suits must be filed within one of three time periods, each with different triggering events. The first begins when the breach occurs. Specifically, under §1113(1), suit must be filed within six years of “the date of the last action which constituted a part of the breach or violation” or, in cases of breach by omission, “the latest date on which the fiduciary could have cured the breach or violation.” We have referred to §1113(1) as a statute of repose, which “effect[s] a legislative judgment that a defendant should be free from liability after the legislatively determined period of time.” California Public Employees’ Retirement System v. ANZ Securities, Inc., 582 U. S. ___, ___ (2017) (slip op., at 5) (internal quotation marks omitted). The second period, which accelerates the filing deadline, begins when the plaintiff gains “actual knowledge” of the breach. Under §1113(2), suit must be filed within three years of “the earliest date on which the plaintiff had actual knowledge of the breach or violation.” Section 1113(2) is a statute of limitations, which “encourage[s] plaintiffs to pursue diligent prosecution of known claims.” Id., at ___ (slip op., at 5) (internal quotation marks omitted). The third period, which applies “in the case of fraud or concealment,” begins when the plaintiff discovers the alleged breach. §1113. In such cases, suit must be filed within six years of “the date of discovery.” Ibid. B Respondent Sulyma worked at Intel Corporation from 2010 to 2012. He participated in two Intel retirement plans, the Intel Retirement Contribution Plan and the Intel 401(k) Savings Plan. Payments into these plans were in turn invested in two funds managed by the Intel Investment Policy Committee.[1] These funds mostly comprised stocks and bonds. After the stock market decline in 2008, however, the committee increased the funds’ shares of alternative assets, such as hedge funds, private equity, and commodities. These assets carried relatively high fees. And as the stock market rebounded, Sulyma’s funds lagged behind others such as index funds. Sulyma filed this suit on behalf of a putative class in October 2015, alleging primarily that the committee and other plan administrators (petitioners here) had breached their fiduciary duties by overinvesting in alternative assets. Petitioners countered that the suit was untimely under §1113(2). Although Sulyma filed it within six years of the alleged breaches, he filed it more than three years after petitioners had disclosed their investment decisions to him. ERISA and its implementing regulations mandate various disclosures to plan participants. See generally 29 U. S. C. §§1021–1031; see also Gobeille v. Liberty Mut. Ins. Co., 577 U. S. ___, ___–___ (2016). Sulyma received numerous disclosures while working at Intel, some explaining the extent to which his retirement plans were invested in alternative assets. In November 2011, for example, he received an e-mail informing him that a Qualified Default Investment Alternative (QDIA) notice was available on a website called NetBenefits, where many of his disclosures were hosted. See App. 149–151; see also 29 CFR §§2550.404c–5(b)–(d) (2019) (QDIA notices); §2520.104b–1(c) (regulating electronic disclosure). This notice broke down the percentages at which his 401(k) fund was invested in stocks, bonds, hedge funds, and commodities. See App. 236. In 2012, he received a summary plan description explaining that the funds were invested in stocks and alternative assets, id., at 227, and referring him to other documents—called fund fact sheets—with the percentages in graphical form. See 29 U. S. C. §§1022, 1024(b) (summary plan descriptions); see also App. 307 (June 2012 fact sheet for his 401(k) plan fund); id., at 338 (June 2012 fact sheet for his retirement contribution plan fund); id., at 277–340 (other fact sheets provided during his tenure at Intel). Also in 2012, he received e-mails directing him to annual disclosures that petitioners provided for both his plans, which showed the underlying funds’ return rates and again directed him to the NetBenefits site for further information. See 29 CFR §2550.404a–5; see also App. 242–243 (retirement contribution plan annual disclosure); id., at 250–251 (401(k) plan annual disclosure). Petitioners submitted records showing that Sulyma visited the NetBenefits site repeatedly during his employment. Id., at 258–276. But he testified in his deposition that he did not “remember reviewing” the above disclosures during his tenure. Id., at 175; see also id., at 183, 193, 196–197. He also stated in a declaration that he was “unaware” while working at Intel “that the monies that [he] had invested through the Intel retirement plans had been invested in hedge funds or private equity.” Id., at 212. He recalled reviewing only account statements sent to him by mail, which directed him to the NetBenefits site and noted that his plans were invested in “short-term/other” assets but did not specify which. See, e.g., id., at 375. The District Court granted summary judgment to petitioners under §1113(2), reasoning that “[i]t would be improper to allow Sulyma’s claims to survive merely because he did not look further into the disclosures made to him.” 2017 WL 1217185, *9 (ND Cal., Mar. 31, 2017). The Ninth Circuit reversed. As relevant here,[2] the court construed “actual knowledge” to mean “what it says: knowledge that is actual, not merely a possible inference from ambiguous circumstances.” 909 F.3d 1069, 1076 (2018) (internal quotation marks omitted). Although Sulyma “had sufficient information available to him to know about the allegedly imprudent investments” more than three years before filing suit, the court held that his testimony created a dispute as to when he actually gained that knowledge. Id., at 1077. Several Circuits have likewise construed §1113(2) to require “knowledge that is actual,” id., at 1076, but one has construed it to require only proof of sufficient disclosure.[3] We granted certiorari, 587 U. S. ___ (2019), to resolve whether the phrase “actual knowledge” does in fact mean “what it says,” 909 F. 3d, at 1076, and hold that it does. II A “We must enforce plain and unambiguous statutory language” in ERISA, as in any statute, “according to its terms.” Hardt v. Reliance Standard Life Ins. Co., 560 U.S. 242, 251 (2010). Although ERISA does not define the phrase “actual knowledge,” its meaning is plain. Dictionaries are hardly necessary to confirm the point, but they do. When Congress passed ERISA, the word “actual” meant what it means today: “existing in fact or reality.” Webster’s Seventh New Collegiate Dictionary 10 (1967); accord, Merriam-Webster’s Collegiate Dictionary 13 (11th ed. 2005) (same); see also American Heritage Dictionary 14 (1973) (“In existence; real; factual”); id., at 18 (5th ed. 2011) (“Existing in reality and not potential, possible, simulated, or false”). So did the word “knowledge,” which meant and still means “the fact or condition of being aware of something.” Webster’s Seventh New Collegiate Dictionary 469 (1967); accord, Merriam-Webster’s Collegiate Dictionary 691 (2005) (same); see also American Heritage Dictionary 725 (1973) (“Familiarity, awareness, or understanding gained through experience or study”); id., at 973 (2011) (same). Thus, to have “actual knowledge” of a piece of information, one must in fact be aware of it. Legal dictionaries give “actual knowledge” the same meaning: “[r]eal knowledge as distinguished from presumed knowledge or knowledge imputed to one.” Ballentine’s Law Dictionary 24 (3d ed. 1969); accord, Black’s Law Dictionary 1043 (11th ed. 2019) (defining “actual knowledge” as “[d]irect and clear knowledge, as distinguished from constructive knowledge”).[4] The qualifier “actual” creates that distinction. In everyday speech, “actual knowledge” might seem redundant; one who claims “knowledge” of a topic likely means to suggest that he actually knows a thing or two about it. But the law will sometimes impute knowledge—often called “constructive” knowledge—to a person who fails to learn something that a reasonably diligent person would have learned. See id., at 1043. Similarly, we held in Merck & Co. v. Reynolds, 559 U.S. 633 (2010), that the word “discovery,” when used in a statute of limitations without qualification, “encompasses not only those facts the plaintiff actually knew, but also those facts a reasonably diligent plaintiff would have known.” Id., at 648. The addition of “actual” in §1113(2) signals that the plaintiff ’s knowledge must be more than “potential, possible, virtual, conceivable, theoretical, hypothetical, or nominal.” Black’s Law Dictionary 53 (4th ed. 1951). Indeed, in Merck, we cited §1113(2) as evidence of the “linguistic distinction” between “ ‘actual knowledge’ ” and the “hypothetical” knowledge that a reasonably diligent plaintiff would have. 559 U. S., at 646–647 (quoting §1113(2); emphasis in original). Congress has drawn the same distinction elsewhere in ERISA. Multiple provisions contain alternate 6-year and 3-year limitations periods, with the 6-year period beginning at “the date on which the cause of action arose” and the 3-year period starting at “the earliest date on which the plaintiff acquired or should have acquired actual knowledge of the existence of such cause of action.” §§1303(e)(6), (f )(5) (emphasis added); accord, §§1370(f )(1)–(2), 1451(f )(1)–(2). ERISA also requires plaintiffs challenging the suspension of benefits under §1085 to do so within “one year after the earliest date on which the plaintiff acquired or should have acquired actual knowledge of the existence of such cause of action.” §1085(e)(9)(I)(iv). Thus, Congress has repeatedly drawn a “linguistic distinction” between what an ERISA plaintiff actually knows and what he should actually know. Merck, 559 U. S., at 647. And when Congress has included both forms of knowledge in a provision limiting ERISA actions, it has done so explicitly. We cannot assume that it meant to do so by implication in §1113(2). Instead we “generally presum[e] that Congress acts intentionally and purposely when it includes particular language in one section of a statute but omits it in another.” BFP v. Resolution Trust Corporation, 511 U.S. 531, 537 (1994) (internal quotation marks omitted). Petitioners dispute the characterization of anything less than actual knowledge as constructive knowledge, arguing that the latter term usually refers to information that a plaintiff must seek out rather than information that is sent to him. But if a plaintiff is not aware of a fact, he does not have “actual knowledge” of that fact however close at hand the fact might be. §1113(2). And Congress has never added to §1113(2) the language it has used in other ERISA limitations provisions to encompass both what a plaintiff actually knows and what he reasonably could know. As presently written, therefore, §1113(2) requires more than evidence of disclosure alone. That all relevant information was disclosed to the plaintiff is no doubt relevant in judging whether he gained knowledge of that information. See Part III, infra. To meet §1113(2)’s “actual knowledge” requirement, however, the plaintiff must in fact have become aware of that information. B Petitioners offer arguments for a broader reading of §1113(2) based on text, context, purpose, and statutory history. All founder on Congress’s choice of the word “actual.” As for text, petitioners do not dispute the normal definitions of “actual,” “knowledge,” or “actual knowledge.” They focus instead on the least conspicuous part of the phrase “had actual knowledge”: the word “had.” §1113(2). Once a plaintiff receives a disclosure, they argue, he “ha[s]” the knowledge that §1113(2) requires because he effectively holds it in his hand. Ibid. In other words, he has the requisite knowledge because he could acquire it with reason- able effort. That turns §1113(2) into what it is plainly not: a constructive-knowledge requirement. Petitioners’ contextual argument fails for the same reason. As they point out, ERISA’s disclosure regime is meant to “ensur[e] that ‘the individual participant knows exactly where he stands with respect to the plan.’ ” Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 118 (1989) (quoting H. R. Rep. No. 93–533, p. 11 (1973)). This is the reason for ERISA’s requirements that disclosures be written for a lay audience. See, e.g., 29 U. S. C. §1022(a). Once plan administrators satisfy their obligations to impart knowledge, petitioners say, §1113(2)’s knowledge requirement is satisfied too. But that is simply not what §1113(2) says. Unlike other ERISA limitations periods—which also form §1113(2)’s context—§1113(2) begins only when a plaintiff actually is aware of the relevant facts, not when he should be. And a given plaintiff will not necessarily be aware of all facts disclosed to him; even a reasonably diligent plaintiff would not know those facts immediately upon receiving the disclosure. Although “the words of a statute must be read in their context,” Davis v. Michigan Dept. of Treasury, 489 U.S. 803, 809 (1989), petitioners’ argument again gives the word “actual” little meaning at all. Petitioners also argue that §1113(2)’s plain meaning undermines its purpose of protecting plan administrators from suits over bygone investment decisions. If a plan participant can simply deny knowledge, they say, administrators will rarely get the benefit of §1113(2). But even if this is true, as it may well be, we cannot say that heeding the clear meaning of the word “actual” renders the statute so “ ‘[in]coherent’ ” that it must be disregarded. Kingdomware Technologies, Inc. v. United States, 579 U. S. ___, ___ (2016) (slip op., at 8). For one thing, plan participants are not the only potential plaintiffs subject to §1113. The Secretary of Labor, for example, may also sue imprudent fiduciaries for the benefit of plan participants. See §1132(a)(2). And the United States represents that the Secretary will have a hard time doing so within §1113(2)’s timeframe if deemed to have actual knowledge of the facts contained in the many reports that the Department receives from ERISA plans each year. See Brief for United States as Amicus Curiae 27–28. Moreover, the statute’s repose period will still protect defendants from suits filed more than six years after the alleged breach. See §1113(1). Petitioners may well be correct that heeding the plain meaning of §1113(2) substantially diminishes the protection that it provides for ERISA fiduciaries, but by the same token, petitioners’ interpretation would greatly reduce §1113(1)’s value for beneficiaries, given the disclosure regime that petitioners themselves emphasize. Choosing between these alternatives is a task for Congress, and we must assume that the language of §1113(2) reflects Congress’s choice. If policy considerations suggest that the current scheme should be altered, Congress must be the one to do it. See, e.g., Azar v. Allina Health Services, 587 U. S. ___, ___ (2019). Finally, petitioners argue that the plain meaning of “actual knowledge” renders an earlier version of §1113(2) incoherent. As originally enacted, the §1113(2) limitations period began either when the plaintiff gained actual knowledge of the alleged breach or when “a report from which [the plaintiff] could reasonably be expected to have obtained knowledge . . . was filed with” the Secretary of Labor. 29 U. S. C. §1113(2) (1976 ed.). That latter, constructive-knowledge clause was later repealed. See Omnibus Budget Reconciliation Act of 1987, §9342(b), 101Stat. 1330–371. According to petitioners, if “actual knowledge” means what it says, then the original version of §1113(2) charged plan participants with learning what was sent to the Secretary but not what was sent to them. The version at issue here, however, is the current one—from which Congress removed any mention of constructive knowledge. “When Congress acts to amend a statute, we presume it intends its amendment to have real and substantial effect.” Intel Corp. v. Advanced Micro Devices, Inc., 542 U.S. 241, 258–259 (2004) (internal quotation marks omitted). Section 1113(2)’s history thus more readily suggests that the current version does in fact require actual knowledge. III Nothing in this opinion forecloses any of the “usual ways” to prove actual knowledge at any stage in the litigation. Farmer v. Brennan, 511 U.S. 825, 842 (1994). Plaintiffs who recall reading particular disclosures will of course be bound by oath to say so in their depositions. On top of that, actual knowledge can be proved through “inference from circumstantial evidence.” Ibid.; see also Staples v. United States, 511 U.S. 600, 615–616, n. 11 (1994) (“[K]nowledge can be inferred from circumstantial evidence”). Evidence of disclosure would no doubt be relevant, as would electronic records showing that a plaintiff viewed the relevant disclosures and evidence suggesting that the plaintiff took action in response to the information contained in them. And though, “[a]t the summary judgment stage, facts must be viewed in the light most favorable to the nonmoving party,” that is true “only if there is a ‘genuine’ dispute as to those facts.” Scott v. Harris, 550 U.S. 372, 380 (2007) (quoting Fed. Rule Civ. Proc. 56(c)). If a plaintiff ’s denial of knowledge is “blatantly contradicted by the record,” “a court should not adopt that version of the facts for purposes of ruling on a motion for summary judgment.” 550 U. S., at 380. Today’s opinion also does not preclude defendants from contending that evidence of “willful blindness” supports a finding of “actual knowledge.” Cf. Global-Tech Appliances, Inc. v. SEB S. A., 563 U.S. 754, 769 (2011). In the case before us, however, petitioners do not argue that “actual knowledge” is established in any of these ways, only that they need not offer any such proof. And that is incorrect. * * * For these reasons, we affirm. It is so ordered. Notes 1 Specifically the Intel Global Diversified Fund, in which his retirement contribution plan was automatically invested, and the Intel Target Date 2045 Fund, which he chose for his 401(k) plan. 2 The court also addressed the separate question of what exactly a plaintiff must actually know about a defendant’s conduct and the relevant law in order for §1113(2) to apply. That question is not before us and we do not address it. 3 Compare Caputo v. Pfizer, Inc., 267 F.3d 181, 194 (CA2 2001); Reich v. Lancaster, 55 F.3d 1034, 1056–1057 (CA5 1995); Gluck v. Unisys Corp., 960 F.2d 1168, 1176 (CA3 1992); Radiology Center, S. C., v. Stifel, Nicolaus & Co., 919 F.2d 1216, 1222 (CA7 1990); Brock v. Nellis, 809 F.2d 753, 754–755 (CA11 1987), with Brown v. Owens Corning Investment Review Comm., 622 F.3d 564, 571 (CA6 2010) (“Actual knowledge does not require proof that the individual Plaintiffs actually saw or read the documents that disclosed the allegedly harmful investments” (internal quotation marks omitted)). 4 Petitioners cite this dictionary’s somewhat puzzling second definition of “actual knowledge,” which it dubs “implied actual knowledge”: “[k]nowledge of information that would lead a reasonable person to inquire further.” Black’s Law Dictionary 1043 (11th ed. 2019). Not even this entry, however, appears to equate “implied actual knowledge” with “actual knowledge” as normally understood. It instead proceeds to reference the common-law “discovery rule,” ibid., under which a limitations period begins when “the plaintiff discovers (or reasonably should have discovered) the injury giving rise to the claim,” id., at 585 (emphasis added); see also Merck & Co. v. Reynolds, 559 U.S. 633, 646 (2010). As we noted in Merck, that rule is broader than “actual knowledge.” Id.,at 647.

591.US.2019_18-1323

Louisiana’s Act 620, which is almost word-for-word identical to the Texas “admitting privileges” law at issue in Whole Woman’s Health v. Hellerstedt, 579 U. S. ___, requires any doctor who performs abortions to hold “active admitting privileges at a hospital . . . located not further than thirty miles from the location at which the abortion is performed or induced,” and defines “active admitting privileges” as being “a member in good standing” of the hospital’s “medical staff . . . with the ability to admit a patient and to provide diagnostic and surgical services to such patient.” In these consolidated cases, five abortion clinics and four abortion providers challenged Act 620 before it was to take effect, alleging that it was unconstitutional because (among other things) it imposed an undue burden on the right of their patients to obtain an abortion. (The plaintiff providers and two additional doctors are referred to as Does 1 through 6.) The plaintiffs asked for a temporary restraining order (TRO), followed by a preliminary injunction to prevent the law from taking effect. The defendant (State) opposed the TRO request but also urged the court not to delay ruling on the preliminary injunction motion, asserting that there was no doubt about the physicians’ standing. Rather than staying the Act’s effective date, the District Court provisionally forbade the State to enforce the Act’s penalties, while directing the plaintiff doctors to continue to seek privileges and to keep the court apprised of their progress. Several months later, after a 6-day bench trial, the District Court declared Act 620 unconstitutional on its face and preliminarily enjoined its enforcement. On remand in light of Whole Woman’s Health, the District Court ruled favorably on the plaintiffs’ request for a permanent injunction on the basis of the record previously developed, finding, among other things, that the law offers no significant health benefit; that conditions on admitting privileges common to hospitals throughout the State have made and will continue to make it impossible for abortion providers to obtain conforming privileges for reasons that have nothing to do with the State’s asserted interests in promoting women’s health and safety; and that this inability places a substantial obstacle in the path of women seeking an abortion. The court concluded that the law imposes an undue burden and is thus unconstitutional. The Fifth Circuit reversed, agreeing with the District Court’s interpretation of the standards that apply to abortion regulations, but disagreeing with nearly every one of the District Court’s factual findings. Held: The judgment is reversed. 905 F.3d 787, reversed. Justice Breyer, joined by Justice Ginsburg, Justice Sotomayor, and Justice Kagan, concluded: 1. The State’s unmistakable concession of standing as part of its effort to obtain a quick decision from the District Court on the merits of the plaintiffs’ undue-burden claims and a long line of well-established precedents foreclose its belated challenge to the plaintiffs’ standing in this Court. Pp. 11–16. 2. Given the District Court’s factual findings and precedents, particularly Whole Woman’s Health, Act 620 violates the Constitution. Pp. 16–40. (a) Under the applicable constitutional standards set forth in the Court’s earlier abortion-related cases, particularly Planned Parenthood of Southeastern Pa. v. Casey, 505 U.S. 833, and Whole Woman’s Health, “ ‘[u]nnecessary health regulations that have the purpose or effect of presenting a substantial obstacle to a woman seeking an abortion impose an undue burden on the right’ ” and are therefore “constitutionally invalid,” Whole Woman’s Health, 579 U. S., at ___. This standard requires courts independently to review the legislative findings upon which an abortion-related statute rests and to weigh the law’s “asserted benefits against the burdens” it imposes on abortion access. Id., at ___. The District Court here, like the trial court in Whole Woman’s Health, faithfully applied these standards. The Fifth Circuit disagreed with the District Court, not so much in respect to the legal standards, but in respect to the factual findings on which the District Court relied in assessing both the burdens that Act 620 imposes and the health-related benefits it might bring. Under well-established legal standards, a district court’s findings of fact “must not be set aside unless clearly erroneous, and the reviewing court must give due regard to the trial court’s opportunity to judge the witnesses’ credibility.” Fed. Rule. Civ. Proc. 52(a)(6). When the district court is “sitting without a jury,” the appellate court “is not to decide factual issues de novo,” Anderson v. Bessemer City, 470 U.S. 564, 573. Provided “the district court’s account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently.” Id., at 573–574. Viewed in light of this standard, the testimony and other evidence contained in the extensive record developed over the 6-day trial support the District Court’s conclusion on Act 620’s constitutionality. Pp. 16–19. (b) Taken together, the District Court’s findings and the evidence underlying them are sufficient to support its conclusion that enforcing the admitting-privileges requirement would drastically reduce the number and geographic distribution of abortion providers, making it impossible for many women to obtain a safe, legal abortion in the State and imposing substantial obstacles on those who could. Pp. 19–35. (1) The evidence supporting the court’s findings in respect to Act 620’s impact on abortion providers is stronger and more detailed than that in Whole Woman’s Health. The District Court supervised Does 1, 2, 5, and 6 for more than 18 months as they tried, and largely failed, to obtain conforming privileges from 13 relevant hospitals; it relied on a combination of direct evidence that some of the doctors’ applications were denied for reasons having nothing to do with their ability to perform abortions safely, and circumstantial evidence—including hospital bylaws with requirements like those considered in Whole Woman’s Health and evidence that showed the role that opposition to abortion plays in some hospitals’ decisions—that explained why other applications were denied despite the doctors’ good-faith efforts. Just as in Whole Woman’s Health, that evidence supported the District Court’s factual finding that Louisiana’s admitting-privileges requirement serves no “relevant credentialing function.” 579 U. S., at ___. The Fifth Circuit's conclusion that Does 2, 5, and 6 acted in bad faith cannot be squared with the clear-error standard of review that applies to the District Court’s contrary findings. Pp. 19–31. (2) The District Court also drew from the record evidence several conclusions in respect to the burden that Act 620 is likely to impose upon women’s ability to access an abortion in Louisiana. It found that enforcing that requirement would prevent Does 1, 2, and 6 from providing abortions altogether. Doe 3 gave uncontradicted, in-court testimony that he would stop performing abortions if he was the last provider in northern Louisiana, so the departure of Does 1 and 2 would also eliminate Doe 3. And Doe 5’s inability to obtain privileges in the Baton Rouge area would leave Louisiana with just one clinic with one provider to serve the 10,000 women annually who seek abortions in the State. Those women not altogether prevented from obtaining an abortion would face “longer waiting times, and increased crowding.” Whole Woman’s Health, 579 U. S., at ___. Delays in obtaining an abortion might increase the risk that a woman will experience complications from the procedure and may make it impossible for her to choose a non-invasive medication abortion. Both expert and lay witnesses testified that the burdens of increased travel to distant clinics would fall disproportionately on poor women, who are least able to absorb them. Pp. 31–35. (c) An examination of the record also shows that the District Court’s findings regarding the law’s asserted benefits are not “clearly erroneous.” The court found that the admitting-privileges requirement serves no “relevant credentialing function.” 250 F. Supp. 3d 27, 87. Hospitals can, and do, deny admitting privileges for reasons unrelated to a doctor’s ability safely to perform abortions, focusing primarily upon a doctor’s ability to perform the inpatient, hospital-based procedures for which the doctor seeks privileges—not outpatient abortions. And nothing in the record indicates that the vetting of applicants for privileges adds significantly to the vetting already provided by the State Board of Medical Examiners. The court’s finding that the admitting-privileges requirement “does not conform to prevailing medical standards and will not improve the safety of abortion in Louisiana,” ibid., is supported by expert and lay trial testimony. And, as in Whole Woman’s Health, the State introduced no evidence “showing that patients have better outcomes when their physicians have admitting privileges” or “of any instance in which an admitting privileges requirement would have helped even one woman obtain better treatment,” 250 F. Supp. 3d., at 64. Pp. 35–38. (d) In light of the record, the District Court’s significant factual findings—both as to burdens and as to benefits—have ample evidentiary support and are not “clearly erroneous.” Thus, the court’s related factual and legal determinations and its ultimate conclusion that Act 620 is unconstitutional are proper. P. 38. The Chief Justice agreed that abortion providers in this case have standing to assert the constitutional rights of their patients and concluded that because Louisiana’s Act 620 imposes a burden on access to abortion just as severe as that imposed by the nearly identical Texas law invalidated four years ago in Whole Woman’s Health v. Hellerstedt, 579 U. S. ___, it cannot stand under principles of stare decisis. Pp. 1–16. Breyer, J., announced the judgment of the Court and delivered an opinion, in which Ginsburg, Sotomayor, and Kagan, JJ., joined. Roberts, C. J., filed an opinion concurring in the judgment. Thomas, J., filed a dissenting opinion. Alito, J., filed a dissenting opinion, in which Gorsuch, J., joined, in which Thomas, J., joined except as to Parts III–C and IV–F, and in which Kavanaugh, J., joined as to Parts I, II, and III. Gorsuch, J., and Kavanaugh, J., filed dissenting opinions. Notes 1 Together with No. 18–1460, Russo, Interim Secretary, Louisiana Department of Health and Hospitals v. June Medical Services L. L. C. et al., also on certiorari to the same court.

; Casey, 505 U. S., at 845 (majority opinion); Akron v. Akron Center for Reproductive Health, Inc., 462 U.S. 416, 440, n. 30 (1983); Planned Parenthood of Central Mo. v. Danforth, 428 U.S. 52, 62 (1976); Doe v. Bolton, 410 U.S. 179, 188–189 (1973). And we have generally permitted plaintiffs to assert third-party rights in cases where the “ ‘enforcement of the challenged restriction against the litigant would result indirectly in the violation of third parties’ rights.’ ” Kowalski, 543 U. S., at 130 (quoting Warth, 422 U. S., at 510); see, e.g., Department of Labor v. Triplett, 494 U.S. 715, 720 (1990) (Scalia, J., for the Court) (attorney raising rights of clients to challenge restrictions on fee arrangements); Craig, 429 U. S., at 192 (convenience store raising rights of young men to challenge sex-based restriction on beer sales); Doe, 410 U. S., at 188 (abortion provider raising the rights of pregnant women to access an abortion); Carey v. Population Services Int’l, 431 U.S. 678 (1977) (distributors of contraceptives raising rights of prospective purchasers to challenge restrictions on sales of contraceptives); Eisenstadt v. Baird, 405 U.S. 438 (1972) (similar); Griswold v. Connecticut, 381 U.S. 479, 481 (1965) (similar); Sullivan v. Little Hunting Park, Inc., 396 U.S. 229 (1969) (white property owner raising rights of black contractual counterparty to challenge discriminatory restrictions on ability to contract); Barrows v. Jackson, 346 U.S. 249 (1953) (similar). In such cases, we have explained, “the obvious claimant” and “the least awkward challenger” is the party upon whom the challenged statute imposes “legal duties and disabilities.” Craig, 429 U. S., at 196–197; see Akron, 462 U. S., at 440, n. 30; Danforth, 428 U. S., at 62; Doe, 410 U. S., at 188. The case before us lies at the intersection of these two lines of precedent. The plaintiffs are abortion providers challenging a law that regulates their conduct. The “threatened imposition of governmental sanctions” for noncompliance eliminates any risk that their claims are abstract or hypothetical. Craig, 429 U. S., at 195. That threat also assures us that the plaintiffs have every incentive to “resist efforts at restricting their operations by acting as advocates of the rights of third parties who seek access to their market or function.” Ibid. And, as the parties who must actually go through the process of applying for and maintaining admitting privileges, they are far better positioned than their patients to address the burdens of compliance. See Singleton, 428 U. S., at 117 (plurality opinion) (observing that “the physician is uniquely qualified to litigate the constitutionality of the State’s interference with, or discrimination against,” a woman’s decision to have an abortion). They are, in other words, “the least awkward” and most “obvious” claimants here. Craig, 429 U. S., at 197. Our dissenting colleagues suggest that this case is different because the plaintiffs have challenged a law ostensibly enacted to protect the women whose rights they are asserting. See post, at 25–26 (opinion of Alito, J.); post, at 7 (opinion of Gorsuch, J.). But that is a common feature of cases in which we have found third-party standing. The restriction on sales of 3.2% beer to young men challenged by a drive-through convenience store in Craig was defended on “public health and safety grounds,” including the premise that young men were particularly susceptible to driving while intoxicated. 429 U. S., at 199–200; see Hager, Gender Discrimination and the Courts: New Ground to Cover, Washington Post, Sept. 26, 1976, p. 139. And the rule requiring approval from the Department of Labor for attorney fee arrangements challenged by a lawyer in Triplett was “designed to protect [their clients] from their improvident contracts, in the interest not only of themselves and their families but of the public.” 494 U. S., at 722 (internal quotation marks omitted). Nor is this the first abortion case to address provider standing to challenge regulations said to protect women. Both the hospitalization requirement in Akron, 462 U. S., at 435, and the hospital-accreditation requirement in Doe, 410 U. S., at 195, were defended as health and safety regulations. And the ban on saline amniocentesis in Danforth was based on the legislative finding “that the technique is deleterious to maternal health.” 428 U. S., at 76 (internal quotation marks omitted). In short, the State’s strategic waiver and a long line of well-established precedents foreclose its belated challenge to the plaintiffs’ standing. We consequently proceed to consider the merits of the plaintiffs’ claims. III A Turning to the merits, we apply the constitutional standards set forth in our earlier abortion-related cases, and in particular in Casey and Whole Woman’s Health. At the risk of repetition, we remind the reader of the standards we described above. In Whole Woman’s Health, we quoted Casey in explaining that “ ‘a statute which, while furthering [a] valid state interest has the effect of placing a substantial obstacle in the path of a woman’s choice cannot be considered a permissible means of serving its legitimate ends.’ ” 579 U. S., at ___ (slip op., at 19) (quoting Casey, 505 U. S., at 877 (plurality opinion)). We added that “ ‘[u]nnecessary health regulations’ ” impose an unconstitutional “ ‘undue burden’ ” if they have “ ‘the purpose or effect of presenting a substantial obstacle to a woman seeking an abortion.’ ” 579 U. S., at ___ (slip op., at 19) (quoting Casey, 505 U. S., at 878; emphasis added). We went on to explain that, in applying these standards, courts must “consider the burdens a law imposes on abortion access together with the benefits those laws confer.” 579 U. S., at ___ – ___ (slip op., at 19–20). We cautioned that courts “must review legislative ‘factfinding under a deferential standard.’ ” Id., at ___ (slip op., at 20) (quoting Gonzales, 550 U. S., at 165). But they “must not ‘place dispositive weight’ on those ‘findings,’ ” for the courts “ ‘retai[n] an independent constitutional duty to review factual findings where constitutional rights are at stake.’ ” 579 U. S., at ___ (slip op., at 20) (quoting Gonzales, 550 U. S., at 165; emphasis deleted). We held in Whole Woman’s Health that the trial court faithfully applied these standards. It “considered the evidence in the record—including expert evidence, presented in stipulations, depositions, and testimony.” 579 U. S., at ___ (slip op., at 21). It “then weighed the asserted benefits” of the law “against the burdens” it imposed on abortion access. Ibid. And it concluded that the balance tipped against the statute’s constitutionality. The District Court in this suit did the same. B The Court of Appeals disagreed with the District Court, not so much in respect to the legal standards that we have just set forth, but because it did not agree with the factual findings on which the District Court relied in assessing both the burdens that Act 620 imposes and the health-related benefits it might bring. Compare, e.g., supra, at 6–9, with supra, at 9–11. We have consequently reviewed the record in detail ourselves. In doing so, we have applied well-established legal standards. We start from the premise that a district court’s findings of fact, “whether based on oral or other evidence, must not be set aside unless clearly erroneous, and the reviewing court must give due regard to the trial court’s opportunity to judge the witnesses’ credibility.” Fed. Rule Civ. Proc. 52(a)(6). In “ ‘applying [this] standard to the findings of a district court sitting without a jury, appellate courts must constantly have in mind that their function is not to decide factual issues de novo.’ ” Anderson v. Bessemer City, 470 U.S. 564, 573 (1985) (quoting Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 123 (1969)). Where “the district court’s account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently.” Anderson, 470 U. S., at 573–574. “A finding that is ‘plausible’ in light of the full record—even if another is equally or more so—must govern.” Cooper v. Harris, 581 U. S. ___, ___ (2017) (slip op., at 4). Our dissenting colleagues suggest that a different, less-deferential standard should apply here because the District Court enjoined the admitting-privileges requirement before it was enforced. See post, at 11–12 (opinion of Alito, J.); post, at 11–13 (opinion of Gorsuch, J.). We are aware of no authority suggesting that appellate scrutiny of factual determinations varies with the timing of a plaintiff ’s lawsuit or a trial court’s decision. And, in any event, the record belies the dissents’ claims that the District Court’s findings in this case were “conjectural” or premature. As we have explained, the District Court’s order on the plaintiffs’ motion for a temporary restraining order suspended only Act 620’s penalties. The plaintiffs were required to continue in their efforts to obtain admitting privileges. See supra, at 5. The District Court supervised those efforts through the trial and beyond. See 250 F. Supp. 3d, at 77. It based its findings on this real-world evidence, not speculative guesswork. Nor can we agree with the suggestion that the timing of the District Court’s decision somehow prejudiced the State. From the start, the State urged that the District Court decide the merits of the plaintiffs’ claims without awaiting a decision on their applications for admitting privileges. See App. 43–44. And, when this case returned to the District Court in August 2016, following our decision in Whole Woman’s Health, the State stipulated that the case was ripe for decision on the record as it stood in June 2015. See supra, at 5–6. In short, we see no legal or practical basis to depart from the familiar standard that applies to all “[f]indings of fact.” Fed. Rule Civ. Proc. 52(a). Under that familiar standard, we find that the testimony and other evidence contained in the extensive record developed over the 6-day trial support the District Court’s ultimate conclusion that, “[e]ven if Act 620 could be said to further women’s health to some marginal degree, the burdens it imposes far outweigh any such benefit, and thus the Act imposes an unconstitutional undue burden.” 250 F. Supp. 3d, at 88. IV The District Court’s Substantial-Obstacle Determination The District Court found that enforcing the admitting-privileges requirement would “result in a drastic reduction in the number and geographic distribution of abortion providers.” Id., at 87. In light of demographic, economic, and other evidence, the court concluded that this reduction would make it impossible for “many women seeking a safe, legal abortion in Louisiana . . . to obtain one” and that it would impose “substantial obstacles” on those who could. Id., at 88. We consider each of these findings in turn. A Act 620’s Effect on Abortion Providers We begin with the District Court’s findings in respect to Act 620’s impact on abortion providers. As we have said, the court found that the Act would prevent Does 1, 2, and 6 from providing abortions. And it found that the Act would bar Doe 5 from working in his Baton Rouge-based clinic, relegating him to New Orleans. See supra, at 7–8. 1 In Whole Woman’s Health, we said that, by presenting “direct testimony” from doctors who had been unable to secure privileges, and “plausible inferences to be drawn from the timing of the clinic closures” around the law’s effective date, the plaintiffs had “satisfied their burden” to establish that the Texas admitting-privileges requirement caused the closure of those clinics. 579 U. S., at ___ (slip op., at 26). We wrote that these inferences were bolstered by the submissions of amici in the medical profession, which “describe[d] the undisputed general fact that hospitals often” will restrict admitting privileges to doctors likely to seek a “certain number of admissions per year.” Id., at ___ (slip op., at 24) (internal quotation marks omitted). The likely effect of such requirements was that abortion providers “would be unable to maintain admitting privileges or obtain those privileges for the future, because the fact that abortions are so safe meant that providers were unlikely to have any patients to admit.” Id., at ___ (slip op., at 25). We also referred to “common prerequisites to obtaining admitting privileges that have nothing to do with ability to perform medical procedures”; for example, requirements that doctors have “treated a high number of patients in the hospital setting in the past year, clinical data requirements, residency requirements, and other discretionary factors.” Ibid. To illustrate how these criteria impacted abortion providers, we noted the example of an obstetrician with 38 years’ experience who had been denied admitting privileges for reasons “ ‘not based on clinical competence considerations.’ ” Ibid. This, we said, showed that the law served no “relevant credentialing function,” but prevented qualified providers from serving women who seek an abortion. Id., at ___ (slip op., at 25). And that, in turn, “help[ed] to explain why the new [law’s admitting-privileges] requirement led to the closure of ” so many Texas clinics. Id., at ___ (slip op., at 24). The evidence on which the District Court relied in this case is even stronger and more detailed. The District Court supervised Does 1, 2, 5, and 6 for over a year and a half as they tried, and largely failed, to obtain conforming privileges from 13 relevant hospitals. See 250 F. Supp. 3d, at 77–78; App. 48–55, 64–82. The court heard direct evidence that some of the doctors’ applications were denied for reasons that had nothing to do with their ability to perform abortions safely. 250 F. Supp. 3d, at 68–70, 76–77; App. 1310, 1435–1436. It also compiled circumstantial evidence that explains why other applications were denied and explains why, given the costs of applying and the reputational risks that accompany rejection, some providers could have chosen in good faith not to apply to every qualifying hospital. Id., at 1135, 1311 (discussing the costs associated with unsuccessful applications). That circumstantial evidence includes documents and testimony that described the processes Louisiana hospitals follow when considering applications for admitting privileges, including requirements like the ones we cited in Whole Woman’s Health that are unrelated to a doctor’s competency to perform abortions. See generally Brief for Medical Staff Professionals as Amici Curiae 11–30 (reviewing the hospital bylaws in the record). The evidence shows, among other things, that the fact that hospital admissions for abortion are vanishingly rare means that, unless they also maintain active OB/GYN practices, abortion providers in Louisiana are unlikely to have any recent in-hospital experience. 250 F. Supp. 3d, at 49. Yet such experience can well be a precondition to obtaining privileges. Doe 2, a board-certified OB/GYN with nearly 40 years’ experience, testified that he had not “done any in-hospital work in ten years” and that just two of his patients in the preceding 5 years had required hospitalization. App. 387, 400. As a result, he was unable to comply with one hospital’s demand that he produce data on “patient admissions and management, consultations and procedures performed” in-hospital before his application could be “processed.” Id., at 1435; see id., at 437–438. Doe 1, a board-certified family doctor with over 10 years’ experience, was similarly unable to “submit documentation of hospital admissions and management of patients.” Id., at 1436. The evidence also shows that many providers, even if they could initially obtain admitting privileges, would be unable to keep them. That is because, unless they have a practice that requires regular in-hospital care, they will lose the privileges for failing to use them. Doe 6, a board-certified OB/GYN practitioner with roughly 50 years’ experience, provides only medication abortions. Id., at 1308. Of the thousands of women he served over the decade before the District Court’s decision, during which he also performed surgical abortions, just two required a direct transfer to a hospital and one of them was treated without being admitted. Id., at 1309. That safety record would make it impossible for Doe 6 to maintain privileges at any of the many Louisiana hospitals that require newly appointed physicians to undergo a process of “focused professional practice evaluation,” in which they are observed by hospital staff as they perform in-hospital procedures. See Record 2635, 2637, 2681, 9054; Brief for Medical Staff Professionals as Amici Curiae 28–29 (describing this practice); cf. Record 10755 (requiring an “on-going review” of practice “in the Operating Room”). And it would likewise disqualify him at hospitals that require physicians to admit a minimum number of patients, either initially or on an ongoing basis. See, e.g., id., at 9040, 9068–9069, 9150–9153; cf. App. 1193, 1182 (provider with no patient contacts in first year assigned to “Affiliate” status, without admitting privileges). The evidence also shows that opposition to abortion played a significant role in some hospitals’ decisions to deny admitting privileges. 250 F. Supp. 3d, at 48–49, 51–53 (collecting evidence). Some hospitals expressly bar anyone with privileges from performing abortions. App. 1180, 1205. Others are unwilling to extend privileges to abortion providers as a matter of discretion. Id., at 1127–1129. For example, Doe 2 testified that he was told not to bother asking for admitting privileges at University Health in Shreveport because of his abortion work. Id., at 383–384. And Doe 1 was told that his abortion work was an impediment to his application. Id., at 1315–1316. Still other hospitals have requirements that abortion providers cannot satisfy because of the hostility they face in Louisiana. Many Louisiana hospitals require applicants to identify a doctor (called a “covering physician”) willing to serve as a backup should the applicant admit a patient and then for some reason become unavailable. See Record 9154, 9374, 9383, 9478, 9667, 10302, 10481, 10637, 10659–10661, 10676. The District Court found “that opposition to abortion can present a major, if not insurmountable hurdle, for an applicant getting the required covering physician.” 250 F. Supp. 3d, at 49; cf. Whole Woman’s Health, 579 U. S., at ___ (slip op., at 25) (citing testimony describing similar problems faced by Texas providers seeking covering physicians). Doe 5 is a board-certified OB/GYN who had been practicing for more than nine years at the time of trial. Of the thousands of abortions he performed in the three years prior to the District Court’s decision, not one required a direct transfer to a hospital. App. 1134. Yet he was unable to secure privileges at three Baton Rouge hospitals because he could not find a covering physician willing to be publicly associated with an abortion provider. Id., at 1335–1336. Doe 3, a board-certified OB/GYN with nearly 45 years of experience, testified that he, too, had difficulty arranging coverage because of his abortion work. Id., at 200–202. Just as in Whole Woman’s Health, the experiences of the individual doctors in this case support the District Court’s factual finding that Louisiana’s admitting-privileges requirement, like that in Texas’ law, serves no “ ‘relevant credentialing function.’ ” 250 F. Supp. 3d, at 87 (quoting Whole Woman’s Health, 579 U. S., at ___ (slip op., at 25). 2 The Court of Appeals found another explanation for the doctors’ inability to obtain privileges more compelling. It conceded that Doe 1 would not be able to obtain admitting privileges in spite of his good-faith attempts. It concluded, however, that Does 2, 5, and 6 had acted in bad faith. 905 F. 3d, at 807. The problem is that the law requires appellate courts to review a trial court’s findings under the deferential clear-error standard we have described. See supra, at 17–18. Our review of the record convinces us that the Court of Appeals misapplied that standard. Justice Alito does not dispute that the District Court’s findings are not “clearly erroneous.” He argues instead that both the District Court and the Court of Appeals applied the wrong legal standard to the record in this case. By asking whether the doctors acted in “good faith,” he contends, the courts below failed to account for the doctors’ supposed “incentive to do as little as” possible to obtain conforming privileges. Post, at 12–14 (dissenting opinion); cf. post, at 11–12 (Gorsuch, J., dissenting). But that is not a legal argument at all. It is simply another way of saying that the doctors acted in bad faith. The District Court, after monitoring the doctors’ efforts for a year and a half, found otherwise. And “[w]hen the record is examined in light of the appropriately deferential standard, it is apparent that it contains nothing that mandates a finding that the District Court’s conclusion was clearly erroneous.” Anderson, 470 U. S., at 577. Doe 2 The District Court found that Doe 2 tried in good faith to get admitting privileges within 30 miles of his Shreveport-area clinic. 250 F. Supp. 3d, at 68. The Court of Appeals thought that conclusion clearly erroneous for three reasons. First, the appeals court suggested that Doe 2 failed to submit the data needed to process his application to Bossier’s Willis-Knighton Health Center. 905 F. 3d, at 808. It is true that Doe 2 submitted no additional information in response to the last letter he received from Willis-Knighton. But the record explains that failure. Doe 2 reasonably believed there was no point in doing so. The hospital’s letter explained that the data Doe 2 had already “submitted supports the outpatient [abortion] procedures you perform[ed].” App. 1435. But, the letter added, this data did “not support your request for hospital privileges” because it did not allow the hospital to “evaluate patient admissions and management, consultations, and procedures performed.” Ibid. Doe 2 testified at trial that he understood this to mean that he would have to submit records of hospital admissions, even though he had not “done any in-hospital work in ten years.” Id., at 387; see id., at 437 (“I’ve explained that that information doesn’t exist”). Doe 2’s understanding was consistent with Willis-Knighton’s similar letter to Doe 1, which explicitly stated that “we require that you submit documentation of hospital admissions and management of patients . . . .” Id., at 1436. The record also shows that Doe 2 could not have maintained the “adequate number of inpatient contacts” Willis-Knighton requires to support continued privileges. Record 9640; see App. 387–390, 404. Justice Alito faults Doe 2 for failing to pursue an application for “courtesy staff ” privileges. See post, at 18–19. For one thing, it is far from clear that courtesy privileges entitle a physician to admit patients, as Act 620 requires. Compare, e.g., Record 9640 with id., at 9643. For another, that would not solve the problem that Doe 2 lacked the required in-hospital experience. Justice Alito wonders whether Willis-Knighton might have conferred courtesy privileges even without that experience. But the factors the hospital considers for both tiers of privileges are facially identical. Id., at 9669. We have no license to reverse a trial court’s factual findings based on speculative inferences from facts not in evidence. Second, the Court of Appeals found Doe 2’s explanation that Christus Schumpert Hospital “would not staff an abortion provider” to be “blatantly contradicted by the record.” 905 F. 3d, at 808. The record, however, contains Christus’ bylaws. They state that “[n]o activity prohibited by” the Ethical and Religious Directives to which the hospital subscribes “shall be engaged in by any Medical Staff appointee or any other person exercising clinical privileges at the Health System.” App. 1180. These directives provide that abortion “is never permitted.” Id., at 1205. And they warn against “the danger of scandal in any association with abortion providers.” Ibid. The State suggests that the Court of Appeals, in speaking of a “contradic[tion],” was referring to the fact that Doe 3 had admitting privileges at Christus, as had Doe 2 at an earlier time. Brief for Respondent 75. Doe 3 testified, however, that he did not know whether Christus was “aware that I was performing abortions” and that he did not “feel like testing the waters there”—i.e., by “asking [Christus] how they would feel” if they were aware that he “was performing abortions.” App. 273. And nothing in the record suggests that Christus, 10 years earlier, was aware of Doe 2’s connection with abortion. Justice Alito imagines a number of ways that Christus may have become aware of Doe 2 or Doe 3’s abortion practice. See post, at 17–18, and n. 10 (dissenting opinion). The State apparently did not see fit to test these theories or probe the doctors’ accounts on cross-examination, however. And the District Court’s finding of good faith is plainly permissible on the record before us. Finally, the Court of Appeals faulted Doe 2 for failing to apply to Minden Hospital. The record also explains that decision. Minden subjects all new appointees to “not less than” six months of “focused professional practice evaluation.” Record 9281; see also id., at 9252. That evaluation requires an assessment of the provider’s in-hospital work. See supra, at 22. Doe 2 could not meet that requirement because, as we have said, Doe 2 does not do in-hospital work, and only two of his patients in the past five years have required hospitalization. App. 400. Moreover, Minden’s bylaws express a preference for applicants whom “members of the current Active Staff of the Hospital” have recommended. Id., at 1211. Doe 2 testified that Minden Hospital was “a smaller hospital,” “very close to the [geographic] limits,” where he “[did]n’t really know anyone.” Id., at 454. He applied to those hospitals where he believed he had the highest likelihood of success. Ibid. Given this evidence, the Fifth Circuit was wrong to conclude that the District Court’s findings in respect to Doe 2 were “clearly erroneous.” See Anderson, 470 U. S., at 575. Doe 5 The District Court found that Doe 5 was unable to obtain admitting privileges at three hospitals in range of his Baton Rouge clinic in spite of his good-faith efforts to satisfy each hospital’s requirement that he find a covering physician. 250 F. Supp. 3d, at 76; see App. 1334–1335 (Women’s Hospital); Record 2953 (Baton Rouge General), 10659–10661 (Lane Regional). The Court of Appeals disagreed. It thought that Doe 5’s efforts reflected a “lackluster approach” because he asked only one doctor to cover him. 905 F. 3d, at 809. The record shows, however, that Doe 5 asked the doctor most likely to respond affirmatively: the doctor with whom Doe 5’s Baton Rouge clinic already had a patient transfer agreement. App. 1135. Yet Doe 5 testified that even this doctor was “too afraid to be my covering physician at the hospital” because, while the transfer agreement could apparently be “kept confidential,” he feared that an agreement to serve as a covering physician would not remain a secret. Id., at 1135–1136. And, if the matter became well known, the doctor whom Doe 5 asked worried that it could make him a target of threats and protests. Ibid. Doe 5 was familiar with the problem. Anti-abortion protests had previously forced him to leave his position as a staff member of a hospital northeast of Baton Rouge. Id., at 1137–1138, 1330. And activists had picketed the school attended by the children of a former colleague, who then stopped performing abortions as a result. Record 14036–14037. With his own experience and their existing relationship in mind, Doe 5 could have reasonably thought that, if this doctor wouldn’t serve as his covering physician, no one would. And it was well within the District Court’s discretion to credit that reading of the record. Cf. Cooper, 581 U. S., at ___ (slip op., at 4). Doe 5’s testimony was internally consistent and consistent with what the District Court called the “mountain of un-contradicted and un-objected to evidence” in the record that supported its general finding “that opposition to abortion can present a major, if not insurmountable hurdle, for an applicant getting the required covering physician,” including Doe 3’s similar experience. 250 F. Supp. 3d, at 51, 49; see id., at 51–53; App. 200–202. The Court of Appeals did not address this general finding or the evidence the District Court relied on to support it, and neither do our dissenting colleagues. Cf. post, at 20–21 (opinion of Alito, J.); post, at 12 (opinion of Gorsuch, J.). The Court of Appeals pointed to what it described as Doe 4’s testimony that “finding a covering physician is not overly burdensome.” 905 F. 3d, at 809. Doe 4’s actual testimony was that he did not believe requiring doctors to obtain a covering physician was “an overburdensome requirement for admitting privileges.” Record 14154. In context, that statement is most naturally read as saying that such a requirement was reasonable, not that it was easy to fulfill. In fact, Doe 4 testified that he had been unable to apply to two hospitals for admitting privileges because he could not find a covering physician. Id., at 14154–14155. Moreover, Doe 4’s statement referred to his efforts to obtain admitting privileges in New Orleans, not in Baton Rouge. Ibid. Doe 5 testified that he could more easily find a covering physician in New Orleans (where he did obtain privileges) because attitudes toward abortion there were less hostile than in Baton Rouge, so the doctors’ testimony would be consistent even under the Fifth Circuit’s view. App. 1335–1336. Once again, the appeals court’s conclusion cannot be squared with the standard of review. Cf. Anderson, 470 U. S., at 575. Doe 6 Finally, the District Court found that, notwithstanding his good-faith efforts, Doe 6 would not be able to obtain admitting privileges within 30 miles of the clinic in New Orleans where he worked. The Court of Appeals did not question Doe 6’s decision not to apply to Tulane Hospital. Nor did it take issue with the District Court’s finding that his application to East Jefferson Hospital had been denied de facto through no fault of his own. 250 F. Supp. 3d, at 77; App. 54. But the appeals court reversed the District Court’s finding on the ground that Doe 6 should have (but did not) apply for admitting privileges at seven other hospitals in New Orleans, including Touro Hospital, which had granted limited privileges to Doe 5. 905 F. 3d, at 809–810. Doe 6 testified that he did not apply to other hospitals because he did not admit a sufficient number of patients to receive active admitting privileges. App. 1310. As we have explained, supra, at 21–22, Doe 6 provides only medication abortions involving no surgical intervention. See App. 1308. The State’s own admitting-privileges expert, Dr. Robert Marier, testified that a doctor in Doe 6’s position would “probably not” be able to obtain “active admitting and surgical privileges” at any hospital. Id., at 884; see 250 F. Supp. 3d, at 44 (finding Dr. Marier “generally well qualified” to express an opinion on “the issue of admitting privileges and hospital credentialing”). The record contains the bylaws of four of the seven hospitals to which the Court of Appeals referred. All four directly support the testimony of Doe 6 and the State’s expert. Three hospitals require doctors who receive admitting privileges to undergo a process of “focused professional practice evaluation.” See Record 2635, 2637, 2681 (Touro Hospital), 9054 (New Orleans East Hospital), 10755 (East Jefferson Hospital). As we have explained, this evaluation requires hospital staff to observe a doctor with admitting privileges while he or she performs a certain number of procedures. See supra, at 22. If the doctor admits no patients (and Doe 6 has no patients requiring admission), there is nothing to observe. Another hospital requires physicians to admit a minimum number of patients, either initially or after receiving admitting privileges. Record 9150–9153 (West Jefferson Hospital). And one requires both. Id., at 9040, 9069 (New Orleans East Hospital). The record apparently is silent as to the remaining three hospitals, but that silence cannot contradict the well-supported testimony of Doe 6 and the State’s expert that Doe 6 would not receive admitting privileges from any of them. Good faith does not require an exercise in futility. We recognize that Doe 5 was able to secure limited admitting privileges at Touro Hospital, to which Doe 6 did not apply. But, unlike Doe 6, Doe 5 primarily performs surgical abortions. App. 1330. And while Doe 5 was a hospital-based physician as recently as 2012, Doe 6 has not held privileges at any hospital since 2005. Id., at 1310, 1329. Doe 5’s success therefore does not directly contradict the evidence that we have described in respect to Doe 6 or render the District Court’s conclusion as to Doe 6 clearly erroneous. And, as we have said, “[a] finding that is ‘plausible’ in light of the full record—even if another is equally or more so—must govern.” Cooper, 581 U. S., at ___ (slip op., at 4). Without actually disputing any of the evidence we have discussed, Justice Alito maintains that the plaintiffs could have introduced still more evidence to support the District Court’s determination. See post, at 20. As we have said, however, “the trial on the merits should be ‘the “main event” . . . rather than a “tryout on the road.” ’ ” Anderson, 470 U. S., at 575. “[T]he parties to a case on appeal have already been forced to concentrate their energies and resources on persuading the trial judge that their account of the facts is the correct one; requiring them to persuade three more judges at the appellate level”—let alone another nine in this Court—“is requiring too much.” Ibid. Other Doctors Finally, Justice Alito and Justice Gorsuch suggest that the District Court failed to account for the possibility that new abortion providers might eventually replace Does 1, 2, 3, 5, and 6. See post, at 11–12 (opinion of Alito, J.); post, at 11–13 (opinion of Gorsuch, J.). But the Court of Appeals did not dispute, and the record supports, the District Court’s additional finding that, for “the same reasons that Does 1, 2, 4, 5, and 6 have had difficulties getting active admitting privileges, reasons unrelated to their competence . . . it is unlikely that the [a]ffected clinics will be able to comply with the Act by recruiting new physicians who have or can obtain admitting privileges.” 250 F. Supp. 3d, at 82. B Act 620’s Impact on Abortion Access The District Court drew from the record evidence, including the factual findings we have just discussed, several conclusions in respect to the burden that Act 620 is likely to impose upon women’s ability to access abortions in Louisiana. To better understand the significance of these conclusions, the reader should keep in mind the geographic distribution of the doctors and their clinics. Figure 1 shows the distribution of doctors and clinics at the time of the District Court’s decision. Figure 2 shows the projected distribution if the admitting-privileges requirement were enforced, as found by the District Court. The figures in parentheses indicate the approximate number of abortions each physician performed annually, according to the District Court. Figure 1 — Distribution of Abortion Clinics and Providers at the Time of the District Court’s Decision Figure 2 — Projected Distribution of Abortion Clinics and Providers Following Enforcement of Act 620 1 As we have seen, enforcing the admitting-privileges requirement would eliminate Does 1, 2, and 6. The District Court credited Doe 3’s uncontradicted, in-court testimony that he would stop performing abortions if he was the last provider in northern Louisiana. 250 F. Supp. 3d, at 79; see App. 263–265. So the departure of Does 1 and 2 would also eliminate Doe 3. That would leave only Doe 5. And Doe 5’s inability to obtain privileges in the Baton Rouge area would leave Louisiana with just one clinic with one provider to serve the 10,000 women annually who seek abortions in the State. 250 F. Supp. 3d, at 80, 87–88; cf. Whole Woman’s Health, 579 U. S., at ___ (slip op., at 26). Working full time in New Orleans, Doe 5 would be able to absorb no more than about 30% of the annual demand for abortions in Louisiana. App. 1134, 1331; see id., at 1129. And because Doe 5 does not perform abortions beyond 18 weeks, women between 18 weeks and the state legal limit of 20 weeks would have little or no way to exercise their constitutional right to an abortion. Id., at 1330–1331. Those women not altogether prevented from obtaining an abortion would face other burdens. As in Whole Woman’s Health, the reduction in abortion providers caused by Act 620 would inevitably mean “longer waiting times, and increased crowding.” 579 U. S., at ___ (slip op., at 26). The District Court heard testimony that delays in obtaining an abortion increase the risk that a woman will experience complications from the procedure and may make it impossible for her to choose a noninvasive medication abortion. App. 220, 290, 312–313; see also id., at 1139, 1305, 1313, 1316, 1323. Even if they obtain an appointment at a clinic, women who might previously have gone to a clinic in Baton Rouge or Shreveport would face increased driving distances. New Orleans is nearly a five hour drive from Shreveport; it is over an hour from Baton Rouge; and Baton Rouge is more than four hours from Shreveport. The impact of those increases would be magnified by Louisiana’s requirement that every woman undergo an ultrasound and receive mandatory counseling at least 24 hours before an abortion. La. Rev. Stat. Ann. §40:1061.10(D). A Shreveport resident seeking an abortion who might previously have obtained care at one of that city’s local clinics would either have to spend nearly 20 hours driving back and forth to Doe 5’s clinic twice, or else find overnight lodging in New Orleans. As the District Court stated, both experts and laypersons testified that the burdens of this increased travel would fall disproportionately on poor women, who are least able to absorb them. App. 106–107, 178, 502–508, 543; see also id., at 311–312. 2 We note that the Court of Appeals also faulted the District Court for factoring Doe 3’s departure into its calculations. The appeals court thought that Doe 3’s personal choice to stop practicing could not be attributed to Act 620. 905 F. 3d, at 810–811. That is beside the point. Even if we pretended as though (contrary to the record evidence) Doe 3 would continue to provide abortions at Shreveport-based Hope Clinic, the record nonetheless supports the District Court’s alternative finding that Act 620’s burdens would remain substantial. See 250 F. Supp. 3d, at 80–81, 84, 87. The record tells us that Doe 3 is presently able to see roughly 1,000–1,500 women annually. Id., at 81; see App. 207, 243–244. Doe 3 testified that this was in addition to “working very, very long hours maintaining [his] private [OB/GYN] practice.” Id., at 265, 1323; see id., at 118, 1147. And, the District Court found that Doe 5 can perform no more than roughly 3,000 abortions annually. See supra, at 33. So even if Doe 3 remained active in Shreveport, the annual demand for abortions in Louisiana would be more than double the capacity. And although the availability of abortions in Shreveport might lessen the driving distances faced by some women, it would still leave thousands of Louisiana women with no practical means of obtaining a safe, legal abortion, and it would not meaningfully address the health risks associated with crowding and delay for those able to secure an appointment with one of the State’s two remaining providers. * * * Taken together, we think that these findings and the evidence that underlies them are sufficient to support the District Court’s conclusion that Act 620 would place substantial obstacles in the path of women seeking an abortion in Louisiana. V Benefits We turn finally to the law’s asserted benefits. The District Court found that there was “ ‘no significant health-related problem that the new law helped to cure.’ ” 250 F. Supp. 3d, at 86 (quoting Whole Woman’s Health, 579 U. S., at ___ (slip op., at 22)). It found that the admitting-privileges requirement “[d]oes [n]ot [p]rotect [w]omen’s [h]ealth,” provides “no significant health benefits,” and makes no improvement to women’s health “compared to prior law.” 250 F. Supp. 3d, at 86 (boldface deleted). Our examination of the record convinces us that these findings are not “clearly erroneous.” First, the District Court found that the admitting-privileges requirement serves no “relevant credentialing function.” Id., at 87 (quoting Whole Woman’s Health, 579 U. S., at ___ (slip op., at 25)). As we have seen, hospitals can, and do, deny admitting privileges for reasons unrelated to a doctor’s ability safely to perform abortions. And Act 620’s requirement that physicians obtain privileges at a hospital within 30 miles of the place where they perform abortions further constrains providers for reasons that bear no relationship to competence. Moreover, while “competency is a factor” in credentialing decisions, 250 F. Supp. 3d, at 46, hospitals primarily focus upon a doctor’s ability to perform the inpatient, hospital-based procedures for which the doctor seeks privileges—not outpatient abortions. App. 877, 1373; see id., at 907; Brief for Medical Staff Professionals as Amici Curiae 26; Brief for American College of Obstetricians and Gynecologists et al. as Amici Curiae 12. Indeed, the State’s admitting-privileges expert, Dr. Robert Marier, testified that, when he served as the Executive Director of Louisiana’s Board of Medical Examiners, he concurred in the Board’s position that a physician was competent to perform first-trimester surgical abortions and to “recognize and address complications from the procedure” so long as they had completed an accredited residency in obstetrics and gynecology or been trained in abortion procedures during another residency—irrespective of their affiliation with any hospital. App. 872–873, 1305; cf. post, at 5–6 (Alito, J., dissenting). And nothing in the record indicates that the background vetting for admitting privileges adds significantly to the vetting that the State Board of Medical Examiners already provides. 250 F. Supp. 3d, at 87; App. 1355–1356, 1358–1359. Second, the District Court found that the admitting-privileges requirement “does not conform to prevailing medical standards and will not improve the safety of abortion in Louisiana.” 250 F. Supp. 3d, at 64; see id., at 64–66. As in Whole Woman’s Health, the expert and lay testimony presented at trial shows that: “Complications from surgical abortion are relatively rare,” and “[t]hey very rarely require transfer to a hospital or emergency room and are generally not serious.” App. 287; see id., at 129; cf. Whole Woman’s Health, 579 U. S., at ___ (slip op., at 22–23). For those patients who do experience complications at the clinic, the transfer agreement required by existing law is “sufficient to ensure continuity of care for patients in an emergency.” App. 1050; see id., at 194, 330–332, 1059. The “standard protocol” when a patient experiences a complication after returning home from the clinic is to send her “to the hospital that is nearest and able to provide the service that the patient needs,” which is not necessarily a hospital within 30 miles of the clinic. Id., at 351; see id., at 115–116, 180, 793; La. Rev. Stat. Ann. §40:1061.10(A)(2)(b)(ii) (requiring abortion providers to furnish patients with the name and telephone number of the hospital nearest to their home); cf. Whole Woman’s Health, 579 U. S., at ___ (slip op., at 23). As in Whole Woman’s Health, the State introduced no evidence “showing that patients have better outcomes when their physicians have admitting privileges” or “of any instance in which an admitting privileges requirement would have helped even one woman obtain better treatment.” 250 F. Supp. 3d, at 64; Whole Woman’s Health, 579 U. S., at ___ – ___ (slip op., at 23–24); see also Centers for Medicare and Medicaid Services, 84 Fed. Reg. 51790–51791 (2019) (“Under modern procedures, emergency responders (and patients themselves) take patients to hospital emergency rooms without regard to prior agreements between particular physicians and particular hospitals”); Brief for American College of Obstetricians and Gynecologists et al. as Amici Curiae 6 (local admitting-privileges requirements for abortion providers offer no medical benefit and do not meaningfully advance continuity of care). VI Conclusion We conclude, in light of the record, that the District Court’s significant factual findings—both as to burdens and as to benefits—have ample evidentiary support. None is “clearly erroneous.” Given the facts found, we must also uphold the District Court’s related factual and legal determinations. These include its determination that Louisiana’s law poses a “substantial obstacle” to women seeking an abortion; its determination that the law offers no significant health-related benefits; and its determination that the law consequently imposes an “undue burden” on a woman’s constitutional right to choose to have an abortion. We also agree with its ultimate legal conclusion that, in light of these findings and our precedents, Act 620 violates the Constitution. VII As a postscript, we explain why we have found unconvincing several further arguments that the State has made. First, the State suggests that the record supports the Court of Appeals’ conclusion that Act 620 poses no substantial obstacle to the abortion decision. See Brief for Respondent 73, 80. This argument misconceives the question before us. “The question we must answer” is “not whether the [Fifth] Circuit’s interpretation of the facts was clearly erroneous, but whether the District Court’s finding[s were] clearly erroneous.” Anderson, 470 U. S., at 577 (emphasis added). As we have explained, we think the District Court’s factual findings here are plausible in light of the record as a whole. Nothing in the State’s briefing furnishes a basis to disturb that conclusion. Second, the State says that the record does not show that Act 620 will burden every woman in Louisiana who seeks an abortion. Brief for Respondent 69–70 (citing United States v. Salerno, 481 U.S. 739, 745 (1987)). True, but beside the point. As we stated in Casey, a State’s abortion-related law is unconstitutional on its face if “it will operate as a substantial obstacle to a woman’s choice to undergo an abortion” in “a large fraction of the cases in which [it] is relevant.” 505 U. S., at 895 (majority opinion). In Whole Woman’s Health, we reaffirmed that standard. We made clear that the phrase refers to a large fraction of “those women for whom the provision is an actual rather than an irrelevant restriction.” 579 U. S., at ___ (slip op., at 39) (quoting Casey, 505 U. S., at 895; brackets omitted). That standard, not an “every woman” standard, is the standard that must govern in this case. Third, the State argues that Act 620 would not make it “nearly impossible” for a woman to obtain an abortion. Brief for Respondent 71–72. But, again, the words “nearly impossible” do not describe the legal standard that governs here. Since Casey, we have repeatedly reiterated that the plaintiff ’s burden in a challenge to an abortion regulation is to show that the regulation’s “purpose or effect” is to “plac[e] a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus.” 505 U. S., at 877 (plurality opinion); see Whole Woman’s Health, 579 U. S., at ___ (slip op., at 8); Gonzales, 550 U. S., at 156; Stenberg, 530 U. S., at 921; Mazurek, 520 U. S., at 971. Finally, the State makes several arguments about the standard of review that it would have us apply in cases where a regulation is found not to impose a substantial obstacle to a woman’s choice. Brief for Respondent 60–66. That, however, is not this case. The record here establishes that Act 620’s admitting-privileges requirement places a substantial obstacle in the path of a large fraction of those women seeking an abortion for whom it is a relevant restriction. * * * This case is similar to, nearly identical with, Whole Woman’s Health. And the law must consequently reach a similar conclusion. Act 620 is unconstitutional. The Court of Appeals’ judgment is erroneous. It is Reversed.

589.US.2019_18-6135

In Clark v. Arizona, 548 U.S. 735, this Court catalogued the diverse strains of the insanity defense that States have adopted to absolve mentally ill defendants of criminal culpability. Two—the cognitive and moral incapacity tests—appear as alternative pathways to acquittal in the landmark English ruling M’Naghten’s Case, 10 Cl. & Fin. 200, 8 Eng. Rep. 718. The moral incapacity test asks whether a defendant’s illness left him unable to distinguish right from wrong with respect to his criminal conduct. Respondent Kansas has adopted the cognitive incapacity test, which examines whether a defendant was able to understand what he was doing when he committed a crime. Specifically, under Kansas law a defendant may raise mental illness to show that he “lacked the culpable mental state required as an element of the offense charged,” Kan. Stat. Ann §21–5209. Kansas does not recognize any additional way that mental illness can produce an acquittal, although a defendant may use evidence of mental illness to argue for a lessened punishment at sentencing. See §§21–6815(c)(1)(C), 21–6625(a). In particular, Kansas does not recognize a moral-incapacity defense. Kansas charged petitioner James Kahler with capital murder after he shot and killed four family members. Prior to trial, he argued that Kansas’s insanity defense violates due process because it permits the State to convict a defendant whose mental illness prevented him from distinguishing right from wrong. The court disagreed and the jury returned a conviction. During the penalty phase, Kahler was free to raise any argument he wished that mental illness should mitigate his sentence, but the jury still imposed the death penalty. The Kansas Supreme Court rejected Kahler’s due process argument on appeal. Held: Due process does not require Kansas to adopt an insanity test that turns on a defendant’s ability to recognize that his crime was morally wrong. Pp. 6–24. (a) A state rule about criminal liability violates due process only if it “offends some principle of justice so rooted in the traditions and conscience our people as to be ranked as fundamental.” Leland v. Oregon, 343 U.S. 790, 798 (internal quotation marks omitted). History is the primary guide for this analysis. The due process standard sets a high bar, and a rule of criminal responsibility is unlikely to be sufficiently entrenched to bind all States to a single approach. As the Court explained in Powell v. Texas, 392 U.S. 514, the scope of criminal responsibility is animated by complex and ever-changing ideas that are best left to the States to evaluate and reevaluate over time. This principle applies with particular force in the context of the insanity defense, which also involves evolving understandings of mental illness. This Court has thus twice declined to constitutionalize a particular version of the insanity defense, see Leland, 343 U.S. 790; Clark, 548 U.S. 735, holding instead that a State’s “insanity rule[ ] is substantially open to state choice,” id., at 752. Pp. 6–9. (b) Against this backdrop, Kahler argues that Kansas has abolished the insanity defense—and, in particular, that it has impermissibly jettisoned the moral-incapacity approach. As a starting point, Kahler is correct that for hundreds of years jurists and judges have recognized that insanity can relieve criminal responsibility. But Kansas recognizes the same: Under Kansas law, mental illness is a defense to culpability if it prevented a defendant from forming the requisite criminal intent; a defendant is permitted to offer whatever evidence of mental health he deems relevant at sentencing; and a judge has discretion to replace a defendant’s prison term with commitment to a mental health facility. So Kahler can prevail only by showing that due process requires States to adopt a specific test of insanity—namely, the moral-incapacity test. He cannot do so. Taken as a whole, the early common law cases and commentaries reveal no settled consensus favoring Kahler’s preferred right-from-wrong rule. Even after M’Naghten gained popularity in the 19th century, States continued to experiment with new approaches. Clark therefore declared: “History shows no deference to M’Naghten that could elevate its formula to the level of fundamental principle.” 548 U. S., at 749–752. The tapestry of approaches States have adopted shows that no single version of the insanity defense has become so ingrained in American law as to rank as “fundamental.” Id., at 749. This result is not surprising. Ibid. The insanity defense sits at the juncture of medical views of mental illness and moral and legal theories of criminal culpability—two areas of conflict and change. Small wonder that no particular test of insanity has developed into a constitutional baseline. And it is not for the courts to insist on any single criterion moving forward. Defining the precise relationship between criminal culpability and mental illness requires balancing complex considerations, among them the workings of the brain, the purposes of criminal law, and the ideas of free will and responsibility. This balance should remain open to revision as new medical knowledge emerges and societal norms evolve. Thus—as the Court recognized previously in Leland, Powell, and Clark—the defense is a project for state governance, not constitutional law. Pp. 10–24. 307 Kan. 374, 410 P.3d 105, affirmed. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Alito, Gorsuch, and Kavanaugh, JJ., joined. Breyer, J., filed a dissenting opinion, in which Ginsburg and Sotomayor, JJ., joined.

This case is about Kansas’s treatment of a criminal defendant’s insanity claim. In Kansas, a defendant can invoke mental illness to show that he lacked the requisite mens rea (intent) for a crime. He can also raise mental illness after conviction to justify either a reduced term of imprisonment or commitment to a mental health facility. But Kansas, unlike many States, will not wholly exonerate a defendant on the ground that his illness prevented him from recognizing his criminal act as morally wrong. The issue here is whether the Constitution’s Due Process Clause forces Kansas to do so—otherwise said, whether that Clause compels the acquittal of any defendant who, because of mental illness, could not tell right from wrong when committing his crime. We hold that the Clause imposes no such requirement. I A In Clark v. Arizona, 548 U.S. 735, 749 (2006), this Court catalogued state insanity defenses, counting four “strains variously combined to yield a diversity of American standards” for when to absolve mentally ill defendants of criminal culpability. The first strain asks about a defendant’s “cognitive capacity”—whether a mental illness left him “unable to understand what he [was] doing” when he committed a crime. Id., at 747, 749. The second examines his “moral capacity”—whether his illness rendered him “un- able to understand that his action [was] wrong.” Ibid. Those two inquiries, Clark explained, appeared as alternative pathways to acquittal in the landmark English ruling M’Naghten’s Case, 10 Cl. & Fin. 200, 8 Eng. Rep. 718 (H. L. 1843), as well as in many follow-on American decisions and statutes: If the defendant lacks either cognitive or moral capacity, he is not criminally responsible for his behavior. Yet a third “building block[ ]” of state insanity tests, gaining popularity from the mid-19th century on, focuses on “volitional incapacity”—whether a defendant’s mental illness made him subject to “irresistible[ ] impulse[s]” or otherwise unable to “control[ ] his actions.” Clark, 548 U. S., at 749, 750, n. 11; see, e.g., Parsons v. State, 81 Ala. 577, 597, 2 So. 854, 866–867 (1887). And bringing up the rear, in Clark’s narration, the “product-of-mental-illness test” broadly considers whether the defendant’s criminal act stemmed from a mental disease. 548 U. S., at 749–750. As Clark explained, even that taxonomy fails to capture the field’s complexity. See id., at 750, n. 11. Most notable here, M’Naghten’s “moral capacity” prong later produced a spinoff, adopted in many States, that does not refer to morality at all. Instead of examining whether a mentally ill defendant could grasp that his act was immoral, some jurisdictions took to asking whether the defendant could understand that his act was illegal. Compare, e.g., People v. Schmidt, 216 N.Y. 324, 333–334, 110 N.E. 945, 947 (1915) (Cardozo, J.) (asking about moral right and wrong), with, e.g., State v. Hamann, 285 N.W.2d 180, 183 (Iowa 1979) (substituting ideas of legal right and wrong). That change in legal standard matters when a mentally ill defendant knew that his act violated the law yet believed it morally justified. See, e.g., Schmidt, 216 N. Y., at 339, 110 N. E., at 949; People v. Serravo, 823 P.2d 128, 135 (Colo. 1992).[1] Kansas law provides that “[i]t shall be a defense to a prosecution under any statute that the defendant, as a result of mental disease or defect, lacked the culpable mental state required as an element of the offense charged.” Kan. Stat. Ann. §21–5209 (2018 Cum. Supp.).[2] Under that statute, a defendant may introduce any evidence of any mental illness to show that he did not have the intent needed to commit the charged crime. Suppose, for example, that the defendant shot someone dead and goes on trial for murder. He may then offer psychiatric testimony that he did not understand the function of a gun or the consequences of its use—more generally stated, “the nature and quality” of his actions. M’Naghten, 10 Cl. & Fin., at 210, 8 Eng. Rep., at 722. And a jury crediting that testimony must acquit him. As everyone here agrees, Kansas law thus uses M’Naghten’s “cognitive capacity” prong—the inquiry into whether a mentally ill defendant could comprehend what he was doing when he committed a crime. See Brief for Petitioner 41; Brief for Respondent 31; Brief for United States as Amicus Curiae 18. If the defendant had no such capacity, he could not form the requisite intent—and thus is not criminally responsible. At the same time, the Kansas statute provides that “[m]ental disease or defect is not otherwise a defense.” §21–5209. In other words, Kansas does not recognize any additional way that mental illness can produce an acquittal.[3] Most important for this case, a defendant’s moral incapacity cannot exonerate him, as it would if Kansas had adopted both original prongs of M’Naghten. Assume, for example, that a defendant killed someone because of an “insane delusion that God ha[d] ordained the sacrifice.” Schmidt, 216 N. Y., at 339, 110 N. E., at 949. The defendant knew what he was doing (killing another person), but he could not tell moral right from wrong; indeed, he thought the murder morally justified. In many States, that fact would preclude a criminal conviction, although it would almost always lead to commitment in a mental health facility. In Kansas, by contrast, evidence of a mentally ill defendant’s moral incapacity—or indeed, of anything except his cognitive inability to form the needed mens rea—can play no role in determining guilt. That partly closed-door policy changes once a verdict is in. At the sentencing phase, a Kansas defendant has wide latitude to raise his mental illness as a reason to judge him not fully culpable and so to lessen his punishment. See §§21–6815(c)(1)(C), 21–6625(a). He may present evidence (of the kind M’Naghten deemed relevant) that his disease made him unable to understand his act’s moral wrongness—as in the example just given of religious delusion. See §21–6625(a). Or he may try to show (in line with M’Naghten’s spinoff ) that the illness prevented him from “appreciat[ing] the [conduct’s] criminality.” §21–6625(a)(6). Or again, he may offer testimony (here invoking volitional incapacity) that he simply could not “conform [his] conduct” to legal restraints. Ibid. Kansas sentencing law thus provides for an individualized determination of how mental illness, in any or all of its aspects, affects culpability. And the same kind of evidence can persuade a court to place a defendant who needs psychiatric care in a mental health facility rather than a prison. See §22–3430. In that way, a defendant in Kansas lacking, say, moral capacity may wind up in the same kind of institution as a like defendant in a State that would bar his conviction. B This case arises from a terrible crime. In early 2009, Karen Kahler filed for divorce from James Kahler and moved out of their home with their two teenage daughters and 9-year-old son. Over the following months, James Kahler became more and more distraught. On Thanksgiving weekend, he drove to the home of Karen’s grandmother, where he knew his family was staying. Kahler entered through the back door and saw Karen and his son. He shot Karen twice, while allowing his son to flee the house. He then moved through the residence, shooting Karen’s grandmother and each of his daughters in turn. All four of his victims died. Kahler surrendered to the police the next day and was charged with capital murder. Before trial, Kahler filed a motion arguing that Kansas’s treatment of insanity claims violates the Fourteenth Amendment’s Due Process Clause. Kansas, he asserted, had “unconstitutionally abolished the insanity defense” by allowing the conviction of a mentally ill person “who cannot tell the difference between right and wrong.” App. 11–12. The trial court denied the motion, leaving Kahler to attempt to show through psychiatric and other testimony that severe depression had prevented him from forming the intent to kill. See id., at 16; §21–5209. The jury convicted Kahler of capital murder. At the penalty phase, the court permitted Kahler to offer additional evidence of his mental illness and to argue in whatever way he liked that it should mitigate his sentence. The jury still decided to impose the death penalty. Kahler appealed, again challenging the constitutionality of Kansas’s approach to insanity claims. The Kansas Supreme Court rejected his argument, relying on an earlier precedential decision. See 307 Kan. 374, 400–401, 410 P.3d 105, 124–125 (2018) (discussing State v. Bethel, 275 Kan. 456, 66 P.3d 840 (2003)). There, the court denied that any single version of the insanity defense is so “ingrained in our legal system” as to count as “fundamental.” Id., at 473, 66 P. 3d, at 851. The court thus found that “[d]ue process does not mandate that a State adopt a particular insanity test.” Ibid. Kahler then asked this Court to decide whether the Due Process Clause requires States to provide an insanity defense that acquits a defendant who could not “distinguish right from wrong” when committing his crime—or, otherwise put, whether that Clause requires States to adopt the moral-incapacity test from M’Naghten. Pet. for Cert. 18. We granted certiorari, 586 U. S. ___ (2019), and now hold it does not.[4] II A A challenge like Kahler’s must surmount a high bar. Under well-settled precedent, a state rule about criminal liability—laying out either the elements of or the defenses to a crime—violates due process only if it “offends some principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental.” Leland v. Oregon, 343 U.S. 790, 798 (1952) (internal quotation marks omitted). Our primary guide in applying that standard is “historical practice.” Montana v. Egelhoff, 518 U.S. 37, 43 (1996) (plurality opinion). And in assessing that practice, we look primarily to eminent common-law authorities (Blackstone, Coke, Hale, and the like), as well as to early English and American judicial decisions. See, e.g., id., at 44–45; Patterson v. New York, 432 U.S. 197, 202 (1977). The question is whether a rule of criminal responsibility is so old and venerable—so entrenched in the central values of our legal system—as to prevent a State from ever choosing another. An affirmative answer, though not unheard of, is rare. See, e.g., Clark, 548 U. S., at 752 (“[T]he conceptualization of criminal offenses” is mostly left to the States). In Powell v. Texas, 392 U.S. 514 (1968), this Court explained why. There, Texas declined to recognize “chronic alcoholism” as a defense to the crime of public drunkenness. Id., at 517 (plurality opinion). The Court upheld that decision, emphasizing the paramount role of the States in setting “standards of criminal responsibility.” Id., at 533. In refusing to impose “a constitutional doctrine” defining those standards, the Court invoked the many “interlocking and overlapping concepts” that the law uses to assess when a person should be held criminally accountable for “his antisocial deeds.” Id., at 535–536. “The doctrines of actus reus, mens rea, insanity, mistake, justification, and duress”—the Court counted them off—reflect both the “evolving aims of the criminal law” and the “changing religious, moral, philosophical, and medical views of the nature of man.” Id., at 536. Or said a bit differently, crafting those doctrines involves balancing and rebalancing over time complex and oft-competing ideas about “social policy” and “moral culpability”—about the criminal law’s “practical effectiveness” and its “ethical foundations.” Id., at 538, 545, 548 (Black, J., concurring). That “constantly shifting adjustment” could not proceed in the face of rigid “[c]onstitution[al] formulas.” Id., at 536–537 (plurality opinion). Within broad limits, Powell thus concluded, “doctrine[s] of criminal responsibility” must remain “the province of the States.” Id., at 534, 536. Nowhere has the Court hewed more closely to that view than in addressing the contours of the insanity defense. Here, uncertainties about the human mind loom large. See, e.g., Ake v. Oklahoma, 470 U.S. 68, 81 (1985) (“[P]sychiatrists disagree widely and frequently on what constitutes mental illness, on [proper] diagnos[es, and] on cure and treatment”). Even as some puzzles get resolved, others emerge. And those perennial gaps in knowledge intersect with differing opinions about how far, and in what ways, mental illness should excuse criminal conduct. See Clark, 548 U. S., at 749–752 (canvassing how those competing views produced a wealth of insanity tests); supra, at 1–2. “This whole problem,” we have noted, “has evoked wide disagreement.” Leland, 343 U. S., at 801. On such unsettled ground, we have hesitated to reduce “experimentation, and freeze [the] dialogue between law and psychiatry into a rigid constitutional mold.” Powell, 392 U. S., at 536–537. Indeed, while addressing the demand for an alcoholism defense in Powell, the Court pronounced—as something close to self-evident—that “[n]othing could be less fruitful” than to define a specific “insanity test in constitutional terms.” Id., at 536. And twice before we have declined to do so. In Leland v. Oregon, a criminal defendant challenged as a violation of due process the State’s use of the moral-incapacity test of insanity—the very test Kahler now asks us to require. See 343 U. S., at 800–801. According to the defendant, Oregon instead had to adopt the volitional-incapacity (or irresistible-impulse) test to comply with the Constitution. See ibid.; supra, at 2. We rejected that argument. “[P]sychiatry,” we first noted, “has made tremendous strides since [the moral-incapacity] test was laid down in M’Naghten’s Case,” implying that the test seemed a tad outdated. 343 U. S., at 800–801. But still, we reasoned, “the progress of science has not reached a point where its learning” would demand “eliminat[ing] the right and wrong test from [the] criminal law.” Id., at 801. And anyway, we continued, the “choice of a test of legal sanity involves not only scientific knowledge but questions of basic policy” about when mental illness should absolve someone of “criminal responsibility.” Ibid. The matter was thus best left to each State to decide on its own. The dissent agreed (while parting from the majority on another ground): “[I]t would be indefensible to impose upon the States[ ] one test rather than another for determining criminal culpability” for the mentally ill, “and thereby to displace a State’s own choice.” Id., at 803 (opinion of Frankfurter, J.). A half-century later, we reasoned similarly in Clark. There, the defendant objected to Arizona’s decision to discard the cognitive-incapacity prong of M’Naghten and leave in place only the moral-incapacity one—essentially the flipside of what Kansas has done. Again, we saw no due process problem. Many States, we acknowledged, allowed a defendant to show insanity through either prong of M’Naghten. See 548 U. S., at 750. But we denied that this approach “represents the minimum that a government must provide.” Id., at 748. In so doing, we invoked the States’ traditional “capacity to define crimes and defenses,” and noted how views of mental illness had been particularly “subject to flux and disagreement.” Id., at 749, 752. And then we surveyed the disparate ways that state laws had historically excused criminal conduct because of mental disease—those “strains variously combined to yield a diversity of American standards.” See id., at 749–752; supra, at 1–2. The takeaway was “clear”: A State’s “insanity rule[ ] is substantially open to state choice.” Clark, 548 U. S., at 752. Reiterating Powell’s statement, Clark held that “no particular” insanity test serves as “a baseline for due process.” 548 U. S., at 752. Or said just a bit differently, that “due process imposes no single canonical formulation of legal insanity.” Id., at 753. B Yet Kahler maintains that Kansas’s treatment of insanity fails to satisfy due process. He sometimes makes his argument in the broadest of strokes, as he did before trial. See supra, at 5. Kansas, he then contends, has altogether “abolished the insanity defense,” in disregard of hundreds of years of historical practice. Brief for Petitioner 39. His central claim, though, is more confined. It is that Kansas has impermissibly jettisoned the moral-incapacity test for insanity. See id., at 12, 23. As earlier noted, both Clark and Leland described that test as coming from M’Naghten. See 548 U. S., at 749; 343 U. S., at 801; supra, at 2, 8. But according to Kahler (and the dissent), the moral-incapacity inquiry emerged centuries before that decision, thus forming part of the English common-law heritage this country inherited. See Brief for Petitioner 21, 42; post, at 4–14 (opinion of Breyer, J.). And the test, he claims, served for all that time—and continuing into the present—as the touchstone of legal insanity: If a defendant could not understand that his act was morally wrong, then he could not be found criminally liable. See Brief for Petitioner 20–23; see also post, at 15. So Kahler concludes that the moral-incapacity standard is a “principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental.” Leland, 343 U. S., at 798; see supra, at 6. In essence—and contra Clark—that test is the “single canonical formulation of legal insanity” and thus the irreducible “baseline for due process.” 548 U. S., at 752–753; see supra, at 9.[5] One point, first, of agreement: Kahler is right that for hundreds of years jurists and judges have recognized insanity (however defined) as relieving responsibility for a crime. “In criminal cases therefore,” Sir William Blackstone wrote, “lunatics are not chargeable for their own acts, if committed when under these incapacities.” 4 Commentaries on the Laws of England 24 (1769). Sir Edward Coke even earlier explained that in criminal cases, “the act and wrong of a mad man shall not be imputed to him.” 2 Institutes of the Laws of England §405, p. 247b (1628) (Coke). And so too Henry de Bracton thought that a “madman” could no sooner be found criminally liable than a child. 2 Bracton on Laws and Customs of England 384 (S. Thorne transl. 1968) (Bracton). That principle of non-culpability appeared in case after case involving allegedly insane defendants, on both sides of the Atlantic. “The defense of insanity[ ] is a defense for all crimes[,] from the highest to the lowest,” said the Court in Old Bailey. Trial of Samuel Burt (July 19, 1786), in 6 Proceedings in the Old Bailey 874 (E. Hodgson ed. 1788) (Old Bailey Proceedings). Repeated Justice Story, when riding circuit: “In general, insanity is an excuse for the commission of every crime, because the party has not the possession of that reason, which includes responsibility.” United States v. Drew, 25 F. Cas. 913 (No. 14,993) (CC Mass. 1828); see also, e.g., State v. Marler, 2 Ala. 43, 49 (1841) (“If the prisoner was insane, he was not an accountable being”); Cornwell v. State, 8 Tenn. 147, 156 (1827) (“[P]erfect madness” will “free a man from punishment for crime”). We have not found a single case to the contrary. But neither do we think Kansas departs from that broad principle. First, Kansas has an insanity defense negating criminal liability—even though not the type Kahler demands. As noted earlier, Kansas law provides that it is “a defense to a prosecution” that “the defendant, as a result of mental disease or defect, lacked the culpable mental state required” for a crime. §21–5209; see supra, at 3. That provision enables a defendant to present psychiatric and other evidence of mental illness to defend himself against a criminal charge. More specifically, the defendant can use that evidence to show that his illness left him without the cognitive capacity to form the requisite intent. See supra, at 3. Recall that such a defense was exactly what the defendant in Clark wanted, in preference to Arizona’s moral-incapacity defense: His (unsuccessful) appeal rested on the trial court’s exclusion of psychiatric testimony to show that he lacked the relevant mens rea. See 548 U. S., at 745–747; supra, at 9. Here, Kahler could do what Clark could not—try to show through such testimony that he had no intent to kill. Of course, Kahler would have preferred Arizona’s kind of insanity defense (just as Clark would have liked Kansas’s). But that does not mean that Kansas (any more than Arizona) failed to offer any insanity defense at all. Second, and significantly, Kansas permits a defendant to offer whatever mental health evidence he deems relevant at sentencing. See §§21–6815(c)(1)(C), 21–6625(a); supra, at 4. A mentally ill defendant may argue there that he is not blameworthy because he could not tell the difference between right and wrong. Or, because he did not know his conduct broke the law. Or, because he could not control his behavior. Or, because of anything else. In other words, any manifestation of mental illness that Kansas’s guilt-phase insanity defense disregards—including the moral incapacity Kahler highlights—can come in later to mitigate culpability and lessen punishment. And that same kind of evidence can persuade a judge to replace any prison term with commitment to a mental health facility. See §22–3430; supra, at 4–5. So as noted above, a defendant arguing moral incapacity may well receive the same treatment in Kansas as in States that would acquit—and, almost certainly, commit—him for that reason. See supra, at 4–5. In sum, Kansas does not bar, but only channels to sentencing, the mental health evidence that falls outside its intent-based insanity defense. When combined with Kansas’s allowance of mental health evidence to show a defendant’s inability to form criminal intent, that sentencing regime defeats Kahler’s charge that the State has “abolish[ed] the insanity defense entirely.”[6] Brief for Petitioner 39. So Kahler can prevail here only if he can show (again, contra Clark) that due process demands a specific test of legal insanity—namely, whether mental illness prevented a defendant from understanding his act as immoral. Kansas, as we have explained, does not use that type of insanity rule. See supra, at 3–4. If a mentally ill defendant had enough cognitive function to form the intent to kill, Kansas law directs a conviction even if he believed the murder morally justified. In Kansas’s judgment, that delusion does not make an intentional killer entirely blameless. See Brief for Respondent 40. Rather than eliminate, it only lessens the defendant’s moral culpability. See ibid. And sentencing is the appropriate place to consider mitigation: The decisionmaker there can make a nuanced evaluation of blame, rather than choose, as a trial jury must, between all and nothing. See ibid. In any event, so Kansas thinks.[7] Those views are contested and contestable; other States—many others—have made a different choice. But Kahler must show more than that. He must show that adopting the moral-incapacity version of the insanity rule is not a choice at all—because, again, that version is “so rooted in the traditions and conscience of our people as to be ranked as fundamental.” Leland, 343 U. S., at 798. And he cannot. The historical record is, on any fair reading, complex—even messy. As we will detail, it reveals early versions of not only Kahler’s proposed standard but also Kansas’s alternative. Early commentators on the common law proposed various formulations of the insanity defense, with some favoring a morality inquiry and others a mens rea approach. Kahler cites William Lambard’s 16th-century treatise defining a “mad man” as one who “hath no knowledge of good nor evil” (the right and wrong of the day). Eirenarcha, ch. 21, p. 218 (1581). He likewise points to William Hawkins’s statement, over a hundred years later, that a “lunatick[ ]” is not punishable because “under a natural disability of distinguishing between good and evil.” 1 Pleas of the Crown §1, p. 2 (1716) (capitalization omitted). Both true enough. But other early versions of the insanity test—and from a more famous trio of jurists—demanded the kind of cognitive impairment that prevented a defendant from understanding the nature of his acts, and thus intending his crime. Henry de Bracton’s 13th-century treatise gave rise to what became known as the “wild beast” test. See J. Biggs, The Guilty Mind 82 (1955). Used for hundreds of years, it likened a “madman” to an “animal[ ] which lack[s] reason” and so could not have “the intention to injure.” Bracton 384; see ibid. (A “madman” cannot commit a crime because “[i]t is will and purpose which mark” misdeeds). Sir Edward Coke similarly linked the definition of insanity to a defendant’s inability to form criminal intent. He described a legally insane person in 1628 as so utterly “without his mind or discretion” that he could not have the needed mens rea. 2 Coke §405, at 247b. So too Lord Matthew Hale a century later. He explained that insanity involves “a total alienation of the mind or perfect madness,” such that a defendant could not act “animo felonico,” meaning with felonious intent. 1 Pleas of the Crown, ch. 4, pp. 30, 37 (1736); see id., at 37 (“[F]or being under a full alienation of mind, he acts not per electionem or intentionem [by choice or intent]”).[8] Quite a few of the old common-law cases similarly stressed the issue of cognitive capacity. To be sure, even these cases included some references to the ability to tell right from wrong (and the dissent eagerly cherry-picks every one of them). But the decisions’ overall focus was less on whether a defendant thought his act moral than on whether he had the ability to do much thinking at all. In the canonical case of Rex v. Arnold, 16 How. St. Tr. 695 (1724), for example, the jury charge descended straight from Bracton: “[I]t is not every kind of frantic humour or something unaccountable in a man’s actions, that points him out to be such a madman as is to be exempted from punishment: it must be a man that is totally deprived of his understanding and memory, and doth not know what he is doing, no more than an infant, than a brute, or a wild beast.” Id., at 764–765. And the court offered an accompanying test linking that lack of reason to mens rea: If a man is “deprived of his reason, and consequently of his intention, he cannot be guilty.” Id., at 764; see ibid. (defining a “madman” as a “person that hath no design”); see also Trial of William Walker (Apr. 21, 1784), in 4 Old Bailey Proceedings 544, 547 (asking whether the defendant had a “distemper of mind which had deprived him of the use of his reason” or instead whether “he knew what he was doing [and] meant to do it”); Beverley’s Case, 4 Co. Rep. 123b, 124b, 76 Eng. Rep. 1118, 1121 (K. B. 1603) (asking whether a man “is deprived of reason and understanding” and so “cannot have a felonious intent”). The House of Lords used much the same standard in Rex v. Lord Ferrers, 19 How. St. Tr. 886 (1760), when sitting in judgment on one of its members. There, the Solicitor General told the Lords to address “the capacity and intention of the noble prisoner.” Id., at 948. Relying heavily on Hale’s treatise, he defined the legally insane as suffering from an “alienation of mind” and a “total[ ] want of reason.” Id., at 947. And in recapping the evidence on that issue, he asked about the defendant’s intention: “Did [Ferrers] proceed with deliberation? Did he know the consequences” of his act? Id., at 948.[9] In such cases, even the language of morality mostly worked in service of the emphasis on cognition and mens rea. The idea was that if a defendant had such a “total[ ] want of reason” as to preclude moral thinking, he could not possibly have formed the needed criminal intent. Id., at 947. Lord Chief Justice Mansfield put the point neatly in Bellingham’s Case, 1 G. Collinson, Treatise on the Law Concerning Idiots, Lunatics, and Other Persons Non Compotes Mentis 636 (1812) (Collinson). He instructed the jury: “If a man were deprived of all power of reasoning, so as not to be able to distinguish whether it was right or wrong to commit the most wicked transaction, he could not certainly do an act against the law. Such a man, so destitute of all power of judgment, could have no intention at all.” Id., at 671. On that account, moral incapacity was a byproduct of the kind of cognitive breakdown that precluded finding mens rea, rather than a self-sufficient test of insanity. See also Rex v. Offord, 5 Car. & P. 168, 169, 172 Eng. Rep. 924, 925 (N. P. 1831) (“express[ing] complete accordance in the observations of th[e] learned Judge” in Bellingham). Or said another way, a mentally ill defendant’s inability to distinguish right from wrong, rather than independently producing an insanity acquittal, served as a sign—almost a kind of evidence—that the defendant lacked the needed criminal intent. Other early common-law cases do not adopt the mens rea approach—but neither can they sustain Kahler’s position. Kahler relies mainly on Hadfield’s Case, 27 How. St. Tr. 1281 (1800), to show that common-law courts would acquit a mentally ill defendant who understood the nature of his act, but believed it moral. See Reply Brief 4. There, the defendant had deliberately set out to assassinate King George III on the view that doing so would bring about the Second Coming. See 27 How. St. Tr., at 1322. The judge instructed the jury that the defendant was so “deranged” as to make acquittal appropriate. Id., at 1353. Maybe, as Kahler argues, that directive stemmed from the defendant’s inability to tell right from wrong. But the judge never used that language, or stated any particular legal standard, so it is hard to know. Still other judges explained insanity to juries by throwing everything against the wall—mixing notions of cognitive incapacity, moral incapacity, and more, without trying to order, prioritize, or even distinguish among them. See, e.g., Regina v. Oxford, 9 Car. & P. 525, 545–548, 173 Eng. Rep. 941, 950 (N. P. 1840); Trial of Francis Parr (Jan. 15, 1787), in 2 Old Bailey Proceedings 228–229; Bowler’s Case, 1 Collinson 674. Those decisions treat the inability to make moral judgments more as part of an all-things-considered assessment of legal insanity, and less as its very definition. But even if some of them belong in Kahler’s corner, that would be far from enough. Taken as a whole, the common-law cases reveal no settled consensus favoring Kahler’s preferred insanity rule. And without that, they cannot support his proposed constitutional baseline. Only with M’Naghten, in 1843, did a court articulate, and momentum grow toward accepting, an insanity defense based independently on moral incapacity. See Clark, 548 U. S., at 749; Leland, 343 U. S., at 801; supra, at 2, 8. The M’Naghten test, as already described, found insanity in either of two circumstances. See supra, at 1–2. A defendant was acquitted if he “labour[ed] under such a defect of reason, from disease of the mind, [1] as not to know the nature and quality of the act he was doing; or, [2] if he did know it, that he did not know he was doing what was wrong.” 10 Cl. & Fin., at 210, 8 Eng. Rep., at 722 (emphasis added). That test disaggregated the concepts of cognitive and moral incapacity, so that each served as a stand-alone defense. And its crisp two-part formulation proved influential, not only in Great Britain but in the United States too. Over the course of the 19th century, many States adopted the test, making it the most popular one in the country. Still, Clark unhesitatingly declared: “History shows no deference to M’Naghten that could elevate its formula to the level of fundamental principle.” 548 U. S., at 749. As Clark elaborated, even M’Naghten failed to unify state insanity defenses. See 548 U. S., at 749–752. States continued to experiment with insanity rules, reflecting what one court called “the infinite variety of forms [of] insanity” and the “difficult and perplexing” nature of the defense. Roberts v. State, 3 Ga. 310, 328, 332 (1847). Some States in the 1800s gravitated to the newly emergent “volitional incapacity” standard, focusing on whether the defendant could at all control his actions. Clark, 548 U. S., at 749; see, e.g., Roberts, 3 Ga., at 331. One court viewed that inquiry as “much more practical” than the “right and wrong test,” which it thought often “speculative and difficult of determination.” State v. Felter, 25 Iowa 67, 82, 84 (1868); see Leland, 343 U. S., at 801 (recognizing such skepticism about the moral-incapacity test); supra, at 8–9. Another prophesied that the volitional test was the one “towards which all the modern authorities in this country[ ] are gradually but surely tending.” Parsons, 81 Ala., at 586, 2 So., at 859. But that test, too, failed to sweep all before it: State innovation proceeded apace. See, e.g., State v. Pike, 49 N. H. 399, 442 (1870) (applying the “product” test, which excuses a defendant whose crime “was the offspring or product of mental disease”); N. D. Cent. Code Ann. §12.1–04.1–01(1)(a) (2012) (replacing the right-from-wrong test with an inquiry into whether the defendant’s act arose from “[a] serious distortion of [his] capacity to recognize reality”). Much as medical views of mental illness changed as time passed, so too did legal views of how to account for that illness when assigning blame. As earlier noted, even the States that adopted M’Naghten soon divided on what its second prong should mean. See supra, at 2–3. Most began by asking, as Kahler does, about a defendant’s ability to grasp that his act was immoral. See, e.g., Wright v. State, 4 Neb. 407, 409 (1876); State v. Spencer, 21 N. J. L. 196, 201 (1846). Thus, Clark labeled M’Naghten’s second prong a test of “moral capacity,” and invoked the oft-used phrase “telling right from wrong” (or in older language, good from evil) to describe its central inquiry. 548 U. S., at 747, 753; see supra, at 2. But over the years, 16 States have reoriented the test to focus on the defendant’s understanding that his act was illegal—that is, legally rather than morally “wrong.”[10] They thereby excluded from the ranks of the insane those who knew an act was criminal but still thought it right. Contrary to Kahler’s (and the dissent’s) contention, that difference matters. See Reply Brief 7 (claiming that “there is little daylight between these inquiries”); post, at 17, 21 (same). The two tests will treat some, even though not all, defendants in opposite ways. And the defendants they will treat differently are exactly those Kahler (and the dissent) focus on: those who know exactly what they are doing (including that it is against the law) but believe it morally justified—because, say, it is commanded by God (or in the dissent’s case, a dog). See Brief for Petitioner 15; post, at 20; Schmidt, 216 N. Y., at 339, 110 N. E., at 949.[11] A famed theorist of criminal law put the point this way: “A kills B knowing that he is killing B, and knowing that it is illegal to kill B, but under an insane delusion that the salvation of the human race will be obtained by . . . the murder of B[.] A’s act is a crime if the word ‘wrong’ [in M’Naghten] means illegal. It is not a crime if the word wrong means morally wrong.” 2 J. Stephen, History of the Criminal Law of England, ch. 19, p. 149 (1883). So constitutionalizing the moral-incapacity standard, as Kahler requests, would require striking down not only the five state laws like Kansas’s (as the dissent at times suggests, see post, at 16), but 16 others as well (as the dissent eventually concedes is at least possible, see post, at 21). And with what justification? The emergence of M’Naghten’s legal variant, far from raising a due process problem, merely confirms what Clark already recognized. Even after its articulation in M’Naghten (much less before), the moral-incapacity test has never commanded the day. Clark, 548 U. S., at 749.[12] Indeed, just decades ago Congress gave serious consideration to adopting a mens rea approach like Kansas’s as the federal insanity rule. See United States v. Pohlot, 827 F.2d 889, 899, and n. 9 (CA3 1987) (describing bipartisan support for that proposal). The Department of Justice at the time favored that version of the insanity test. Perhaps more surprisingly, the American Medical Association did too. And the American Psychiatric Association took no position one way or the other. Although Congress chose in the end to adhere to the M’Naghten rule, the debate over the bill itself reveals continuing division over the proper scope of the insanity defense. Nor is that surprising, given the nature of the inquiry. As the American Psychiatric Association once noted, “insanity is a matter of some uncertainty.” Insanity Defense Work Group, Statement on the Insanity Defense, 140 Am. J. Psych. 681, 685 (1983). Across both time and place, doctors and scientists have held many competing ideas about mental illness. And that is only the half of it. Formulating an insanity defense also involves choosing among theories of moral and legal culpability, themselves the subject of recurrent controversy. At the juncture between those two spheres of conflict and change, small wonder there has not been the stasis Kahler sees—with one version of the insanity defense entrenched for hundreds of years. And it is not for the courts to insist on any single criterion going forward. We have made the point before, in Leland, Powell, and Clark. See supra, at 7–9. Just a brief reminder: “[F]ormulating a constitutional rule would reduce, if not eliminate, [the States’] fruitful experimentation, and freeze the developing productive dialogue between law and psychiatry into a rigid constitutional mold.” Powell, 392 U. S., at 536–537. Or again: In a sphere of “flux and disagreement,” with “fodder for reasonable debate about what the cognate legal and medical tests should be,” due process imposes no one view of legal insanity. Clark, 548 U. S., at 752–753. Defining the precise relationship between criminal culpability and mental illness involves examining the workings of the brain, the purposes of the criminal law, the ideas of free will and responsibility. It is a project demanding hard choices among values, in a context replete with uncertainty, even at a single moment in time. And it is a project, if any is, that should be open to revision over time, as new medical knowledge emerges and as legal and moral norms evolve. Which is all to say that it is a project for state governance, not constitutional law. We therefore decline to require that Kansas adopt an insanity test turning on a defendant’s ability to recognize that his crime was morally wrong. Contrary to Kahler’s view, Kansas takes account of mental health at both trial and sentencing. It has just not adopted the particular insanity defense Kahler would like. That choice is for Kansas to make—and, if it wishes, to remake and remake again as the future unfolds. No insanity rule in this country’s heritage or history was ever so settled as to tie a State’s hands centuries later. For that reason, we affirm the judgment below. It is so ordered. Notes 1 Another complicating factor in Clark’s classification scheme is that States “limit, in varying degrees, which sorts of mental illness” can support an insanity claim. Clark v. Arizona, 548 U.S. 735, 750, n. 11 (2006). So even two States using the same test for judging culpability may apply it to differently sized sets of offenders. See infra, at 21, n. 11. 2 At the time of the crime in this case, a materially identical provision was codified at §22–3220 (2007). 3 Four other States similarly exonerate a mentally ill defendant only when he cannot understand the nature of his actions and so cannot form the requisite mens rea. See Alaska Stat. §§12.47.010(a), 12.47.020 (2018); Idaho Code Ann. §§18–207(1), (3) (2016); Mont. Code Ann. §46–14–102 (2019); Utah Code §76–2–305 (2017). 4 Kahler also asked us to decide whether the Eighth Amendment requires that States make available the moral-incapacity defense. See Pet. for Cert. 18. But that claim is not properly before us. Kahler did not raise the argument below, and the Kansas courts therefore did not address it. 5 Although the dissent at times claims to the contrary, its argument is the same. Given the clear direction of our precedent, the dissent must purport to grant the States “leeway” in defining legal insanity. Post, at 1. But the entirety of the dissent’s historical analysis focuses on the moral-incapacity standard—attempting to show, just as Kahler does, that it both preceded and succeeded M’Naghten. See post, at 4–17. And in line with that narration, the dissent insists on moral understanding as the indispensable criterion of legal sanity—the sine qua non of criminal responsibility. See, e.g., post, at 1, 3–4, 8–9, 18–21. Indeed, the dissent offers only one way the States have actual “leeway” to change their insanity rules: They can “expand upon M’Naghten’s principles” by finding that even some who have moral capacity are insane. Post, at 22. But that is just to say that moral capacity is the constitutional floor—again, exactly what Kahler argues. 6 We here conclude only that Kansas’s scheme does not abolish the insanity defense. We say nothing, one way or the other, about whether any other scheme might do so. 7 The dissent is therefore wrong to suggest that Kansas’s law has become untethered from moral judgments about culpability. See post, at 1, 3, 16–22. No doubt, Kansas’s moral judgments differ from the dissent’s. Again, Kansas believes that an intentional killer is not wholly blameless, even if, for example, he thought his actions commanded by God. The dissent, in contrast, considers Kansas’s view benighted (as maybe some in the majority do too). But that is not a dispute, as the dissent suggests, about whether morality should play a role in assigning legal responsibility. It is instead a disagreement about what morality entails—that is, about when a defendant is morally culpable for an act like murder. See State v. Bethel, 275 Kan. 456, 465–471, 66 P.3d 840, 847–850 (2003) (accepting Kansas’s view that “moral blameworthiness” is linked to a defendant’s intent to kill, rather than to his ability to tell right from wrong). And we have made clear, from Leland to Powell to Clark, that courts do not get to make such judgments. See supra, at 7–9. Instead, the States have broad discretion to decide who counts as blameworthy, and to weigh that along with other factors in defining the elements of, and defenses to, crimes. 8 The dissent tries to recruit these three jurists to the side of the moral-incapacity test, see post, at 5–7, but cannot succeed. Even the carefully curated passages the dissent quotes focus on cognitive capability rather than moral judgment. See, e.g., post, at 5–6 (asking whether a defendant had “sense and reason” or “understanding and liberty of will”). In so doing, they refer to the defendant’s ability to form the requisite mens rea, or felonious intent. See Clark, 548 U. S., at 747; supra, at 1–3. The dissent still insists all is not lost because (it says) mens rea itself hinged at common law on a defendant’s “moral understanding.” Post, at 8–9. Here, the dissent infers from the use of “good-from-evil” language in various common-law treatises and cases that moral blameworthiness must have defined the mens rea inquiry. See ibid. But to begin with—and to repeat the point made in the text—the most influential treatises used little of that language, emphasizing instead the need for a defendant to intend his act in the ordinary sense of the term. And as we will explain, the joint presence of references to mens rea and moral understanding in other common-law sources involving insanity does not show that most jurists saw the two concepts as one and the same. See infra, at 16–19. Some may well have viewed mens rea through a moral prism; but others emphasized cognitive understanding in using that term; and still others combined the moral and cognitive in diverse ways. Which is to say that the record is far more complicated than the dissent lets on, with jurists invoking, both within particular sources and across all of them, a variety of ways to resolve insanity claims. And under our long-established precedent, that motley sort of history cannot provide the basis for a successful due process claim. 9 Even in the face of these instructions, the dissent claims that Arnold and Ferrers actually used the moral-incapacity test. See post, at 9–11. The assertion is based on some “good and evil” language (in Ferrers, mostly from witnesses) appearing in the case reports. But scholars generally agree, in line with our view, that Arnold and Ferrers “demonstrate how strictly” courts viewed “the criteria of insanity.” 1 N. Walker, Crime and Insanity in England 53 (1968) (noting that the two decisions “have often been cited” for that proposition). Kahler himself does not dispute the point; indeed, he essentially concedes our reading. Rather than try to make the decisions say something they do not, he argues only that they were “outlier[s]” and “could hardly have been less typical.” Brief for Petitioner 22, n. 5; Reply Brief 4 (internal quotation marks omitted). But that contrasting response fares no better. As even the dissent agrees, these were the “seminal” common-law decisions relating to insanity—indeed, two of only a small number in that period to make it into official reports. Post, at 9. 10 See State v. Skaggs, 120 Ariz. 467, 472, 586 P.2d 1279, 1284 (1978); Wallace v. State, 766 So. 2d 364, 367 (Fla. App. 2000); State v. Hamann, 285 N.W.2d 180, 184 (Iowa 1979); Commonwealth v. Lawson, 475 Mass. 806, 811, 62 N. E. 3d 22, 28 (2016); State v. Worlock, 117 N. J. 596, 610–611, 569 A.2d 1314, 1322 (1990); People v. Wood, 12 N.Y.2d 69, 76, 187 N.E.2d 116, 121–122 (1962); State v. Carreiro, 2013–Ohio–1103, 988 N.E.2d 21, 27 (App.); McElroy v. State, 242 S.W. 883, 884 (Tenn. 1922); McAfee v. State, 467 S.W.3d 622, 636 (Tex. Crim. App. 2015); State v. Crenshaw, 98 Wash. 2d 789, 794–795, 659 P.2d 488, 492–493 (1983); Ark. Code Ann. §5–2–301(6) (2017); Ill. Comp. Stat., ch. 720, §5/6–2(a) (West 2016); Ky. Rev. Stat. Ann. §504.020(1) (West 2016); Md. Crim. Proc. Code Ann. §3–109(a) (2018); Ore. Rev. Stat. §161.295(1) (2019); Vt. Stat. Ann., Tit. 13, §4801(a)(1) (2019). 11 The great judge (later Justice) whom the dissent cites to suggest there is no real difference between the legal wrong and moral wrong tests wrote a lengthy opinion whose point was the opposite. Consider a case, Judge Cardozo said: “A mother kills her infant child to whom she has been devotedly attached. She knows the nature and quality of the act; she knows that the law condemns it; but she is inspired by an insane delusion that God has appeared to her and ordained the sacrifice.” People v. Schmidt, 216 N.Y. 324, 339, 110 N.E. 945, 949 (1915). If the legal wrong test were used, Judge Cardozo continued, “it would be the duty of a jury to hold her responsible for the crime.” Ibid. But not if the focus was, as in the original M’Naghten test, on moral wrong. And that difference led the New York Court of Appeals to hold that the trial court’s jury instruction was in error. See 216 N. Y., at 340, 110 N. E., at 950. The additional cases the dissent cites to downplay the distinction between moral and legal wrong in fact follow Schmidt in recognizing when they diverge. See Worlock, 117 N. J., at 611, 569 A. 2d, at 1322 (explaining that “the distinction between moral and legal wrong may be critical” when, for example, a defendant “knowingly kill[s] another in obedience to a command from God”); Crenshaw, 98 Wash. 2d, at 798, 659 P. 2d, at 494 (acknowledging Schmidt’s view that even when a defendant “knows that the law and society condemn [her] act,” she should not be held responsible if “her free will has been subsumed by her belief in [a] deific decree”). 12 The diversity of American approaches to insanity is also evident in the States’ decisions about which kinds of mental illness can support the defense. See Clark, 548 U. S., at 750, n. 11; supra, at 3, n. 1. Some States limit the defense to those with a “severe” mental disease. See, e.g., Ala. Code §13A–3–1 (2015). Others prohibit its assertion by defendants with specific mental disorders. See, e.g., Ariz. Rev. Stat. Ann. §13–502 (2010) (“psychosexual” or “impulse control disorders”); Ore. Rev. Stat. §161.295(2) (“personality disorders”). In particular, many States follow the Model Penal Code in prohibiting psychopaths from raising the defense. See ALI, Model Penal Code §4.01(2), p. 163 (1985); e.g., Ind. Code §35–41–3–6(b) (2019) (“abnormality manifested only by repeated unlawful or otherwise antisocial conduct”). All those limitations apply even when the defendant’s mental illness prevented him from recognizing that his crime was immoral. In that way too, many States have departed from the principle that Kahler (along with the dissent) claims the Constitution commands.

589.US.2019_17-834

The Immigration Reform and Control Act of 1986 (IRCA) makes it unlawful to hire an alien knowing that he or she is unauthorized to work in the United States. 8 U. S. C. §§1324a(a)(1), (h)(3). IRCA requires employers to comply with a federal employment verification system. §1324a(b). Using a federal work-authorization form (I–9), they “must attest” that they have “verified” that any new employee, regardless of citizenship or nationality, “is not an unauthorized alien” by examining approved documents, e.g., a United States passport or an alien registration card, §1324a(b)(1)(A). IRCA concomitantly requires all employees to complete an I–9 by their first day of employment and to attest that they are authorized to work. §1324a(b)(2). Every employee must also provide certain personal information, including name, address, birth date, Social Security number, e-mail address, and telephone number. It is a federal crime for an employee to provide false information on an I–9 or to use fraudulent documents to show work authorization. See 18 U. S. C. §§1028, 1546. But it is not a federal crime for an alien to work without authorization, and state laws criminalizing such conduct are preempted. Arizona v. United States, 567 U.S. 387, 403–407. The I–9 forms and appended documentation, as well as the employment verification system, may only be used for enforcement of the Immigration and Nationality Act or other specified federal prohibitions. See §§1324a(b)(5), (d)(2)(F). IRCA does not directly address the use of an employee’s federal and state tax-withholding forms, the W–4 and K–4 respectively. Finally, IRCA expressly “preempt[s] any State or local law imposing civil or criminal sanctions . . . upon those who employ, or recruit or refer for a fee for employment, unauthorized aliens.” §1324a(h)(2). Kansas makes it a crime to commit “identity theft” or engage in fraud to obtain a benefit. Respondents, three unauthorized aliens, were tried for fraudulently using another person’s Social Security number on the W–4’s and K–4’s that they submitted upon obtaining employment. They had used the same Social Security numbers on their I–9 forms. Respondents were convicted, and the Kansas Court of Appeals affirmed. A divided Kansas Supreme Court reversed, concluding that §1324a(b)(5) expressly prohibits a State from using any information contained within an I–9 as the basis for a state law identity theft prosecution of an alien who uses another’s Social Security information in an I–9. The court deemed irrelevant the fact that this information was also included in the W–4 and K–4. One justice concurred based on implied preemption. Held: 1. The Kansas statutes under which respondents were convicted are not expressly preempted. IRCA’s express preemption provision applies only to employers and those who recruit or refer prospective employees and is thus plainly inapplicable. The Kansas Supreme Court instead relied on §1324a(b)(5), which broadly restricts any use of an I–9, information “contained in” an I–9, and any documents appended to an I–9, reasoning that respondents’ W–4’s and K–4’s used the same false Social Security numbers contained in their I–9’s. The theory that no information placed on an I–9 could ever be used by any entity or person for any reason—other than the handful of federal statutes mentioned in §1324a(b)(5)—is contrary to standard English usage. A tangible object can be “contained in” only one place at any point in time, but information may be “contained in” many different places. The mere fact that an I–9 contains an item of information, such as a name or address, does not mean that information “contained in” the I–9 is used whenever that name or address is used elsewhere. Nothing in §1324a(b)(5)’s text supports the Kansas Supreme Court’s limiting interpretation to prosecuting aliens for using a false identity to establish “employment eligibility.” And respondents’ express preemption argument cannot be saved by §1324a(d)(2)(F), which prohibits use of the federal employment verification system “for law enforcement purposes other than” enforcement of IRCA and the same handful of federal statutes mentioned in §1324a(b)(5). This argument fails because it rests on a misunderstanding of the meaning of the federal “employment verification system.” The sole function of that system is to establish that an employee is not barred from working in this country. The completion of tax-withholding documents plays no part in the process of determining whether a person is authorized to work. Pp. 10–14. 2. Respondents’ argument that Kansas’s laws are preempted by implication is also rejected. Pp. 15–20. (a) The laws do not fall into a field that is implicitly reserved exclusively for federal regulation, including respondents’ claimed field of “fraud on the federal verification system.” The submission of tax- withholding forms is neither part of, nor “related” to, the verification system. Employees may complete their W–4’s, K–4’s, and I–9’s at roughly the same time, but IRCA plainly does not foreclose all state regulation of information required as a precondition of employment. In arguing that the State’s statutes require proof that the accused engaged in the prohibited conduct for the purpose of getting a “benefit,” respondents conflate the benefit that results from complying with the federal employment verification system with the benefit of actually getting a job. Submitting W–4’s and K–4’s helped respondents get jobs, but it did not assist them in showing that they were authorized to work in this country. Federal law does not create a comprehensive and unified system regarding the information that a State may require employees to provide. Pp. 15–17. (b) There is likewise no ground for holding that the Kansas statutes at issue conflict with federal law. It is certainly possible to comply with both IRCA and the Kansas statutes, and respondents do not suggest otherwise. They instead maintain that the Kansas statutes, as applied in their prosecutions, stand as “an obstacle to the accomplishment and execution of the full purposes” of IRCA—one of which is purportedly that the initiation of any legal action against an unauthorized alien for using a false identity in applying for employment should rest exclusively within the prosecutorial discretion of federal authorities. Respondents analogize their case to Arizona v. United States, 567 U. S., at 404–407, where the Court concluded that a state law making it a crime for an unauthorized alien to obtain employment conflicted with IRCA, which does not criminalize that conduct. But here, Congress made no decision that an unauthorized alien who uses a false identity on tax-withholding forms should not face criminal prosecution, and it has made using fraudulent information on a W–4 a federal crime. Moreover, in the present cases, there is certainly no suggestion that the Kansas prosecutions frustrated any federal interests. Federal authorities played a role in all three cases, and the Federal Government fully supports Kansas’s position in this Court. In the end, however, the possibility that federal enforcement priorities might be upset is not enough to provide a basis for preemption. The Supremacy Clause gives priority to “the Laws of the United States,” not the criminal law enforcement priorities or preferences of federal officers. Art. VI, cl. 2. Pp. 18–20. 306 Kan. 1113, 401 P.3d 588 (first judgment); 306 Kan. 1100, 401 P.3d 155 (second judgment); and 306 Kan. 1107, 401 P.3d 159 (third judgment), reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a concurring opinion, in which Gorsuch, J., joined. Breyer, J., filed an opinion concurring in part and dissenting in part, in which Ginsburg, Sotomayor, and Kagan, JJ., joined. Notes 1 Together with Kansas v. Morales (see this Court’s Rule 12.4) and Kansas v. Ochoa-Lara (see this Court’s Rule 12.4), also on certiorari to the same court.

Kansas law makes it a crime to commit “identity theft” or engage in fraud to obtain a benefit. Respondents—three aliens who are not authorized to work in this country—were convicted under these provisions for fraudulently using another person’s Social Security number on state and federal tax-withholding forms that they submitted when they obtained employment. The Supreme Court of Kansas held that a provision of the Immigration Reform and Control Act of 1986 (IRCA), 100Stat. 3359, expressly preempts the Kansas statutes at issue insofar as they provide a basis for these prosecutions. We reject this reading of the provision in question, as well as respondents’ alternative arguments based on implied preemption. We therefore reverse. I A The foundation of our laws on immigration and naturalization is the Immigration and Nationality Act (INA), 66Stat. 163, as amended, 8 U. S. C. §1101 et seq., which sets out the “ ‘terms and conditions of admission to the country and the subsequent treatment of aliens lawfully in the country.’ ” Chamber of Commerce of United States of America v. Whiting, 563 U.S. 582, 587 (2011). As initially enacted, the INA did not prohibit the employment of illegal aliens, and this Court held that federal law left room for the States to regulate in this field. See De Canas v. Bica, 424 U.S. 351, 353 (1976). With the enactment of IRCA, Congress took a different approach. IRCA made it unlawful to hire an alien knowing that he or she is unauthorized to work in the United States. 8 U. S. C. §§1324a(a)(1)(A), (h)(3). To enforce this prohibition, IRCA requires employers to comply with a federal employment verification system. §1324a(b). Using a federal work-authorization form (I–9), employers “must attest” that they have “verified” that an employee “is not an unauthorized alien” by examining approved documents such as a United States passport or alien registration card. §1324a(b)(1)(A); see also §§1324a(b)(1)(B)–(D); 8 CFR §274a.2(a)(2) (2019) (establishing Form I–9). This requirement applies to the hiring of any individual regardless of citizenship or nationality. 8 U. S. C. §1324a(b)(1). Employers who fail to comply may face civil and criminal sanctions. See §§1324a(e)(4), (f ); 8 CFR §274 A. 10. IRCA instructs employers to retain copies of their I–9 forms and allows employers to make copies of the documents submitted by employees to show their authorization to work. 8 U. S. C. §§1324a(b)(3)–(4). IRCA concomitantly imposes duties on all employees, regardless of citizenship. No later than their first day of employment, all employees must complete an I–9 and attest that they fall into a category of persons who are authorized to work in the United States. §1324a(b)(2); 8 CFR §274a.2(b)(1)(i)(A). In addition, under penalty of perjury, every employee must provide certain personal information—specifically: name, residence address, birth date, Social Security number, e-mail address, and telephone number. It is a federal crime for an employee to provide false information on an I–9 or to use fraudulent documents to show authorization to work. See 18 U. S. C. §§1028, 1546. Federal law does not make it a crime for an alien to work without authorization, and this Court has held that state laws criminalizing such conduct are preempted. Arizona v. United States, 567 U.S. 387, 403–407 (2012). But if an alien works illegally, the alien’s immigration status may be adversely affected. See 8 U. S. C. §§1255(c)(2), (8), 1227(a)(1)(C)(i). While IRCA imposes these requirements on employers and employees, it also limits the use of I–9 forms. A provision entitled “Limitation on use of attestation form,” §1324a(b)(5), provides that I–9 forms and “any information contained in or appended to such form[s] may not be used for purposes other than for enforcement of ” the INA or other specified provisions of federal law, including those prohibiting the making of a false statement in a federal mat- ter ( 18 U. S. C. §1001), identity theft (§1028), immigration-document fraud (§1546), and perjury (§1621). In addition, 8 U. S. C. §1324a(d)(2)(F) prohibits use of “the employ- ment verification system” “for law enforcement purposes,” apart from the enforcement of the aforementioned federal statutes. Although IRCA expressly regulates the use of I–9’s and documents appended to that form, no provision of IRCA directly addresses the use of other documents, such as federal and state tax-withholding forms, that an employee may complete upon beginning a new job. A federal regulation provides that all employees must furnish their employers with a signed withholding exemption certificate when they start a new job, but federal law apparently does not require the discharge of an employee who fails to do so. See 26 CFR §§31.3402(f )(2)–1, (5)–1 (2019). Instead, the regulation provides that if an employee fails to provide a signed W–4, the employer must treat the employee “as a single person claiming no exemptions.” §31.3402(f )(2)–1(a). The submission of a fraudulent W–4, however, is a federal crime. 26 U. S. C. §7205. Kansas uses a tax-withholding form (K–4) that is similar to the federal form. Kan. Stat. Ann. §79–3298 (2018 Cum. Supp.); Kansas Dept. of Revenue, Notice 07–07: New K–4 Form for State Withholding (Sept. 5, 2007), www.ortho don.com/home/document/KS-WithholdingForm.pdf; Kansas Dept. of Revenue, Kansas Withholding Form K–4, www.ks revenue.org/k4info.html. Employees must attest to the veracity of the information under penalty of perjury. Form K–4, Kansas Employee’s Withholding Allowance Certificate (rev. Nov. 2018), www.ksrevenue.org/pdf/k-4.pdf; Kan. Stat. Ann. §21–5903; see also Kansas Dept. of Revenue, Tax Fraud Enforcement, www.ksrevenue.org/taxfraud.html. Finally, IRCA contains a provision that expressly “preempt[s] any State or local law imposing civil or criminal sanctions (other than through licensing and similar laws) upon those who employ, or recruit or refer for a fee for employment, unauthorized aliens.” 8 U. S. C. §1324a(h)(2) (emphasis added). This provision makes no mention of state or local laws that impose criminal or civil sanctions on employees or applicants for employment. See ibid. B Like other States, Kansas has laws against fraud, forgeries, and identity theft. These statutes apply to citizens and aliens alike and are not limited to conduct that occurs in connection with employment. The Kansas identity-theft statute criminalizes the “using” of any “personal identifying information” belonging to another person with the intent to “[d]efraud that person, or anyone else, in order to receive any benefit.” Kan. Stat. Ann. §21–6107(a)(1). “[P]ersonal identifying information” includes, among other things, a person’s name, birth date, driver’s license number, and Social Security number. §21–6107(e)(2). Kansas courts have interpreted the statute to cover the use of another person’s Social Security number to receive the benefits of employment. See State v. Meza, 38 Kan. App. 2d 245, 247–250, 165 P.3d 298, 301–302 (2007). Kansas’s false-information statute criminalizes, among other things, “making, generating, distributing or drawing” a “written instrument” with knowledge that it “falsely states or represents some material matter” and “with intent to defraud, obstruct the detection of a theft or felony offense or induce official action.” §21–5824. The respondents in the three cases now before us are aliens who are not authorized to work in this country but nevertheless secured employment by using the identity of other persons on the I–9 forms that they completed when they applied for work. They also used these same false identities when they completed their W–4’s and K–4’s. All three respondents were convicted under one or both of the Kansas laws just mentioned for fraudulently using another person’s Social Security number on tax-withholding forms. We summarize the pertinent facts related to these three prosecutions. C Ramiro Garcia. In August 2012, a local patrol officer stopped Garcia for speeding and learned that Garcia had been previously contacted by a financial crimes detective about possible identity theft. App. 39–44, 89–91; 306 Kan. 1113, 1114, 401 P.3d 588, 590 (2017). Local authorities obtained the documents that Garcia had completed when he began work at a restaurant, and a joint state-federal investigation discovered that Garcia had used another person’s Social Security number on his I–9, W–4, and K–4 forms. The State then charged Garcia with identity theft. The complaint alleged that, when he began work at the restaurant, he used another person’s Social Security number with the intent to defraud and in order to receive a benefit. App. 9–10. Donaldo Morales. A joint state-federal investigation of Morales began after the Kansas Department of Labor notified a Social Security agent that an employee at a local restaurant was using a Social Security number that did not match the identifying information in the department’s files. 306 Kan. 1100, 1101, 401 P.3d 155, 156 (2017); App. to Pet. for Cert. 73; App. 124–125, 168–170. A federal agent contacted the restaurant and learned that Morales had used another person’s Social Security number on his I–9, W–4, and K–4 forms. The federal agent arrested Morales, who then admitted that he had bought the Social Security number from someone he met in a park. App. 171–172; 306 Kan., at 1101–1102, 401 P. 3d, at 156; App. to Pet. for Cert.73. This information was turned over to state prosecutors, who charged Morales with identity theft and making false information. App. 124–125; 306 Kan., at 1101, 401 P. 3d, at 156. Guadalupe Ochoa-Lara. Ochoa-Lara came to the attention of a joint state-federal task force after officers learned that he had used a Social Security number issued to someone else when he leased an apartment. 306 Kan. 1107, 1108–1109, 401 P.3d 159, 160–161 (2017). The individual to whom this number was lawfully assigned advised the investigating officers that she had no knowledge that another person was using her number, and she later told authorities that income that she had not earned had been reported under her number. Id., at 1109, 401 P. 3d, at 160. After contacting the restaurant where Ochoa-Lara worked, investigators determined that he had also used the same Social Security number to complete his I–9 and W–4 forms. Ibid. The State charged Ochoa-Lara with identity theft and making false information for using another’s Social Security number on those documents. D In all three cases, respondents argued before trial that IRCA preempted their prosecutions. They relied on 8 U. S. C. §1324a(b)(5), which, as noted, provides that I–9 forms and “any information contained in or appended to such form[s] may not be used for purposes other than for enforcement of ” the INA or other listed federal statutes. In response, the State dismissed the charges that were based on I–9’s and agreed not to rely on the I–9’s at trial. The State maintained, however, that §1324a(b)(5) did not apply to the respondents’ use of false Social Security numbers on the tax-withholding forms. The trial courts allowed the State to proceed with the charges based on those forms. The State entered the K–4’s and W–4’s into evidence against Garcia and Morales, and Ochoa-Lara stipulated to using a stolen Social Security number on a W–4. App. 109–110; 306 Kan., at 1108–1109, 401 P. 3d, at 160–161.[1] Respondents were convicted, and three separate panels of the Kansas Court of Appeals affirmed their convictions. A divided Kansas Supreme Court reversed, concluding that “the plain and unambiguous language of 8 U. S. C. §1324a(b)(5)” expressly prohibits a State from using “any information contained within [an] I–9 as the bas[i]s for a state law identity theft prosecution of an alien who uses another’s Social Security information in an I–9.” 306 Kan., at 1130–1131, 401 P.3d at 599 (emphasis deleted). The court added that “[t]he fact that this information was included in the W–4 and K–4 did not alter the fact that it was also part of the I–9.” Id., at 1131, 401 P. 3d, at 599. In deciding the appeal on these grounds, the court appears to have embraced the proposition that any fact to which an employee attests in an I–9 is information that is “contained in” the I–9 and is thus subject to the restrictions imposed by §1324a(b)(5), namely, that this fact cannot be used by anyone for any purpose other than the few listed in that provision. Nevertheless, the court suggested that its holding did not sweep this broadly but was instead limited to the prosecution of aliens for using a false identity to establish “employment eligibility.” Id., at 1126, 1131, 401 P. 3d, at 596, 600. Justice Luckert concurred based on implied, not express, preemption. In her view, IRCA occupies “the field” within which the prosecutions at issue fell, namely, “the use of false documents, including those using the identity of others, when an unauthorized alien seeks employment.” Id. at 1136, 401 P. 3d, at 602. Justice Luckert also opined that the Kansas statutes, as applied in these cases, conflict with IRCA because they “usur[p] federal enforcement discretion” regarding the treatment of aliens who obtain employment even though they are barred from doing so under federal law. Ibid., 401 P. 3d, at 603. Two members of the court, Justices Biles and Stegall, dissented, and we granted review. 586 U. S. ___ (2019). II The Supremacy Clause provides that the Constitution, federal statutes, and treaties constitute “the supreme Law of the Land.” Art. VI, cl. 2. The Clause provides “a rule of decision” for determining whether federal or state law applies in a particular situation. Armstrong v. Exceptional Child Center, Inc., 575 U.S. 320, 324 (2015). If federal law “imposes restrictions or confers rights on private actors” and “a state law confers rights or imposes restrictions that conflict with the federal law,” “the federal law takes precedence and the state law is preempted.” Murphy v. National Collegiate Athletic Assn., 584 U. S. ___, ___ (2018) (slip op., at 22). In all cases, the federal restrictions or rights that are said to conflict with state law must stem from either the Constitution itself or a valid statute enacted by Congress. “There is no federal preemption in vacuo,” without a constitutional text, federal statute, or treaty made under the authority of the United States. Puerto Rico Dept. of Consumer Affairs v. ISLA Petroleum Corp., 485 U.S. 495, 503 (1988); see also Whiting, 563 U. S., at 599 (preemption cannot be based on “a ‘freewheeling judicial inquiry into whether a state statute is in tension with federal objectives.’ ”) (citation omitted); Virginia Uranium, Inc. v. Warren, 587 U. S. ___, ___ (2019) (lead opinion of Gorsuch, J.) (slip op., at 3) (“Invoking some brooding federal interest or appealing to a judicial policy preference” does not show preemption). In some cases, a federal statute may expressly preempt state law. See Pacific Gas & Elec. Co. v. State Energy Resources Conservation and Development Comm’n, 461 U.S. 190, 203 (1983) (“It is well established that within constitutional limits Congress may preempt state authority by so stating in express terms.”). But it has long been established that preemption may also occur by virtue of restrictions or rights that are inferred from statutory law. See, e.g., Osborn v. Bank of United States, 9 Wheat. 738, 865 (1824) (rejecting argument that a federal exemption from state regulation “not being expressed, ought not to be implied by the Court”). And recent cases have often held state laws to be impliedly preempted. See, e.g., Arizona 567 U. S., at 400–408; Kurns v. Railroad Friction Products Corp., 565 U.S. 625, 630–631 (2012); PLIVA, Inc. v. Mensing, 564 U.S. 604, 617–618 (2011). In these cases, respondents do not contend that the Kansas statutes under which they were convicted are preempted in their entirety. Instead, they argue that these laws must yield only insofar as they apply to an unauthorized alien’s use of false documents on forms submitted for the purpose of securing employment. In making this argument, respondents invoke all three categories of preemption identified in our cases. They defend the Kansas Supreme Court’s holding that provisions of IRCA expressly bar their prosecutions. And they also argue that the decision below is supported by “field” or “conflict” preemption or some combination of the two. We consider these arguments in turn. III We begin with the argument that the state criminal statutes under which respondents were convicted are expressly preempted. As noted, IRCA contains a provision that expressly preempts state law, but it is plainly inapplicable here. That provision applies only to the imposition of criminal or civil liability on employers and those who receive a fee for recruiting or referring prospective employees. 8 U. S. C. §1324a(h)(2). It does not mention state or local laws that impose criminal or civil sanctions on employees or applicants for employment. The Kansas Supreme Court did not base its holding on this provision but instead turned to §1324a(b)(5), which is far more than a preemption provision. This provision broadly restricts any use of an I–9, information contained in an I–9, and any documents appended to an I–9. Thus, unlike a typical preemption provision, it applies not just to the States but also to the Federal Government and all private actors. The Kansas Supreme Court thought that the prosecutions in these cases ran afoul of this provision because the charges were based on respondents’ use in their W–4’s and K–4’s of the same false Social Security numbers that they also inserted on their I–9’s. Taken at face value, this theory would mean that no information placed on an I–9— including an employee’s name, residence address, date of birth, telephone number, and e-mail address—could ever be used by any entity or person for any reason. This interpretation is flatly contrary to standard English usage. A tangible object can be “contained in” only one place at any point in time, but an item of information is different. It may be “contained in” many different places, and it is not customary to say that a person uses information that is contained in a particular source unless the person makes use of that source. Consider a person’s e-mail address, one of the bits of information that is called for on an I–9. A person’s e-mail address may be “contained in” a great many places. Individuals often provide their e-mail addresses to a wide circle of friends, acquaintances, online vendors, work-related contacts, and others. In addition, the records of every recipient of an e-mail from a particular person will contain that address.[2] In ordinary speech, no one would say that a person who uses an e-mail address has used information that is contained in all these places. Suppose that John used his e-mail address five years ago to purchase a pair of shoes and that the vendor has that address in its files. Suppose that John now sends an e-mail to Mary and that Mary sends an e-mail reply. No one would say that Mary has used information contained in the files of the shoe vendor. Or consider this bit of information: that the first man set foot on the moon on July 20, 1969.[3] That fact was reported in newspapers around the world, from Neil Armstrong’s hometown newspaper, the Wapakoneta (Ohio) Daily News[4] to the Soviet newspaper Izvestia.[5] Suppose that an elementary school student writes a report in which she states that the first man walked on the moon in 1969. No one would say that the student used information contained in the Wapakoneta Daily News or Izvestia if she never saw those publications. But it would be natural to say that the student used information contained in a book in the school library if that is where she got the information for her report. Accordingly, the mere fact that an I–9 contains an item of information, such as a name or address, does not mean that information “contained in” the I–9 is used whenever that name or address is later employed. If this were not so, strange consequences would ensue. Recall that 8 U. S. C. §1324a(b)(5) applies to the Federal Government. Under 26 U. S. C. §7205, it is a crime to willfully supply false information on a W–4, and this provision is not among those listed in 8 U. S. C. §1324a(b)(5). Thus, if an individual provided the same false information on an I–9 and a W–4, the Federal Government could not prosecute this individual under 26 U. S. C. §7205 even if the Government made no use whatsoever of the I–9. And that is just the beginning. Suppose that an employee truthfully states on his I–9 that his name is Jim Smith. Under the interpretation of 8 U. S. C. §1324a(b)(5) that the Kansas Supreme Court seemingly adopted, no one could use Jim’s name for any purpose. If he robbed a bank, prosecutors could not use his name in an indictment. His employer could not cut a paycheck using that name. His sister could not use his name to mail him a birthday card. The Kansas Supreme Court tried to fend off these consequences by suggesting that its interpretation applied only to the prosecution of aliens for using a false identity to establish “employment eligibility.” 306 Kan., at 1126, 401 P. 3d, at 596. But there is no trace of these limitations in the text of §1324a(b)(5). The point need not be belabored any further: The argument that §1324a(b)(5) expressly bars respondents’ prosecutions cannot be defended. Apparently recognizing this, respondents turn to §1324a(d)(2)(F), which prohibits use of the federal employment verification system[6] “for law enforcement purposes other than” enforcement of IRCA and the same handful of federal statutes mentioned in §1324a(b)(5): 18 U. S. C. §1001 (false statements), §1028 (identity theft), §1546 (immigration-document fraud), and §1621 (perjury). This argument fails because it rests on a misunderstanding of the meaning of the federal “employment verification system.” The sole function of that system is to establish that an employee is not barred from working in this country due to alienage. As described in §1324a(b), the system includes the steps that an employee must take to establish that he or she is not prohibited from working, the steps that an employer must take to verify the employee’s status, and certain related matters—such as the preservation and copy- ing of records that are used to show authorization to work. The federal employment verification system does not include things that an employee must or may do to satisfy requirements unrelated to work authorization. And completing tax-withholding documents plays no part in the process of determining whether a person is authorized to work.[7] Instead, those documents are part of the apparatus used to enforce federal and state income tax laws.[8] For all these reasons, there is no express preemption in these cases. IV We therefore proceed to consider respondents’ alternative argument that the Kansas laws, as applied, are preempted by implication. This argument, like all preemption arguments, must be grounded “in the text and structure of the statute at issue.” CSX Transp., Inc. v. Easterwood, 507 U.S. 658, 664 (1993). A Respondents contend, first, that the Kansas statutes, as applied, fall into a field that is implicitly reserved exclusively for federal regulation. In rare cases, the Court has found that Congress “legislated so comprehensively” in a particular field that it “left no room for supplementary state legislation,” R. J. Reynolds Tobacco Co. v. Durham County, 479 U.S. 130, 140 (1986), but that is certainly not the situation here. In order to determine whether Congress has implicitly ousted the States from regulating in a particular field, we must first identify the field in which this is said to have occurred. In their merits brief in this Court, respondents’ primary submission is that IRCA preempts “the field of fraud on the federal employment verification system,” Brief for Respondents 41 (quotation altered), but this argument fails because, as already explained, the submission of tax- withholding forms is not part of that system. At some points in their brief, respondents define the supposedly preempted field more broadly as the “field relating to the federal employment verification system,” id., at 42 (emphasis added); see also id., at 40, but this formulation does not rescue the argument. The submission of tax- withholding forms is fundamentally unrelated to the federal employment verification system because, as explained, those forms serve entirely different functions. The employment verification system is designed to prevent the employment of unauthorized aliens, whereas tax-withholding forms help to enforce income tax laws. And using another person’s Social Security number on tax forms threatens harm that has no connection with immigration law. For instance, using another person’s Social Security number on tax-withholding forms affects the wages reported to federal and state tax authorities. In addition, many benefits—such as those for disability, unemployment, and retirement—are tied to an individual’s work status and income. Inaccurate data also affect the accuracy of a State’s tax information.[9] It is true that employees generally complete their W–4’s and K–4’s at roughly the same time as their I–9’s, but IRCA plainly does not foreclose all state regulation of information that must be supplied as a precondition of employment. New employees may be required by law to provide all sorts of information that has nothing to do with authorization to work in the United States, such as information about age (for jobs with a minimum age requirement), educational degrees, licensing, criminal records, drug use, and personal information needed for a background check. IRCA surely does not preclude States from requiring and regulating the submission of all such information. Respondents suggest that federal law precludes their prosecutions because both the Kansas identity-theft statute and the Kansas false-information statute require proof that the accused engaged in the prohibited conduct for the purpose of getting a “benefit.” Their argument is as follows. Since the benefit alleged by the prosecution in these cases was getting a job, and since the employment verification system concerns authorization to work, the theory of respondents’ prosecutions is related to that system. This argument conflates the benefit that results from complying with the federal employment verification system (verifying authorization to work in the United States) with the benefit of actually getting a job. Submitting W–4’s and K–4’s helped respondents get jobs, but this did not in any way assist them in showing that they were authorized to work in this country. Thus, respondents’ “relating to” argument must be rejected, as must the even broader definitions of the putatively preempted field advanced by respondents at earlier points in this litigation. Contrary to respondents’ suggestion, IRCA certainly does not bar all state regulation regarding the “use of false documents . . . when an unauthorized alien seeks employment.” Brief in Opposition 21. Nor does IRCA exclude a State from the entire “field of employment verification.” Id., at 22. For example, IRCA certainly does not prohibit a public school system from requiring applicants for teaching positions to furnish legitimate teaching certificates. And it does not prevent a police department from verifying that a prospective officer does not have a record of abusive behavior. Respondents argue that field preemption in these cases “follows directly” from our decision in Arizona, 567 U.S. 387, Brief for Respondents 45–46, but that is not so. In Arizona, relying on our prior decision in Hines v. Davidowitz, 312 U.S. 52 (1941), we held that federal immigration law occupied the field of alien registration. 567 U. S., at 400–402. “Federal law,” we observed, “makes a single sovereign responsible for maintaining a comprehensive and unified system to keep track of aliens within the Nation’s borders.” Id., at 401–402. But federal law does not create a comprehensive and unified system regarding the information that a State may require employees to provide. In sum, there is no basis for finding field preemption in these cases. B We likewise see no ground for holding that the Kansas statutes at issue conflict with federal law. It is certainly possible to comply with both IRCA and the Kansas statutes, and respondents do not suggest otherwise. They instead maintain that the Kansas statutes, as applied in their prosecutions, stand as “an obstacle to the accomplishment and execution of the full purposes” of IRCA—one of which is purportedly that the initiation of any legal action against an unauthorized alien for using a false identity in applying for employment should rest exclusively within the prosecutorial discretion of federal authorities. Brief for Respondents 49–55. Allowing Kansas to bring prosecutions like these, according to respondents, would risk upsetting federal enforcement priorities and frustrating federal objectives, such as obtaining the cooperation of unauthorized aliens in making bigger cases. Ibid. Respondents analogize these cases to our holding in Arizona, 567 U. S., at 404–407—that a state law making it a crime for an unauthorized alien to obtain employment conflicted with IRCA, which does not criminalize that conduct—but respondents’ analogy is unsound. In Arizona, the Court inferred that Congress had made a considered decision that it was inadvisable to criminalize the conduct in question. In effect, the Court concluded that IRCA implicitly conferred a right to be free of criminal (as opposed to civil) penalties for working illegally, and thus a state law making it a crime to engage in that conduct conflicted with this federal right. Nothing similar is involved here. In enacting IRCA, Congress did not decide that an unauthorized alien who uses a false identity on tax-withholding forms should not face criminal prosecution. On the contrary, federal law makes it a crime to use fraudulent information on a W–4. 26 U. S. C. §7205. The mere fact that state laws like the Kansas provisions at issue overlap to some degree with federal criminal provisions does not even begin to make a case for conflict preemption. From the beginning of our country, criminal law enforcement has been primarily a responsibility of the States, and that remains true today. In recent times, the reach of federal criminal law has expanded, and there are now many instances in which a prosecution for a particular course of conduct could be brought by either federal or state prosecutors. Our federal system would be turned upside down if we were to hold that federal criminal law preempts state law whenever they overlap, and there is no basis for inferring that federal criminal statutes preempt state laws whenever they overlap. Indeed, in the vast majority of cases where federal and state laws overlap, allowing the States to prosecute is entirely consistent with federal interests. In the present cases, there is certainly no suggestion that the Kansas prosecutions frustrated any federal interests. Federal authorities played a role in all three cases, and the Federal Government fully supports Kansas’s position in this Court. In the end, however, the possibility that federal enforcement priorities might be upset is not enough to provide a basis for preemption. The Supremacy Clause gives priority to “the Laws of the United States,” not the criminal law enforcement priorities or preferences of federal officers. Art. VI, cl. 2. Finally, contrary to respondents’ suggestion, these cases are very different from Buckman Co. v. Plaintiffs’ Legal Comm., 531 U.S. 341 (2001), and Wisconsin Dept. of Industry v. Gould Inc., 475 U.S. 282 (1986). In Buckman Co., the preempted state tort claim for fraud on the Food and Drug Administration threatened serious disruption of the sensitive and highly technical process of approving medical devices. 531 U. S., at 347–353. In these cases, the state prosecutions posed no comparable risk. In Gould, the decision rested on a special preemption rule governing state laws regulating matters that the National Labor Relations Act “protects, prohibits, or arguably protects.” 475 U. S., at 286–289; San Diego Building Trades Council v. Garmon, 359 U.S. 236, 246 (1959). No similar rule is operative or appropriate here. * * * For these reasons, the judgments of the Supreme Court of Kansas are reversed, and these cases are remanded for further proceedings not inconsistent with this opinion. It is so ordered. Notes 1 In Morales’s bench trial, the State also introduced into evidence his I–9 and a photocopy of a permanent resident card and Social Security card that was appended to his I–9. App. 152–154, 178–179. The trial court, however, explicitly assured Morales that it would not make any findings based on the I–9, and defense counsel did not further object to the introduction of the I–9 into evidence. Id., at 150–151. Before the state appellate courts, Morales did not argue that admitting the I–9 and photocopy was error. Nor did his brief in opposition to certiorari argue that the admission of these exhibits provided a ground for relief under federal law. See this Court’s Rule 15.2. 2 Of course, a considerate sender may remember to put the addresses in the BCC line. 3 Twentieth Century Almanac 405 (R. Ferrell & J. Bowman eds. 1984); NASA, The First Person on the Moon (last updated Apr. 9, 2009), www.nasa.gov/audience/forstudents/k-4/stories/first-person-on-moon.html. 4 Neil Steps on the Moon, Wapakoneta Daily News, July 21, 1969, p. 1, https://blogs.loc.gov/headlinesandheroes/2019/08/newspaper-coverage-of-one-giant-leap-for-mankind. 5 See The First Steps: Luna Took the Envoys of the Earth, Izvestia, Moscow Evening ed., Jul. 21, 1969, p. 1 (transl.); NASA, Astronautics and Aeronautics, 1969: Chronology on Science, Technology, and Policy 233 (NASA SP–4014 1970); see also McFall-Johnsen, Newspaper Front Pages From 50 Years Ago Reveal How the World Reacted to the Apollo 11 Moon Landing, Business Insider US, July 20, 2019, http://www.businessinsider.com/apollo-11-moon-landing-newspaper-front-pages-2019-7/. 6 This provision refers to “the system,” but it is apparent that this means “the employment verification system,” which is described in some detail in §1324a(b). There is no other system to which this reference could plausibly refer. 7 Moreover, these documents are not always submitted when an employee begins a job. Instead, new W–4’s and K–4’s may be, and often are, completed at later dates when an employee wishes to make changes that affect the amount of withholding. 26 CFR §31.3402(f )(2)-1; IRS, Publication 505: Tax Withholding and Estimated Tax 3 (May 15, 2019) (“During the year, changes may occur . . . . When this happens, you may need to give your employer a new Form W–4 . . . . Otherwise, if you want to change your withholding allowances for any reason, you can generally do that whenever you wish”); Kansas Dept. of Revenue, Kansas Withholding Form K–4, www.ksrevenue.org/k4info.html. 8 Respondents also contend that 18 U. S. C. §1546(c) expressly preempts the relevant Kansas statutes as applied in their prosecutions, but it is impossible to see any basis for that argument in the statutory text. This subsection, which is part of a provision that criminalizes certain conduct relating to immigration and authorization to work, provides that the section “does not prohibit any lawfully authorized investigative, protective, or intelligence activity” of a federal or state law enforcement agency, a federal intelligence agency, or others engaged in certain activity relating to the prosecution of organized crime. How this provision can be seen as expressly barring respondents’ prosecutions is a mystery. 9 See, e.g., Kansas Dept. of Revenue, Annual Reports, www.ksrevenue.org/prannualreport.html; Kansas Dept. of Revenue, Tax Fraud Enforcement, www.ksrevenue.org/taxfraud.html.

589.US.2019_18-556

A Kansas deputy sheriff ran a license plate check on a pickup truck, discovering that the truck belonged to respondent Glover and that Glover’s driver’s license had been revoked. The deputy pulled the truck over because he assumed that Glover was driving. Glover was in fact driving and was charged with driving as a habitual violator. He moved to suppress all evidence from the stop, claiming that the deputy lacked reasonable suspicion. The District Court granted the motion, but the Court of Appeals reversed. The Kansas Supreme Court in turn reversed, holding that the deputy violated the Fourth Amendment by stopping Glover without reasonable suspicion of criminal activity. Held: When the officer lacks information negating an inference that the owner is driving the vehicle, an investigative traffic stop made after running a vehicle’s license plate and learning that the registered owner’s driver’s license has been revoked is reasonable under the Fourth Amendment. Pp. 3–10. (a) An officer may initiate a brief investigative traffic stop when he has “a particularized and objective basis” to suspect legal wrongdoing. United States v. Cortez, 449 U.S. 411, 417. The level of suspicion required is less than that necessary for probable cause and “depends on ‘ “the factual and practical considerations of everyday life on which reasonable and prudent men, not legal technicians, act.” ’ ” Prado Navarette v. California, 572 U.S. 393, 402. Courts must therefore permit officers to make “commonsense judgments and inferences about human behavior.” Illinois v. Wardlow, 528 U.S. 119, 125. P. 3. (b) Here, the deputy’s commonsense inference that the owner of a vehicle was likely the vehicle’s driver provided more than reasonable suspicion to initiate the stop. That inference is not made unreasonable merely because a vehicle’s driver is not always its registered owner or because Glover had a revoked license. Though common sense suffices to justify the officer’s inference, empirical studies demonstrate that drivers with suspended or revoked licenses frequently continue to drive. And Kansas’ license-revocation scheme, which covers drivers who have already demonstrated a disregard for the law or are categorically unfit to drive, reinforces the reasonableness of the inference that an individual with a revoked license will continue to drive. Pp. 4–6. (c) Glover’s counterarguments are unpersuasive. He argues that the deputy’s inference was unreasonable because it was not grounded in his law enforcement training or experience. Such a requirement, however, is inconsistent with this Court’s Fourth Amendment jurisprudence. See, e.g., Navarette, 572 U. S., at 402. It would also place the burden on police officers to justify their inferences by referring to training materials or experience, and it would foreclose their ability to rely on common sense obtained outside of their work duties. Glover’s argument that Kansas’ view would permit officers to base reasonable suspicion exclusively on probabilities also carries little force. Officers, like jurors, may rely on probabilities in the reasonable suspicion context. See, e.g., United States v. Sokolow, 490 U.S. 1, 8–9. Moreover, the deputy here did more than that: He combined facts obtained from a database and commonsense judgments to form a reasonable suspicion that a specific individual was potentially engaged in specific criminal activity. Pp. 6–8. (d) The scope of this holding is narrow. The reasonable suspicion standard “ ‘takes into account the totality of the circumstances.’ ” Navarette, 572 U. S., at 397. The presence of additional facts might dispel reasonable suspicion, but here, the deputy possessed no information sufficient to rebut the reasonable inference that Glover was driving his own truck. P. 9. 308 Kan. 590, 422 P.3d 64, reversed and remanded. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Alito, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Kagan, J., filed a concurring opinion, in which Ginsburg, J., joined. Sotomayor, J., filed a dissenting opinion.

This case presents the question whether a police officer violates the Fourth Amendment by initiating an investigative traffic stop after running a vehicle’s license plate and learning that the registered owner has a revoked driver’s license. We hold that when the officer lacks information negating an inference that the owner is the driver of the vehicle, the stop is reasonable. I Kansas charged respondent Charles Glover, Jr., with driving as a habitual violator after a traffic stop revealed that he was driving with a revoked license. See Kan. Stat. Ann. §8–285(a)(3) (2001). Glover filed a motion to suppress all evidence seized during the stop, claiming that the officer lacked reasonable suspicion. Neither Glover nor the police officer testified at the suppression hearing. Instead, the parties stipulated to the following facts: “1. Deputy Mark Mehrer is a certified law enforcement officer employed by the Douglas County Kansas Sheriff ’s Office. 2. On April 28, 2016, Deputy Mehrer was on routine patrol in Douglas County when he observed a 1995 Chevrolet 1500 pickup truck with Kansas plate 295ATJ. 3. Deputy Mehrer ran Kansas plate 295ATJ through the Kansas Department of Revenue’s file service. The registration came back to a 1995 Chevrolet 1500 pickup truck. 4. Kansas Department of Revenue files indicated the truck was registered to Charles Glover Jr. The files also indicated that Mr. Glover had a revoked driver’s license in the State of Kansas. 5. Deputy Mehrer assumed the registered owner of the truck was also the driver, Charles Glover Jr. 6. Deputy Mehrer did not observe any traffic infractions, and did not attempt to identify the driver [of] the truck. Based solely on the information that the registered owner of the truck was revoked, Deputy Mehrer initiated a traffic stop. 7. The driver of the truck was identified as the defendant, Charles Glover Jr.” App. to Pet. for Cert. 60–61. The District Court granted Glover’s motion to suppress. The Court of Appeals reversed, holding that “it was reasonable for [Deputy] Mehrer to infer that the driver was the owner of the vehicle” because “there were specific and articulable facts from which the officer’s common-sense inference gave rise to a reasonable suspicion.” 54 Kan. App. 2d 377, 385, 400 P.3d 182, 188 (2017). The Kansas Supreme Court reversed. According to the court, Deputy Mehrer did not have reasonable suspicion because his inference that Glover was behind the wheel amounted to “only a hunch” that Glover was engaging in criminal activity. 308 Kan. 590, 591, 422 P.3d 64, 66 (2018). The court further explained that Deputy Mehrer’s “hunch” involved “applying and stacking unstated assumptions that are unreasonable without further factual basis,” namely, that “the registered owner was likely the primary driver of the vehicle” and that “the owner will likely disregard the suspension or revocation order and continue to drive.” Id., at 595–597, 422 P. 3d, at 68–70. We granted Kansas’ petition for a writ of certiorari, 587 U. S. ___ (2019), and now reverse. II Under this Court’s precedents, the Fourth Amendment permits an officer to initiate a brief investigative traffic stop when he has “a particularized and objective basis for suspecting the particular person stopped of criminal activity.” United States v. Cortez, 449 U.S. 411, 417–418 (1981); see also Terry v. Ohio, 392 U.S. 1, 21–22 (1968). “Although a mere ‘hunch’ does not create reasonable suspicion, the level of suspicion the standard requires is considerably less than proof of wrongdoing by a preponderance of the evidence, and obviously less than is necessary for probable cause.” Prado Navarette v. California, 572 U.S. 393, 397 (2014) (quotation altered); United States v. Sokolow, 490 U.S. 1, 7 (1989). Because it is a “less demanding” standard, “reasonable suspicion can be established with information that is different in quantity or content than that required to establish probable cause.” Alabama v. White, 496 U.S. 325, 330 (1990). The standard “depends on the factual and practical considerations of everyday life on which reasonable and prudent men, not legal technicians, act.” Navarette, supra, at 402 (quoting Ornelas v. United States, 517 U.S. 690, 695 (1996) (emphasis added; internal quotation marks omitted)). Courts “cannot reasonably demand scientific certainty . . . where none exists.” Illinois v. Wardlow, 528 U.S. 119, 125 (2000). Rather, they must permit officers to make “commonsense judgments and inferences about human behavior.” Ibid.; see also Navarette, supra, at 403 (noting that an officer “ ‘need not rule out the possibility of innocent conduct’ ”). III We have previously recognized that States have a “vital interest in ensuring that only those qualified to do so are permitted to operate motor vehicles [and] that licensing, registration, and vehicle inspection requirements are being observed.” Delaware v. Prouse, 440 U.S. 648, 658 (1979). With this in mind, we turn to whether the facts known to Deputy Mehrer at the time of the stop gave rise to reason- able suspicion. We conclude that they did. Before initiating the stop, Deputy Mehrer observed an individual operating a 1995 Chevrolet 1500 pickup truck with Kansas plate 295ATJ. He also knew that the registered owner of the truck had a revoked license and that the model of the truck matched the observed vehicle. From these three facts, Deputy Mehrer drew the commonsense inference that Glover was likely the driver of the vehicle, which provided more than reasonable suspicion to initiate the stop. The fact that the registered owner of a vehicle is not always the driver of the vehicle does not negate the reason- ableness of Deputy Mehrer’s inference. Such is the case with all reasonable inferences. The reasonable suspicion inquiry “falls considerably short” of 51% accuracy, see United States v. Arvizu, 534 U.S. 266, 274 (2002), for, as we have explained, “[t]o be reasonable is not to be perfect,” Heien v. North Carolina, 574 U.S. 54, 60 (2014). Glover’s revoked license does not render Deputy Mehrer’s inference unreasonable either. Empirical studies demonstrate what common experience readily reveals: Drivers with revoked licenses frequently continue to drive and therefore to pose safety risks to other motorists and pedestrians. See, e.g., 2 T. Neuman et al., National Coop. Hwy. Research Program Report 500: A Guide for Addressing Collisions Involving Unlicensed Drivers and Drivers With Suspended or Revoked Licenses, p. III–1 (2003) (noting that 75% of drivers with suspended or revoked licenses continue to drive); National Hwy. and Traffic Safety Admin., Research Note: Driver License Compliance Status in Fatal Crashes 2 (Oct. 2014) (noting that approximately 19% of motor vehicle fatalities from 2008–2012 “involved drivers with invalid licenses”). Although common sense suffices to justify this inference, Kansas law reinforces that it is reasonable to infer that an individual with a revoked license may continue driving. The State’s license-revocation scheme covers drivers who have already demonstrated a disregard for the law or are categorically unfit to drive. The Division of Vehicles of the Kansas Department of Revenue (Division) “shall” revoke a driver’s license upon certain convictions for involuntary manslaughter, vehicular homicide, battery, reckless driving, fleeing or attempting to elude a police officer, or conviction of a felony in which a motor vehicle is used. Kan. Stat. Ann. §§8–254(a), 8–252. Reckless driving is defined as “driv[ing] any vehicle in willful or wanton disregard for the safety of persons or property.” §8–1566(a). The Division also has discretion to revoke a license if a driver “[h]as been convicted with such frequency of serious offenses against traffic regulations governing the movement of vehicles as to indicate a disrespect for traffic laws and a disregard for the safety of other persons on the highways,” “has been convicted of three or more moving traffic violations committed on separate occasions within a 12-month period,” “is incompetent to drive a motor vehicle,” or “has been convicted of a moving traffic violation, committed at a time when the person’s driving privileges were restricted, suspended[,] or revoked.” §§8–255(a)(1)–(4). Other reasons include violating license restrictions, §8–245(c), being under house arrest, §21–6609(c), and being a habitual violator, §8–286, which Kansas defines as a resident or nonresident who has been convicted three or more times within the past five years of certain enumerated driving offenses, §8–285. The concerns motivating the State’s various grounds for revocation lend further credence to the inference that a registered owner with a revoked Kansas driver’s license might be the one driving the vehicle. IV Glover and the dissent respond with two arguments as to why Deputy Mehrer lacked reasonable suspicion. Neither is persuasive. A First, Glover and the dissent argue that Deputy Mehrer’s inference was unreasonable because it was not grounded in his law enforcement training or experience. Nothing in our Fourth Amendment precedent supports the notion that, in determining whether reasonable suspicion exists, an officer can draw inferences based on knowledge gained only through law enforcement training and experience. We have repeatedly recognized the opposite. In Navarette, we noted a number of behaviors—including driving in the median, crossing the center line on a highway, and swerving—that as a matter of common sense provide “sound indicia of drunk driving.” 572 U. S., at 402. In Wardlow, we made the unremarkable observation that “[h]eadlong flight—wherever it occurs—is the consummate act of evasion” and therefore could factor into a police officer’s reasonable suspicion determination. 528 U. S., at 124. And in Sokolow, we recognized that the defendant’s method of payment for an airplane ticket contributed to the agents’ reasonable suspicion of drug trafficking because we “fe[lt] confident” that “[m]ost business travelers . . . purchase airline tickets by credit card or check” rather than cash. 490 U. S., at 8–9. So too here. The inference that the driver of a car is its registered owner does not require any specialized training; rather, it is a reasonable inference made by ordinary people on a daily basis. The dissent reads our cases differently, contending that they permit an officer to use only the common sense derived from his “experiences in law enforcement.” Post, at 5 (opinion of Sotomayor, J.). Such a standard defies the “common sense” understanding of common sense, i.e., information that is accessible to people generally, not just some specialized subset of society. More importantly, this standard appears nowhere in our precedent. In fact, we have stated that reasonable suspicion is an “abstract” concept that cannot be reduced to “a neat set of legal rules,” Arvizu, 534 U. S., at 274 (internal quotation marks omitted), and we have repeatedly rejected courts’ efforts to impose a rigid structure on the concept of reasonableness, ibid.; Sokolow, 490 U. S., at 7–8. This is precisely what the dissent’s rule would do by insisting that officers must be treated as bifurcated persons, completely precluded from drawing factual inferences based on the commonly held knowledge they have acquired in their everyday lives. The dissent’s rule would also impose on police the burden of pointing to specific training materials or field experiences justifying reasonable suspicion for the myriad infractions in municipal criminal codes. And by removing common sense as a source of evidence, the dissent would considerably narrow the daylight between the showing required for probable cause and the “less stringent” showing required for reasonable suspicion. Prouse, 440 U. S., at 654; see White, 496 U. S., at 330. Finally, it would impermissibly tie a traffic stop’s validity to the officer’s length of service. See Devenpeck v. Alford, 543 U.S. 146, 154 (2004). Such requirements are inconsistent with our Fourth Amendment jurisprudence, and we decline to adopt them here. In reaching this conclusion, we in no way minimize the significant role that specialized training and experience routinely play in law enforcement investigations. See, e.g., Arvizu, 534 U. S., at 273–274. We simply hold that such experience is not required in every instance. B Glover and the dissent also contend that adopting Kansas’ view would eviscerate the need for officers to base reasonable suspicion on “specific and articulable facts” particularized to the individual, see Terry, 392 U. S., at 21, because police could instead rely exclusively on probabilities. Their argument carries little force. As an initial matter, we have previously stated that officers, like jurors, may rely on probabilities in the reasonable suspicion context. See Sokolow, 490 U. S., at 8–9; Cortez, 449 U. S., at 418. Moreover, as explained above, Deputy Mehrer did not rely exclusively on probabilities. He knew that the license plate was linked to a truck matching the observed vehicle and that the registered owner of the vehicle had a revoked license. Based on these minimal facts, he used common sense to form a reasonable suspicion that a specific individual was potentially engaged in specific criminal activity—driving with a revoked license. Traffic stops of this nature do not delegate to officers “broad and unlimited discretion” to stop drivers at random. United States v. Brignoni-Ponce, 422 U.S. 873, 882 (1975). Nor do they allow officers to stop drivers whose conduct is no different from any other driver’s. See Brown v. Texas, 443 U.S. 47, 52 (1979). Accordingly, combining database information and commonsense judgments in this context is fully consonant with this Court’s Fourth Amendment precedents.[1] V This Court’s precedents have repeatedly affirmed that “ ‘the ultimate touchstone of the Fourth Amendment is “reasonableness.” ’ ” Heien, 574 U. S., at 60 (quoting Riley v. California, 573 U.S. 373, 381 (2014)). Under the totality of the circumstances of this case, Deputy Mehrer drew an entirely reasonable inference that Glover was driving while his license was revoked. We emphasize the narrow scope of our holding. Like all seizures, “[t]he officer’s action must be ‘justified at its inception.’ ” Hiibel v. Sixth Judicial Dist. Court of Nev., Humboldt Cty., 542 U.S. 177, 185 (2004) (quoting United States v. Sharpe, 470 U.S. 675, 682 (1985)). “The standard takes into account the totality of the circumstances—the whole picture.” Navarette, 572 U. S., at 397 (internal quotation marks omitted). As a result, the presence of additional facts might dispel reasonable suspicion. See Terry, supra, at 28. For example, if an officer knows that the registered owner of the vehicle is in his mid-sixties but observes that the driver is in her mid-twenties, then the totality of the circumstances would not “raise a suspicion that the particular individual being stopped is engaged in wrongdoing.” Cortez, 449 U. S., at 418; Ornelas, 517 U. S., at 696 (“ ‘[e]ach case is to be decided on its own facts and circumstances’ ” (quoting Ker v. California, 374 U.S. 23, 33 (1963))). Here, Deputy Mehrer possessed no exculpatory information—let alone sufficient information to rebut the reasonable inference that Glover was driving his own truck—and thus the stop was justified.[2] * * * For the foregoing reasons, we reverse the judgment of the Kansas Supreme Court, and we remand the case for further proceedings not inconsistent with this opinion. It is so ordered. Notes 1 The dissent contends that this approach “pave[s] the road to finding reasonable suspicion based on nothing more than a demographic profile.” Post, at 6–7 (opinion of Sotomayor, J.). To alleviate any doubt, we reiterate that the Fourth Amendment requires, and Deputy Mehrer had, an individualized suspicion that a particular citizen was engaged in a particular crime. Such a particularized suspicion would be lacking in the dissent’s hypothetical scenario, which, in any event, is already prohibited by our precedents. See United States v. Brignoni-Ponce, 422 U.S. 873, 876 (1975) (holding that it violated the Fourth Amendment to stop and “question [a vehicle’s] occupants [about their immigration status] when the only ground for suspicion [was] that the occupants appear[ed] to be of Mexican ancestry”). 2 The dissent argues that this approach impermissibly places the burden of proof on the individual to negate the inference of reasonable suspicion. Post, at 3. Not so. As the above analysis makes clear, it is the information possessed by the officer at the time of the stop, not any information offered by the individual after the fact, that can negate the inference.

590.US.2019_18-1059

During former New Jersey Governor Chris Christie’s 2013 reelection campaign, his Deputy Chief of Staff, Bridget Anne Kelly, avidly courted Democratic mayors for their endorsements, but Fort Lee’s mayor refused to back the Governor’s campaign. Determined to punish the mayor, Kelly, Port Authority Deputy Executive Director William Baroni, and another Port Authority official, David Wildstein, decided to reduce from three to one the number of lanes long reserved at the George Washington Bridge’s toll plaza for Fort Lee’s morning commuters. To disguise their efforts at political retribution, Wildstein devised a cover story: The lane realignment was for a traffic study. As part of that cover story, the defendants asked Port Authority traffic engineers to collect some numbers about the effect of the changes. At the suggestion of a Port Authority manager, they also agreed to pay an extra toll collector overtime so that Fort Lee’s one remaining lane would not be shut down if the collector on duty needed a break. The lane realignment caused four days of gridlock in Fort Lee, and only ended when the Port Authority’s Executive Director learned of the scheme. Baroni and Kelly were convicted in federal court of wire fraud, fraud on a federally funded program or entity (the Port Authority), and conspiracy to commit each of those crimes. The Third Circuit affirmed. Held: Because the scheme here did not aim to obtain money or property, Baroni and Kelly could not have violated the federal-program fraud or wire fraud laws. The federal wire fraud statute makes it a crime to effect (with the use of the wires) “any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises.” 18 U. S. C. §1343. Similarly, the federal-program fraud statute bars “obtain[ing] by fraud” the “property” (including money) of a federally funded program or entity. §666(a)(1)(A). These statutes are “limited in scope to the protection of property rights,” and do not authorize federal prosecutors to “set[ ] standards of disclosure and good government for local and state officials.” McNally v. United States, 483 U.S. 350, 360. So under either provision, the Government had to show not only that Baroni and Kelly engaged in deception, but that an object of their fraud was money or property. Cleveland v. United States, 531 U.S. 12, 26. The Government argues that the scheme had the object of obtaining the Port Authority’s money or property in two ways. First, the Government claims that Baroni and Kelly sought to commandeer part of the Bridge itself by taking control of its physical lanes. Second, the Government asserts that the defendants aimed to deprive the Port Authority of the costs of compensating the traffic engineers and back-up toll collectors. For different reasons, neither of these theories can sustain the verdicts. Baroni’s and Kelly’s realignment of the access lanes was an exercise of regulatory power—a reallocation of the lanes between different groups of drivers. This Court has already held that a scheme to alter such a regulatory choice is not one to take the government’s property. Id., at 23. And while a government’s right to its employees’ time and labor is a property interest, the prosecution must also show that it is an “object of the fraud.” Pasquantino v. United States, 544 U.S. 349, 355. Here, the time and labor of the Port Authority employees were just the implementation costs of the defendants’ scheme to reallocate the Bridge’s lanes—an incidental (even if foreseen) byproduct of their regulatory object. Neither defendant sought to obtain the services that the employees provided. Pp. 6–13. 909 F.3d 550, reversed and remanded. Kagan, J., delivered the opinion for a unanimous Court.

For four days in September 2013, traffic ground to a halt in Fort Lee, New Jersey. The cause was an unannounced realignment of 12 toll lanes leading to the George Washington Bridge, an entryway into Manhattan administered by the Port Authority of New York and New Jersey. For decades, three of those access lanes had been reserved during morning rush hour for commuters coming from the streets of Fort Lee. But on these four days—with predictable consequences—only a single lane was set aside. The public officials who ordered that change claimed they were reducing the number of dedicated lanes to conduct a traffic study. In fact, they did so for a political reason—to punish the mayor of Fort Lee for refusing to support the New Jersey Governor’s reelection bid. Exposure of their behavior led to the criminal convictions we review here. The Government charged the responsible officials under the federal statutes prohibiting wire fraud and fraud on a federally funded program or entity. See 18 U. S. C. §§1343, 666(a)(1)(A). Both those laws target fraudulent schemes for obtaining property. See §1343 (barring fraudulent schemes “for obtaining money or property”); §666(a)(1)(A) (making it a crime to “obtain[ ] by fraud . . . property”). The jury convicted the defendants, and the lower courts upheld the verdicts. The question presented is whether the defendants committed property fraud. The evidence the jury heard no doubt shows wrongdoing—deception, corruption, abuse of power. But the federal fraud statutes at issue do not criminalize all such conduct. Under settled precedent, the officials could violate those laws only if an object of their dishonesty was to obtain the Port Authority’s money or property. The Government contends it was, because the officials sought both to “commandeer” the Bridge’s access lanes and to divert the wage labor of the Port Authority employees used in that effort. Tr. of Oral Arg. 58. We disagree. The realignment of the toll lanes was an exercise of regulatory power—something this Court has already held fails to meet the statutes’ property requirement. And the employees’ labor was just the incidental cost of that regulation, rather than itself an object of the officials’ scheme. We therefore reverse the convictions. I The setting of this case is the George Washington Bridge. Running between Fort Lee and Manhattan, it is the busiest motor-vehicle bridge in the world. Twelve lanes with tollbooths feed onto the Bridge’s upper level from the Fort Lee side. Decades ago, the then-Governor of New Jersey committed to a set allocation of those lanes for the morning commute. And (save for the four days soon described) his plan has lasted to this day. Under the arrangement, nine of the lanes carry traffic coming from nearby highways. The three remaining lanes, designated by a long line of traffic cones laid down each morning, serve only cars coming from Fort Lee. The case’s cast of characters are public officials who worked at or with the Port Authority and had political ties to New Jersey’s then-Governor Chris Christie. The Port Authority is a bi-state agency that manages bridges, tunnels, airports, and other transportation facilities in New York and New Jersey. At the time relevant here, William Baroni was its Deputy Executive Director, an appointee of Governor Christie and the highest ranking New Jersey official in the agency. Together with the Executive Director (a New York appointee), he oversaw “all aspects of the Port Authority’s business,” including operation of the George Washington Bridge. App. 21 (indictment). David Wildstein (who became the Government’s star witness) functioned as Baroni’s chief of staff. And Bridget Anne Kelly was a Deputy Chief of Staff to Governor Christie with special responsibility for managing his relations with local officials. She often worked hand-in-hand with Baroni and Wildstein to deploy the Port Authority’s resources in ways that would encourage mayors and other local figures to support the Governor. The fateful lane change arose out of one mayor’s resistance to such blandishments. In 2013, Governor Christie was up for reelection, and he wanted to notch a large, bipartisan victory as he ramped up for a presidential campaign. On his behalf, Kelly avidly courted Democratic mayors for their endorsements—among them, Mark Sokolich of Fort Lee. As a result, that town received some valuable benefits from the Port Authority, including an expensive shuttle-bus service. But that summer, Mayor Sokolich informed Kelly’s office that he would not back the Governor’s campaign. A frustrated Kelly reached out to Wildstein for ideas on how to respond. He suggested that getting rid of the dedicated Fort Lee lanes on the Bridge’s toll plaza would cause rush-hour traffic to back up onto local streets, leading to gridlock there. Kelly agreed to the idea in an admirably concise e-mail: “Time for some traffic problems in Fort Lee.” App. 917 (trial exhibit). In a later phone conversation, Kelly confirmed to Wildstein that she wanted to “creat[e] a traffic jam that would punish” Mayor Sokolich and “send him a message.” Id., at 254 (Wildstein testimony). And after Wildstein relayed those communications, Baroni gave the needed sign-off. To complete the scheme, Wildstein then devised “a cover story”—that the lane change was part of a traffic study, intended to assess whether to retain the dedicated Fort Lee lanes in the future. Id., at 264. Wildstein, Baroni, and Kelly all agreed to use that “public policy” justification when speaking with the media, local officials, and the Port Authority’s own employees. Id., at 265. And to give their story credibility, Wildstein in fact told the Port Authority’s engineers to collect “some numbers on how[ ] far back the traffic was delayed.” Id., at 305. That inquiry bore little resemblance to the Port Authority’s usual traffic studies. According to one engineer’s trial testimony, the Port Authority never closes lanes to study traffic patterns, because “computer-generated model[ing]” can itself predict the effect of such actions. Id., at 484 (testimony of Umang Patel); see id., at 473–474 (similar testimony of Victor Chung). And the information that the Port Authority’s engineers collected on this singular occasion was mostly “not useful” and “discarded.” Id., at 484–485 (Patel testimony). Nor did Wildstein or Baroni show any interest in the data. They never asked to review what the engineers had found; indeed, they learned of the results only weeks later, after a journalist filed a public-records request. So although the engineers spent valuable time assessing the lane change, their work was to no practical effect. Baroni, Wildstein, and Kelly also agreed to incur another cost—for extra toll collectors—in pursuit of their object. Wildstein’s initial thought was to eliminate all three dedicated lanes by not laying down any traffic cones, thus turning the whole toll plaza into a free-for-all. But the Port Authority’s chief engineer told him that without the cones “there would be a substantial risk of sideswipe crashes” involving cars coming into the area from different directions. Id., at 284 (Wildstein testimony). So Wildstein went back to Baroni and Kelly and got their approval to keep one lane reserved for Fort Lee traffic. That solution, though, raised another complication. Ordinarily, if a toll collector on a Fort Lee lane has to take a break, he closes his booth, and drivers use one of the other two lanes. Under the one-lane plan, of course, that would be impossible. So the Bridge manager told Wildstein that to make the scheme work, “an extra toll collector” would always have to be “on call” to relieve the regular collector when he went on break. Id., at 303. Once again, Wildstein took the news to Baroni and Kelly. Baroni thought it was “funny,” remarking that “only at the Port Authority would [you] have to pay a toll collector to just sit there and wait.” Ibid. Still, he and Kelly gave the okay. The plan was now ready, and on September 9 it went into effect. Without advance notice and on the (traffic-heavy) first day of school, Port Authority employees placed traffic cones two lanes further to the right than usual, restricting cars from Fort Lee to a single lane. Almost immediately, the town’s streets came to a standstill. According to the Fort Lee Chief of Police, the traffic rivaled that of 9/11, when the George Washington Bridge had shut down. School buses stood in place for hours. An ambulance struggled to reach the victim of a heart attack; police had trouble responding to a report of a missing child. Mayor Sokolich tried to reach Baroni, leaving a message that the call was about an “urgent matter of public safety.” Id., at 323. Yet Baroni failed to return that call or any other: He had agreed with Wildstein and Kelly that they should all maintain “radio silence.” Id., at 270. A text from the Mayor to Baroni about the locked-in school buses—also unanswered—went around the horn to Wildstein and Kelly. The last replied: “Is it wrong that I am smiling?” Id., at 990 (Kelly text message). The three merrily kept the lane realignment in place for another three days. It ended only when the Port Authority’s Executive Director found out what had happened and reversed what he called their “abusive decision.” Id., at 963 (e-mail of Patrick Foye). The fallout from the scheme was swift and severe. Baroni, Kelly, and Wildstein all lost their jobs. More to the point here, they all ran afoul of federal prosecutors. Wildstein pleaded guilty to conspiracy charges and agreed to cooperate with the Government. Baroni and Kelly went to trial on charges of wire fraud, fraud on a federally funded program or entity (the Port Authority), and conspiracy to commit each of those crimes. The jury found both of them guilty on all counts. The Court of Appeals for the Third Circuit affirmed, rejecting Baroni’s and Kelly’s claim that the evidence was insufficient to support their convictions. See United States v. Baroni, 909 F.3d 550, 560–579 (2018). We granted certiorari. 588 U. S. ___ (2019). II The Government in this case needed to prove property fraud. The federal wire fraud statute makes it a crime to effect (with use of the wires) “any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises.” 18 U. S. C. §1343. Construing that disjunctive language as a unitary whole, this Court has held that “the money-or- property requirement of the latter phrase” also limits the former. McNally v. United States, 483 U.S. 350, 358 (1987). The wire fraud statute thus prohibits only deceptive “schemes to deprive [the victim of] money or property.” Id., at 356. Similarly, the federal-program fraud statute bars “obtain[ing] by fraud” the “property” (including money) of a federally funded program or entity like the Port Authority. §666(a)(1)(A). So under either provision, the Government had to show not only that Baroni and Kelly engaged in deception, but that an “object of the[ir] fraud [was] ‘property.’ ” Cleveland v. United States, 531 U.S. 12, 26 (2000).[1] That requirement, this Court has made clear, prevents these statutes from criminalizing all acts of dishonesty by state and local officials. Some decades ago, courts of appeals often construed the federal fraud laws to “proscribe[ ] schemes to defraud citizens of their intangible rights to honest and impartial government.” McNally, 483 U. S., at 355. This Court declined to go along. The fraud statutes, we held in McNally, were “limited in scope to the protection of property rights.” Id., at 360. They did not authorize federal prosecutors to “set[ ] standards of disclosure and good government for local and state officials.” Ibid. Congress responded to that decision by enacting a law barring fraudulent schemes “to deprive another of the intangible right of honest services”—regardless of whether the scheme sought to divest the victim of any property. §1346. But the vagueness of that language led this Court to adopt “a limiting construction,” confining the statute to schemes involving bribes or kickbacks. Skilling v. United States, 561 U.S. 358, 405, 410 (2010). We specifically rejected a proposal to construe the statute as encompassing “undisclosed self-dealing by a public official,” even when he hid financial interests. Id., at 409. The upshot is that federal fraud law leaves much public corruption to the States (or their electorates) to rectify. Cf. N. J. Stat. Ann. §2C:30–2 (West 2016) (prohibiting the unauthorized exercise of official functions). Save for bribes or kickbacks (not at issue here), a state or local official’s fraudulent schemes violate that law only when, again, they are “for obtaining money or property.” 18 U. S. C. §1343; see §666(a)(1)(A) (similar). The Government acknowledges this much, but thinks Baroni’s and Kelly’s convictions remain valid. According to the Government’s theory of the case, Baroni and Kelly “used a lie about a fictional traffic study” to achieve their goal of reallocating the Bridge’s toll lanes. Brief for United States 43. The Government accepts that the lie itself—i.e., that the lane change was part of a traffic study, rather than political payback—could not get the prosecution all the way home. See id., at 43–44. As the Government recognizes, the deceit must also have had the “object” of obtaining the Port Authority’s money or property. Id., at 44. The scheme met that requirement, the Government argues, in two ways. First, the Government claims that Baroni and Kelly sought to “commandeer[ ]” part of the Bridge itself—to “take control” of its “physical lanes.” Tr. of Oral Arg. 58–59. Second, the Government asserts that the two defendants aimed to deprive the Port Authority of the costs of compensating the traffic engineers and back-up toll collectors who performed work relating to the lane realignment. On either theory, the Government insists, Baroni’s and Kelly’s scheme targeted “a ‘species of valuable right [or] interest’ that constitutes ‘property’ under the fraud statutes.” Brief for United States 22 (quoting Pasquantino v. United States, 544 U.S. 349, 356 (2005)). We cannot agree. As we explain below, the Government could not have proved—on either of its theories, though for different reasons—that Baroni’s and Kelly’s scheme was “directed at the [Port Authority’s] property.” Brief for United States 44. Baroni and Kelly indeed “plotted to reduce [Fort Lee’s] lanes.” Id., at 34. But that realignment was a quintessential exercise of regulatory power. And this Court has already held that a scheme to alter such a regulatory choice is not one to appropriate the government’s property. See Cleveland, 531 U. S., at 23. By contrast, a scheme to usurp a public employee’s paid time is one to take the government’s property. But Baroni’s and Kelly’s plan never had that as an object. The use of Port Authority employees was incidental to—the mere cost of implementing—the sought-after regulation of the Bridge’s toll lanes. Start with this Court’s decision in Cleveland, which reversed another set of federal fraud convictions based on the distinction between property and regulatory power. The defendant there had engaged in a deceptive scheme to influence, to his own benefit, Louisiana’s issuance of gaming licenses. The Government argued that his fraud aimed to deprive the State of property by altering its licensing decisions. This Court rejected the claim. The State’s “intangible rights of allocation, exclusion, and control”—its prerogatives over who should get a benefit and who should not—do “not create a property interest.” Ibid. Rather, the Court stated, those rights “amount to no more and no less than” the State’s “sovereign power to regulate.” Ibid.; see id., at 20 (“[T]he State’s core concern” in allocating gaming licenses “is regulatory”). Or said another way: The defendant’s fraud “implicate[d] the Government’s role as sovereign” wielding “traditional police powers”—not its role “as property holder.” Id., at 23–24. And so his conduct, however deceitful, was not property fraud. The same is true of the lane realignment. Through that action, Baroni and Kelly changed the traffic flow onto the George Washington Bridge’s tollbooth plaza. Contrary to the Government’s view, the two defendants did not “commandeer” the Bridge’s access lanes (supposing that word bears its normal meaning). They (of course) did not walk away with the lanes; nor did they take the lanes from the Government by converting them to a non-public use. Rather, Baroni and Kelly regulated use of the lanes, as officials responsible for roadways so often do—allocating lanes as between different groups of drivers. To borrow Cleveland’s words, Baroni and Kelly exercised the regulatory rights of “allocation, exclusion, and control”—deciding that drivers from Fort Lee should get two fewer lanes while drivers from nearby highways should get two more. They did so, according to all the Government’s evidence, for bad reasons; and they did so by resorting to lies. But still, what they did was alter a regulatory decision about the toll plaza’s use—in effect, about which drivers had a “license” to use which lanes. And under Cleveland, that run-of-the-mine exercise of regulatory power cannot count as the taking of property. A government’s right to its employees’ time and labor, by contrast, can undergird a property fraud prosecution. Suppose that a mayor uses deception to get “on-the-clock city workers” to renovate his daughter’s new home. United States v. Pabey, 664 F.3d 1084, 1089 (CA7 2011). Or imagine that a city parks commissioner induces his employees into doing gardening work for political contributors. See United States v. Delano, 55 F.3d 720, 723 (CA2 1995). As both defendants agree, the cost of those employees’ services would qualify as an economic loss to a city, sufficient to meet the federal fraud statutes’ property requirement. See Brief for Respondent Baroni 27; Tr. of Oral Arg. 16. No less than if the official took cash out of the city’s bank account would he have deprived the city of a “valuable entitlement.” Pasquantino, 544 U. S., at 357. But that property must play more than some bit part in a scheme: It must be an “object of the fraud.” Id., at 355; see Brief for United States 44; supra, at 6–7. Or put differently, a property fraud conviction cannot stand when the loss to the victim is only an incidental byproduct of the scheme.[2] In the home-and-garden examples cited above, that constraint raised no problem: The entire point of the fraudsters’ plans was to obtain the employees’ services. But now consider the difficulty if the prosecution in Cleveland had raised a similar employee-labor argument. As the Government noted at oral argument here, the fraud on Louisiana’s licensing system doubtless imposed costs calculable in employee time: If nothing else, some state worker had to process each of the fraudster’s falsified applications. But still, the Government acknowledged, those costs were “[i]ncidental.” Tr. of Oral Arg. 63. The object of the scheme was never to get the employees’ labor: It was to get gaming licenses. So the labor costs could not sustain the conviction for property fraud. See id., at 62–63. This case is no different. The time and labor of Port Authority employees were just the implementation costs of the defendants’ scheme to reallocate the Bridge’s access lanes. Or said another way, the labor costs were an incidental (even if foreseen) byproduct of Baroni’s and Kelly’s regulatory object. Neither defendant sought to obtain the services that the employees provided. The back-up toll collectors—whom Baroni joked would just “sit there and wait”—did nothing he or Kelly thought useful. App. 303; see supra, at 5. Indeed, those workers came onto the scene only because the Port Authority’s chief engineer managed to restore one of Fort Lee’s lanes to reduce the risk of traffic accidents. See supra, at 5. In the defendants’ original plan, which scrapped all reserved lanes, there was no reason for extra toll collectors. And similarly, Baroni and Kelly did not hope to obtain the data that the traffic engineers spent their time collecting. By the Government’s own account, the traffic study the defendants used for a cover story was a “sham,” and they never asked to see its results. Brief for United States 4, 32; see supra, at 5. Maybe, as the Government contends, all of this work was “needed” to realize the final plan—“to accomplish what [Baroni and Kelly] were trying to do with the [B]ridge.” Tr. of Oral Arg. 60. Even if so, it would make no difference. Every regulatory decision (think again of Cleveland, see supra, at 11) requires the use of some employee labor. But that does not mean every scheme to alter a regulation has that labor as its object. Baroni’s and Kelly’s plan aimed to impede access from Fort Lee to the George Washington Bridge. The cost of the employee hours spent on implementing that plan was its incidental byproduct. To rule otherwise would undercut this Court’s oft- repeated instruction: Federal prosecutors may not use property fraud statutes to “set[ ] standards of disclosure and good government for local and state officials.” McNally, 483 U. S., at 360; see supra, at 7. Much of governance involves (as it did here) regulatory choice. If U. S. Attorneys could prosecute as property fraud every lie a state or local official tells in making such a decision, the result would be—as Cleveland recognized—“a sweeping expansion of federal criminal jurisdiction.” 531 U. S., at 24. And if those prosecutors could end-run Cleveland just by pointing to the regulation’s incidental costs, the same ballooning of federal power would follow. In effect, the Federal Government could use the criminal law to enforce (its view of ) integrity in broad swaths of state and local policymaking. The property fraud statutes do not countenance that outcome. They do not “proscribe[ ] schemes to defraud citizens of their intangible rights to honest and impartial government.” McNally, 483 U. S., at 355; see supra, at 7. They bar only schemes for obtaining property. III As Kelly’s own lawyer acknowledged, this case involves an “abuse of power.” Tr. of Oral Arg. 19. For no reason other than political payback, Baroni and Kelly used deception to reduce Fort Lee’s access lanes to the George Washington Bridge—and thereby jeopardized the safety of the town’s residents. But not every corrupt act by state or local officials is a federal crime. Because the scheme here did not aim to obtain money or property, Baroni and Kelly could not have violated the federal-program fraud or wire fraud laws. We therefore reverse the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 The conspiracy verdicts raise no separate issue. None of the parties doubts that those convictions stand or fall with the substantive offenses. If there was property fraud here, there was also conspiracy to commit it. But if not, not. 2 Without that rule, as Judge Easterbrook has elaborated, even a practical joke could be a federal felony. See United States v. Walters, 997 F.2d 1219, 1224 (CA7 1993). His example goes: “A [e-mails] B an invitation to a surprise party for their mutual friend C. B drives his car to the place named in the invitation,” thus expending the cost of gasoline. Ibid. “But there is no party; the address is a vacant lot; B is the butt of a joke.” Ibid. Wire fraud? No. And for the reason Judge Easterbrook gave: “[T]he victim’s loss must be an objective of the [deceitful] scheme rather than a byproduct of it.” Id., at 1226.

591.US.2019_19-431

The Patient Protection and Affordable Care Act of 2010 (ACA) requires covered employers to provide women with “preventive care and screenings” without “any cost sharing requirements,” and relies on Preventive Care Guidelines (Guidelines) “supported by the Health Resources and Services Administration” (HRSA) to determine what “preventive care and screenings” includes. 42 U. S. C. §300gg–13(a)(4). Those Guidelines mandate that health plans provide coverage for all Food and Drug Administration approved contraceptive methods. When the Departments of Health and Human Services, Labor, and the Treasury (Departments) incorporated the Guidelines, they also gave HRSA the discretion to exempt religious employers, such as churches, from providing contraceptive coverage. Later, the Departments also promulgated a rule accommodating qualifying religious organizations that allowed them to opt out of coverage by self-certifying that they met certain criteria to their health insurance issuer, which would then exclude contraceptive coverage from the employer’s plan and provide participants with separate payments for contraceptive services without imposing any cost-sharing requirements. Religious entities challenged the rules under the Religious Freedom Restoration Act of 1993 (RFRA). In Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, this Court held that the contraceptive mandate substantially burdened the free exercise of closely held corporations with sincerely held religious objections to providing their employees with certain methods of contraception. And in Zubik v. Burwell, 578 U. S. ___, the Court opted to remand without deciding the RFRA question in cases challenging the self-certification accommodation so that the parties could develop an approach that would accommodate employers’ concerns while providing women full and equal coverage. Under Zubik’s direction and in light of Hobby Lobby’s holding, the Departments promulgated two interim final rules (IFRs). The first significantly expanded the church exemption to include an employer that “objects . . . based on its sincerely held religious beliefs,” “to its establishing, maintaining, providing, offering, or arranging [for] coverage or payments for some or all contraceptive services.” 82 Fed. Reg. 47812. The second created a similar “moral exemption” for employers with sincerely held moral objections to providing some or all forms of contraceptive coverage. The Departments requested post-promulgation comments on both IFRs. Pennsylvania sued, alleging that the IFRs were procedurally and substantively invalid under the Administrative Procedure Act (APA). After the Departments issued final rules, responding to post-promulgation comments but leaving the IFRs largely intact, New Jersey joined Pennsylvania’s suit. Together they filed an amended complaint, alleging that the rules were substantively unlawful because the Departments lacked statutory authority under either the ACA or RFRA to promulgate the exemptions. They also argued that the rules were procedurally defective because the Departments failed to comply with the APA’s notice and comment procedures. The District Court issued a preliminary nationwide injunction against the implementation of the final rules, and the Third Circuit affirmed. Held: 1. The Departments had the authority under the ACA to promulgate the religious and moral exemptions. Pp. 14–22. (a) As legal authority for both exemptions, the Departments invoke §300gg–13(a)(4), which states that group health plans must provide women with “preventive care and screenings . . . as provided for in comprehensive guidelines supported by [HRSA].” The pivotal phrase, “as provided for,” grants sweeping authority to HRSA to define the preventive care that applicable health plans must cover. That same grant of authority empowers it to identify and create exemptions from its own Guidelines. The “fundamental principle of statutory interpretation that ‘absent provision[s] cannot be supplied by the courts,’ ” Rotkiske v. Klemm, 589 U. S. ___, ___ applies not only to adding terms not found in the statute, but also to imposing limits on an agency’s discretion that are not supported by the text, see Watt v. Energy Action Ed. Foundation, 454 U.S. 151, 168. Concerns that the exemptions thwart Congress’ intent by making it significantly harder for interested women to obtain seamless access to contraception without cost-sharing cannot justify supplanting the text’s plain meaning. Even if such concerns are legitimate, they are more properly directed at the regulatory mechanism that Congress put in place. Pp. 14–18. (b) Because the ACA provided a basis for both exemptions, the Court need not decide whether RFRA independently compelled the Departments’ solution. However, the argument that the Departments could not consider RFRA at all is without merit. It is clear from the face of the statute that the contraceptive mandate is capable of violating RFRA. The ACA does not explicitly exempt RFRA, and the regulations implementing the contraceptive mandate qualify as “Federal law” or “the implementation of [Federal] law” under RFRA. §2000bb–3(a). Additionally, this Court stated in Hobby Lobby that the mandate violated RFRA as applied to entities with complicity-based objections. And both Hobby Lobby and Zubik instructed the Departments to consider RFRA going forward. Moreover, in light of the basic requirements of the rulemaking process, the Departments’ failure to discuss RFRA at all when formulating their solution would make them susceptible to claims that the rules were arbitrary and capricious for failing to consider an important aspect of the problem. Pp. 19–22. 2. The rules promulgating the exemptions are free from procedural defects. Pp. 22–26. (a) Respondents claim that because the final rules were preceded by a document entitled “Interim Final Rules with Request for Comments” instead of “General Notice of Proposed Rulemaking,” they are procedurally invalid under the APA. The IFRs’ request for comments readily satisfied the APA notice requirements. And even assuming that the APA requires an agency to publish a document entitled “notice of proposed rulemaking,” there was no “prejudicial error” here, 5 U. S. C. §706. Pp. 22–24. (b) Pointing to the fact that the final rules made only minor alterations to the IFRs, respondents also contend that the final rules are procedurally invalid because nothing in the record suggests that the Departments maintained an open mind during the post-promulgation process. The “open-mindedness” test has no basis in the APA. Each of the APA’s procedural requirements was satisfied: The IFRs provided sufficient notice, §553(b); the Departments “g[a]ve interested persons an opportunity to participate in the rule making through submission of written data, views or arguments,” §553(c); the final rules contained “a concise general statement of their basis and purpose,” ibid.; and they were published more than 30 days before they became effective, §553(d). Pp. 24–26. 930 F.3d 543, reversed and remanded. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Alito, Gorsuch, and Kavanaugh, JJ., joined. Alito, J., filed a concurring opinion, in which Gorsuch, J., joined. Kagan, J., filed an opinion concurring in the judgment, in which Breyer, J., joined. Ginsburg, J., filed a dissenting opinion, in which Sotomayor, J., joined. Notes 1 Together with 19–454, Trump, President of the United States, et al. v. Pennsylvania et al., on certiorari to the same Court.

In these consolidated cases, we decide whether the Government created lawful exemptions from a regulatory requirement implementing the Patient Protection and Affordable Care Act of 2010 (ACA), 124Stat. 119. The requirement at issue obligates certain employers to provide contraceptive coverage to their employees through their group health plans. Though contraceptive coverage is not required by (or even mentioned in) the ACA provision at issue, the Government mandated such coverage by promulgating interim final rules (IFRs) shortly after the ACA’s passage. This requirement is known as the contraceptive mandate. After six years of protracted litigation, the Departments of Health and Human Services, Labor, and the Treasury (Departments)—which jointly administer the relevant ACA provision[1]—exempted certain employers who have religious and conscientious objections from this agency-created mandate. The Third Circuit concluded that the Departments lacked statutory authority to promulgate these exemptions and affirmed the District Court’s nationwide preliminary injunction. This decision was erroneous. We hold that the Departments had the authority to provide exemptions from the regulatory contraceptive requirements for employers with religious and conscientious objections. We accordingly reverse the Third Circuit’s judgment and remand with instructions to dissolve the nationwide preliminary injunction. I The ACA’s contraceptive mandate—a product of agency regulation—has existed for approximately nine years. Litigation surrounding that requirement has lasted nearly as long. In light of this extensive history, we begin by summarizing the relevant background. A The ACA requires covered employers to offer “a group health plan or group health insurance coverage” that provides certain “minimum essential coverage.” 26 U. S. C. §5000A(f )(2); §§4980H(a), (c)(2). Employers who do not comply face hefty penalties, including potential fines of $100 per day for each affected employee. §§4980D(a)–(b); see also Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 696–697 (2014). These cases concern regulations promulgated under a provision of the ACA that requires covered employers to provide women with “preventive care and screenings” without “any cost sharing requirements.” 42 U. S. C. §300gg–13(a)(4).[2] The statute does not define “preventive care and screenings,” nor does it include an exhaustive or illustrative list of such services. Thus, the statute itself does not explicitly require coverage for any specific form of “preventive care.” Hobby Lobby, 573 U. S., at 697. Instead, Congress stated that coverage must include “such additional preventive care and screenings . . . as provided for in comprehensive guidelines supported by the Health Resources and Services Administration” (HRSA), an agency of the Department of Health and Human Services (HHS). §300gg–13(a)(4). At the time of the ACA’s enactment, these guidelines were not yet written. As a result, no specific forms of preventive care or screenings were (or could be) referred to or incorporated by reference. Soon after the ACA’s passage, the Departments began promulgating rules related to §300gg–13(a)(4). But in doing so, the Departments did not proceed through the notice and comment rulemaking process, which the Administrative Procedure Act (APA) often requires before an agency’s regulation can “have the force and effect of law.” Perez v. Mortgage Bankers Assn., 575 U.S. 92, 96 (2015) (internal quotation marks omitted); see also 5 U. S. C. §553. Instead, the Departments invoked the APA’s good cause exception, which permits an agency to dispense with notice and comment and promulgate an IFR that carries immediate legal force. §553(b)(3)(B). The first relevant IFR, promulgated in July 2010, primarily focused on implementing other aspects of §300gg–13. 75 Fed. Reg. 41728. The IFR indicated that HRSA planned to develop its Preventive Care Guidelines (Guidelines) by August 2011. Ibid. However, it did not mention religious exemptions or accommodations of any kind. As anticipated, HRSA released its first set of Guidelines in August 2011. The Guidelines were based on recommendations compiled by the Institute of Medicine (now called the National Academy of Medicine), “a nonprofit group of volunteer advisers.” Hobby Lobby, 573 U. S., at 697. The Guidelines included the contraceptive mandate, which required health plans to provide coverage for all contraceptive methods and sterilization procedures approved by the Food and Drug Administration as well as related education and counseling. 77 Fed. Reg. 8725 (2012). The same day the Guidelines were issued, the Departments amended the 2010 IFR. 76 Fed. Reg. 46621 (2011). When the 2010 IFR was originally published, the Departments began receiving comments from numerous religious employers expressing concern that the Guidelines would “impinge upon their religious freedom” if they included contraception. Id., at 46623. As just stated, the Guidelines ultimately did contain contraceptive coverage, thus making the potential impact on religious freedom a reality. In the amended IFR, the Departments determined that “it [was] appropriate that HRSA . . . tak[e] into account the [mandate’s] effect on certain religious employers” and concluded that HRSA had the discretion to do so through the creation of an exemption. Ibid. The Departments then determined that the exemption should cover religious employers, and they set out a four-part test to identify which employers qualified. The last criterion required the entity to be a church, an integrated auxiliary, a convention or association of churches, or “the exclusively religious activities of any religious order.” Ibid. HRSA created an exemption for these employers the same day. 78 Fed. Reg. 39871 (2013). Because of the narrow focus on churches, this first exemption is known as the church exemption. The Guidelines were scheduled to go into effect for plan years beginning on August 1, 2012. 77 Fed. Reg. 8725–8726. But in February 2012, before the Guidelines took effect, the Departments promulgated a final rule that temporarily prevented the Guidelines from applying to certain religious nonprofits. Specifically, the Departments stated their intent to promulgate additional rules to “accommodat[e] non-exempted, non-profit organizations’ religious objections to covering contraceptive services.” Id., at 8727. Until that rulemaking occurred, the 2012 rule also provided a temporary safe harbor to protect such employers. Ibid. The safe harbor covered nonprofits “whose plans have consistently not covered all or the same subset of contraceptive services for religious reasons.”[3] Thus, the nonprofits who availed themselves of this safe harbor were not subject to the contraceptive mandate when it first became effective. The Departments promulgated another final rule in 2013 that is relevant to these cases in two ways. First, after reiterating that §300gg–13(a)(4) authorizes HRSA “to issue guidelines in a manner that exempts group health plans established or maintained by religious employers,” the Departments “simplif[ied]” and “clarif[ied]” the definition of a religious employer. 78 Fed. Reg. 39873.[4] Second, pursuant to that same authority, the Departments provided the anticipated accommodation for eligible religious organizations, which the regulation defined as organizations that “(1) [o]ppos[e] providing coverage for some or all of the contraceptive services . . . on account of religious objections; (2) [are] organized and operat[e] as . . . nonprofit entit[ies]; (3) hol[d] [themselves] out as . . . religious organization[s]; and (4) self-certif[y] that [they] satisf[y] the first three criteria.” Id., at 39874. The accommodation required an eligible organization to provide a copy of the self-certification form to its health insurance issuer, which in turn would exclude contraceptive coverage from the group health plan and provide payments to beneficiaries for contraceptive services separate from the health plan. Id., at 39878. The Departments stated that the accommodation aimed to “protec[t]” religious organizations “from having to contract, arrange, pay, or refer for [contraceptive] coverage” in a way that was consistent with and did not violate the Religious Freedom Restoration Act of 1993 (RFRA), 107Stat. 1488, 42 U. S. C. §2000bb et seq. 78 Fed. Reg. 39871, 39886–39887. This accommodation is referred to as the self-certification accommodation. B Shortly after the Departments promulgated the 2013 final rule, two religious nonprofits run by the Little Sisters of the Poor (Little Sisters) challenged the self-certification accommodation. The Little Sisters “are an international congregation of Roman Catholic women religious” who have operated homes for the elderly poor in the United States since 1868. See Mission Statement: Little Sisters of the Poor, http://www.littlesistersofthepoor.org/mission-statement. They feel called by their faith to care for their elderly residents regardless of “faith, finances, or frailty.” Brief for Residents and Families of Residents at Homes of the Little Sisters of the Poor as Amici Curiae 14. The Little Sisters endeavor to treat all residents “as if they were Jesus [Christ] himself, cared for as family, and treated with dignity until God calls them to his home.” Complaint ¶14 in Little Sisters of the Poor Home for the Aged, Denver, Colo. v. Sebelius, No. 1:13–cv–02611 (D Colo.), p. 5 (Complaint). Consistent with their Catholic faith, the Little Sisters hold the religious conviction “that deliberately avoiding reproduction through medical means is immoral.” Little Sisters of the Poor Home for the Aged, Denver, Colo. v. Burwell, 794 F.3d 1151, 1167 (CA10 2015). They challenged the self-certification accommodation, claiming that completing the certification form would force them to violate their religious beliefs by “tak[ing] actions that directly cause others to provide contraception or appear to participate in the Departments’ delivery scheme.” Id., at 1168. As a result, they alleged that the self-certification accommodation violated RFRA. Under RFRA, a law that substantially burdens the exercise of religion must serve “a compelling governmental interest” and be “the least restrictive means of furthering that compelling governmental interest.” §§2000bb–1(a)–(b). The Court of Appeals disagreed that the self-certification accommodation substantially burdened the Little Sisters’ free exercise rights and thus rejected their RFRA claim. Little Sisters, 794 F. 3d, at 1160. The Little Sisters were far from alone in raising RFRA challenges to the self-certification accommodation. Religious nonprofit organizations and educational institutions across the country filed a spate of similar lawsuits, most resulting in rulings that the accommodation did not violate RFRA. See, e.g., East Texas Baptist Univ. v. Burwell, 793 F.3d 449 (CA5 2015); Geneva College v. Secretary, U. S. Dept. of Health and Human Servs., 778 F.3d 422 (CA3 2015); Priests for Life v. United States Dept. of Health and Human Servs., 772 F.3d 229 (CADC 2014); Michigan Catholic Conference v. Burwell, 755 F.3d 372 (CA6 2014); University of Notre Dame v. Sebelius, 743 F.3d 547 (CA7 2014); but see Sharpe Holdings, Inc. v. United States Dept. of Health and Human Servs., 801 F.3d 927 (CA8 2015); Dordt College v. Burwell, 801 F.3d 946 (CA8 2015). We granted certiorari in cases from four Courts of Appeals to decide the RFRA question. Zubik v. Burwell, 578 U. S. ___, ___ (2016) (per curiam). Ultimately, however, we opted to remand the cases without deciding that question. In supplemental briefing, the Government had “confirm[ed]” that “ ‘contraceptive coverage could be provided to petitioners’ employees, through petitioners’ insurance companies, without any . . . notice from petitioners.’ ” Id., at ___ (slip op., at 3). Petitioners, for their part, had agreed that such an approach would not violate their free exercise rights. Ibid. Accordingly, because all parties had accepted that an alternative approach was “feasible,” ibid., we directed the Government to “accommodat[e] petitioners’ religious exercise while at the same time ensuring that women covered by petitioners’ health plans receive full and equal health coverage, including contraceptive coverage,” id., at ___ (slip op., at 4) (internal quotation marks omitted). C Zubik was not the only relevant ruling from this Court about the contraceptive mandate. As the Little Sisters and numerous others mounted their challenges to the self-certification accommodation, a host of other entities challenged the contraceptive mandate itself as a violation of RFRA. See, e.g., Hobby Lobby Stores, Inc. v. Sebelius, 723 F.3d 1114 (CA10 2013) (en banc); Korte v. Sebelius, 735 F.3d 654 (CA7 2013); Gilardi v. United States Dept. of Health and Human Servs., 733 F.3d 1208 (CADC 2013); Conestoga Wood Specialties Corp. v. Secretary of U. S. Dept. of Health and Human Servs., 724 F.3d 377 (CA3 2013); Autocam Corp. v. Sebelius, 730 F.3d 618 (CA6 2013). This Court granted certiorari in two cases involving three closely held corporations to decide whether the mandate violated RFRA. Hobby Lobby, 573 U.S. 682. The individual respondents in Hobby Lobby opposed four methods of contraception covered by the mandate. They sincerely believed that human life begins at conception and that, because the challenged methods of contraception risked causing the death of a human embryo, providing those methods of contraception to employees would make the employers complicit in abortion. Id., at 691, 720. We held that the mandate substantially burdened respondents’ free exercise, explaining that “[if] the owners comply with the HHS mandate, they believe they will be facilitating abortions, and if they do not comply, they will pay a very heavy price.” Id., at 691. “If these consequences do not amount to a substantial burden,” we stated, “it is hard to see what would.” Ibid. We also held that the mandate did not utilize the least restrictive means, citing the self-certification accommodation as a less burdensome alternative. Id., at 730–731. Thus, as the Departments began the task of reformulating rules related to the contraceptive mandate, they did so not only under Zubik’s direction to accommodate religious exercise, but also against the backdrop of Hobby Lobby’s pronouncement that the mandate, standing alone, violated RFRA as applied to religious entities with complicity-based objections. D In 2016, the Departments attempted to strike the proper balance a third time, publishing a request for information on ways to comply with Zubik. 81 Fed. Reg. 47741. This attempt proved futile, as the Departments ultimately concluded that “no feasible approach” had been identified. Dept. of Labor, FAQs About Affordable Care Act Implementation Part 36, p. 4 (2017). The Departments maintained their position that the self-certification accommodation was consistent with RFRA because it did not impose a substantial burden and, even if it did, it utilized the least restrictive means of achieving the Government’s interests. Id., at 4–5. In 2017, the Departments tried yet again to comply with Zubik, this time by promulgating the two IFRs that served as the impetus for this litigation. The first IFR significantly broadened the definition of an exempt religious employer to encompass an employer that “objects . . . based on its sincerely held religious beliefs,” “to its establishing, maintaining, providing, offering, or arranging [for] coverage or payments for some or all contraceptive services.” 82 Fed. Reg. 47812 (2017). Among other things, this definition included for-profit and publicly traded entities. Because they were exempt, these employers did not need to participate in the accommodation process, which nevertheless remained available under the IFR. Id., at 47806. As with their previous regulations, the Departments once again invoked §300gg–13(a)(4) as authority to promulgate this “religious exemption,” stating that it “include[d] the ability to exempt entities from coverage requirements announced in HRSA’s Guidelines.” Id., at 47794. Additionally, the Departments announced for the first time that RFRA compelled the creation of, or at least provided the discretion to create, the religious exemption. Id., at 47800–47806. As the Departments explained: “We know from Hobby Lobby that, in the absence of any accommodation, the contraceptive-coverage requirement imposes a substantial burden on certain objecting employers. We know from other lawsuits and public comments that many religious entities have objections to complying with the [self-certification] accommodation based on their sincerely held religious beliefs.” Id., at 47806. The Departments “believe[d] that the Court’s analysis in Hobby Lobby extends, for the purposes of analyzing a substantial burden, to the burdens that an entity faces when it religiously opposes participating in the [self-certification] accommodation process.” Id., at 47800. They thus “conclude[d] that it [was] appropriate to expand the exemption to other . . . organizations with sincerely held religious beliefs opposed to contraceptive coverage.” Id., at 47802; see also id., at 47810–47811. The second IFR created a similar “moral exemption” for employers—including nonprofits and for-profits with no publicly traded components—with “sincerely held moral” objections to providing some or all forms of contraceptive coverage. Id., at 47850, 47861–47862. Citing congressional enactments, precedents from this Court, agency practice, and state laws that provided for conscience protections, id., at 47844–47847, the Departments invoked their authority under the ACA to create this exemption, id., at 47844. The Departments requested post-promulgation comments on both IFRs. Id., at 47813, 47854. E Within a week of the 2017 IFRs’ promulgation, the Commonwealth of Pennsylvania filed an action seeking declaratory and injunctive relief. Among other claims, it alleged that the IFRs were procedurally and substantively invalid under the APA. The District Court held that the Commonwealth was likely to succeed on both claims and granted a preliminary nationwide injunction against the IFRs. The Federal Government appealed. While that appeal was pending, the Departments issued rules finalizing the 2017 IFRs. See 83 Fed. Reg. 57536 (2018); 83 Fed. Reg. 57592, codified at 45 CFR pt. 147 (2018). Though the final rules left the exemptions largely intact, they also responded to post-promulgation comments, explaining their reasons for neither narrowing nor expanding the exemptions beyond what was provided for in the IFRs. See 83 Fed. Reg. 57542–57545, 57598–57603. The final rule creating the religious exemption also contained a lengthy analysis of the Departments’ changed position regarding whether the self-certification process violated RFRA. Id., at 57544–57549. And the Departments explained that, in the wake of the numerous lawsuits challenging the self-certification accommodation and the failed attempt to identify alternative accommodations after the 2016 request for information, “an expanded exemption rather than the existing accommodation is the most appropriate administrative response to the substantial burden identified by the Supreme Court in Hobby Lobby.” Id., at 57544–57545. After the final rules were promulgated, the State of New Jersey joined Pennsylvania’s suit and, together, they filed an amended complaint. As relevant, the States—respondents here—once again challenged the rules as substantively and procedurally invalid under the APA. They alleged that the rules were substantively unlawful because the Departments lacked statutory authority under either the ACA or RFRA to promulgate the exemptions. Respondents also asserted that the IFRs were not adequately justified by good cause, meaning that the Departments impermissibly used the IFR procedure to bypass the APA’s notice and comment procedures. Finally, respondents argued that the purported procedural defects of the IFRs likewise infected the final rules. The District Court issued a nationwide preliminary injunction against the implementation of the final rules the same day the rules were scheduled to take effect. The Federal Government appealed, as did one of the homes operated by the Little Sisters, which had in the meantime intervened in the suit to defend the religious exemption.[5] The appeals were consolidated with the previous appeal, which had been stayed. The Third Circuit affirmed. In its view, the Departments lacked authority to craft the exemptions under either statute. The Third Circuit read 42 U. S. C. §300gg–13(a)(4) as empowering HRSA to determine which services should be included as preventive care and screenings, but not to carve out exemptions from those requirements. It also concluded that RFRA did not compel or permit the religious exemption because, under Third Circuit precedent that was vacated and remanded in Zubik, the Third Circuit had concluded that the self-certification accommodation did not impose a substantial burden on free exercise. As for respondents’ procedural claim, the court held that the Departments lacked good cause to bypass notice and comment when promulgating the 2017 IFRs. In addition, the court determined that, because the IFRs and final rules were “virtually identical,” “[t]he notice and comment exercise surrounding the Final Rules [did] not reflect any real open-mindedness.” Pennsylvania v. President of United States, 930 F.3d 543, 568–569 (2019). Though it rebuked the Departments for their purported attitudinal deficiencies, the Third Circuit did not identify any specific public comments to which the agency did not appropriately respond. Id., at 569, n. 24.[6] We granted certiorari. 589 U. S. ___ (2020). II Respondents contend that the 2018 final rules providing religious and moral exemptions to the contraceptive mandate are both substantively and procedurally invalid. We begin with their substantive argument that the Departments lacked statutory authority to promulgate the rules. A The Departments invoke 42 U. S. C. §300gg–13(a)(4) as legal authority for both exemptions. This provision of the ACA states that, “with respect to women,” “[a] group health plan and a health insurance issuer offering group or individual health insurance coverage shall, at a minimum provide . . . such additional preventive care and screenings not described in paragraph (1) as provided for in comprehensive guidelines supported by [HRSA].” The Departments maintain, as they have since 2011, that the phrase “as provided for” allows HRSA both to identify what preventive care and screenings must be covered and to exempt or accommodate certain employers’ religious objections. See 83 Fed. Reg. 57540–57541; see also post, at 3 (Kagan, J., concurring in judgment). They also argue that, as with the church exemption, their role as the administering agencies permits them to guide HRSA in its discretion by “defining the scope of permissible exemptions and accommodations for such guidelines.” 82 Fed. Reg. 47794. Respondents, on the other hand, contend that §300gg–13(a)(4) permits HRSA to only list the preventive care and screenings that health plans “shall . . . provide,” not to exempt entities from covering those identified services. Because that asserted limitation is found nowhere in the statute, we agree with the Departments. “Our analysis begins and ends with the text.” Octane Fitness, LLC v. ICON Health & Fitness, Inc., 572 U.S. 545, 553 (2014). Here, the pivotal phrase is “as provided for.” To “provide” means to supply, furnish, or make available. See Webster’s Third New International Dictionary 1827 (2002) (Webster’s Third); American Heritage Dictionary 1411 (4th ed. 2000); 12 Oxford English Dictionary 713 (2d ed. 1989). And, as the Departments explained, the word “as” functions as an adverb modifying “provided,” indicating “the manner in which” something is done. 83 Fed. Reg. 57540. See also Webster’s Third 125; 1 Oxford English Dictionary, at 673; American Heritage Dictionary 102 (5th ed. 2011). On its face, then, the provision grants sweeping authority to HRSA to craft a set of standards defining the preventive care that applicable health plans must cover. But the statute is completely silent as to what those “comprehensive guidelines” must contain, or how HRSA must go about creating them. The statute does not, as Congress has done in other statutes, provide an exhaustive or illustrative list of the preventive care and screenings that must be included. See, e.g., 18 U. S. C. §1961(1); 28 U. S. C. §1603(a). It does not, as Congress did elsewhere in the same section of the ACA, set forth any criteria or standards to guide HRSA’s selections. See, e.g., 42 U. S. C. §300gg–13(a)(3) (requiring “evidence-informed preventive care and screenings” (emphasis added)); §300gg–13(a)(1) (“evidence-based items or services”). It does not, as Congress has done in other contexts, require that HRSA consult with or refrain from consulting with any party in the formulation of the Guidelines. See, e.g., 16 U. S. C. §1536(a)(1); 23 U. S. C. §138. This means that HRSA has virtually unbridled discretion to decide what counts as preventive care and screenings. But the same capacious grant of authority that empowers HRSA to make these determinations leaves its discretion equally unchecked in other areas, including the ability to identify and create exemptions from its own Guidelines. Congress could have limited HRSA’s discretion in any number of ways, but it chose not to do so. See Ali v. Federal Bureau of Prisons, 552 U.S. 214, 227 (2008); see also Rotkiske v. Klemm, 589 U. S. ___, ___ (2019) (slip op., at 6); Husted v. A. Philip Randolph Institute, 584 U. S. ___, ___ (2018) (slip op., at 16). Instead, it enacted “ ‘expansive language offer[ing] no indication whatever’ ” that the statute limits what HRSA can designate as preventive care and screenings or who must provide that coverage. Ali, 552 U. S., at 219–220 (quoting Harrison v. PPG Industries, Inc., 446 U.S. 578, 589 (1980)). “It is a fundamental principle of statutory interpretation that ‘absent provision[s] cannot be supplied by the courts.’ ” Rotkiske, 589 U. S., at ___ (slip op., at 5) (quoting A. Scalia & B. Garner, Reading Law: The Interpretation of Legal Texts 94 (2012)); Nichols v. United States, 578 U. S. ___, ___ (2016) (slip op., at 6). This principle applies not only to adding terms not found in the statute, but also to imposing limits on an agency’s discretion that are not supported by the text. See Watt v. Energy Action Ed. Foundation, 454 U.S. 151, 168 (1981). By introducing a limitation not found in the statute, respondents ask us to alter, rather than to interpret, the ACA. See Nichols, 578 U. S., at ___ (slip op., at 6). By its terms, the ACA leaves the Guidelines’ content to the exclusive discretion of HRSA. Under a plain reading of the statute, then, we conclude that the ACA gives HRSA broad discretion to define preventive care and screenings and to create the religious and moral exemptions.[7] The dissent resists this conclusion, asserting that the Departments’ interpretation thwarts Congress’ intent to provide contraceptive coverage to the women who are interested in receiving such coverage. See post, at 1, 21 (opinion of Ginsburg, J.). It also argues that the exemptions will make it significantly harder for interested women to obtain seamless access to contraception without cost sharing, post, at 15–17, which we have previously “assume[d]” is a compelling governmental interest, Hobby Lobby, 573 U. S., at 728; but see post, at 10–12 (Alito, J., concurring). The Departments dispute that women will be adversely impacted by the 2018 exemptions. 82 Fed. Reg. 47805. Though we express no view on this disagreement, it bears noting that such a policy concern cannot justify supplanting the text’s plain meaning. See Gitlitz v. Commissioner, 531 U.S. 206, 220 (2001). “It is not for us to rewrite the statute so that it covers only what we think is necessary to achieve what we think Congress really intended.” Lewis v. Chicago, 560 U.S. 205, 215 (2010). Moreover, even assuming that the dissent is correct as an empirical matter, its concerns are more properly directed at the regulatory mechanism that Congress put in place to protect this assumed governmental interest. As even the dissent recognizes, contraceptive coverage is mentioned nowhere in §300gg–13(a)(4), and no language in the statute itself even hints that Congress intended that contraception should or must be covered. See post, at 4–5 (citing legislative history and amicus briefs). Thus, contrary to the dissent’s protestations, it was Congress, not the Departments, that declined to expressly require contraceptive coverage in the ACA itself. See 83 Fed. Reg. 57540. And, it was Congress’ deliberate choice to issue an extraordinarily “broad general directiv[e]” to HRSA to craft the Guidelines, without any qualifications as to the substance of the Guidelines or whether exemptions were permissible. Mistretta v. United States, 488 U.S. 361, 372 (1989). Thus, it is Congress, not the Departments, that has failed to provide the protection for contraceptive coverage that the dissent seeks.[8] No party has pressed a constitutional challenge to the breadth of the delegation involved here. Cf. Gundy v. United States, 588 U. S. ___ (2019). The only question we face today is what the plain language of the statute authorizes. And the plain language of the statute clearly allows the Departments to create the preventive care standards as well as the religious and moral exemptions.[9] B The Departments also contend, consistent with the reasoning in the 2017 IFR and the 2018 final rule establishing the religious exemption, that RFRA independently compelled the Departments’ solution or that it at least authorized it.[10] In light of our holding that the ACA provided a basis for both exemptions, we need not reach these arguments.[11] We do, however, address respondents’ argument that the Departments could not even consider RFRA as they formulated the religious exemption from the contraceptive mandate. Particularly in the context of these cases, it was appropriate for the Departments to consider RFRA. As we have explained, RFRA “provide[s] very broad protection for religious liberty.” Hobby Lobby, 573 U. S., at 693. In RFRA’s congressional findings, Congress stated that “governments should not substantially burden religious exercise,” a right described by RFRA as “unalienable.” 42 U. S. C. §§2000bb(a)(1), (3). To protect this right, Congress provided that the “[g]overnment shall not substantially burden a person’s exercise of religion even if the burden results from a rule of general applicability” unless “it demonstrates that application of the burden . . . is in furtherance of a compelling governmental interest; and . . . is the least restrictive means of furthering that compelling governmental interest.” §§2000bb–1(a)–(b). Placing Congress’ intent beyond dispute, RFRA specifies that it “applies to all Federal law, and the implementation of that law, whether statutory or otherwise.” §2000bb–3(a). RFRA also permits Congress to exclude statutes from RFRA’s protections. §2000bb–3(b). It is clear from the face of the statute that the contraceptive mandate is capable of violating RFRA. The ACA does not explicitly exempt RFRA, and the regulations implementing the contraceptive mandate qualify as “Federal law” or “the implementation of [Federal] law.” §2000bb–3(a); cf. Chrysler Corp. v. Brown, 441 U.S. 281, 297–298 (1979). Additionally, we expressly stated in Hobby Lobby that the contraceptive mandate violated RFRA as applied to entities with complicity-based objections. 573 U. S., at 736. Thus, the potential for conflict between the contraceptive mandate and RFRA is well settled. Against this backdrop, it is unsurprising that RFRA would feature prominently in the Departments’ discussion of exemptions that would not pose similar legal problems. Moreover, our decisions all but instructed the Departments to consider RFRA going forward. For instance, though we held that the mandate violated RFRA in Hobby Lobby, we left it to the Federal Government to develop and implement a solution. At the same time, we made it abundantly clear that, under RFRA, the Departments must accept the sincerely held complicity-based objections of religious entities. That is, they could not “tell the plaintiffs that their beliefs are flawed” because, in the Departments’ view, “the connection between what the objecting parties must do . . . and the end that they find to be morally wrong . . . is simply too attenuated.” Hobby Lobby, 573 U. S., at 723–724. Likewise, though we did not decide whether the self-certification accommodation ran afoul of RFRA in Zubik, we directed the parties on remand to “accommodat[e]” the free exercise rights of those with complicity-based objections to the self-certification accommodation. 578 U. S., at ___ (slip op., at 4). It is hard to see how the Departments could promulgate rules consistent with these decisions if they did not overtly consider these entities’ rights under RFRA. This is especially true in light of the basic requirements of the rulemaking process. Our precedents require final rules to “articulate a satisfactory explanation for [the] action including a rational connection between the facts found and the choice made.” Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U.S. 29, 43 (1983) (internal quotation marks omitted). This requirement allows courts to assess whether the agency has promulgated an arbitrary and capricious rule by “entirely fail[ing] to consider an important aspect of the problem [or] offer[ing] an explanation for its decision that runs counter to the evidence before [it].” Ibid.; see also Department of Commerce v. New York, 588 U. S. ___, ___–___ (2019) (Breyer, J., concurring in part and dissenting in part) (slip op., at 3–4); Genuine Parts Co. v. EPA, 890 F.3d 304, 307 (CADC 2018); Pacific Coast Federation of Fishermen’s Assns. v. United States Bur. of Reclamation, 426 F.3d 1082, 1094 (CA9 2005). Here, the Departments were aware that Hobby Lobby held the mandate unlawful as applied to religious entities with complicity-based objections. 82 Fed. Reg. 47799; 83 Fed. Reg. 57544–57545. They were also aware of Zubik’s instructions. 82 Fed. Reg. 47799. And, aside from our own decisions, the Departments were mindful of the RFRA concerns raised in “public comments and . . . court filings in dozens of cases—encompassing hundreds of organizations.” Id., at 47802; see also id., at 47806. If the Departments did not look to RFRA’s requirements or discuss RFRA at all when formulating their solution, they would certainly be susceptible to claims that the rules were arbitrary and capricious for failing to consider an important aspect of the problem.[12] Thus, respondents’ argument that the Departments erred by looking to RFRA as a guide when framing the religious exemption is without merit. III Because we hold that the Departments had authority to promulgate the exemptions, we must next decide whether the 2018 final rules are procedurally invalid. Respondents present two arguments on this score. Neither is persuasive. A Unless a statutory exception applies, the APA requires agencies to publish a notice of proposed rulemaking in the Federal Register before promulgating a rule that has legal force. See 5 U. S. C. §553(b). Respondents point to the fact that the 2018 final rules were preceded by a document entitled “Interim Final Rules with Request for Comments,” not a document entitled “General Notice of Proposed Rulemaking.” They claim that since this was insufficient to satisfy §553(b)’s requirement, the final rules were procedurally invalid. Respondents are incorrect. Formal labels aside, the rules contained all of the elements of a notice of proposed rulemaking as required by the APA. The APA requires that the notice of proposed rulemaking contain “reference to the legal authority under which the rule is proposed” and “either the terms or substance of the proposed rule or a description of the subjects and issues involved.” §§553(b)(2)–(3). The request for comments in the 2017 IFRs readily satisfies these requirements. That request detailed the Departments’ view that they had legal authority under the ACA to promulgate both exemptions, 82 Fed. Reg. 47794, 47844, as well as authority under RFRA to promulgate the religious exemption, id., at 47800–47806. And respondents do not—and cannot—argue that the IFRs failed to air the relevant issues with sufficient detail for respondents to understand the Departments’ position. See supra, at 10–11. Thus, the APA notice requirements were satisfied. Even assuming that the APA requires an agency to publish a document entitled “notice of proposed rulemaking” when the agency moves from an IFR to a final rule, there was no “prejudicial error” here. §706. We have previously noted that the rule of prejudicial error is treated as an “administrative law . . . harmless error rule,” National Assn. of Home Builders v. Defenders of Wildlife, 551 U.S. 644, 659–660 (2007) (internal quotation marks omitted). Here, the Departments issued an IFR that explained its position in fulsome detail and “provide[d] the public with an opportunity to comment on whether [the] regulations . . . should be made permanent or subject to modification.” 82 Fed. Reg. 47815; see also id., at 47852, 47855. Respondents thus do not come close to demonstrating that they experienced any harm from the title of the document, let alone that they have satisfied this harmless error rule. “The object [of notice and comment], in short, is one of fair notice,” Long Island Care at Home, Ltd. v. Coke, 551 U.S. 158, 174 (2007), and respondents certainly had such notice here. Because the IFR complied with the APA’s requirements, this claim fails.[13] B Next, respondents contend that the 2018 final rules are procedurally invalid because “nothing in the record signal[s]” that the Departments “maintained an open mind throughout the [post-promulgation] process.” Brief for Respondents 27. As evidence for this claim, respondents point to the fact that the final rules made only minor alterations to the IFRs, leaving their substance unchanged. The Third Circuit applied this “open-mindedness” test, concluding that because the final rules were “virtually identical” to the IFRs, the Departments lacked the requisite “flexible and open-minded attitude” when they promulgated the final rules. 930 F. 3d, at 569 (internal quotation marks omitted). We decline to evaluate the final rules under the open-mindedness test. We have repeatedly stated that the text of the APA provides the “ ‘maximum procedural requirements’ ” that an agency must follow in order to promulgate a rule. Perez, 575 U. S., at 100 (quoting Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc., 435 U.S. 519, 524 (1978)). Because the APA “sets forth the full extent of judicial authority to review executive agency action for procedural correctness,” FCC v. Fox Television Stations, Inc., 556 U.S. 502, 513 (2009), we have repeatedly rejected courts’ attempts to impose “judge-made procedur[es]” in addition to the APA’s mandates, Perez, 575 U. S., at 102; see also Pension Benefit Guaranty Corporation v. LTV Corp., 496 U.S. 633, 654–655 (1990); Vermont Yankee, 435 U. S., at 549. And like the procedures that we have held invalid, the open-mindedness test violates the “general proposition that courts are not free to impose upon agencies specific procedural requirements that have no basis in the APA.” LTV Corp., 496 U. S., at 654. Rather than adopting this test, we focus our inquiry on whether the Departments satisfied the APA’s objective criteria, just as we have in previous cases. We conclude that they did. Section 553(b) obligated the Departments to provide adequate notice before promulgating a rule that has legal force. As explained supra, at 22–23, the IFRs provided sufficient notice. Aside from these notice requirements, the APA mandates that agencies “give interested persons an opportunity to participate in the rule making through submission of written data, views, or arguments,” §553(c); states that the final rules must include “a concise general statement of their basis and purpose,” ibid.; and requires that final rules must be published 30 days before they become effective, §553(d). The Departments complied with each of these statutory procedures. They “request[ed] and encourag[ed] public comments on all matters addressed” in the rules—i.e., the basis for the Departments’ legal authority, the rationales for the exemptions, and the detailed discussion of the exemptions’ scope. 82 Fed. Reg. 47813, 47854. They also gave interested parties 60 days to submit comments. Id., at 47792, 47838. The final rules included a concise statement of their basis and purpose, explaining that the rules were “necessary to protect sincerely held” moral and religious objections and summarizing the legal analysis supporting the exemptions. 83 Fed. Reg. 57592; see also id., at 57537–57538. Lastly, the final rules were published on November 15, 2018, but did not become effective until January 14, 2019—more than 30 days after being published. Id., at 57536, 57592. In sum, the rules fully complied with “ ‘the maximum procedural requirements [that] Congress was willing to have the courts impose upon agencies in conducting rulemaking procedures.’ ” Perez, 575 U. S., at 102 (quoting Vermont Yankee, 435 U. S., at 524). Accordingly, respondents’ second procedural challenge also fails.[14] * * * For over 150 years, the Little Sisters have engaged in faithful service and sacrifice, motivated by a religious calling to surrender all for the sake of their brother. “[T]hey commit to constantly living out a witness that proclaims the unique, inviolable dignity of every person, particularly those whom others regard as weak or worthless.” Complaint ¶14. But for the past seven years, they—like many other religious objectors who have participated in the litigation and rulemakings leading up to today’s decision—have had to fight for the ability to continue in their noble work without violating their sincerely held religious beliefs. After two decisions from this Court and multiple failed regulatory attempts, the Federal Government has arrived at a solution that exempts the Little Sisters from the source of their complicity-based concerns—the administratively imposed contraceptive mandate. We hold today that the Departments had the statutory authority to craft that exemption, as well as the contemporaneously issued moral exemption. We further hold that the rules promulgating these exemptions are free from procedural defects. Therefore, we reverse the judgment of the Court of Appeals and remand the cases for further proceedings consistent with this opinion. It is so ordered. Notes 1 See 42 U. S. C. §300gg–92; 29 U. S. C. §1191c; 26 U. S. C. §9833. 2 The ACA exempts “grandfathered” plans from 42 U. S. C. §300gg–13(a)(4)—i.e., “those [plans] that existed prior to March 23, 2010, and that have not made specified changes after that date.” Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 699 (2014). See §§18011(a), (e); 29 CFR §2590.715–1251 (2019). As of 2018, an estimated 16 percent of employees “with employer-sponsored coverage were enrolled in a grandfathered group health plan.” 84 Fed. Reg. 5971 (2019). 3 Dept. of Health and Human Servs., Center for Consumer Information and Insurance Oversight, Centers for Medicare & Medicaid Services, Guidance on the Temporary Enforcement Safe Harbor for Certain Employers, Group Health Plans and Group Health Insurance Issuers With Respect to the Requirement To Cover Contraceptive Services Without Cost Sharing Under Section 2713 of the Public Health Service Act, Section 715(a)(1) of the Employee Retirement Income Security Act, and Section 9815(a)(1) of the Internal Revenue Code, p. 2 (2013). 4 The Departments took this action to prevent an unduly narrow interpretation of the church exemption, in which “an otherwise exempt plan [was] disqualified because the employer’s purposes extend[ed] beyond the inculcation of religious values or because the employer . . . serve[d] people of different religious faiths.” 78 Fed. Reg. 39874. But see post, at 12–13 (Ginsburg, J., dissenting) (arguing that the church exemption only covered houses of worship). 5 The Little Sisters moved to intervene in the District Court to defend the 2017 religious-exemption IFR, but the District Court denied that motion. The Third Circuit reversed. After that reversal, the Little Sisters appealed the District Court’s preliminary injunction of the 2017 IFRs, and that appeal was consolidated with the Federal Government’s appeal. 6 The Third Circuit also determined sua sponte that the Little Sisters lacked appellate standing to intervene because a District Court in Colorado had permanently enjoined the contraceptive mandate as applied to plans in which the Little Sisters participate. This was error. Under our precedents, at least one party must demonstrate Article III standing for each claim for relief. An intervenor of right must independently demonstrate Article III standing if it pursues relief that is broader than or different from the party invoking a court’s jurisdiction. See Town of Chester v. Laroe Estates, Inc., 581 U. S. ___, ___ (2017) (slip op., at 6). Here, the Federal Government clearly had standing to invoke the Third Circuit’s appellate jurisdiction, and both the Federal Government and the Little Sisters asked the court to dissolve the injunction against the religious exemption. The Third Circuit accordingly erred by inquiring into the Little Sisters’ independent Article III standing. 7 Though not necessary for this analysis, our decisions in Zubik v. Burwell, 578 U. S. ___ (2016) (per curiam), and Hobby Lobby, 573 U.S. 682, implicitly support the conclusion that §300gg–13(a)(4) empowered HRSA to create the exemptions. As respondents acknowledged at oral argument, accepting their interpretation of the ACA would require us to conclude that the Departments had no authority under the ACA to promulgate the initial church exemption, see Tr. of Oral Arg. 69–71, 91, which by extension would mean that the Departments lacked authority for the 2013 self-certification accommodation. That reading of the ACA would create serious tension with Hobby Lobby, which pointed to the self-certification accommodation as an example of a less restrictive means available to the Government, 573 U. S., at 730–731, and Zubik, which expressly directed the Departments to “accommodat[e]” petitioners’ religious exercise, 578 U. S., at ___ (slip op., at 4). It would be passing strange for this Court to direct the Departments to make such an accommodation if it thought the ACA did not authorize one. In addition, we are not aware of, and the dissent does not point to, a single case predating Hobby Lobby or Zubik in which the Departments took the position that they could not adopt a different approach because they lacked the statutory authority under the ACA to do so. 8 HRSA has altered its Guidelines multiple times since 2011, always proceeding without notice and comment. See 82 Fed. Reg. 47813–47814; 83 Fed. Reg. 8487; 85 Fed. Reg. 722–723 (2020). Accordingly, if HRSA chose to exercise that discretion to remove contraception coverage from the next iteration of its Guidelines, it would arguably nullify the contraceptive mandate altogether without proceeding through notice and comment. The combination of the agency practice of proceeding without notice and comment and HRSA’s discretion to alter the Guidelines, though not necessary for our analysis, provides yet another indication of Congress’ failure to provide strong protections for contraceptive coverage. 9 The dissent does not attempt to argue that the self-certification accommodation can coexist with its interpretation of the ACA. As for the church exemption, the dissent claims that it is rooted in the First Amendment’s respect for church autonomy. See post, at 12–13. But the dissent points to no case, brief, or rule in the nine years since the church exemption’s implementation in which the Departments defended its validity on that ground. The most the dissent can point to is a stray comment in the rule that expanded the self-certification accommodation to closely held corporations in the wake of Hobby Lobby. See post, at 13 (quoting 80 Fed. Reg. 41325 (2015)). 10 The dissent claims that “all agree” that the exemption is not supported by the Free Exercise Clause. Post, at 2. A constitutional claim is not presented in these cases, and we express no view on the merits of that question. 11 The dissent appears to agree that the Departments had authority under RFRA to “cure” any RFRA violations caused by its regulations. See post, at 14, n. 16 (disclaiming the view that agencies must wait for courts to determine a RFRA violation); see also supra, at 5 (explaining that the safe harbor and commitment to developing an accommodation occurred prior to the Guidelines going into effect). The dissent also does not—as it cannot—dispute our directive in Zubik. 12 Here, too, the Departments have consistently taken the position that their rules had to account for RFRA in response to comments that the rules would violate that statute. See Dept. of Labor, FAQs About Affordable Care Act Implementation Part 36, pp. 4–5 (2017) (2016 Request for Information); 78 Fed. Reg. 39886–39887 (2013 rule); 77 Fed. Reg. 8729 (2012 final rule). As the 2017 IFR explained, the Departments simply reached a different conclusion on whether the accommodation satisfied RFRA. See 82 Fed. Reg. 47800–40806 (summarizing the previous ways in which the Departments accounted for RFRA and providing a lengthy explanation for the changed position). 13 We note as well that the Departments promulgated many other IFRs in addition to the three related to the contraceptive mandate. See, e.g., 75 Fed. Reg. 27122 (dependent coverage); id., at 34538 (grandfathered health plans); id., at 37188 (pre-existing conditions). 14 Because we conclude that the IFRs’ request for comment satisfiesthe APA’s rulemaking requirements, we need not reach respondents’ additional argument that the Departments lacked good cause to promulgate the 2017 IFRs.

591.US.2019_18-1501

To punish securities fraud, the Securities and Exchange Commission is authorized to seek “equitable relief” in civil proceedings, 15 U. S. C. §78u(d)(5). In Kokesh v. SEC, 581 U. S. ___, this Court held that a disgorgement order in a Securities and Exchange Commission (SEC) enforcement action constitutes a “penalty” for purposes of the applicable statute of limitations. The Court did not, however, address whether disgorgement can qualify as “equitable relief” under §78u(d)(5), given that equity historically excludes punitive sanctions. Petitioners Charles Liu and Xin Wang solicited foreign nationals to invest in the construction of a cancer-treatment center, but, an SEC investigation revealed, misappropriated much of the funds in violation of the terms of a private offering memorandum. The SEC brought a civil action against petitioners, seeking, as relevant here, disgorgement equal to the full amount petitioners had raised from investors. Petitioners argued that the disgorgement remedy failed to account for their legitimate business expenses, but the District Court disagreed and ordered petitioners jointly and severally liable for the full amount. The Ninth Circuit affirmed. Held: A disgorgement award that does not exceed a wrongdoer’s net profits and is awarded for victims is equitable relief permissible under §78u(d)(5). Pp. 5–20. (a) In interpreting statutes that provide for “equitable relief,” this Court analyzes whether a particular remedy falls into “those categories of relief that were typically available in equity.” Mertens v. Hewitt Associates, 508 U.S. 248, 256. Relevant here are two principles of equity jurisprudence. Equity practice has long authorized courts to strip wrongdoers of their ill-gotten gains. And to avoid transforming that remedy into a punitive sanction, courts restricted it to an individual wrongdoer’s net profits to be awarded for victims. Pp. 5–14. (1) Whether it is called restitution, an accounting, or disgorgement, the equitable remedy that deprives wrongdoers of their net profits from unlawful activity reflects both the foundational principle that “it would be inequitable that [a wrongdoer] should make a profit out of his own wrong,” Root v. Railway Co., 105 U.S. 189, 207, and the countervailing equitable principle that the wrongdoer should not be punished by “pay[ing] more than a fair compensation to the person wronged,” Tilghman v. Proctor, 125 U.S. 136, 145–146. The remedy has been a mainstay of equity courts, and is not limited to cases involving a breach of trust or fiduciary duty, see Root, 105 U. S., at 214. Pp. 6–9. (2) To avoid transforming a profits award into a penalty, equity courts restricted the remedy in various ways. A constructive trust was often imposed on wrongful gains for wronged victims. See, e.g., Burdell v. Denig, 92 U.S. 716, 720. Courts also generally awarded profits-based remedies against individuals or partners engaged in concerted wrongdoing, not against multiple wrongdoers under a joint-and-several liability theory. See, e.g., Ambler v. Whipple, 20 Wall. 546, 559. Finally, courts limited awards to the net profits from wrongdoing after deducting legitimate expenses. See, e.g., Rubber Co. v. Goodyear, 9 Wall. 788, 804. Pp. 9–12. (3) Congress incorporated these longstanding equitable principles into §78u(d)(5), but courts have occasionally awarded disgorgement in ways that test the bounds of equity practice. Petitioners claim that disgorgement is necessarily a penalty under Kokesh, and thus not available at equity. But Kokesh expressly declined to reach that question. The Government contends that the SEC’s interpretation has Congress’ tacit support. But Congress does not enlarge the breadth of an equitable, profit-based remedy simply by using the term “disgorgement” in various statutes. Pp. 12–14. (b) Petitioners briefly claim that their disgorgement award crosses the bounds of traditional equity practice by failing to return funds to victims, imposing joint-and-several liability, and declining to deduct business expenses from the award. Because the parties did not fully brief these narrower questions, the Court does not decide them here. But certain principles may guide the lower courts’ assessment of these arguments on remand. Pp. 14–20. (1) Section 78u(d)(5) provides limited guidance as to whether the practice of depositing a defendant’s gains with the Treasury satisfies its command that any remedy be “appropriate or necessary for the benefit of investors,” and the equitable nature of the profits remedy generally requires the SEC to return a defendant’s gains to wronged investors. The parties, however, do not identify a specific order in this case directing any proceeds to the Treasury. If one is entered on remand, the lower courts may evaluate in the first instance whether that order would be for the benefit of investors and consistent with equitable principles. Pp. 14–17. (2) Imposing disgorgement liability on a wrongdoer for benefits that accrue to his affiliates through joint-and-several liability runs against the rule in favor of holding defendants individually liable. See Belknap v. Schild, 161 U.S. 10, 25–26. The common law did, however, permit liability for partners engaged in concerted wrongdoing. See, e.g., Ambler, 20 Wall., at 559. On remand, the Ninth Circuit may determine whether the facts are such that petitioners can, consistent with equitable principles, be found liable for profits as partners in wrongdoing or whether individual liability is required. Pp. 17–18. (3) Courts may not enter disgorgement awards that exceed the gains “made upon any business or investment, when both the receipts and payments are taken into the account.” Goodyear, 9 Wall., at 804. When the “entire profit of a business or undertaking” results from the wrongdoing, a defendant may be denied “inequitable deductions.” Root, 105 U. S., at 203. Accordingly, courts must deduct legitimate expenses before awarding disgorgement under §78u(d)(5). The District Court below did not ascertain whether any of petitioners’ expenses were legitimate. On remand, the lower courts should examine whether including such expenses in a profits-based remedy is consistent with the equitable principles underlying §78u(d)(5). Pp. 18–20. 754 Fed. Appx. 505, vacated and remanded. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Alito, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a dissenting opinion.

In Kokesh v. SEC, 581 U. S. ___ (2017), this Court held that a disgorgement order in a Securities and Exchange Commission (SEC) enforcement action imposes a “penalty” for the purposes of 28 U. S. C. §2462, the applicable statute of limitations. In so deciding, the Court reserved an antecedent question: whether, and to what extent, the SEC may seek “disgorgement” in the first instance through its power to award “equitable relief ” under 15 U. S. C. §78u(d)(5), a power that historically excludes punitive sanctions. The Court holds today that a disgorgement award that does not exceed a wrongdoer’s net profits and is awarded for victims is equitable relief permissible under §78u(d)(5). The judgment is vacated, and the case is remanded for the courts below to ensure the award was so limited. I A Congress authorized the SEC to enforce the Securities Act of 1933, 48Stat. 74, as amended, 15 U. S. C. §77a et seq., and the Securities Exchange Act of 1934, 48Stat. 881, as amended, 15 U. S. C. §78a et seq., and to punish securities fraud through administrative and civil proceedings. In administrative proceedings, the SEC can seek limited civil penalties and “disgorgement.” See §77h–1(e) (“In any cease-and-desist proceeding under subsection (a), the Commission may enter an order requiring accounting and disgorgement”); see also §77h–1(g) (“Authority to impose money penalties”). In civil actions, the SEC can seek civil penalties and “equitable relief.” See, e.g., §78u(d)(5) (“In any action or proceeding brought or instituted by the Commission under any provision of the securities laws, . . . any Federal court may grant . . . any equitable relief that may be appropriate or necessary for the benefit of investors”); see also §78u(d)(3) (“Money penalties in civil actions” (quotation modified)). Congress did not define what falls under the umbrella of “equitable relief.” Thus, courts have had to consider which remedies the SEC may impose as part of its §78u(d)(5) powers. Starting with SEC v. Texas Gulf Sulphur Co., 446 F.2d 1301 (CA2 1971), courts determined that the SEC had authority to obtain what it called “restitution,” and what in substance amounted to “profits” that “merely depriv[e ]” a defendant of “the gains of . . . wrongful conduct.” Id., at 1307–1308. Over the years, the SEC has continued to request this remedy, later referred to as “disgorgement,”[1] and courts have continued to award it. See SEC v. Commonwealth Chemical Securities, Inc., 574 F.2d 90, 95 (CA2 1978) (explaining that, when a court awards “[d]isgorgement of profits in an action brought by the SEC,” it is “exercising the chancellor’s discretion to prevent unjust enrichment”); see also SEC v. Blatt, 583 F.2d 1325, 1335 (CA5 1978); SEC v. Washington Cty. Util. Dist., 676 F.2d 218, 227 (CA6 1982). In Kokesh, this Court determined that disgorgement constituted a “penalty” for the purposes of 28 U. S. C. §2462, which establishes a 5-year statute of limitations for “an action, suit or proceeding for the enforcement of any civil fine, penalty, or forfeiture.” The Court reached this conclusion based on several considerations, namely, that disgorgement is imposed as a consequence of violating public laws, it is assessed in part for punitive purposes, and in many cases, the award is not compensatory. 581 U. S., at ___–___ (slip op., at 7–9). But the Court did not address whether a §2462 penalty can nevertheless qualify as “equitable relief ” under §78u(d)(5), given that equity never “lends its aid to enforce a forfeiture or penalty.” Marshall v. Vicksburg, 15 Wall. 146, 149 (1873). The Court cautioned, moreover, that its decision should not be interpreted “as an opinion on whether courts possess authority to order disgorgement in SEC enforcement proceedings.” Kokesh, 581 U. S., at ___, n. 3 (slip op., at 5, n. 3). This question is now squarely before the Court. B The SEC action and disgorgement award at issue here arise from a scheme to defraud foreign nationals. Petitioners Charles Liu and his wife, Xin (Lisa) Wang, solicited nearly $27 million from foreign investors under the EB–5 Immigrant Investor Program (EB–5 Program). 754 Fed. Appx. 505, 506 (CA9 2018) (case below). The EB–5 Program, administered by the U. S. Citizenship and Immigration Services, permits noncitizens to apply for permanent residence in the United States by investing in approved commercial enterprises that are based on “proposals for promoting economic growth.” See USCIS, EB–5 Immigrant Investor Program, https://www.uscis.gov/eb-5. Investments in EB–5 projects are subject to the federal securities laws. Liu sent a private offering memorandum to prospective investors, pledging that the bulk of any contributions would go toward the construction costs of a cancer-treatment center. The memorandum specified that only amounts collected from a small administrative fee would fund “ ‘legal, accounting and administration expenses.’ ” 754 Fed. Appx., at 507. An SEC investigation revealed, however, that Liu spent nearly $20 million of investor money on ostensible marketing expenses and salaries, an amount far more than what the offering memorandum permitted and far in excess of the administrative fees collected. 262 F. Supp. 3d 957, 960–964 (CD Cal. 2017). The investigation also revealed that Liu diverted a sizable portion of those funds to personal accounts and to a company under Wang’s control. Id., at 961, 964. Only a fraction of the funds were put toward a lease, property improvements, and a proton-therapy machine for cancer treatment. Id., at 964–965. The SEC brought a civil action against petitioners, alleging that they violated the terms of the offering documents by misappropriating millions of dollars. The District Court found for the SEC, granting an injunction barring petitioners from participating in the EB–5 Program and imposing a civil penalty at the highest tier authorized. Id., at 975, 976. It also ordered disgorgement equal to the full amount petitioners had raised from investors, less the $234,899 that remained in the corporate accounts for the project. Id., at 975–976. Petitioners objected that the disgorgement award failed to account for their business expenses. The District Court disagreed, concluding that the sum was a “reasonable approximation of the profits causally connected to [their] violation.” Ibid. The court ordered petitioners jointly and severally liable for the full amount that the SEC sought. App. to Pet. for Cert. 62a. The Ninth Circuit affirmed. It acknowledged that Kokesh “expressly refused to reach” the issue whether the District Court had the authority to order disgorgement. 754 Fed. Appx., at 509. The court relied on Circuit precedent to conclude that the “proper amount of disgorgement in a scheme such as this one is the entire amount raised less the money paid back to the investors.” Ibid.; see also SEC v. JT Wallenbrock & Assocs., 440 F.3d 1109, 1113, 1114 (CA9 2006) (reasoning that it would be “unjust to permit the defendants to offset . . . the expenses of running the very business they created to defraud . . . investors”). We granted certiorari to determine whether §78u(d)(5) authorizes the SEC to seek disgorgement beyond a defendant’s net profits from wrongdoing. 589 U. S. ___ (2019). II Our task is a familiar one. In interpreting statutes like §78u(d)(5) that provide for “equitable relief,” this Court analyzes whether a particular remedy falls into “those categories of relief that were typically available in equity.” Mertens v. Hewitt Associates, 508 U.S. 248, 256 (1993); see also CIGNA Corp. v. Amara, 563 U.S. 421, 439 (2011); Montanile v. Board of Trustees of Nat. Elevator Industry Health Benefit Plan, 577 U.S. 136, 142 (2016). The “basic contours of the term are well known” and can be discerned by consulting works on equity jurisprudence. Great-West Life & Annuity Ins. Co. v. Knudson, 534 U.S. 204, 217 (2002). These works on equity jurisprudence reveal two principles. First, equity practice long authorized courts to strip wrongdoers of their ill-gotten gains, with scholars and courts using various labels for the remedy. Second, to avoid transforming an equitable remedy into a punitive sanction, courts restricted the remedy to an individual wrongdoer’s net profits to be awarded for victims. A Equity courts have routinely deprived wrongdoers of their net profits from unlawful activity, even though that remedy may have gone by different names. Compare, e.g., 1 D. Dobbs, Law of Remedies §4.3(5), p. 611 (1993) (“Accounting holds the defendant liable for his profits”), with id., §4.1(1), at 555 (referring to “restitution” as the relief that “measures the remedy by the defendant’s gain and seeks to force disgorgement of that gain”); see also Restatement (Third) of Restitution and Unjust Enrichment §51, Comment a, p. 204 (2010) (Restatement (Third)) (“Restitution measured by the defendant’s wrongful gain is frequently called ‘disgorgement.’ Other cases refer to an ‘accounting’ or an ‘accounting for profits’ ”); 1 J. Pomeroy, Equity Jurisprudence §101, p. 112 (4th ed. 1918) (describing an accounting as an equitable remedy for the violation of strictly legal primary rights). No matter the label, this “profit-based measure of unjust enrichment,” Restatement (Third) §51, Comment a, at 204, reflected a foundational principle: “[I]t would be inequitable that [a wrongdoer] should make a profit out of his own wrong,” Root v. Railway Co., 105 U.S. 189, 207 (1882). At the same time courts recognized that the wrongdoer should not profit “by his own wrong,” they also recognized the countervailing equitable principle that the wrongdoer should not be punished by “pay[ing] more than a fair compensation to the person wronged.” Tilghman v. Proctor, 125 U.S. 136, 145–146 (1888). Decisions from this Court confirm that a remedy tethered to a wrongdoer’s net unlawful profits, whatever the name, has been a mainstay of equity courts. In Porter v. Warner Holding Co., 328 U.S. 395 (1946), the Court interpreted a section of the Emergency Price Control Act of 1942 that encompassed a “comprehensiv[e]” grant of “equitable jurisdiction.” Id., at 398. “[O]nce [a District Court’s] equity jurisdiction has been invoked” under that provision, the Court concluded, “a decree compelling one to disgorge profits . . . may properly be entered.” Id., at 398–399. Subsequent cases confirm the “ ‘protean character’ of the profits-recovery remedy.” Petrella v. Metro-Goldwyn-Mayer, Inc., 572 U.S. 663, 668, n. 1 (2014). In Tull v. United States, 481 U.S. 412 (1987), the Court described “disgorgement of improper profits” as “traditionally considered an equitable remedy.” Id., at 424. While the Court acknowledged that disgorgement was a “limited form of penalty” insofar as it takes money out of the wrongdoer’s hands, it nevertheless compared disgorgement to restitution that simply “ ‘restor[es] the status quo,’ ” thus situating the remedy squarely within the heartland of equity. Ibid.[2] In Great-West, the Court noted that an “accounting for profits” was historically a “form of equitable restitution.” 534 U. S., at 214, n. 2. And in Kansas v. Nebraska, 574 U.S. 445 (2015), a “ ‘basically equitable’ ” original jurisdiction proceeding, the Court ordered disgorgement of Nebraska’s gains from exceeding its allocation under an interstate water compact. Id., at 453, 475. Most recently, in SCA Hygiene Products Aktiebolag v. First Quality Baby Products, LLC, 580 U. S. ___ (2017), the Court canvassed pre-1938 patent cases invoking equity jurisdiction. It noted that many cases sought an “accounting,” which it described as an equitable remedy requiring disgorgement of ill-gotten profits. Id., at ___ (slip op., at 11). This Court’s “transsubstantive guidance on broad and fundamental” equitable principles, Romag Fasteners, Inc. v. Fossil Group, Inc., 590 U. S. ___, ___ (2020) (slip op., at 5), thus reflects the teachings of equity treatises that identify a defendant’s net profits as a remedy for wrongdoing. Contrary to petitioners’ argument, equity courts did not limit this remedy to cases involving a breach of trust or of fiduciary duty. Brief for Petitioners 28–29. As petitioners acknowledge, courts authorized profits-based relief in patent-infringement actions where no such trust or special relationship existed. Id., at 29; see also Root, 105 U. S., at 214 (“[I]t is nowhere said that the patentee’s right to an account is based upon the idea that there is a fiduciary relation created between him and the wrong-doer by the fact of infringement”). Petitioners attempt to distinguish these patent cases by suggesting that an “accounting” was appropriate only because Congress explicitly conferred that remedy by statute in 1870. Brief for Petitioners 29 (citing the Act of July 8, 1870, §55, 16Stat. 206). But patent law had not previously deviated from the general principles outlined above: This Court had developed the rule that a plaintiff may “recover the amount of . . . profits that the defendants have made by the use of his invention” through “a series of decisions under the patent act of 1836, which simply conferred upon the courts of the United States general equity jurisdiction . . . in cases arising under the patent laws.” Tilghman, 125 U. S., at 144. The 1836 statute, in turn, incorporated the substance of an earlier statute from 1819 which granted courts the ability to “proceed according to the course and principles of courts of equity” to “prevent the violation of patent-rights.” Root, 105 U. S., at 193. Thus, as these cases demonstrate, equity courts habitually awarded profits-based remedies in patent cases well before Congress explicitly authorized that form of relief. B While equity courts did not limit profits remedies to particular types of cases, they did circumscribe the award in multiple ways to avoid transforming it into a penalty outside their equitable powers. See Marshall, 15 Wall., at 149. For one, the profits remedy often imposed a constructive trust on wrongful gains for wronged victims. The remedy itself thus converted the wrongdoer, who in many cases was an infringer, “into a trustee, as to those profits, for the owner of the patent which he infringes.” Burdell v. Denig, 92 U.S. 716, 720 (1876). In “converting the infringer into a trustee for the patentee as regards the profits thus made,” the chancellor “estimat[es] the compensation due from the infringer to the patentee.” Packet Co. v. Sickles, 19 Wall. 611, 617–618 (1874); see also Clews v. Jamieson, 182 U.S. 461, 480 (1901) (describing an accounting as involving a “ ‘distribution of the trust moneys among all the beneficiaries who are entitled to share therein’ ” in an action against the governing committee of a stock exchange). Equity courts also generally awarded profits-based remedies against individuals or partners engaged in concerted wrongdoing, not against multiple wrongdoers under a joint-and-several liability theory. See Ambler v. Whipple, 20 Wall. 546, 559 (1874) (ordering an accounting against a partner who had “knowingly connected himself with and aided in . . . fraud”). In Elizabeth v. Pavement Co., 97 U.S. 126 (1878), for example, a city engaged contractors to install pavement in a manner that infringed a third party’s patent. The patent holder brought a suit in equity to recover profits from both the city and its contractors. The Court held that only the contractors (the only parties to make a profit) were responsible, even though the parties answered jointly. Id., at 140; see also ibid. (rejecting liability for an individual officer who merely acted as an agent of the defendant and received a salary for his work). The rule against joint-and-several liability for profits that have accrued to another appears throughout equity cases awarding profits. See, e.g., Belknap v. Schild, 161 U.S. 10, 25–26 (1896) (“The defendants, in any such suit, are therefore liable to account for such profits only as have accrued to themselves from the use of the invention, and not for those which have accrued to another, and in which they have no participation”); Keystone Mfg. Co. v. Adams, 151 U.S. 139, 148 (1894) (reversing profits award that was based not on what defendant had made from infringement but on what third persons had made from the use of the invention); Jennings v. Carson, 4 Cranch 2, 21 (1807) (holding that an order requiring restitution could not apply to “those who were not in possession of the thing to be restored” and “had no power over it”) (citing Penhallow v. Doane’s Administrators, 3 Dall. 54 (1795) (reversing a restitution award in admiralty that ordered joint damages in excess of what each defendant received)). Finally, courts limited awards to the net profits from wrongdoing, that is, “the gain made upon any business or investment, when both the receipts and payments are taken into the account.” Rubber Co. v. Goodyear, 9 Wall. 788, 804 (1870); see also Livingston v. Woodworth, 15 How. 546, 559–560 (1854) (restricting an accounting remedy “to the actual gains and profits . . . during the time” the infringing machine “was in operation and during no other period” to avoid “convert[ing] a court of equity into an instrument for the punishment of simple torts”); Seymour v. McCormick, 16 How. 480, 490 (1854) (rejecting a blanket rule that infringing one component of a machine warranted a remedy measured by the full amounts of the profits earned from the machine); Mowry v. Whitney, 14 Wall. 620, 649 (1872) (vacating an accounting that exceeded the profits from infringement alone); Wooden-Ware Co. v. United States, 106 U.S. 432, 434–435 (1882) (explaining that an innocent trespasser is entitled to deduct labor costs from the gains obtained by wrongfully harvesting lumber). The Court has carved out an exception when the “entire profit of a business or undertaking” results from the wrongful activity. Root, 105 U. S., at 203. In such cases, the Court has explained, the defendant “will not be allowed to diminish the show of profits by putting in unconscionable claims for personal services or other inequitable deductions.” Ibid. In Goodyear, for example, the Court affirmed an accounting order that refused to deduct expenses under this rule. The Court there found that materials for which expenses were claimed were bought for the purposes of the infringement and “extraordinary salaries” appeared merely to be “dividends of profit under another name.” 9 Wall., at 803; see also Callaghan v. Myers, 128 U.S. 617, 663–664 (1888) (declining to deduct a defendant’s personal and living expenses from his profits from copyright violations, but distinguishing the expenses from salaries of officers in a corporation). Setting aside that circumstance, however, courts consistently restricted awards to net profits from wrongdoing after deducting legitimate expenses. Such remedies, when assessed against only culpable actors and for victims, fall comfortably within “those categories of relief that were typically available in equity.” Mertens, 508 U. S., at 256. C By incorporating these longstanding equitable principles into §78u(d)(5), Congress prohibited the SEC from seeking an equitable remedy in excess of a defendant’s net profits from wrongdoing. To be sure, the SEC originally endeavored to conform its disgorgement remedy to the common-law limitations in §78u(d)(5). Over the years, however, courts have occasionally awarded disgorgement in three main ways that test the bounds of equity practice: by ordering the proceeds of fraud to be deposited in Treasury funds instead of disbursing them to victims, imposing joint-and-several disgorgement liability, and declining to deduct even legitimate expenses from the receipts of fraud.[3] The SEC’s disgorgement remedy in such incarnations is in considerable tension with equity practices. Petitioners go further. They claim that this Court effectively decided in Kokesh that disgorgement is necessarily a penalty, and thus not the kind of relief available at equity. Brief for Petitioners 19–20, 22–26. Not so. Kokesh expressly declined to pass on the question. 581 U. S., at ___, n. 3 (slip op., at 5, n. 3). To be sure, the Kokesh Court evaluated a version of the SEC’s disgorgement remedy that seemed to exceed the bounds of traditional equitable principles. But that decision has no bearing on the SEC’s ability to conform future requests for a defendant’s profits to the limits outlined in common-law cases awarding a wrongdoer’s net gains. The Government, for its part, contends that the SEC’s interpretation of the equitable disgorgement remedy has Congress’ tacit support, even if it exceeds the bounds of equity practice. Brief for Respondent 13–21. It points to the fact that Congress has enacted a number of other statutes referring to “disgorgement.” That argument attaches undue significance to Congress’ use of the term. It is true that Congress has authorized the SEC to seek “disgorgement” in administrative actions. 15 U. S. C. §77h–1(e) (“In any cease-and-desist proceeding under subsection (a), the Commission may enter an order requiring accounting and disgorgement”). But it makes sense that Congress would expressly name the equitable powers it grants to an agency for use in administrative proceedings. After all, agencies are unlike federal courts where, “[u]nless otherwise provided by statute, all . . . inherent equitable powers . . . are available for the proper and complete exercise of that jurisdiction.” Porter, 328 U. S., at 398. Congress does not enlarge the breadth of an equitable, profit-based remedy simply by using the term “disgorgement” in various statutes. The Government argues that under the prior-construction principle, Congress should be presumed to have been aware of the scope of “disgorgement” as interpreted by lower courts and as having incorporated the (purportedly) prevailing meaning of the term into its subsequent enactments. Brief for Respondent 24. But “that canon has no application” where, among other things, the scope of disgorgement was “far from ‘settled.’ ” Armstrong v. Exceptional Child Center, Inc., 575 U.S. 320, 330 (2015). At bottom, even if Congress employed “disgorgement” as a shorthand to cross-reference the relief permitted by §78u(d)(5), it did not silently rewrite the scope of what the SEC could recover in a way that would contravene limitations embedded in the statute. After all, such “statutory reference[s]” to a remedy grounded in equity “must, absent other indication, be deemed to contain the limitations upon its availability that equity typically imposes.” Great-West, 534 U. S., at 211, n. 1. Accordingly, Congress’ own use of the term “disgorgement” in assorted statutes did not expand the contours of that term beyond a defendant’s net profits—a limit established by longstanding principles of equity. III Applying the principles discussed above to the facts of this case, petitioners briefly argue that their disgorgement award is unlawful because it crosses the bounds of traditional equity practice in three ways: It fails to return funds to victims, it imposes joint-and-several liability, and it declines to deduct business expenses from the award. Because the parties focused on the broad question whether any form of disgorgement may be ordered and did not fully brief these narrower questions, we do not decide them here. We nevertheless discuss principles that may guide the lower courts’ assessment of these arguments on remand. A Section 78u(d)(5) restricts equitable relief to that which “may be appropriate or necessary for the benefit of investors.” The SEC, however, does not always return the entirety of disgorgement proceeds to investors, instead depositing a portion of its collections in a fund in the Treasury. See SEC, Division of Enforcement, 2019 Ann. Rep. 16–17, https://www.sec.gov/files/enforcement-annual-report-2019.pdf. Congress established that fund in the Dodd-Frank Wall Street Reform and Consumer Protection Act for disgorgement awards that are not deposited in “disgorgement fund[s]” or otherwise “distributed to victims.” 124 Stat. 1844. The statute provides that these sums may be used to pay whistleblowers reporting securities fraud and to fund the activities of the Inspector General. Ibid. Here, the SEC has not returned the bulk of funds to victims, largely, it contends, because the Government has been unable to collect them.[4] The statute provides limited guidance as to whether the practice of depositing a defendant’s gains with the Treasury satisfies the statute’s command that any remedy be “appropriate or necessary for the benefit of investors.” The equitable nature of the profits remedy generally requires the SEC to return a defendant’s gains to wronged investors for their benefit. After all, the Government has pointed to no analogous common-law remedy permitting a wrongdoer’s profits to be withheld from a victim indefinitely without being disbursed to known victims. Cf. Root, 105 U. S., at 214–215 (comparing the accounting remedy to a breach-of-trust action, where a court would require the defendant to “refund the amount of profit which they have actually realized”). The Government maintains, however, that the primary function of depriving wrongdoers of profits is to deny them the fruits of their ill-gotten gains, not to return the funds to victims as a kind of restitution. See, e.g., SEC, Report Pursuant to Section 308(C) of the Sarbanes Oxley Act of 2002, p. 3, n. 2 (2003) (taking the position that disgorgement is not intended to make investors whole, but rather to deprive wrongdoers of ill-gotten gains); see also 6 T. Hazen, Law of Securities Regulation §16.18, p. 8 (rev. 7th ed. 2016) (concluding that the remedial nature of the disgorgement remedy does not mean that it is essentially compensatory and concluding that the “primary function of the remedy is to deny the wrongdoer the fruits of ill-gotten gains”). Under the Government’s theory, the very fact that it conducted an enforcement action satisfies the requirement that it is “appropriate or necessary for the benefit of investors.” But the SEC’s equitable, profits-based remedy must do more than simply benefit the public at large by virtue of depriving a wrongdoer of ill-gotten gains. To hold otherwise would render meaningless the latter part of §78u(d)(5). Indeed, this Court concluded similarly in Mertens when analyzing statutory language accompanying the term “equitable remedy.” 508 U. S., at 253 (interpreting the term “appropriate equitable relief ”). There, the Court found that the additional statutory language must be given effect since the section “does not, after all, authorize . . . ‘equitable relief ’ at large.” Ibid. As in Mertens, the phrase “appropriate or necessary for the benefit of investors” must mean something more than depriving a wrongdoer of his net profits alone, else the Court would violate the “cardinal principle of interpretation that courts must give effect, if possible, to every clause and word of a statute.” Parker Drilling Management Services, Ltd. v. Newton, 587 U. S. ___, ___ (2019) (slip op., at 9) (internal quotation marks omitted). The Government additionally suggests that the SEC’s practice of depositing disgorgement funds with the Treasury may be justified where it is infeasible to distribute the collected funds to investors.[5] Brief for Respondent 37. It is an open question whether, and to what extent, that practice nevertheless satisfies the SEC’s obligation to award relief “for the benefit of investors” and is consistent with the limitations of §78u(d)(5). The parties have not identified authorities revealing what traditional equitable principles govern when, for instance, the wrongdoer’s profits cannot practically be disbursed to the victims. But we need not address the issue here. The parties do not identify a specific order in this case directing any proceeds to the Treasury. If one is entered on remand, the lower courts may evaluate in the first instance whether that order would indeed be for the benefit of investors as required by §78u(d)(5) and consistent with equitable principles. B The SEC additionally has sought to impose disgorgement liability on a wrongdoer for benefits that accrue to his affiliates, sometimes through joint-and-several liability, in a manner sometimes seemingly at odds with the common-law rule requiring individual liability for wrongful profits. See, e.g., SEC v. Contorinis, 743 F.3d 296, 302 (CA2 2014) (holding that a defendant could be forced to disgorge not only what he “personally enjoyed from his exploitation of inside information, but also the profits of such exploitation that he channeled to friends, family, or clients”); SEC v. Clark, 915 F.2d 439, 454 (CA9 1990) (“It is well settled that a tipper can be required to disgorge his tippee’s profits”); SEC v. Whittemore, 659 F.3d 1, 10 (CADC 2011) (approving joint-and-several disgorgement liability where there is a close relationship between the defendants and collaboration in executing the wrongdoing). That practice could transform any equitable profits-focused remedy into a penalty. Cf. Marshall, 15 Wall., at 149. And it runs against the rule to not impose joint liability in favor of holding defendants “liable to account for such profits only as have accrued to themselves . . . and not for those which have accrued to another, and in which they have no participation.” Belknap, 161 U. S., at 25–26; see also Elizabeth v. Pavement Co., 97 U.S. 126 (1878). The common law did, however, permit liability for partners engaged in concerted wrongdoing. See, e.g., Ambler, 20 Wall., at 559. The historic profits remedy thus allows some flexibility to impose collective liability. Given the wide spectrum of relationships between participants and beneficiaries of unlawful schemes—from equally culpable codefendants to more remote, unrelated tipper-tippee arrangements—the Court need not wade into all the circumstances where an equitable profits remedy might be punitive when applied to multiple individuals. Here, petitioners were married. 754 Fed. Appx. 505; 262 F. Supp. 3d, at 960–961. The Government introduced evidence that Liu formed business entities and solicited investments, which he misappropriated. Id., at 961. It also presented evidence that Wang held herself out as the president, and a member of the management team, of an entity to which Liu directed misappropriated funds. Id., at 964. Petitioners did not introduce evidence to suggest that one spouse was a mere passive recipient of profits. Nor did they suggest that their finances were not commingled, or that one spouse did not enjoy the fruits of the scheme, or that other circumstances would render a joint-and-several disgorgement order unjust. Cf. SEC v. Hughes Capital Corp., 124 F.3d 449, 456 (CA3 1997) (finding that codefendant spouse was liable for unlawful proceeds where they funded her “lavish lifestyle”). We leave it to the Ninth Circuit on remand to determine whether the facts are such that petitioners can, consistent with equitable principles, be found liable for profits as partners in wrongdoing or whether individual liability is required. C Courts may not enter disgorgement awards that exceed the gains “made upon any business or investment, when both the receipts and payments are taken into the account.” Goodyear, 9 Wall., at 804; see also Restatement (Third) §51, Comment h, at 216 (reciting the general rule that a defendant is entitled to a deduction for all marginal costs incurred in producing the revenues that are subject to disgorgement). Accordingly, courts must deduct legitimate expenses before ordering disgorgement under §78u(d)(5). A rule to the contrary that “make[s] no allowance for the cost and expense of conducting [a] business” would be “inconsistent with the ordinary principles and practice of courts of chancery.” Tilghman, 125 U. S., at 145–146; cf. SEC v. Brown, 658 F.3d 858, 861 (CA8 2011) (declining to deduct even legitimate expenses like payments to innocent third-party employees and vendors). The District Court below declined to deduct expenses on the theory that they were incurred for the purposes of furthering an entirely fraudulent scheme. It is true that when the “entire profit of a business or undertaking” results from the wrongdoing, a defendant may be denied “inequitable deductions” such as for personal services. Root, 105 U. S., at 203. But that exception requires ascertaining whether expenses are legitimate or whether they are merely wrongful gains “under another name.” Goodyear, 9 Wall., at 803. Doing so will ensure that any disgorgement award falls within the limits of equity practice while preventing defendants from profiting from their own wrong. Root, 105 U. S., at 207. Although it is not necessary to set forth more guidance addressing the various circumstances where a defendant’s expenses might be considered wholly fraudulent, it suffices to note that some expenses from petitioners’ scheme went toward lease payments and cancer-treatment equipment. Such items arguably have value independent of fueling a fraudulent scheme. We leave it to the lower court to examine whether including those expenses in a profits-based remedy is consistent with the equitable principles underlying §78u(d)(5). * * * For the foregoing reasons, we vacate the judgment below and remand the case to the Ninth Circuit for further proceedings consistent with this opinion. It is so ordered. Notes 1 Courts have noted the relatively recent vintage of the term “disgorgement.” See, e.g., SEC v. Cavanaugh, 445 F.3d 105, 116, n. 24 (CA2 2006). The dissent contends that this recency in terminology alone removes disgorgement from the class of traditional equitable remedies, post, at 4 (opinion of Thomas, J.), despite seeming to recognize disgorgement’s parallels to restitution-based awards well within that class, post, at 4–5. It is no surprise that the dissent notes such parallels, given this Court’s acknowledgment that “disgorgement of improper profits” is “a remedy only for restitution” that is “traditionally considered . . . equitable.” Tull v. United States, 481 U.S. 412, 424 (1987); see also infra, at 7. The dissent also observes the solid equitable roots of an accounting for profits, post, at 3; accord, infra, at 6 (discussing the equitable origins of the accounting remedy), a remedy closely resembling disgorgement, see infra, at 8–9. In any event, casting aside a form of relief solely “based on the particular label affixed to [it] would ‘elevate form over substance,’ ” Aetna Health Inc. v. Davila, 542 U.S. 200, 214 (2004), leaving unresolved the question before us: whether the underlying profits-based award conforms to equity practice. 2 The dissent acknowledges that this Court has “referred to disgorgement as an equitable remedy in some of its prior decisions.” Post, at 6 (citing Feltner v. Columbia Pictures Television, Inc., 523 U.S. 340, 352 (1998)). While the dissent attempts to discount those cases for having “merely referred to the term” only “in passing,” post, at 6, those cases expressly “characterized as equitable . . . actions for disgorgement of improper profits” in analyzing whether certain remedies were traditionally available in equity, Feltner, 523 U. S., at 352 (citing Teamsters v. Terry, 494 U.S. 558, 570 (1990) (“characteriz[ing] damages as equitable where they are restitutionary, such as in ‘action[s] for disgorgement of improper profits’ ”); Tull, 481 U. S., at 424). 3 See, e.g., SEC v. Clark, 915 F.2d 439, 441, 454 (CA9 1990) (requiring defendant to disgorge the profits that his stockbroker made from unlawful trades); SEC v. Brown, 658 F.3d 858, 860–861 (CA8 2011) (per curiam) (ordering joint-and-several disgorgement of funds collected from investors and concluding that “ ‘the overwhelming weight of authority hold[s] that securities law violators may not offset their disgorgement liability with business expenses’ ”); SEC v. Contorinis, 743 F.3d 296, 304–306 (CA2 2014) (requiring defendant to disgorge benefits conferred on close associates). 4 According to the Government, petitioners “transferred the bulk of their misappropriated funds to China, defied the district court’s order to repatriate those funds, and fled the United States.” Brief for Respondent 36. 5 We express no view as to whether the SEC has offered adequate proof of failed attempts to return funds to investors here. To the extent that feasibility is relevant at all to equitable principles, we observe that lower courts are well equipped to evaluate the feasibility of returning funds to victims of fraud. See, e.g., SEC v. Lund, 570 F. Supp. 1397, 1404–1405 (CD Cal. 1983) (appointing a magistrate judge to determine whether it was feasible to locate victims of financial wrongdoing).

590.US.2019_18-8369

The Prison Litigation Reform Act of 1995 (PLRA) established what has become known as the three-strikes rule, which generally prevents a prisoner from bringing suit in forma pauperis (IFP) if he has had three or more prior suits “dismissed on the grounds that [they were] frivolous, malicious, or fail[ed] to state a claim upon which relief may be granted.” 28 U. S. C. §1915(g). Petitioner Arthur Lomax, an inmate in a Colorado prison, filed this suit against respondent prison officials to challenge his expulsion from the facility’s sex-offender treatment program. He also moved for IFP status, but he had already brought three unsuccessful legal actions during his time in prison. If the dispositions of those cases qualify as strikes under Section 1915(g), Lomax may not now proceed IFP. The courts below concluded that they did, rejecting Lomax’s argument that two of the dismissals should not count as strikes because they were without prejudice. Held: Section 1915(g)’s three-strikes provision refers to any dismissal for failure to state a claim, whether with prejudice or without. This case begins, and pretty much ends, with Section 1915(g)’s text. The provision’s broad language covers all dismissals for failure to state a claim, whether issued with or without prejudice to a plaintiff’s ability to reassert his claim in a later action. A strike-call under Section 1915(g) thus hinges exclusively on the basis for the dismissal, regardless of the decision’s prejudicial effect. To reach the opposite result would require reading the word “dismissed” in Section 1915(g) as “dismissed with prejudice.” Doing so would also introduce inconsistencies into the PLRA, which has three other provisions mentioning “dismiss[als]” for “fail[ure] to state a claim.” §§1915(e)(2)(B)(ii), 1915A(b); 42 U. S. C. §1997e(c). As the parties agree, those provisions do not deprive courts of the ability to dismiss suits without prejudice. Lomax nonetheless maintains that Section 1915(g)’s phrase “dismissed [for] fail[ure] to state a claim” is a “legal term of art” referring only to dismissals with prejudice. To support this view, he points to Federal Rule of Civil Procedure 41(b), which tells courts to treat a dismissal “as an adjudication on the merits”—meaning a dismissal with prejudice—where the dismissal order does not specify. But Rule 41(b) is necessary precisely because “dismissed for failure to state a claim” refers to dismissals both with and without prejudice. The existence of the rule thus undercuts Lomax’s position. Lomax also argues that the Court should interpret the phrase “failure to state a claim” based on the other two grounds for dismissal listed in Section 1915(g). But contra Lomax’s view, courts can and sometimes do dismiss at least frivolous actions without prejudice. Still more fundamentally, interpreting the phrase “failure to state a claim” based on the pre-existing terms “frivolous” and “malicious” would defeat the PLRA’s expansion of the statute beyond what was already there. Pp. 3–7. 754 Fed. Appx. 756, affirmed. Kagan, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Alito, Sotomayor, Gorsuch, and Kavanaugh, JJ., joined, and in which Thomas, J., joined as to all but footnote 4.

if the prisoner has, on 3 or more prior occasions, while incarcerated or detained in any facility, brought an action or appeal in a court of the United States that was dismissed on the grounds that it is frivolous, malicious, or fails to state a claim upon which relief may be granted, unless the prisoner is under imminent danger of serious physical injury.” 28 U. S. C. §1915(g). 3 Two of the cases were dismissed under Heck v. Humphrey, 512 U.S. 477 (1994), which holds that a claim challenging the validity of a conviction or sentence under 42 U. S. C. §1983 “does not accrue until the conviction or sentence has been invalidated.” 512 U. S., at 490. In concluding that those two Heck dismissals were for failure to state a claim, the District Court followed Circuit precedent. See Smith v. Veterans Admin., 636 F.3d 1306, 1312 (CA10 2011). Not all Courts of Appeals accept that view. See, e.g., Mejia v. Harrington, 541 Fed. Appx. 709, 710 (CA7 2013). But Lomax did not raise that issue, and we therefore do not address it. 4 Four Circuits treat dismissals without prejudice for failure to state a claim as strikes. See Orr v. Clements, 688 F.3d 463, 465 (CA8 2012); Paul v. Marberry, 658 F.3d 702, 704 (CA7 2011); O’Neal v. Price, 531 F.3d 1146, 1154 (CA9 2008); Day v. Maynard, 200 F.3d 665, 667 (CA10 1999) (per curiam). Two Circuits do the opposite. See Millhouse v. Heath, 866 F.3d 152, 162–163 (CA3 2017); McLean v. United States, 566 F.3d 391, 396–397 (CA4 2009). 5 Note, however, that the provision does not apply when a court gives a plaintiff leave to amend his complaint. Courts often take that path if there is a chance that amendment can cure a deficient complaint. See Fed. Rule Civ. Proc. 15(a) (discussing amendments to pleadings). In that event, because the suit continues, the court’s action falls outside of Section 1915(g) and no strike accrues. See Brief for Respondents 31–35 (noting that flexible amendment practices “ensure that potentially meritorious prisoner suits are not hastily dismissed with a strike”); Brief for United States as Amicus Curiae 27–28 (similar); Tr. of Oral Arg. 32–34, 44 (similar).

590.US.2019_18-1086

Petitioners (collectively Lucky Brand) and respondent (Marcel) both use the word “Lucky” as part of their marks on jeans and other apparel. Marcel received a trademark registration for the phrase “Get Lucky,” and Lucky Brand uses the registered trademark “Lucky Brand” and other marks with the word “Lucky.” This has led to nearly 20 years of litigation, proceeding in three rounds. The first round resulted in a 2003 settlement agreement in which Lucky Brand agreed to stop using the phrase “Get Lucky” and Marcel agreed to release any claims regarding Lucky Brand’s use of its own trademarks. In the second round (2005 Action), Lucky Brand sued Marcel and its licensee for violating its trademarks. Marcel filed several counterclaims turning, as relevant here, on Lucky Brand’s alleged continued use of “Get Lucky,” but it did not claim that Lucky Brand’s use of its own marks alone infringed the “Get Lucky” mark. In both a motion to dismiss the counterclaims and an answer to them, Lucky Brand argued that the counterclaims were barred by the settlement agreement, but it did not invoke that defense later in the proceedings. The court in the 2005 Action permanently enjoined Lucky Brand from copying or imitating Marcel’s “Get Lucky” mark, and a jury found against Lucky Brand on Marcel’s remaining counterclaims. In the third round (2011 Action), Marcel sued Lucky Brand for continuing to infringe the “Get Lucky” mark, but it did not reprise its 2005 allegation about Lucky Brand’s use of the “Get Lucky” phrase. After protracted litigation, Lucky Brand moved to dismiss, arguing—for the first time since early in the 2005 Action—that Marcel had released its claims in the settlement agreement. Marcel countered that Lucky Brand could not invoke the release defense because it could have pursued that defense in the 2005 Action but did not. The District Court granted Lucky Brand’s motion to dismiss. The Second Circuit vacated and remanded, concluding that “defense preclusion” prohibited Lucky Brand from raising an unlitigated defense that it should have raised earlier. Held: Because Marcel’s 2011 Action challenged different conduct—and raised different claims—from the 2005 Action, Marcel cannot preclude Lucky Brand from raising new defenses. Pp. 6–12. (a) This case asks whether so-called “defense preclusion” is a valid application of res judicata: a term comprising the doctrine of issue preclusion, which precludes a party from relitigating an issue actually decided in a prior action and necessary to the judgment, and the doctrine of claim preclusion, which prevents parties from raising issues that could have been raised and decided in a prior action. Any preclusion of defenses must, at a minimum, satisfy the strictures of issue preclusion or claim preclusion. See, e.g., Davis v. Brown, 94 U.S. 423, 428. Here, issue preclusion does not apply, so the causes of action must share a “common nucleus of operative fact[s]” for claim preclusion to apply, Restatement (Second) of Judgments §24, Comment b, p. 199. Pp. 6–8. (b) Because the two suits here involved different marks and different conduct occurring at different times, they did not share a “common nucleus of operative facts.” The 2005 claims depended on Lucky Brand’s alleged use of “Get Lucky.” But in the 2011 Action, Marcel alleged that the infringement was Lucky Brand’s use of its other marks containing the word “Lucky,” not any use of “Get Lucky” itself. The conduct in the 2011 Action also occurred after the conclusion of the 2005 Action. But claim preclusion generally “ ‘does not bar claims that are predicated on events that postdate the filing of the initial complaint,’ ” Whole Woman’s Health v. Hellerstedt, 579 U. S. ___, ___, because events occurring after a plaintiff files suit often give rise to new “operative facts” creating a new claim to relief. Pp. 8–10. (c) Marcel claims that treatises and this Court’s cases support a version of “defense preclusion” that extends to the facts of this case. But none of those authorities describe scenarios applicable here, and they are unlikely to stand for anything more than that traditional claim or issue preclusion principles may bar defenses raised in a subsequent suit—principles that do not bar Lucky Brand’s release defense here. Pp. 10–12. 898 F.3d 232, reversed and remanded. Sotomayor, J., delivered the opinion for a unanimous Court.

This case arises from protracted litigation between petitioners Lucky Brand Dungarees, Inc., and others (collectively Lucky Brand) and respondent Marcel Fashions Group, Inc. (Marcel). In the latest lawsuit between the two, Lucky Brand asserted a defense against Marcel that it had not pressed fully in a preceding suit between the parties. This Court is asked to determine whether Lucky Brand’s failure to litigate the defense in the earlier suit barred Lucky Brand from invoking it in the later suit. Because the parties agree that, at a minimum, the preclusion of such a defense in this context requires that the two suits share the same claim to relief—and because we find that the two suits here did not—Lucky Brand was not barred from raising its defense in the later action. I Marcel and Lucky Brand both sell jeans and other apparel. Both entities also use the word “Lucky” as part of their marks on clothing. In 1986, Marcel received a federal trademark registration for “Get Lucky”; a few years later, in 1990, Lucky Brand began selling apparel using the registered trademark “Lucky Brand” and other marks that include the word “Lucky.” 779 F.3d 102, 105 (CA2 2015). Three categories of marks are at issue in this case: Marcel’s “Get Lucky” mark; Lucky Brand’s “Lucky Brand” mark; and various other marks owned by Lucky Brand that contain the word “Lucky.” These trademarks have led to nearly 20 years of litigation between the two companies, proceeding in three rounds. A In 2001—the first round—Marcel sued Lucky Brand, alleging that Lucky Brand’s use of the phrase “Get Lucky” in advertisements infringed Marcel’s trademark. In 2003, the parties signed a settlement agreement. As part of the deal, Lucky Brand agreed to stop using the phrase “Get Lucky.” App. 191. In exchange, Marcel agreed to release any claims regarding Lucky Brand’s use of its own trademarks. Id., at 191–192. B The ink was barely dry on the settlement agreement when, in 2005, the parties began a second round of litigation (2005 Action). Lucky Brand filed suit, alleging that Marcel and its licensee violated its trademarks by copying its designs and logos in a new clothing line. As relevant here, Marcel filed several counterclaims that all turned, in large part, on Lucky Brand’s alleged continued use of “Get Lucky”: One batch of allegations asserted that Lucky Brand had continued to use Marcel’s “Get Lucky” mark in violation of the settlement agreement, while others alleged that Lucky Brand’s use of the phrase “Get Lucky” and “Lucky Brand” together was “confusingly similar to”—and thus infringed––Marcel’s “Get Lucky” mark. Defendants’ Answer, Affirmative Defenses, and Counterclaims to Plaintiffs’ Complaint in No. 1:05–cv–06757 (SDNY), Doc. 40–2, p. 39; see id., at 34–41. None of Marcel’s counterclaims alleged that Lucky Brand’s use of its own marks alone—i.e., independent of any alleged use of “Get Lucky”—infringed Marcel’s “Get Lucky” mark. Lucky Brand moved to dismiss the counterclaims, alleging that they were barred by the release provision of the settlement agreement. After the District Court denied the motion without prejudice, Lucky Brand noted the release defense once more in its answer to Marcel’s counterclaims. But as the 2005 Action proceeded, Lucky Brand never again invoked the release defense. The 2005 Action concluded in two phases. First, as a sanction for misconduct during discovery, the District Court concluded that Lucky Brand violated the settlement agreement by continuing to use “Get Lucky” and permanently enjoined Lucky Brand from copying or imitating Marcel’s “Get Lucky” mark. Order Granting Partial Summary Judgment and Injunction in No. 1:05–cv–06757, Doc. 183; see also App. 203–204. The injunction did not enjoin, or even mention, Lucky Brand’s use of any other marks or phrases containing the word “Lucky.” Order Granting Partial Summary Judgment and Injunction, Doc. 183. The case then proceeded to trial. The jury found against Lucky Brand on Marcel’s remaining counterclaims—those that alleged infringement from Lucky Brand’s continued use of the “Get Lucky” catchphrase alongside its own marks. See Brief for Respondent 52. C In April 2011, the third round of litigation began: Marcel filed an action against Lucky Brand (2011 Action), maintaining that Lucky Brand continued to infringe Marcel’s “Get Lucky” mark and, in so doing, contravened the judgment issued in the 2005 Action. This complaint did not reprise Marcel’s earlier allegation (in the 2005 Action) that Lucky Brand continued to use the “Get Lucky” phrase. Marcel argued only that Lucky Brand’s continued, post-2010 use of Lucky Brand’s own marks—some of which used the word “Lucky”—infringed Marcel’s “Get Lucky” mark in a manner that (according to Marcel) was previously found infringing.[1] Marcel requested that the District Court enjoin Lucky Brand from using any of Lucky Brand’s marks containing the word “Lucky.” The District Court granted Lucky Brand summary judgment, concluding that Marcel’s claims in the 2011 Action were essentially the same as its counterclaims in the 2005 Action. But the Court of Appeals for the Second Circuit disagreed. 779 F.3d 102. The court concluded that Marcel’s claims in the 2011 Action were distinct from those it had asserted in the 2005 Action, because the claims at issue in the 2005 Action were “for earlier infringements.” Id., at 110. As the court noted, “[w]inning a judgment . . . does not deprive the plaintiff of the right to sue” for the defendant’s “subsequent similar violations.” Id., at 107. The Second Circuit further rejected Marcel’s request to hold Lucky Brand in contempt for violating the injunction issued in the 2005 Action. The court noted that the conduct at issue in the 2011 Action was Lucky Brand’s use of its own marks—not the use of the phrase “Get Lucky.” By contrast, the 2005 injunction prohibited Lucky Brand from using the “Get Lucky” mark—not Lucky Brand’s own marks that happened to contain the word “Lucky.” Id., at 111. Moreover, the court reasoned that the jury in the 2005 Action had been “free to find infringement of Marcel’s ‘Get Lucky’ mark based solely on Lucky Brand’s use of [the phrase] ‘Get Lucky.’ ” Id., at 112. The court vacated and remanded for further proceedings. On remand to the District Court, Lucky Brand moved to dismiss, arguing—for the first time since its motion to dismiss and answer in the 2005 Action—that Marcel had released its claims by entering the settlement agreement. Marcel countered that Lucky Brand was precluded from invoking the release defense, because it could have pursued the defense fully in the 2005 Action but had neglected to do so. The District Court granted Lucky Brand’s motion to dismiss, holding that it could assert its release defense and that the settlement agreement indeed barred Marcel’s claims. The Second Circuit vacated and remanded, concluding that a doctrine it termed “defense preclusion” prohibited Lucky Brand from raising the release defense in the 2011 Action. 898 F.3d 232 (2018). Noting that a different category of preclusion—issue preclusion—may be wielded against a defendant, see Parklane Hosiery Co. v. Shore, 439 U.S. 322 (1979), the court reasoned that the same should be true of claim preclusion: A defendant should be precluded from raising an unlitigated defense that it should have raised earlier. The panel then held that “defense preclusion” bars a party from raising a defense where: “(i) a previous action involved an adjudication on the merits”; “(ii) the previous action involved the same parties”; “(iii) the defense was either asserted or could have been asserted, in the prior action”; and “(iv) the district court, in its discretion, concludes that preclusion of the defense is appropriate.” 898 F. 3d, at 241. Finding each factor satisfied in this case, the panel vacated the District Court’s judgment. We granted certiorari, 588 U. S. ___ (2019), to resolve differences among the Circuits regarding when, if ever, claim preclusion applies to defenses raised in a later suit. Compare 898 F. 3d, at 241, with Hallco Mfg. Co. v. Foster, 256 F.3d 1290, 1297–1298 (CA Fed. 2001); McKinnon v. Blue Cross and Blue Shield of Alabama, 935 F.2d 1187, 1192 (CA11 1991). II A This case asks whether so-called “defense preclusion” is a valid application of res judicata: a term that now comprises two distinct doctrines regarding the preclusive effect of prior litigation. 18 C. Wright, H. Miller, & E. Cooper, Federal Practice and Procedure §4402 (3d ed. 2016) (Wright & Miller). The first is issue preclusion (sometimes called collateral estoppel), which precludes a party from relitigating an issue actually decided in a prior case and necessary to the judgment. Allen v. McCurry, 449 U.S. 90, 94 (1980); see Parklane Hosiery, 439 U. S., at 326, n. 5. The second doctrine is claim preclusion (sometimes itself called res judicata). Unlike issue preclusion, claim preclusion prevents parties from raising issues that could have been raised and decided in a prior action—even if they were not actually litigated. If a later suit advances the same claim as an earlier suit between the same parties, the earlier suit’s judgment “prevents litigation of all grounds for, or defenses to, recovery that were previously available to the parties, regardless of whether they were asserted or determined in the prior proceeding.” Brown v. Felsen, 442 U.S. 127, 131 (1979); see also Wright & Miller §4407. Suits involve the same claim (or “cause of action”) when they “ ‘aris[e] from the same transaction,’ ” United States v. Tohono O’odham Nation, 563 U.S. 307, 316 (2011) (quoting Kremer v. Chemical Constr. Corp., 456 U.S. 461, 482, n. 22 (1982)), or involve a “common nucleus of operative facts,” Restatement (Second) of Judgments §24, Comment b, p. 199 (1982) (Restatement (Second)). Put another way, claim preclusion “describes the rules formerly known as ‘merger’ and ‘bar.’ ” Taylor v. Sturgell, 553 U.S. 880, 892, n. 5 (2008). “If the plaintiff wins, the entire claim is merged in the judgment; the plaintiff cannot bring a second independent action for additional relief, and the defendant cannot avoid the judgment by offering new defenses.” Wright & Miller §4406. But “[i]f the second lawsuit involves a new claim or cause of action, the parties may raise assertions or defenses that were omitted from the first lawsuit even though they were equally relevant to the first cause of action.” Ibid. As the Second Circuit itself seemed to recognize, see 898 F. 3d, at 236–237, this Court has never explicitly recognized “defense preclusion” as a standalone category of res judicata, unmoored from the two guideposts of issue preclusion and claim preclusion. Instead, our case law indicates that any such preclusion of defenses must, at a minimum, satisfy the strictures of issue preclusion or claim preclusion. See, e.g., Davis v. Brown, 94 U.S. 423, 428 (1877) (holding that where two lawsuits involved different claims, preclusion operates “only upon the matter actually at issue and determined in the original action”).[2] The parties thus agree that where, as here, issue preclusion does not apply, a defense can be barred only if the “causes of action are the same” in the two suits—that is, where they share a “ ‘common nucleus of operative fact[s].’ ” Brief for Respondent 2, 27, 31, 50; accord, Reply Brief 3. B Put simply, the two suits here were grounded on different conduct, involving different marks, occurring at different times. They thus did not share a “common nucleus of operative facts.” Restatement (Second) §24, Comment b, at 199. To start, claims to relief may be the same for the purposes of claim preclusion if, among other things, “ ‘a different judgment in the second action would impair or destroy rights or interests established by the judgment entered in the first action.’ ” Wright & Miller §4407. Here, however, the 2011 Action did not imperil the judgment of the 2005 Action because the lawsuits involved both different conduct and different trademarks. In the 2005 Action, Marcel alleged that Lucky Brand infringed Marcel’s “Get Lucky” mark both by directly imitating its “Get Lucky” mark and by using the “Get Lucky” slogan alongside Lucky Brand’s other marks in a way that created consumer confusion. Brief for Respondent 52. Marcel appears to admit, thus, that its claims in the 2005 Action depended on Lucky Brand’s alleged use of “Get Lucky.” Id., at 9–10 (“Marcel’s reverse-confusion theory [in the 2005 Action] depended, in part, on Lucky’s continued imitation of the GET LUCKY mark”). By contrast, the 2011 Action did not involve any alleged use of the “Get Lucky” phrase. Indeed, Lucky Brand had been enjoined in the 2005 Action from using “Get Lucky,” and in the 2011 Action, Lucky Brand was found not to have violated that injunction. 779 F. 3d, at 111–112. The parties thus do not argue that Lucky Brand continued to use “Get Lucky” after the 2005 Action concluded, and at oral argument, counsel for Marcel appeared to confirm that Marcel’s claims in the 2011 Action did not allege that Lucky Brand continued to use “Get Lucky.” Tr. of Oral Arg. 46. Instead, Marcel alleged in the 2011 Action that Lucky Brand committed infringement by using Lucky Brand’s own marks containing the word “Lucky”—not the “Get Lucky” mark itself. Plainly, then, the 2011 Action challenged different conduct, involving different marks. Not only that, but the complained-of conduct in the 2011 Action occurred after the conclusion of the 2005 Action. Claim preclusion generally “does not bar claims that are predicated on events that postdate the filing of the initial complaint.” Whole Woman’s Health v. Hellerstedt, 579 U. S. ___, ___ (2016) (slip op., at 12) (internal quotation marks omitted); Lawlor v. National Screen Service Corp., 349 U.S. 322, 327–328 (1955) (holding that two suits were not “based on the same cause of action,” because “[t]he conduct presently complained of was all subsequent to” the prior judgment and it “cannot be given the effect of extinguishing claims which did not even then exist and which could not possibly have been sued upon in the previous case”). This is for good reason: Events that occur after the plaintiff files suit often give rise to new “[m]aterial operative facts” that “in themselves, or taken in conjunction with the antecedent facts,” create a new claim to relief. Restatement (Second) §24, Comment f, at 203; 18 J. Moore, D. Coquillette, G. Joseph, G. Vairo, & C. Varner, Federal Practice §131.22[1], p. 131–55, n. 1 (3d ed. 2019) (citing cases where “[n]ew facts create[d a] new claim”). This principle takes on particular force in the trademark context, where the enforceability of a mark and likelihood of confusion between marks often turns on extrinsic facts that change over time. As Lucky Brand points out, liability for trademark infringement turns on marketplace realities that can change dramatically from year to year. Brief for Petitioners 42–45. It is no surprise, then, that the Second Circuit held that Marcel’s 2011 Action claims were not barred by the 2005 Action. By the same token, the 2005 Action could not bar Lucky Brand’s 2011 defenses. At bottom, the 2011 Action involved different marks, different legal theories, and different conduct—occurring at different times. Because the two suits thus lacked a “common nucleus of operative facts,” claim preclusion did not and could not bar Lucky Brand from asserting its settlement agreement defense in the 2011 Action. III Resisting this conclusion, Marcel points to treatises and this Court’s cases, arguing that they support a version of “defense preclusion” doctrine that extends to the facts of this case. Brief for Respondent 24–26. But these authorities do no such thing. As an initial matter, regardless of what those authorities might imply about “defense preclusion,” none of them describe scenarios applicable here. Moreover, we doubt that these authorities stand for anything more than that traditional claim- or issue-preclusion principles may bar defenses raised in a subsequent suit—principles that, as explained above, do not bar Lucky Brand’s release defense here. Take, for example, cases that involve either judgment enforcement or a collateral attack on a prior judgment. Id., at 26–35. In the former scenario, a party takes action to enforce a prior judgment already issued against another; in the latter, a party seeks to avoid the effect of a prior judgment by bringing a suit to undo it. If, in either situation, a different outcome in the second action “would nullify the initial judgment or would impair rights established in the initial action,” preclusion principles would be at play. Restatement (Second) §22(b), at 185; Wright & Miller §4414. In both scenarios, courts simply apply claim preclusion or issue preclusion to prohibit a claim or defense that would attack a previously decided claim.[3] But these principles do not preclude defendants from asserting defenses to new claims, which is precisely what Marcel would have us do here. In any event, judgment-enforcement and collateral-attack scenarios are far afield from the circumstances of this case. Lucky Brand’s defense in the 2011 Action did not threaten the judgment issued in the 2005 Action or, as Marcel argues, “achieve the same practical result” that the above-mentioned principles seek to avoid. Brief for Respondent 31–32. Indeed, while the judgment in the 2005 Action plainly prohibited Lucky Brand from using “Get Lucky,” it did not do the same with respect to Lucky Brand’s continued, standalone use of its own marks containing the word “Lucky”—the only conduct at issue in the 2011 Action. Put simply, Lucky Brand’s defense to new claims in the 2011 Action did not risk impairing the 2005 judgment. Nor do cases like Beloit v. Morgan, 7 Wall. 619 (1869), aid Marcel. See Brief for Respondent 32–33. To be sure, Beloit held that a defendant in a second suit over bonds “of the same issue” was precluded from raising a defense it had not raised in the first suit. 7 Wall., at 620. But the Court there explained that the judgment in the first suit “established conclusively the original validity of the securities described in the bill, and the liability of the town to pay them.” Id., at 623. In other words, by challenging the validity of all bonds of the same issue, the defense in the second suit would have threatened the validity of the judgment in the first suit. The same cannot be said of the defense raised in the 2011 Action vis-à-vis the judgment in the 2005 Action. * * * At bottom, Marcel’s 2011 Action challenged different conduct—and raised different claims—from the 2005 Action. Under those circumstances, Marcel cannot preclude Lucky Brand from raising new defenses. The judgment of the Second Circuit is therefore reversed, and the case is remanded for proceedings consistent with this opinion. It is so ordered. Notes 1 See Complaint for Injunctive Relief and Trademark Infringement in No. 1:11–cv–05523 (SDNY), Doc. 1, ¶15 (“Despite the entry of the [2005 Action judgment], [Lucky Brand] ha[s] continued to willfully . . . infringe [Marcel’s] GET LUCKY mark by using the Lucky Brand marks in the identical manner and form and on the same goods for which [it] w[as] found liable for infringement”); id., ¶20 (“Despite the entry of the” 2005 Action judgment, Lucky Brand has “continued its uninterrupted and willful use of the Lucky Brand marks and any other trademarks including the word ‘Lucky’ ”). 2 There may be good reasons to question any application of claim preclusion to defenses. It has been noted that in suits involving successive claims against the same defendant, courts often “assum[e] that the defendant may raise defenses in the second action that were not raised in the first, even though they were equally available and relevant in both actions.” Wright & Miller §4414. This is because “[v]arious considerations, other than actual merits, may govern” whether to bring a defense, “such as the smallness of the amount or the value of the property in controversy, the difficulty of obtaining the necessary evidence, the expense of the litigation, and [a party’s] own situation.” Cromwell v. County of Sac, 94 U.S. 351, 356 (1877). Here, however, this Court need not determine when (if ever) applying claim preclusion to defenses may be appropriate, because a necessary predicate—identity of claims—is lacking. 3 One might ask: If any preclusion of defenses (under the claim-preclusion rubric) requires identity of claims in two suits, how could the second similar suit have avoided standard claim preclusion in the first place? Different contexts may yield different answers. In a judgment-enforcement context, the answer may be that claim preclusion applies only “to a final judgment rendered in an action separate from that in which the doctrine is asserted.” 18 J. Moore, D. Coquillette, G. Joseph, G. Vairo, & C. Varner, Federal Practice §131.31[1], p. 131–116 (3d ed. 2019) (emphasis added). Thus—although claim preclusion does apply to a later, standalone suit seeking relief that could have been obtained in the first—it “is not applicable to . . . efforts to obtain supplemental relief in the original action, or direct attacks on the judgment.” Ibid (footnote deleted). The upshot is that—even if a court deems the underlying core of operative facts to be the same—a plaintiff in that circumstance is not precluded from enforcing its rights with respect to continuing wrongful conduct.

590.US.2019_18-1023

The Patient Protection and Affordable Care Act established online exchanges where insurers could sell their healthcare plans. The now-expired “Risk Corridors” program aimed to limit the plans’ profits and losses during the exchanges’ first three years (2014 through 2016). See §1342, 124Stat. 211. Section 1342 set out a formula for computing a plan’s gains or losses at the end of each year, providing that eligible profitable plans “shall pay” the Secretary of the Department of Health and Human Services (HHS), while the Secretary “shall pay” eligible unprofitable plans. The Act neither appropriated funds for these yearly payments nor limited the amounts that the Government might pay. Nor was the program required to be budget neutral. Each year, the Government owed more money to unprofitable insurers than profitable insurers owed to the Government, resulting in a total deficit of more than $12 billion. And at the end of each year, the appropriations bills for the Centers for Medicare and Medicaid Services (CMS) included a rider preventing CMS from using the funds for Risk Corridors payments. Petitioners—four health-insurance companies that claim losses under the program—sued the Federal Government for damages in the Court of Federal Claims. Invoking the Tucker Act, they alleged that §1342 obligated the Government to pay the full amount of their losses as calculated by the statutory formula and sought a money judgment for the unpaid sums owed. Only one petitioner prevailed in the trial courts, and the Federal Circuit ruled for the Government in each appeal, holding that §1342 had initially created a Government obligation to pay the full amounts, but that the subsequent appropriations riders impliedly “repealed or suspended” that obligation. Held: 1. The Risk Corridors statute created a Government obligation to pay insurers the full amount set out in §1342’s formula. Pp. 9–16. (a) The Government may incur an obligation directly through statutory language, without also providing details about how the obligation must be satisfied. See United States v. Langston, 118 U.S. 389. Pp. 9–11. (b) Section 1342 imposed a legal duty of the United States that could mature into a legal liability through the insurers’ participation in the exchanges. This conclusion flows from the express terms and context of §1342, which imposed an obligation by using the mandatory term “shall.” The section’s mandatory nature is underscored by the adjacent provisions, which differentiate between when the HHS Secretary “shall” take certain actions and when she “may” exercise discretion. See §§1341(b)(2), 1343(b). Section 1342 neither requires the Risk Corridors program to be budget-neutral nor suggests that the Secretary’s payments to unprofitable plans pivoted on profitable plans’ payments to the Secretary or that a partial payment would satisfy the Government’s whole obligation. It thus must be given its plain meaning: The Government “shall pay” the sum prescribed by §1342. Pp. 11–13. (c) Contrary to the Government’s contention, neither the Appropriations Clause nor the Anti-Deficiency Act addresses whether Congress itself can create or incur an obligation directly by statute. Nor does §1342’s obligation-creating language turn on whether Congress expressly provided budget authority before appropriating funds. The Government’s arguments also conflict with well-settled principles of statutory interpretation. That §1342 contains no language limiting the obligation to the availability of appropriations, while Congress expressly used such limiting language in other Affordable Care Act provisions, indicates that Congress intended a different meaning in §1342. Pp. 13–16. 2. Congress did not impliedly repeal the obligation through its appropriations riders. Pp. 16–23. (a) Because “ ‘repeals by implication are not favored,’ ” Morton v. Mancari, 417 U.S. 535, 549, this Court will regard each of two statutes effective unless Congress’ intention to repeal is “ ‘clear and manifest,’ ” or the laws are “irreconcilable,” id., at 550–551. In the appropriations context, this requires the Government to show “something more than the mere omission to appropriate a sufficient sum.” United States v. Vulte, 233 U.S. 509, 515. As Langston and Vulte confirm, the appropriations riders here did not manifestly repeal or discharge the Government’s uncapped obligation, see Langston, 118 U. S., at 394, and do not indicate “any other purpose than the disbursement of a sum of money for the particular fiscal years,” Vulte, 233 U. S., at 514. Nor is there any indication that HHS and CMS thought that the riders clearly expressed an intent to repeal. Pp. 16–19. (b) Appropriations measures have been found irreconcilable with statutory obligations to pay, but the riders here did not use the kind of “shall not take effect” language decisive in United States v. Will, 449 U.S. 200, 222–223, or purport to “suspen[d]” §1342 prospectively or to foreclose funds from “any other Act” “notwithstanding” §1342’s money-mandating text, United States v. Dickerson, 310 U.S. 554, 556–557. They also did not reference §1342’s payment formula, let alone “irreconcilabl[y]” change it, United States v. Mitchell, 109 U.S. 146, 150, or provide that payments from profitable plans would be “ ‘in full compensation’ ” of the Government’s obligation to unprofitable plans, United States v. Fisher, 109 U.S. 143, 150. Pp. 19–21. (c) The legislative history cited by the Federal Circuit is also unpersuasive. Pp. 22–23. 3. Petitioners properly relied on the Tucker Act to sue for damages in the Court of Federal Claims. Pp. 23–30. (a) The United States has waived its immunity for certain damages suits in the Court of Federal Claims through the Tucker Act. Because that Act does not create “substantive rights,” United States v. Navajo Nation, 556 U.S. 287, 290, a plaintiff must premise her damages action on “other sources of law,” like “statutes or contracts,” ibid., provided those statutes “ ‘can fairly be interpreted as mandating compensation by the Federal Government for the damage sustained,’ ” United States v. White Mountain Apache Tribe, 537 U.S. 465, 472. The Act does, however, yield when the obligation-creating statute provides its own detailed remedies or when the Administrative Procedure Act provides an avenue for relief. Pp. 23–26. (b) Petitioners clear each hurdle: The Risk Corridors statute is fairly interpreted as mandating compensation for damages, and neither exception to the Tucker Act applies. Section 1342’s mandatory “ ‘shall pay’ language” falls comfortably within the class of statutes that permit recovery of money damages in the Court of Federal Claims. This finding is bolstered by §1342’s focus on compensating insurers for past conduct. And there is no separate remedial scheme supplanting the Court of Federal Claims’ power to adjudicate petitioners’ claims. See United States v. Bormes, 568 U.S. 6, 12. Nor does the Administrative Procedure Act bar petitioners’ Tucker Act suit. In contrast to Bowen v. Massachusetts, 487 U.S. 879, a Medicaid case where the State sued the HHS Secretary under the Administrative Procedure Act in district court, petitioners here seek not prospective, nonmonetary relief to clarify future obligations but specific sums already calculated, past due, and designed to compensate for completed labors. The Risk Corridors statute and Tucker Act allow them that remedy. And because the Risk Corridors program expired years ago, this litigation presents no special concern, as Bowen did, about managing a complex ongoing relationship or tracking ever-changing accounting sheets. Pp. 26–30. No. 18–1023 and No. 18–1028 (second judgment), 729 Fed. Appx. 939; No. 18–1028 (first judgment), 892 F.3d 1311; No. 18–1038, 892 F.3d 1184, reversed and remanded. Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Kagan, and Kavanaugh, JJ., joined, and in which Thomas and Gorsuch, JJ., joined as to all but Part III–C. Alito, J., filed a dissenting opinion. Notes 1 Together with No. 18–1028, Moda Health Plan, Inc. v. United States (see this Court’s Rule 12.4) and Blue Cross and Blue Shield of North Carolina v. United States (see this Court’s Rule 12.4); and No. 18–1038, Land of Lincoln Mutual Health Insurance Co. v. United States, also on certiorari to the same court.

purposes” (citing Greenlee County v. United States, 487 F.3d 871, 877 (CA Fed. 2007))). It also agrees with our analysis broadly, having held that “shall pay” language “generally makes a statute money-mandating” under the Tucker Act. Id., at 877 (internal quotation marks omitted). Conversely, the Court of Appeals has concluded that a statute is not money mandating where the Government enjoys “complete discretion” in determining whether (and whom) to pay. See, e.g., Doe v. United States, 463 F.3d 1314, 1324 (2006) (noting that the statutoryterm, “may,” creates a rebuttable presumption that the “statute creates discretion”). 15 The dissent concedes that there may “be some sharply defined categories of claims that may be properly asserted” through the TuckerAct “simply as a matter of precedent.” Post, at 6, and nn. 3, 4 (citing takings, breach-of-contract, failure-to-pay-compensation, and breach-of-fiduciary-duty claims as examples). Petitioners’ claim—breach of anunambiguous statutory promise to pay for services rendered to theGovernment—fits easily within those precedents. The only differences the dissent seems to assert here are that the dollar figure is higher and that petitioners do not deserve a “bailout” for their “bet” that the Federal Government would comply with federal law. Post, at 2, 3, 7; but cf., e.g., 79 Fed. Reg. 30260 (assuring insurers with “concerns that risk corridors collections may not be sufficient to fully fund risk corridors payments” that the Government would still pay). Our analysis in Tucker Act cases has never revolved on such results-oriented reasoning. 16 Having found that the Risk Corridors statute is a money-mandating provision for which a Tucker Act suit lies, we need not resolve petitioners’ alternative arguments for recovery based on an implied-in-fact contract theory or under the Takings Clause.

591.US.2019_18-9526

The Major Crimes Act (MCA) provides that, within “the Indian country,” “[a]ny Indian who commits” certain enumerated offenses “shall be subject to the same law and penalties as all other persons committing any of [those] offenses, within the exclusive jurisdiction of the United States.” 18 U. S. C. §1153(a). “Indian country” includes “all land within the limits of any Indian reservation under the jurisdiction of the United States Government.” §1151. Petitioner Jimcy McGirt was convicted by an Oklahoma state court of three serious sexual offenses. He unsuccessfully argued in state postconviction proceedings that the State lacked jurisdiction to prosecute him because he is an enrolled member of the Seminole Nation and his crimes took place on the Creek Reservation. He seeks a new trial, which, he contends, must take place in federal court. Held: For MCA purposes, land reserved for the Creek Nation since the 19th century remains “Indian country.” Pp. 3–42. (a) Congress established a reservation for the Creek Nation. An 1833 Treaty fixed borders for a “permanent home to the whole Creek Nation of Indians,” 7Stat. 418, and promised that the United States would “grant a patent, in fee simple, to the Creek nation of Indians for the [assigned] land” to continue “so long as they shall exist as a nation, and continue to occupy the country hereby assigned to them,” id., at 419. The patent formally issued in 1852. Though the early treaties did not refer to the Creek lands as a “reservation,” similar language in treaties from the same era has been held sufficient to create a reservation, see, e.g., Menominee Tribe v. United States, 391 U.S. 404, 405, and later Acts of Congress—referring to the “Creek reservation”—leave no room for doubt, see, e.g., 17Stat. 626. In addition, an 1856 Treaty promised that “no portion” of Creek lands “would ever be embraced or included within, or annexed to, any Territory or State,” 11Stat. 700, and that the Creeks would have the “unrestricted right of self-government,” with “full jurisdiction” over enrolled Tribe members and their property, id., at 704. Pp. 3–6. (b) Congress has since broken more than a few promises to the Tribe. Nevertheless, the Creek Reservation persists today. Pp. 6–28. (1) Once a federal reservation is established, only Congress can diminish or disestablish it. Doing so requires a clear expression of congressional intent. Pp. 6–8. (2) Oklahoma claims that Congress ended the Creek Reservation during the so-called “allotment era”—a period when Congress sought to pressure many tribes to abandon their communal lifestyles and parcel their lands into smaller lots owned by individual tribal members. Missing from the allotment-era agreement with the Creek, see 31Stat. 862–864, however, is any statute evincing anything like the “present and total surrender of all tribal interests” in the affected lands. And this Court has already rejected the argument that allotments automatically ended reservations. Pp. 8–13. (3) Oklahoma points to other ways Congress intruded on the Creeks’ promised right to self-governance during the allotment era, including abolishing the Creeks’ tribal courts, 30Stat. 504–505, and requiring Presidential approval for certain tribal ordinances, 31Stat. 872. But these laws fall short of eliminating all tribal interest in the contested lands. Pp. 13–17. (4) Oklahoma ultimately claims that historical practice and demographics are enough by themselves to prove disestablishment. This Court has consulted contemporaneous usages, customs, and practices to the extent they shed light on the meaning of ambiguous statutory terms, but Oklahoma points to no ambiguous language in any of the relevant statutes that could plausibly be read as an act of cession. Such extratextual considerations are of “ ‘limited interpretive value,’ ” Nebraska v. Parker, 577 U.S. 481, ___, and the “least compelling” form of evidence, South Dakota v. Yankton Sioux Tribe, 522 U.S. 329, 356. In the end, Oklahoma resorts to the State’s long historical practice of prosecuting Indians in state court for serious crimes on the contested lands, various statements made during the allotment era, and the speedy and persistent movement of white settlers into the area. But these supply little help with the law’s meaning and much potential for mischief. Pp. 17–28. (c) In the alternative, Oklahoma contends that Congress never established a reservation but instead created a “dependent Indian community.” To hold that the Creek never had a reservation would require willful blindness to the statutory language and a belief that the land patent the Creek received somehow made their tribal sovereignty easier to divest. Congress established a reservation, not a dependent Indian community, for the Creek Nation. Pp. 28–31. (d) Even assuming that the Creek land is a reservation, Oklahoma argues that the MCA has never applied in eastern Oklahoma. It claims that the Oklahoma Enabling Act, which transferred all non-federal cases pending in the territorial courts to Oklahoma’s state courts, made the State’s courts the successors to the federal territorial courts’ sweeping authority to try Indians for crimes committed on reservations. That argument, however, rests on state prosecutorial practices that defy the MCA, rather than on the law’s plain terms. Pp. 32–36. (e) Finally, Oklahoma warns of the potential consequences that will follow a ruling against it, such as unsettling an untold number of convictions and frustrating the State’s ability to prosecute crimes in the future. This Court is aware of the potential for cost and conflict around jurisdictional boundaries. But Oklahoma and its tribes have proven time and again that they can work successfully together as partners, and Congress remains free to supplement its statutory directions about the lands in question at any time. Pp. 36–42. Reversed. Gorsuch, J., delivered the opinion of the Court, in which Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined. Roberts, C. J., filed a dissenting opinion, in which Alito and Kavanaugh, JJ., joined, and in which Thomas, J., joined, except as to footnote 9. Thomas, J., filed a dissenting opinion.

On the far end of the Trail of Tears was a promise. Forced to leave their ancestral lands in Georgia and Alabama, the Creek Nation received assurances that their new lands in the West would be secure forever. In exchange for ceding “all their land, East of the Mississippi river,” the U. S. government agreed by treaty that “[t]he Creek country west of the Mississippi shall be solemnly guarantied to the Creek Indians.” Treaty With the Creeks, Arts. I, XIV, Mar. 24, 1832, 7Stat. 366, 368 (1832 Treaty). Both parties settled on boundary lines for a new and “permanent home to the whole Creek nation,” located in what is now Oklahoma. Treaty With the Creeks, preamble, Feb. 14, 1833, 7Stat. 418 (1833 Treaty). The government further promised that “[no] State or Territory [shall] ever have a right to pass laws for the government of such Indians, but they shall be allowed to govern themselves.” 1832 Treaty, Art. XIV, 7Stat. 368. Today we are asked whether the land these treaties promised remains an Indian reservation for purposes of federal criminal law. Because Congress has not said otherwise, we hold the government to its word. I At one level, the question before us concerns Jimcy McGirt. Years ago, an Oklahoma state court convicted him of three serious sexual offenses. Since then, he has argued in postconviction proceedings that the State lacked jurisdiction to prosecute him because he is an enrolled member of the Seminole Nation of Oklahoma and his crimes took place on the Creek Reservation. A new trial for his conduct, he has contended, must take place in federal court. The Oklahoma state courts hearing Mr. McGirt’s arguments rejected them, so he now brings them here. Mr. McGirt’s appeal rests on the federal Major Crimes Act (MCA). The statute provides that, within “the Indian country,” “[a]ny Indian who commits” certain enumerated offenses “against the person or property of another Indian or any other person” “shall be subject to the same law and penalties as all other persons committing any of the above offenses, within the exclusive jurisdiction of the United States.” 18 U. S. C. §1153(a). By subjecting Indians to federal trials for crimes committed on tribal lands, Congress may have breached its promises to tribes like the Creek that they would be free to govern themselves. But this particular incursion has its limits—applying only to certain enumerated crimes and allowing only the federal government to try Indians. State courts generally have no jurisdiction to try Indians for conduct committed in “Indian country.” Negonsott v. Samuels, 507 U.S. 99, 102–103 (1993). The key question Mr. McGirt faces concerns that last qualification: Did he commit his crimes in Indian country? A neighboring provision of the MCA defines the term to include, among other things, “all land within the limits of any Indian reservation under the jurisdiction of the United States Government, notwithstanding the issuance of any patent, and, including rights-of-way running through the reservation.” §1151(a). Mr. McGirt submits he can satisfy this condition because he committed his crimes on land reserved for the Creek since the 19th century. The Creek Nation has joined Mr. McGirt as amicus curiae. Not because the Tribe is interested in shielding Mr. McGirt from responsibility for his crimes. Instead, the Creek Nation participates because Mr. McGirt’s personal interests wind up implicating the Tribe’s. No one disputes that Mr. McGirt’s crimes were committed on lands described as the Creek Reservation in an 1866 treaty and federal statute. But, in seeking to defend the state-court judgment below, Oklahoma has put aside whatever procedural defenses it might have and asked us to confirm that the land once given to the Creeks is no longer a reservation today. At another level, then, Mr. McGirt’s case winds up as a contest between State and Tribe. The scope of their dispute is limited; nothing we might say today could unsettle Oklahoma’s authority to try non-Indians for crimes against non-Indians on the lands in question. See United States v. McBratney, 104 U.S. 621, 624 (1882). Still, the stakes are not insignificant. If Mr. McGirt and the Tribe are right, the State has no right to prosecute Indians for crimes committed in a portion of Northeastern Oklahoma that includes most of the city of Tulsa. Responsibility to try these matters would fall instead to the federal government and Tribe. Recently, the question has taken on more salience too. While Oklahoma state courts have rejected any suggestion that the lands in question remain a reservation, the Tenth Circuit has reached the opposite conclusion. Murphy v. Royal, 875 F.3d 896, 907–909, 966 (2017). We granted certiorari to settle the question. 589 U. S. ___ (2019). II Start with what should be obvious: Congress established a reservation for the Creeks. In a series of treaties, Congress not only “solemnly guarantied” the land but also “establish[ed] boundary lines which will secure a country and permanent home to the whole Creek Nation of Indians.” 1832 Treaty, Art. XIV, 7Stat. 368; 1833 Treaty, preamble, 7Stat. 418. The government’s promises weren’t made gratuitously. Rather, the 1832 Treaty acknowledged that “[t]he United States are desirous that the Creeks should remove to the country west of the Mississippi” and, in service of that goal, required the Creeks to cede all lands in the East. Arts. I, XII, 7Stat. 366, 367. Nor were the government’s promises meant to be delusory. Congress twice assured the Creeks that “[the] Treaty shall be obligatory on the contracting parties, as soon as the same shall be ratified by the United States.” 1832 Treaty, Art. XV, id., at 368; see 1833 Treaty, Art. IX, 7Stat. 420 (“agreement shall be binding and obligatory” upon ratification). Both treaties were duly ratified and enacted as law. Because the Tribe’s move west was ostensibly voluntary, Congress held out another assurance as well. In the statute that precipitated these negotiations, Congress authorized the President “to assure the tribe . . . that the United States will forever secure and guaranty to them . . . the country so exchanged with them.” Indian Removal Act of 1830, §3, 4Stat. 412. “[A]nd if they prefer it,” the bill continued, “the United States will cause a patent or grant to be made and executed to them for the same; Provided always, that such lands shall revert to the United States, if the Indians become extinct, or abandon the same.” Ibid. If agreeable to all sides, a tribe would not only enjoy the government’s solemn treaty promises; it would hold legal title to its lands. It was an offer the Creek accepted. The 1833 Treaty fixed borders for what was to be a “permanent home to the whole Creek nation of Indians.” 1833 Treaty, preamble, 7Stat. 418. It also established that the “United States will grant a patent, in fee simple, to the Creek nation of Indians for the land assigned said nation by this treaty.” Art. III, id., at 419. That grant came with the caveat that “the right thus guaranteed by the United States shall be continued to said tribe of Indians, so long as they shall exist as a nation, and continue to occupy the country hereby assigned to them.” Ibid. The promised patent formally issued in 1852. See Woodward v. De Graffenried, 238 U.S. 284, 293–294 (1915). These early treaties did not refer to the Creek lands as a “reservation”—perhaps because that word had not yet acquired such distinctive significance in federal Indian law. But we have found similar language in treaties from the same era sufficient to create a reservation. See Menominee Tribe v. United States, 391 U.S. 404, 405 (1968) (grant of land “ ‘for a home, to be held as Indian lands are held,’ ” established a reservation). And later Acts of Congress left no room for doubt. In 1866, the United States entered yet another treaty with the Creek Nation. This agreement reduced the size of the land set aside for the Creek, compensating the Tribe at a price of 30 cents an acre. Treaty Between the United States and the Creek Nation of Indians, Art. III, June 14, 1866, 14Stat. 786. But Congress explicitly restated its commitment that the remaining land would “be forever set apart as a home for said Creek Nation,” which it now referred to as “the reduced Creek reservation.” Arts. III, IX, id., at 786, 788.[1] Throughout the late 19th century, many other federal laws also expressly referred to the Creek Reservation. See, e.g., Treaty Between United States and Cherokee Nation of Indians, Art. IV, July 19, 1866, 14Stat. 800 (“Creek reservation”); Act of Mar. 3, 1873, ch. 322, 17Stat. 626; (multiple references to the “Creek reservation” and “Creek India[n] Reservation”); 11 Cong. Rec. 2351 (1881) (discussing “the dividing line between the Creek reservation and their ceded lands”); Act of Feb. 13, 1891, 26Stat. 750 (describing a cession by referencing the “West boundary line of the Creek Reservation”). There is a final set of assurances that bear mention, too. In the Treaty of 1856, Congress promised that “no portion” of the Creek Reservation “shall ever be embraced or included within, or annexed to, any Territory or State.” Art. IV, 11Stat. 700. And within their lands, with exceptions, the Creeks were to be “secured in the unrestricted right of self-government,” with “full jurisdiction” over enrolled Tribe members and their property. Art. XV, id., at 704. So the Creek were promised not only a “permanent home” that would be “forever set apart”; they were also assured a right to self-government on lands that would lie outside both the legal jurisdiction and geographic boundaries of any State. Under any definition, this was a reservation. III A While there can be no question that Congress established a reservation for the Creek Nation, it’s equally clear that Congress has since broken more than a few of its promises to the Tribe. Not least, the land described in the parties’ treaties, once undivided and held by the Tribe, is now fractured into pieces. While these pieces were initially distributed to Tribe members, many were sold and now belong to persons unaffiliated with the Nation. So in what sense, if any, can we say that the Creek Reservation persists today? To determine whether a tribe continues to hold a reservation, there is only one place we may look: the Acts of Congress. This Court long ago held that the Legislature wields significant constitutional authority when it comes to tribal relations, possessing even the authority to breach its own promises and treaties. Lone Wolf v. Hitchcock, 187 U.S. 553, 566–568 (1903). But that power, this Court has cautioned, belongs to Congress alone. Nor will this Court lightly infer such a breach once Congress has established a reservation. Solem v. Bartlett, 465 U.S. 463, 470 (1984). Under our Constitution, States have no authority to reduce federal reservations lying within their borders. Just imagine if they did. A State could encroach on the tribal boundaries or legal rights Congress provided, and, with enough time and patience, nullify the promises made in the name of the United States. That would be at odds with the Constitution, which entrusts Congress with the authority to regulate commerce with Native Americans, and directs that federal treaties and statutes are the “supreme Law of the Land.” Art. I, §8; Art. VI, cl. 2. It would also leave tribal rights in the hands of the very neighbors who might be least inclined to respect them. Likewise, courts have no proper role in the adjustment of reservation borders. Mustering the broad social consensus required to pass new legislation is a deliberately hard business under our Constitution. Faced with this daunting task, Congress sometimes might wish an inconvenient reservation would simply disappear. Short of that, legislators might seek to pass laws that tiptoe to the edge of disestablishment and hope that judges—facing no possibility of electoral consequences themselves—will deliver the final push. But wishes don’t make for laws, and saving the political branches the embarrassment of disestablishing a reservation is not one of our constitutionally assigned prerogatives. “[O]nly Congress can divest a reservation of its land and diminish its boundaries.” Solem, 465 U. S., at 470. So it’s no matter how many other promises to a tribe the federal government has already broken. If Congress wishes to break the promise of a reservation, it must say so. History shows that Congress knows how to withdraw a reservation when it can muster the will. Sometimes, legislation has provided an “[e]xplicit reference to cession” or an “unconditional commitment . . . to compensate the Indian tribe for its opened land.” Ibid. Other times, Congress has directed that tribal lands shall be “ ‘restored to the public domain.’ ” Hagen v. Utah, 510 U.S. 399, 412 (1994) (emphasis deleted). Likewise, Congress might speak of a reservation as being “ ‘discontinued,’ ” “ ‘abolished,’ ” or “ ‘vacated.’ ” Mattz v. Arnett, 412 U.S. 481, 504, n. 22 (1973). Disestablishment has “never required any particular form of words,” Hagen, 510 U. S., at 411. But it does require that Congress clearly express its intent to do so, “[c]ommon[ly with an] ‘[e]xplicit reference to cession or other language evidencing the present and total surrender of all tribal interests.’ ” Nebraska v. Parker, 577 U.S. 481, ___–___ (2016) (slip op., at 6). B In an effort to show Congress has done just that with the Creek Reservation, Oklahoma points to events during the so-called “allotment era.” Starting in the 1880s, Congress sought to pressure many tribes to abandon their communal lifestyles and parcel their lands into smaller lots owned by individual tribe members. See 1 F. Cohen, Handbook of Federal Indian Law §1.04 (2012) (Cohen), discussing General Allotment Act of 1887, ch. 119, 24Stat. 388. Some allotment advocates hoped that the policy would create a class of assimilated, landowning, agrarian Native Americans. See Cohen §1.04; F. Hoxie, A Final Promise: The Campaign To Assimilate 18–19 (2001). Others may have hoped that, with lands in individual hands and (eventually) freely alienable, white settlers would have more space of their own. See id., at 14–15; cf. General Allotment Act of 1887, §5, 24Stat. 389–390. The Creek were hardly exempt from the pressures of the allotment era. In 1893, Congress charged the Dawes Commission with negotiating changes to the Creek Reservation. Congress identified two goals: Either persuade the Creek to cede territory to the United States, as it had before, or agree to allot its lands to Tribe members. Act of Mar. 3, 1893, ch. 209, §16, 27Stat. 645–646. A year later, the Commission reported back that the Tribe “would not, under any circumstances, agree to cede any portion of their lands.” S. Misc. Doc. No. 24, 53d Cong., 3d Sess., 7 (1894). At that time, before this Court’s decision in Lone Wolf, Congress may not have been entirely sure of its power to terminate an established reservation unilaterally. Perhaps for that reason, perhaps for others, the Commission and Congress took this report seriously and turned their attention to allotment rather than cession.[2] The Commission’s work culminated in an allotment agreement with the Tribe in 1901. Creek Allotment Agreement, ch. 676, 31Stat. 861. With exceptions for certain pre-existing town sites and other special matters, the Agreement established procedures for allotting 160-acre parcels to individual Tribe members who could not sell, transfer, or otherwise encumber their allotments for a number of years. §§3, 7, id., at 862–864 (5 years for any portion, 21 years for the designated “homestead” portion). Tribe members were given deeds for their parcels that “convey[ed] to [them] all right, title, and interest of the Creek Nation.” §23, id., at 867–868. In 1908, Congress relaxed these alienation restrictions in some ways, and even allowed the Secretary of the Interior to waive them. Act of May 27, 1908, ch. 199, §1, 35Stat. 312. One way or the other, individual Tribe members were eventually free to sell their land to Indians and non-Indians alike. Missing in all this, however, is a statute evincing anything like the “present and total surrender of all tribal interests” in the affected lands. Without doubt, in 1832 the Creek “cede[d]” their original homelands east of the Mississippi for a reservation promised in what is now Oklahoma. 1832 Treaty, Art. I, 7Stat. 366. And in 1866, they “cede[d] and convey[ed]” a portion of that reservation to the United States. Treaty With the Creek, Art. III, 14Stat. 786. But because there exists no equivalent law terminating what remained, the Creek Reservation survived allotment. In saying this we say nothing new. For years, States have sought to suggest that allotments automatically ended reservations, and for years courts have rejected the argument. Remember, Congress has defined “Indian country” to include “all land within the limits of any Indian reservation . . . notwithstanding the issuance of any patent, and, including any rights-of-way running through the reservation.” 18 U. S. C. §1151(a). So the relevant statute expressly contemplates private land ownership within reservation boundaries. Nor under the statute’s terms does it matter whether these individual parcels have passed hands to non-Indians. To the contrary, this Court has explained repeatedly that Congress does not disestablish a reservation simply by allowing the transfer of individual plots, whether to Native Americans or others. See Mattz, 412 U. S., at 497 (“[A]llotment under the . . . Act is completely consistent with continued reservation status”); Seymour v. Superintendent of Wash. State Penitentiary, 368 U.S. 351, 356–358 (1962) (holding that allotment act “did no more than open the way for non-Indian settlers to own land on the reservation”); Parker, 577 U. S., at ___ (slip op., at 7) (“[T]he 1882 Act falls into another category of surplus land Acts: those that merely opened reservation land to settlement. . . . Such schemes allow non-Indian settlers to own land on the reservation” (internal quotation marks omitted)). It isn’t so hard to see why. The federal government issued its own land patents to many homesteaders throughout the West. These patents transferred legal title and are the basis for much of the private land ownership in a number of States today. But no one thinks any of this diminished the United States’s claim to sovereignty over any land. To accomplish that would require an act of cession, the transfer of a sovereign claim from one nation to another. 3 E. Washburn, American Law of Real Property *521–*524. And there is no reason why Congress cannot reserve land for tribes in much the same way, allowing them to continue to exercise governmental functions over land even if they no longer own it communally. Indeed, such an arrangement seems to be contemplated by §1151(a)’s plain terms. Cf. Seymour, 368 U. S., at 357–358.[3] Oklahoma reminds us that allotment was often the first step in a plan ultimately aimed at disestablishment. As this Court explained in Mattz, Congress’s expressed policy at the time “was to continue the reservation system and the trust status of Indian lands, but to allot tracts to individual Indians for agriculture and grazing.” 412 U. S., at 496. Then, “[w]hen all the lands had been allotted and the trust expired, the reservation could be abolished.” Ibid. This plan was set in motion nationally in the General Allotment Act of 1887, and for the Creek specifically in 1901. No doubt, this is why Congress at the turn of the 20th century “believed to a man” that “the reservation system would cease” “within a generation at most.” Solem, 465 U. S., at 468. Still, just as wishes are not laws, future plans aren’t either. Congress may have passed allotment laws to create the conditions for disestablishment. But to equate allotment with disestablishment would confuse the first step of a march with arrival at its destination.[4] Ignoring this distinction would run roughshod over many other statutes as well. In some cases, Congress chose not to wait for allotment to run its course before disestablishing a reservation. When it deemed that approach appropriate, Congress included additional language expressly ending reservation status. So, for example, in 1904, Congress allotted reservations belonging to the Ponca and Otoe Tribes, reservations also lying within modern-day Oklahoma, and then provided “further, That the reservation lines of the said . . . reservations . . . are hereby abolished.” Act of Apr. 21, 1904, §8, 33Stat. 217–218 (emphasis deleted); see also DeCoteau v. District County Court for Tenth Judicial Dist., 420 U.S. 425, 439–440, n. 22 (1975) (collecting other examples). Tellingly, however, nothing like that can be found in the nearly contemporary 1901 Creek Allotment Agreement or the 1908 Act. That doesn’t make these laws special. Rather, in using the language that they did, these allotment laws tracked others of the period, parceling out individual tracts, while saving the ultimate fate of the land’s reservation status for another day.[5] C If allotment by itself won’t work, Oklahoma seeks to prove disestablishment by pointing to other ways Congress intruded on the Creek’s promised right to self-governance during the allotment era. It turns out there were many. For example, just a few years before the 1901 Creek Allotment Agreement, and perhaps in an effort to pressure the Tribe to the negotiating table, Congress abolished the Creeks’ tribal courts and transferred all pending civil and criminal cases to the U. S. Courts of the Indian Territory. Curtis Act of 1898, §28, 30Stat. 504–505. Separately, the Creek Allotment Agreement provided that tribal ordinances “affecting the lands of the Tribe, or of individuals after allotment, or the moneys or other property of the Tribe, or of the citizens thereof ” would not be valid until approved by the President of the United States. §42, 31Stat. 872. Plainly, these laws represented serious blows to the Creek. But, just as plainly, they left the Tribe with significant sovereign functions over the lands in question. For example, the Creek Nation retained the power to collect taxes, operate schools, legislate through tribal ordinances, and, soon, oversee the federally mandated allotment process. §§39, 40, 42, id., at 871–872; Buster v. Wright, 135 F. 947, 949–950, 953–954 (CA8 1905). And, in its own way, the congressional incursion on tribal legislative processes only served to prove the power: Congress would have had no need to subject tribal legislation to Presidential review if the Tribe lacked any authority to legislate. Grave though they were, these congressional intrusions on pre-existing treaty rights fell short of eliminating all tribal interests in the land. Much more ominously, the 1901 allotment agreement ended by announcing that the Creek tribal government “shall not continue” past 1906, although the agreement quickly qualified that statement, adding the proviso “subject to such further legislation as Congress may deem proper.” §46, 31Stat. 872. Thus, while suggesting that the tribal government might end in 1906, Congress also necessarily understood it had not ended in 1901. All of which was consistent with the Legislature’s general practice of taking allotment as a first, not final, step toward disestablishment and dissolution. When 1906 finally arrived, Congress adopted the Five Civilized Tribes Act. But instead of dissolving the tribal government as some may have expected, Congress “deem[ed] proper” a different course, simply cutting away further at the Tribe’s autonomy. Congress empowered the President to remove and replace the principal chief of the Creek, prohibited the tribal council from meeting more than 30 days a year, and directed the Secretary of the Interior to assume control of tribal schools. §§6, 10, 28, 34Stat. 139–140, 148. The Act also provided for the handling of the Tribe’s funds, land, and legal liabilities in the event of dissolution. §§11, 27, id., at 141, 148. Despite these additional incursions on tribal authority, however, Congress expressly recognized the Creek’s “tribal existence and present tribal governmen[t]” and “continued [them] in full force and effect for all purposes authorized by law.” §28, id., at 148. In the years that followed, Congress continued to adjust its arrangements with the Tribe. For example, in 1908, the Legislature required Creek officials to turn over all “tribal properties” to the Secretary of the Interior. Act of May 27, 1908, §13, 35Stat. 316. The next year, Congress sought the Creek National Council’s release of certain money claims against the U. S. government. Act of Mar. 3, 1909, ch. 263, 35Stat. 781, 805. And, further still, Congress offered the Creek Nation a one-time opportunity to file suit in the federal Court of Claims for “any and all legal and equitable claims arising under or growing out of any treaty or agreement between the United States and the Creek Indian Nation.” Act of May 24, 1924, ch. 181, 43Stat. 139; see, e.g., United States v. Creek Nation, 295 U.S. 103 (1935). But Congress never withdrew its recognition of the tribal government, and none of its adjustments would have made any sense if Congress thought it had already completed that job. Indeed, with time, Congress changed course completely. Beginning in the 1920s, the federal outlook toward Native Americans shifted “away from assimilation policies and toward more tolerance and respect for traditional aspects of Indian culture.” 1 Cohen §1.05. Few in 1900 might have foreseen such a profound “reversal of attitude” was in the making or expected that “new protections for Indian rights,” including renewed “support for federally defined tribalism,” lurked around the corner. Ibid.; see also M. Scherer, Imperfect Victories: The Legal Tenacity of the Omaha Tribe, 1945–1995, pp. 2–4 (1999). But that is exactly what happened. Pursuant to this new national policy, in 1936, Congress authorized the Creek to adopt a constitution and bylaws, see Act of June 26, 1936, §3, 49Stat. 1967, enabling the Creek government to resume many of its previously suspended functions. Muscogee (Creek) Nation v. Hodel, 851 F.2d 1439, 1442–1447 (CADC 1988).[6] The Creek Nation has done exactly that. In the intervening years, it has ratified a new constitution and established three separate branches of government. Ibid.; see Muscogee Creek Nation (MCN) Const., Arts. V, VI, and VII. Today the Nation is led by a democratically elected Principal Chief, Second Chief, and National Council; operates a police force and three hospitals; commands an annual budget of more than $350 million; and employs over 2,000 people. Brief for Muscogee (Creek) Nation as Amicus Curiae 36–39. In 1982, the Nation passed an ordinance reestablishing the criminal and civil jurisdiction of its courts. See Hodel, 851 F. 2d, at 1442, 1446–1447 (confirming Tribe’s authority to do so). The territorial jurisdiction of these courts extends to any Indian country within the Tribe’s territory as defined by the Treaty of 1866. MCN Stat. 27, §1–102(A). And the State of Oklahoma has afforded full faith and credit to its judgments since at least 1994. See Barrett v. Barrett, 878 P.2d 1051, 1054 (Okla. 1994); Full Faith and Credit of Tribal Courts, Okla. State Cts. Network (Apr. 18, 2019), https://www.oscn.net/applications/oscn/DeliverDocument. asp?CiteID=458214. Maybe some of these changes happened for altruistic reasons, maybe some for other reasons. It seems, for example, that at least certain Members of Congress hesitated about disestablishment in 1906 because they feared any reversion of the Creek lands to the public domain would trigger a statutory commitment to hand over portions of these lands to already powerful railroad interests. See, e.g., 40 Cong. Rec. 2976 (1906) (Sen. McCumber); Id., at 3053 (Sen. Aldrich). Many of those who advanced the reorganization efforts of the 1930s may have done so more out of frustration with efforts to assimilate Native Americans than any disaffection with assimilation as the ultimate goal. See 1 Cohen §1.05; Scherer, Imperfect Victories, at 2–4. But whatever the confluence of reasons, in all this history there simply arrived no moment when any Act of Congress dissolved the Creek Tribe or disestablished its reservation. In the end, Congress moved in the opposite direction.[7] D Ultimately, Oklahoma is left to pursue a very different sort of argument. Now, the State points to historical practices and demographics, both around the time of and long after the enactment of all the relevant legislation. These facts, the State submits, are enough by themselves to prove disestablishment. Oklahoma even classifies and categorizes how we should approach the question of disestablishment into three “steps.” It reads Solem as requiring us to examine the laws passed by Congress at the first step, contemporary events at the second, and even later events and demographics at the third. On the State’s account, we have so far finished only the first step; two more await. This is mistaken. When interpreting Congress’s work in this arena, no less than any other, our charge is usually to ascertain and follow the original meaning of the law before us. New Prime Inc. v. Oliveira, 586 U. S. ___, ___ (2019) (slip op., at 6). That is the only “step” proper for a court of law. To be sure, if during the course of our work an ambiguous statutory term or phrase emerges, we will sometimes consult contemporaneous usages, customs, and practices to the extent they shed light on the meaning of the language in question at the time of enactment. Ibid. But Oklahoma does not point to any ambiguous language in any of the relevant statutes that could plausibly be read as an Act of disestablishment. Nor may a court favor contemporaneous or later practices instead of the laws Congress passed. As Solem explained, “[o]nce a block of land is set aside for an Indian reservation and no matter what happens to the title of individual plots within the area, the entire block retains its reservation status until Congress explicitly indicates otherwise.” 465 U. S., at 470 (citing United States v. Celestine, 215 U.S. 278, 285 (1909)). Still, Oklahoma reminds us that other language in Solem isn’t so constrained. In particular, the State highlights a passage suggesting that “[w]here non-Indian settlers flooded into the opened portion of a reservation and the area has long since lost its Indian character, we have acknowledged that de facto, if not de jure, diminishment may have occurred.” 465 U. S., at 471. While acknowledging that resort to subsequent demographics was “an unorthodox and potentially unreliable method of statutory interpretation,” the Court seemed nonetheless taken by its “obvious practical advantages.” Id., at 472, n. 13, 471. Out of context, statements like these might suggest historical practices or current demographics can suffice to disestablish or diminish reservations in the way Oklahoma envisions. But, in the end, Solem itself found these kinds of arguments provided “no help” in resolving the dispute before it. Id., at 478. Notably, too, Solem suggested that whatever utility historical practice or demographics might have was “demonstrated” by this Court’s earlier decision in Rosebud Sioux Tribe v. Kneip, 430 U.S. 584 (1977). See Solem, 465 U. S., at 470, n. 10. And Rosebud Sioux hardly endorsed the use of such sources to find disestablishment. Instead, based on the statute at issue there, the Court came “to the firm conclusion that congressional intent” was to diminish the reservation in question. 430 U. S., at 603. At that point, the Tribe sought to cast doubt on the clear import of the text by citing subsequent historical events—and the Court rejected the Tribe’s argument exactly because this kind of evidence could not overcome congressional intent as expressed in a statute. Id., at 604–605. This Court has already sought to clarify that extratextual considerations hardly supply the blank check Oklahoma supposes. In Parker, for example, we explained that “[e]vidence of the subsequent treatment of the disputed land . . . has ‘limited interpretive value.’ ” 577 U. S., at ___ (slip op., at 11) (quoting South Dakota v. Yankton Sioux Tribe, 522 U.S. 329, 355 (1998)).[8] Yankton Sioux called it the “least compelling” form of evidence. Id., at 356. Both cases emphasized that what value such evidence has can only be interpretative—evidence that, at best, might be used to the extent it sheds light on what the terms found in a statute meant at the time of the law’s adoption, not as an alternative means of proving disestablishment or diminishment. To avoid further confusion, we restate the point. There is no need to consult extratextual sources when the meaning of a statute’s terms is clear. Nor may extratextual sources overcome those terms. The only role such materials can properly play is to help “clear up . . . not create” ambiguity about a statute’s original meaning. Milner v. Department of Navy, 562 U.S. 562, 574 (2011). And, as we have said time and again, once a reservation is established, it retains that status “until Congress explicitly indicates otherwise.” Solem, 465 U. S., at 470 (citing Celestine, 215 U. S., at 285); see also Yankton Sioux, 522 U. S., at 343 (“[O]nly Congress can alter the terms of an Indian treaty by diminishing a reservation, and its intent to do so must be clear and plain”) (citation and internal quotation marks omitted). The dissent charges that we have failed to take account of the “compelling reasons” for considering extratextual evidence as a matter of course. Post, at 11–12. But Oklahoma and the dissent have cited no case in which this Court has found a reservation disestablished without first concluding that a statute required that result. Perhaps they wish this case to be the first. To follow Oklahoma and the dissent down that path, though, would only serve to allow States and courts to finish work Congress has left undone, usurp the legislative function in the process, and treat Native American claims of statutory right as less valuable than others. None of that can be reconciled with our normal interpretive rules, let alone our rule that disestablishment may not be lightly inferred and treaty rights are to be construed in favor, not against, tribal rights. Solem, 465 U. S., at 472.[9] To see the perils of substituting stories for statutes, we need look no further than the stories we are offered in the case before us. Put aside that the Tribe could tell more than a few stories of its own: Take just the evidence on which Oklahoma and the dissent wish to rest their case. First, they point to Oklahoma’s long historical prosecutorial practice of asserting jurisdiction over Indians in state court, even for serious crimes on the contested lands. If the Creek lands really were part of a reservation, the argument goes, all of these cases should have been tried in federal court pursuant to the MCA. Yet, until the Tenth Circuit’s Murphy decision a few years ago, no court embraced that possibility. See Murphy, 875 F.3d 896. Second, they offer statements from various sources to show that “everyone” in the late 19th and early 20th century thought the reservation system—and the Creek Nation—would be disbanded soon. Third, they stress that non-Indians swiftly moved on to the reservation in the early part of the last century, that Tribe members today constitute a small fraction of those now residing on the land, and that the area now includes a “vibrant city with expanding aerospace, healthcare, technology, manufacturing, and transportation sectors.” Brief for Petitioner in Carpenter v. Murphy, O. T. 2018, No. 17–1107, p. 15. All this history, we are told, supplies “compelling” evidence about the lands in question. Maybe so, but even taken on its own terms none of this evidence tells the story we are promised. Start with the State’s argument about its longstanding practice of asserting jurisdiction over Native Americans. Oklahoma proceeds on the implicit premise that its historical practices are unlikely to have defied the mandates of the federal MCA. That premise, though, appears more than a little shaky. In conjunction with the MCA, §1151(a) not only sends to federal court certain major crimes committed by Indians on reservations. Two doors down, in §1151(c), the statute does the same for major crimes committed by Indians on “Indian allotments, the Indian titles of which have not been extinguished.” Despite this direction, however, Oklahoma state courts erroneously entertained prosecutions for major crimes by Indians on Indian allotments for decades, until state courts finally disavowed the practice in 1989. See State v. Klindt, 782 P.2d 401, 404 (Okla. Crim. App. 1989) (overruling Ex parte Nowabbi, 60 Okla. Crim. III, 61 P.2d 1139 (1936)); see also United States v. Sands, 968 F.2d 1058, 1062–1063 (CA10 1992). And if the State’s prosecution practices disregarded §1151(c) for so long, it’s unclear why we should take those same practices as a reliable guide to the meaning and application of §1151(a). Things only get worse from there. Why did Oklahoma historically think it could try Native Americans for any crime committed on restricted allotments or anywhere else? Part of the explanation, Oklahoma tells us, is that it thought the eastern half of the State was always categorically exempt from the terms of the federal MCA. So whether a crime was committed on a restricted allotment, a reservation, or land that wasn’t Indian country at all, to Oklahoma it just didn’t matter. In the State’s view, when Congress adopted the Oklahoma Enabling Act that paved the way for its admission to the Union, it carved out a special exception to the MCA for the eastern half of the State where the Creek lands can be found. By Oklahoma’s own admission, then, for decades its historical practices in the area in question didn’t even try to conform to the MCA, all of which makes the State’s past prosecutions a meaningless guide for determining what counted as Indian country. As it turns out, too, Oklahoma’s claim to a special exemption was itself mistaken, yet one more error in historical practice that even the dissent does not attempt to defend. See Part V, infra.[10] To be fair, Oklahoma is far from the only State that has overstepped its authority in Indian country. Perhaps often in good faith, perhaps sometimes not, others made similar mistakes in the past. But all that only underscores further the danger of relying on state practices to determine the meaning of the federal MCA. See, e.g., Negonsett, 507 U. S., at 106–107 (“[I]n practice, Kansas had exercised jurisdiction over all offenses committed on Indian reservations involving Indians” (quoting memorandum from Secretary of the Interior, H. R. Rep. No. 1999, 76th Cong., 3d Sess., 4 (1940)); Scherer, Imperfect Victories, at 18 (describing “nationwide jurisdictional confusion” as a result of the MCA); Cohen §6.04(4)(a) (“Before 1942 the state of New York regularly exercised or claimed the right to exercise jurisdiction over the New York reservations, but a federal court decision in that year raised questions about the validity of state jurisdiction”); Brief for United States as Amicus Curiae in Carpenter v. Murphy, O. T. 2018, No. 17–1107, pp. 7a–8a (Letter from Secretary of the Interior, Mar. 27, 1963) (noting that many States have asserted criminal jurisdiction over Indians without an apparent basis in a federal law).[11] Oklahoma next points to various statements during the allotment era which, it says, show that even the Creek understood their reservation was under threat. And there’s no doubt about that. By 1893, the leadership of the Creek Nation saw what the federal government had in mind: “They [the federal government] do not deny any of our rights under treaty, but say they will go to the people themselves and confer with them and urge upon them the necessity of a change in their present condition, and upon their refusal will force a change upon them.” P. Porter & A. McKellop, Printed Statement of Creek Delegates, reprinted in Creek Delegation Documents 8–9 (Feb. 9, 1893). Not a decade later, and as a result of these forced changes, the leadership recognized that “ ‘[i]t would be difficult, if not impossible to successfully operate the Creek government now.’ ” App. to Brief for Respondent 8a (Message to Creek National Council (May 7, 1901), reprinted in The Indian Journal (May 10, 1901)). Surely, too, the future looked even bleaker: “ ‘The remnant of a government now accorded to us can be expected to be maintained only until all settlements of our landed and other interests growing out of treaty stipulations with the government of the United States shall have been settled.’ ” Ibid. But note the nature of these statements. The Creek Nation recognized that the federal government will seek to get popular support or otherwise would force change. Likewise, the Tribe’s government would continue for only so long. These were prophesies, and hardly groundbreaking ones at that. After all, the 1901 Creek Allotment Agreement explicitly said that the tribal government “shall not continue” past 1906. §46, 31Stat. 872. So what might statements like these tell us that isn’t already evident from the statutes themselves? Oklahoma doesn’t suggest they shed light on the meaning of some disputed and ambiguous statutory direction. More nearly, the State seeks to render the Creek’s fears self-fulfilling.[12] We are also asked to consider commentary from those outside the Tribe. In particular, the dissent reports that the federal government “operated” on the “understanding” that the reservation was disestablished. Post, at 32. In support of its claim, the dissent highlights a 1941 statement from Felix Cohen. Then serving as an official at the Interior Department, Cohen opined that “ ‘all offenses by or against Indians’ in the former Indian Territory ‘are subject to State laws.’ ” Ibid. (quoting App. to Supp. Reply Brief for Petitioner in Carpenter v. Murphy, O. T. 2018, No. 17–1107, p. 1a (Memorandum for Commissioner of Indian Affairs (July 11, 1941)). But that statement is incorrect. As we have just seen, Oklahoma’s courts acknowledge that the State lacks jurisdiction over Indian crimes on Indian allotments. See Klindt, 782 P. 2d, at 403–404. And the dissent does not dispute that Oklahoma is without authority under the MCA to try Indians for crimes committed on restricted allotments and any reservation. All of which highlights the pitfalls of elevating commentary over the law.[13] Finally, Oklahoma points to the speedy and persistent movement of white settlers onto Creek lands throughout the late 19th and early 20th centuries. But this history proves no more helpful in discerning statutory meaning. Maybe, as Oklahoma supposes, it suggests that some white settlers in good faith thought the Creek lands no longer constituted a reservation. But maybe, too, some didn’t care and others never paused to think about the question. Certain historians have argued, for example, that the loss of Creek land ownership was accelerated by the discovery of oil in the region during the period at issue here. A number of the federal officials charged with implementing the laws of Congress were apparently openly conflicted, holding shares or board positions in the very oil companies who sought to deprive Indians of their lands. A. Debo, And Still the Waters Run 86–87, 117–118 (1940). And for a time Oklahoma’s courts appear to have entertained sham competency and guardianship proceedings that divested Tribe members of oil rich allotments. Id., at 104–106, 233–234; Brief for Historians et al. as Amici Curiae 26–30. Whatever else might be said about the history and demographics placed before us, they hardly tell a story of unalloyed respect for tribal interests.[14] In the end, only one message rings true. Even the carefully selected history Oklahoma and the dissent recite is not nearly as tidy as they suggest. It supplies us with little help in discerning the law’s meaning and much potential for mischief. If anything, the persistent if unspoken message here seems to be that we should be taken by the “practical advantages” of ignoring the written law. How much easier it would be, after all, to let the State proceed as it has always assumed it might. But just imagine what it would mean to indulge that path. A State exercises jurisdiction over Native Americans with such persistence that the practice seems normal. Indian landowners lose their titles by fraud or otherwise in sufficient volume that no one remembers whose land it once was. All this continues for long enough that a reservation that was once beyond doubt becomes questionable, and then even farfetched. Sprinkle in a few predictions here, some contestable commentary there, and the job is done, a reservation is disestablished. None of these moves would be permitted in any other area of statutory interpretation, and there is no reason why they should be permitted here. That would be the rule of the strong, not the rule of law. IV Unable to show that Congress disestablished the Creek Reservation, Oklahoma next tries to turn the tables in a completely different way. Now, it contends, Congress never established a reservation in the first place. Over all the years, from the federal government’s first guarantees of land and self-government in 1832 and through the litany of promises that followed, the Tribe never received a reservation. Instead, what the Tribe has had all this time qualifies only as a “dependent Indian community.” Even if we were to accept Oklahoma’s bold feat of reclassification, however, it’s hardly clear the State would win this case. “Reservation[s]” and “Indian allotments, the Indian titles to which have not been extinguished,” qualify as Indian country under subsections (a) and (c) of §1151. But “dependent Indian communities” also qualify as Indian country under subsection (b). So Oklahoma lacks jurisdiction to prosecute Mr. McGirt whether the Creek lands happen to fall in one category or another. About this, Oklahoma is at least candid. It admits the entire point of its reclassification exercise is to avoid Solem’s rule that only Congress may disestablish a reservation. And to achieve that, the State has to persuade us not only that the Creek lands constitute a “dependent Indian community” rather than a reservation. It also has to convince us that we should announce a rule that dependent Indian community status can be lost more easily than reservation status, maybe even by the happenstance of shifting demographics. To answer this argument, it’s enough to address its first essential premise. Holding that the Creek never had a reservation would require us to stand willfully blind before a host of federal statutes. Perhaps that is why the Solicitor General, who supports Oklahoma’s disestablishment argument, refuses to endorse this alternative effort. It also may be why Oklahoma introduced this argument for affirmance only for the first time in this Court. And it may be why the dissent makes no attempt to defend Oklahoma here. What are we to make of the federal government’s repeated treaty promises that the land would be “solemnly guarantied to the Creek Indians,” that it would be a “permanent home,” “forever set apart,” in which the Creek would be “secured in the unrestricted right of self-government”? What about Congress’s repeated references to a “Creek reservation” in its statutes? No one doubts that this kind of language normally suffices to establish a federal reservation. So what could possibly make this case different? Oklahoma’s answer only gets more surprising. The reason that the Creek’s lands are not a reservation, we’re told, is that the Creek Nation originally held fee title. Recall that the Indian Removal Act authorized the President not only to “solemnly . . . assure the tribe . . . that the United States will forever secure and guaranty to them . . . the country so exchanged with them,” but also, “if they prefer it, . . . the United States will cause a patent or grant to be made and executed to them for the same.” 4Stat. 412. Recall that the Creek insisted on this additional protection when negotiating the Treaty of 1833, and in fact received a land patent pursuant to that treaty some 19 years later. In the eyes of Oklahoma, the Tribe’s choice on this score was a fateful one. By asking for (and receiving) fee title to their lands, the Creek inadvertently made their tribal sovereignty easier to divest rather than harder. The core of Oklahoma’s argument is that a reservation must be land “reserved from sale.” Celestine, 215 U. S., at 285. Often, that condition is satisfied when the federal government promises to hold aside a particular piece of federally owned land in trust for the benefit of the Tribe. And, admittedly, the Creek’s arrangement was different, because the Tribe held “fee simple title, not the usual Indian right of occupancy.” United States v. Creek Nation, 295 U.S. 103, 109 (1935). Still, as we explained in Part II, the land was reserved from sale in the very real sense that the government could not “give the tribal lands to others, or to appropriate them to its own purposes,” without engaging in “ ‘an act of confiscation.’ ” Id., at 110. It’s hard to see, too, how any difference between these two arrangements might work to the detriment of the Tribe. Just as we have never insisted on any particular form of words when it comes to disestablishing a reservation, we have never done so when it comes to establishing one. See Minnesota v. Hitchcock, 185 U.S. 373, 390 (1902) (“[I]n order to create a reservation it is not necessary that there should be a formal cession or a formal act setting apart a particular tract. It is enough that from what has been there results a certain defined tract appropriated to certain purposes”). As long as 120 years ago, the federal court for the Indian Territory recognized all this and rightly rejected the notion that fee title is somehow inherently incompatible with reservation status. Maxey v. Wright, 54 S.W. 807, 810 (Indian Terr. 1900). By now, Oklahoma’s next move will seem familiar. Seeking to sow doubt around express treaty promises, it cites some stray language from a statute that does not control here, a piece of congressional testimony there, and the scattered opinions of agency officials everywhere in between. See, e.g., Act of July 31, 1882, ch. 360, 22Stat. 179 (referring to Creek land as “Indian country” as opposed to an “Indian reservation”); S. Doc. No. 143, 59th Cong., 1st. Sess., 33 (1906) (Chief of Choctaw Nation—which had an arrangement similar to the Creek’s—testified that both Tribes “object to being classified with the reservation Indians”); Dept. of Interior, Census Office, Report on Indians Taxed and Indians Not Taxed in the U. S. 284 (1894) (Creeks and neighboring Tribes were “not on the ordinary Indian reservation, but on lands patented to them by the United States”). Oklahoma stresses that this Court even once called the Creek lands a “dependent Indian community,” though it used that phrase in passing and only to show that the Tribe’s “property and affairs were subject to the control and management of that government”—a point that would also be true if the lands were a reservation. Creek Nation, 295 U. S., at 109. Unsurprisingly given the Creek Nation’s nearly 200-year occupancy of these lands, both sides have turned up a few clues suggesting the label “reservation” either did or did not apply. One thing everyone can agree on is this history is long and messy. But the most authoritative evidence of the Creek’s relationship to the land lies not in these scattered references; it lies in the treaties and statutes that promised the land to the Tribe in the first place. And, if not for the Tribe’s fee title to its land, no one would question that these treaties and statutes created a reservation. So the State’s argument inescapably boils down to the untenable suggestion that, when the federal government agreed to offer more protection for tribal lands, it really provided less. All this time, fee title was nothing more than another trap for the wary. V That leaves Oklahoma to attempt yet another argument in the alternative. We alluded to it earlier in Part III. Now, the State accepts for argument’s sake that the Creek land is a reservation and thus “Indian country” for purposes of the Major Crimes Act. It accepts, too, that this would normally mean serious crimes by Indians on the Creek Reservation would have to be tried in federal court. But, the State tells us, none of that matters; everything the parties have briefed and argued so far is beside the point. It’s all irrelevant because it turns out the MCA just doesn’t apply to the eastern half of Oklahoma, and it never has. That federal law may apply to other States, even to the western half of Oklahoma itself. But eastern Oklahoma is and has always been exempt. So whether or not the Creek have a reservation, the State’s historic practices have always been correct and it remains free to try individuals like Mr. McGirt in its own courts. Notably, the dissent again declines to join Oklahoma in its latest twist. And, it turns out, for good reason. In support of its argument, Oklahoma points to statutory artifacts from its territorial history. The State of Oklahoma was formed from two territories: the Oklahoma Territory in the west and Indian Territory in the east. Originally, it seems criminal prosecutions in the Indian Territory were split between tribal and federal courts. See Act of May 2, 1890, §30, 26Stat. 94. But, in 1897, Congress abolished that scheme, granting the U. S. Courts of the Indian Territory “exclusive jurisdiction” to try “all criminal causes for the punishment of any offense.” Act of June 7, 1897, 30Stat. 83. These federal territorial courts applied federal law and state law borrowed from Arkansas “to all persons . . . irrespective of race.” Ibid. A year later, Congress abolished tribal courts and transferred all pending criminal cases to U. S. courts of the Indian Territory. Curtis Act of 1898, §28, 30Stat. 504–505. And, Oklahoma says, sending Indians to federal court and all others to state court would be inconsistent with this established and enlightened policy of applying the same law in the same courts to everyone. Here again, however, arguments along these and similar lines have been “frequently raised” but rarely “accepted.” United States v. Sands, 968 F.2d 1058, 1061 (CA10 1992) (Kelly, J.). “The policy of leaving Indians free from state jurisdiction and control is deeply rooted in this Nation’s history.” Rice v. Olson, 324 U.S. 786, 789 (1945). Chief Justice Marshall, for example, held that Indian Tribes were “distinct political communities, having territorial boundaries, within which their authority is exclusive . . . which is not only acknowledged, but guarantied by the United States,” a power dependent on and subject to no state authority. Worcester v. Georgia, 6 Pet. 515, 557 (1832); see also McClanahan v. Arizona Tax Comm’n, 411 U.S. 164, 168–169 (1973). And in many treaties, like those now before us, the federal government promised Indian Tribes the right to continue to govern themselves. For all these reasons, this Court has long “require[d] a clear expression of the intention of Congress” before the state or federal government may try Indians for conduct on their lands. Ex parte Crow Dog, 109 U.S. 556, 572 (1883). Oklahoma cannot come close to satisfying this standard. In fact, the only law that speaks expressly here speaks against the State. When Oklahoma won statehood in 1907, the MCA applied immediately according to its plain terms. That statute, as phrased at the time, provided exclusive federal jurisdiction over qualifying crimes by Indians in “any Indian reservation” located within “the boundaries of any State.” Act of Mar. 3, 1885, ch. 341, §9, 23Stat. 385 (emphasis added); see also 18 U. S. C. §1151 (defining “Indian country” even more broadly). By contrast, every one of the statutes the State directs us to merely discusses the assignment of cases among courts in the Indian Territory. They say nothing about the division of responsibilities between federal and state authorities after Oklahoma entered the Union. And however enlightened the State may think it was for territorial law to apply to all persons irrespective of race, some Tribe members may see things differently, given that the same policy entailed the forcible closure of tribal courts in defiance of treaty terms. Left to hunt for some statute that might have rendered the MCA inapplicable in Oklahoma after statehood, the best the State can find is the Oklahoma Enabling Act. Congress adopted that law in preparation for Oklahoma’s admission in 1907. Among its many provisions sorting out the details associated with Oklahoma’s transition to statehood, the Enabling Act transferred all nonfederal cases pending in territorial courts to Oklahoma’s new state courts. Act of June 16, 1906, §20, 34Stat. 277; see also Act of Mar. 4, 1907, §3, 34Stat. 1287 (clarifying treatment of cases to which United States was a party). The State says this transfer made its courts the inheritors of the federal territorial courts’ sweeping authority to try Indians for crimes committed on reservations. But, at best, this tells only half the story. The Enabling Act not only sent all nonfederal cases pending in territorial courts to state court. It also transferred pending cases that arose “under the Constitution, laws, or treaties of the United States” to federal district courts. §16, 34Stat. 277. Pending criminal cases were thus transferred to federal court if the prosecution would have belonged there had the Territory been a State at the time of the crime. §1, 34Stat. 1287 (amending the Enabling Act). Nor did the statute make any distinction between cases arising in the former eastern (Indian) and western (Oklahoma) territories. So, simply put, the Enabling Act sent state-law cases to state court and federal-law cases to federal court. And serious crimes by Indians in Indian country were matters that arose under the federal MCA and thus properly belonged in federal court from day one, wherever they arose within the new State. Maybe that’s right, Oklahoma acknowledges, but that’s not what happened. Instead, for many years the State continued to try Indians for crimes committed anywhere within its borders. But what can that tell us? The State identifies not a single ambiguous statutory term in the MCA that its actions might illuminate. And, as we have seen, its own courts have acknowledged that the State’s historic practices deviated in meaningful ways from the MCA’s terms. See supra, at 22–23. So, once more, it seems Oklahoma asks us to defer to its usual practices instead of federal law, something we will not and may never do. That takes Oklahoma down to its last straw when it comes to the MCA. If Oklahoma lacks the jurisdiction to try Native Americans it has historically claimed, that means at the time of its entry into the Union no one had the power to try minor Indian-on-Indian crimes committed in Indian country. This much follows, Oklahoma reminds us, because the MCA provides federal jurisdiction only for major crimes, and no tribal forum existed to try lesser cases after Congress abolished the tribal courts in 1898. Curtis Act, §28, 30Stat. 504–505. Whatever one thinks about the plausibility of other discontinuities between federal law and state practice, the State says, it is unthinkable that Congress would have allowed such a significant “jurisdictional gap” to open at the moment Oklahoma achieved statehood. But what the State considers unthinkable turns out to be easily imagined. Jurisdictional gaps are hardly foreign to this area of the law. See, e.g., Duro v. Reina, 495 U.S. 676, 704–706 (1990) (Brennan, J., dissenting). Many tribal courts across the country were absent or ineffective during the early part of the last century, yielding just the sort of gaps Oklahoma would have us believe impossible. Indeed, this might be why so many States joined Oklahoma in prosecuting Indians without proper jurisdiction. The judicial mind abhors a vacuum, and the temptation for state prosecutors to step into the void was surely strong. See supra, at 23–24. With time, too, Congress has filled many of the gaps Oklahoma worries about. One way Congress has done so is by reauthorizing tribal courts to hear minor crimes in Indian country. Congress chose exactly this course for the Creeks and others in 1936. Act of June 26, 1936, §3, 49Stat. 1967; see also Hodel, 851 F. 2d, at 1442–1446. Another option Congress has employed is to allow affected Indian tribes to consent to state criminal jurisdiction. 25 U. S. C. §§1321(a), 1326. Finally, Congress has sometimes expressly expanded state criminal jurisdiction in targeted bills addressing specific States. See, e.g., 18 U. S. C. §3243 (creating jurisdiction for Kansas); Act of May 31, 1946, ch. 279, 60Stat. 229 (same for a reservation in North Dakota); Act of June 30, 1948, ch. 759, 62Stat. 1161 (same for certain reservations in Iowa); 18 U. S. C. §1162 (creating jurisdiction for six additional States). But Oklahoma doesn’t claim to have complied with the requirements to assume jurisdiction voluntarily over Creek lands. Nor has Congress ever passed a law conferring jurisdiction on Oklahoma. As a result, the MCA applies to Oklahoma according to its usual terms: Only the federal government, not the State, may prosecute Indians for major crimes committed in Indian country. VI In the end, Oklahoma abandons any pretense of law and speaks openly about the potentially “transform[ative]” effects of a loss today. Brief for Respondent 43. Here, at least, the State is finally rejoined by the dissent. If we dared to recognize that the Creek Reservation was never disestablished, Oklahoma and dissent warn, our holding might be used by other tribes to vindicate similar treaty promises. Ultimately, Oklahoma fears that perhaps as much as half its land and roughly 1.8 million of its residents could wind up within Indian country. It’s hard to know what to make of this self-defeating argument. Each tribe’s treaties must be considered on their own terms, and the only question before us concerns the Creek. Of course, the Creek Reservation alone is hardly insignificant, taking in most of Tulsa and certain neighboring communities in Northeastern Oklahoma. But neither is it unheard of for significant non-Indian populations to live successfully in or near reservations today. See, e.g., Brief for National Congress of American Indians Fund as Amicus Curiae 26–28 (describing success of Tacoma, Washington, and Mount Pleasant, Michigan); see also Parker, 577 U. S., at ___–___ (slip op., at 10–12) (holding Pender, Nebraska, to be within Indian country despite tribe’s absence from the disputed territory for more than 120 years). Oklahoma replies that its situation is different because the affected population here is large and many of its residents will be surprised to find out they have been living in Indian country this whole time. But we imagine some members of the 1832 Creek Tribe would be just as surprised to find them there. What are the consequences the State and dissent worry might follow from an adverse ruling anyway? Primarily, they argue that recognizing the continued existence of the Creek Reservation could unsettle an untold number of convictions and frustrate the State’s ability to prosecute crimes in the future. But the MCA applies only to certain crimes committed in Indian country by Indian defendants. A neighboring statute provides that federal law applies to a broader range of crimes by or against Indians in Indian country. See 18 U. S. C. §1152. States are otherwise free to apply their criminal laws in cases of non-Indian victims and defendants, including within Indian country. See McBratney, 104 U. S., at 624. And Oklahoma tells us that somewhere between 10% and 15% of its citizens identify as Native American. Given all this, even Oklahoma admits that the vast majority of its prosecutions will be unaffected whatever we decide today. Still, Oklahoma and the dissent fear, “[t]housands” of Native Americans like Mr. McGirt “wait in the wings” to challenge the jurisdictional basis of their state-court convictions. Brief for Respondent 3. But this number is admittedly speculative, because many defendants may choose to finish their state sentences rather than risk reprosecution in federal court where sentences can be graver. Other defendants who do try to challenge their state convictions may face significant procedural obstacles, thanks to well-known state and federal limitations on postconviction review in criminal proceedings.[15] In any event, the magnitude of a legal wrong is no reason to perpetuate it. When Congress adopted the MCA, it broke many treaty promises that had once allowed tribes like the Creek to try their own members. But, in return, Congress allowed only the federal government, not the States, to try tribal members for major crimes. All our decision today does is vindicate that replacement promise. And if the threat of unsettling convictions cannot save a precedent of this Court, see Ramos v. Louisiana, 590 U. S. ___, ___–___ (2020) (plurality opinion) (slip op., at 23–26), it certainly cannot force us to ignore a statutory promise when no precedent stands before us at all. What’s more, a decision for either party today risks upsetting some convictions. Accepting the State’s argument that the MCA never applied in Oklahoma would preserve the state-court convictions of people like Mr. McGirt, but simultaneously call into question every federal conviction obtained for crimes committed on trust lands and restricted Indian allotments since Oklahoma recognized its jurisdictional error more than 30 years ago. See supra, at 22. It’s a consequence of their own arguments that Oklahoma and the dissent choose to ignore, but one which cannot help but illustrate the difficulty of trying to guess how a ruling one way or the other might affect past cases rather than simply proceeding to apply the law as written. Looking to the future, Oklahoma warns of the burdens federal and tribal courts will experience with a wider jurisdiction and increased caseload. But, again, for every jurisdictional reaction there seems to be an opposite reaction: recognizing that cases like Mr. McGirt’s belong in federal court simultaneously takes them out of state court. So while the federal prosecutors might be initially understaffed and Oklahoma prosecutors initially overstaffed, it doesn’t take a lot of imagination to see how things could work out in the end. Finally, the State worries that our decision will have significant consequences for civil and regulatory law. The only question before us, however, concerns the statutory definition of “Indian country” as it applies in federal criminal law under the MCA, and often nothing requires other civil statutes or regulations to rely on definitions found in the criminal law. Of course, many federal civil laws and regulations do currently borrow from §1151 when defining the scope of Indian country. But it is far from obvious why this collateral drafting choice should be allowed to skew our interpretation of the MCA, or deny its promised benefits of a federal criminal forum to tribal members. It isn’t even clear what the real upshot of this borrowing into civil law may be. Oklahoma reports that recognizing the existence of the Creek Reservation for purposes of the MCA might potentially trigger a variety of federal civil statutes and rules, including ones making the region eligible for assistance with homeland security, 6 U. S. C. §§601, 606, historical preservation, 54 U. S. C. §302704, schools, 20 U. S. C. §1443, highways, 23 U. S. C. §120, roads, §202, primary care clinics, 25 U. S. C. §1616e–1, housing assistance, §4131, nutritional programs, 7 U. S. C. §§2012, 2013, disability programs, 20 U. S. C. §1411, and more. But what are we to make of this? Some may find developments like these unwelcome, but from what we are told others may celebrate them. The dissent isn’t so sanguine—it assures us, without further elaboration, that the consequences will be “drastic precisely because they depart from . . . more than a century [of] settled understanding.” Post, at 37. The prediction is a familiar one. Thirty years ago the Solicitor General warned that “[l]aw enforcement would be rendered very difficult” and there would be “grave uncertainty regarding the application” of state law if courts departed from decades of “long-held understanding” and recognized that the federal MCA applies to restricted allotments in Oklahoma. Brief for United States as Amicus Curiae in Oklahoma v. Brooks, O.T. 1988, No. 88–1147, pp. 2, 9, 18, 19. Yet, during the intervening decades none of these predictions panned out, and that fact stands as a note of caution against too readily crediting identical warnings today. More importantly, dire warnings are just that, and not a license for us to disregard the law. By suggesting that our interpretation of Acts of Congress adopted a century ago should be inflected based on the costs of enforcing them today, the dissent tips its hand. Yet again, the point of looking at subsequent developments seems not to be determining the meaning of the laws Congress wrote in 1901 or 1906, but emphasizing the costs of taking them at their word. Still, we do not disregard the dissent’s concern for reliance interests. It only seems to us that the concern is misplaced. Many other legal doctrines—procedural bars, res judicata, statutes of repose, and laches, to name a few—are designed to protect those who have reasonably labored under a mistaken understanding of the law. And it is precisely because those doctrines exist that we are “fre[e] to say what we know to be true . . . today, while leaving questions about . . . reliance interest[s] for later proceedings crafted to account for them.” Ramos, 590 U. S., at ___ (plurality opinion) (slip op., at 24). In reaching our conclusion about what the law demands of us today, we do not pretend to foretell the future and we proceed well aware of the potential for cost and conflict around jurisdictional boundaries, especially ones that have gone unappreciated for so long. But it is unclear why pessimism should rule the day. With the passage of time, Oklahoma and its Tribes have proven they can work successfully together as partners. Already, the State has negotiated hundreds of intergovernmental agreements with tribes, including many with the Creek. See Okla. Stat., Tit. 74, §1221 (2019 Cum. Supp.); Oklahoma Secretary of State, Tribal Compacts and Agreements, www.sos.ok.gov/tribal.aspx. These agreements relate to taxation, law enforcement, vehicle registration, hunting and fishing, and countless other fine regulatory questions. See Brief for Tom Cole et al. as Amici Curiae 13–19. No one before us claims that the spirit of good faith, “comity and cooperative sovereignty” behind these agreements, id., at 20, will be imperiled by an adverse decision for the State today any more than it might be by a favorable one.[16] And, of course, should agreement prove elusive, Congress remains free to supplement its statutory directions about the lands in question at any time. It has no shortage of tools at its disposal. * The federal government promised the Creek a reservation in perpetuity. Over time, Congress has diminished that reservation. It has sometimes restricted and other times expanded the Tribe’s authority. But Congress has never withdrawn the promised reservation. As a result, many of the arguments before us today follow a sadly familiar pattern. Yes, promises were made, but the price of keeping them has become too great, so now we should just cast a blind eye. We reject that thinking. If Congress wishes to withdraw its promises, it must say so. Unlawful acts, performed long enough and with sufficient vigor, are never enough to amend the law. To hold otherwise would be to elevate the most brazen and longstanding injustices over the law, both rewarding wrong and failing those in the right. The judgment of the Court of Criminal Appeals of Oklahoma is Reversed. Notes 1 The dissent by The Chief Justice (hereinafter the dissent) suggests that the Creek’s intervening alliance with the Confederacy “ ‘unsettled’ ” and “ ‘forfeit[ed]’ ” the longstanding promises of the United States. Post, at 3. But the Treaty of 1866 put an end to any Civil War hostility, promising mutual amnesty, “perpetual peace and friendship,” and guaranteeing the Tribe the “quiet possession of their country.” Art. I, 14Stat. 786. Though this treaty expressly reduced the size of the Creek Reservation, the Creek were compensated for the lost territory, and otherwise “retained” their unceded portion. Art. III, ibid. Contrary to the dissent’s implication, nothing in the Treaty of 1866 purported to repeal prior treaty promises. Cf. Art. XII, id., at 790 (the United States expressly “reaffirms and reassumes all obligations of treaty stipulations with the Creek nation entered into before” the Civil War). 2 The dissent stresses, repeatedly, that the Dawes Commission was charged with seeking to extinguish the reservation. Post, at 18, 24. Yet, the dissent fails to mention the Commission’s various reports acknowledging that those efforts were unsuccessful precisely because the Creek refused to cede their lands. 3 The dissent not only fails to acknowledge these features of the statute and our precedents. It proceeds in defiance of them, suggesting that by moving to eliminate communal title and relaxing restrictions on alienation, “Congress destroyed the foundation of [the Creek Nation’s] sovereignty.” Post, at 18–19. But this Court long ago rejected the notion that the purchase of lands by non-Indians is inconsistent with reservation status. See Seymour, 368 U. S., at 357–358. 4 The dissent seemingly conflates these steps in other ways, too, by implying that the passage of an allotment Act itself extinguished title. Post, at 18–19. The reality proved more complicated. Allotment of the Creek lands did not occur overnight, but dragged on for years, well past Oklahoma’s statehood, until Congress finally prohibited any further allotments more than 15 years later. Act of Mar. 2, 1917, 39Stat. 986. 5 The dissent doesn’t purport to find any of the hallmarks of diminishment in the Creek Allotment Agreement. Instead, the dissent tries to excuse their absence by saying that it would have made “little sense” to find such language in an Act transferring the Tribe’s lands to private owners. Post, at 14. But the dissent’s account is impossible to reconcile with history and precedent. As we have noted, plenty of allotment agreements during this era included precisely the language of cession and compensation that the dissent says it would make “little sense” to find there. And this Court has confirmed time and again that allotment agreements without such language do not necessarily disestablish or diminish the reservation at issue. See Mattz v. Arnett, 412 U.S. 481, 497 (1973); Seymour v. Superintendent of Wash. State Penitentiary, 368 U.S. 351, 358 (1962). The dissent’s only answer is to suggest that allotment combined with other statutes limiting the Creek Nation’s governing authority amounted to disestablishment—in other words that it’s the arguments in the next section that really do the work. 6 The dissent calls it “fantasy” to suggest that Congress evinced “any unease about extinguishing the Creek domain” because Congress “did what it set out to do: transform a reservation into a State.” Post, at 22–23. The dissent stresses, too, that the Creek were afforded U. S. citizenship and the right to vote. Post, at 20. But the only thing implausible here is the suggestion that “creat[ing] a new State” or enfranchising Native Americans implies an “intent to terminate” any and all reservations within a State’s boundaries. Post, at 15. This Court confronted—and rejected—that sort of argument long ago in United States v. Sandoval, 231 U.S. 28, 47–48 (1913). The dissent treats that case as a one-off: special because “the tribe in Sandoval, the Pueblo Indians of New Mexico, retained a rare communal title to their lands.” Post, at 21, n. 4. But Sandoval is not only a case about the Pueblos; it is a foundational precedent recognizing that Congress can welcome Native Americans to participate in a broader political community without sacrificing their tribal sovereignty. 7 The dissent ultimately concedes what Oklahoma will not: that no “individual congressional action or piece of evidence, standing alone, disestablished the Creek reservation.” Post, at 9–10. Instead we’re told we must consider “all of the relevant Acts of Congress together, viewed in light of contemporaneous and subsequent contextual evidence.” Ibid. So, once again, the dissent seems to suggest that it’s the arguments in the next section that will get us across the line to disestablishment. 8 The dissent suggests Parker meant to say only that evidence of subsequent treatment had limited interpretative value “in that case.” Post, at 12. But the dissent includes just a snippet of the relevant passage. Read in full, there is little room to doubt Parker invoked a general rule: “This subsequent demographic history cannot overcome our conclusion that Congress did not intend to diminish the reservation in 1882. And it is not our rule to ‘rewrite’ the 1882 Act in light of this subsequent demographic history. DeCoteau, 420 U. S., at 447. After all, evidence of the changing demographics of disputed land is ‘the least compelling’ evidence in our diminishment analysis, for ‘[e]very surplus land Act necessarily resulted in a surge of non-Indian settlement and degraded the “Indian character” of the reservation, yet we have repeatedly stated that not every surplus land Act diminished the affected reservation.’ Yankton Sioux, 522 U. S., at 356. . . . Evidence of the subsequent treatment of the disputed land by Government officials likewise has ‘limited interpretive value.’ Id., at 355.” 577 U. S., at ___ (slip op., at 11). 9 In an effort to support its very different course, the dissent stitches together quotes from Rosebud Sioux Tribe v. Knelp, 430 U.S. 584 (1977), and South Dakota v. Yankton Sioux Tribe, 522 U.S. 329 (1998). Post, at 10–11. But far from supporting the dissent, both cases emphasize that “[t]he focus of our inquiry is congressional intent,” Rosebud, 430 U. S., at 588, n. 4; see also Yankton Sioux, 522 U. S., at 343, and merely acknowledge that extratextual sources may help resolve ambiguity about Congress’s directions. The dissent’s appeal to Solem fares no better. As we have seen, the extratextual sources in Solem only confirmed what the relevant statute already suggested—that the reservation in question was not diminished or disestablished. 465 U. S., at 475–476. 10 The dissent tries to avoid this inconvenient history by distinguishing fee allotments from reservations, noting that the two categories are legally distinct and geographically incommensurate. Post, at 27. But this misses the point: The reason that Oklahoma thought it could prosecute Indians for crimes on restricted allotments applied with equal force to reservations. And it hardly “stretches the imagination” to think that reason was wrong, post, at 28, when the dissent itself does not dispute our rejection of it in Part V. 11 Unable to answer Oklahoma’s admitted error about the very federal criminal statute before us, the dissent travels far afield, pointing to the fact an Oklahoma court heard a civil case in 1915 about an inheritance—involving members of a different Tribe—as “evidence” Congress disestablished the Creek Reservation. See post, at 21 (citing Palmer v. Cully, 52 Okla. 454, 455–465, 153 P. 154, 155–157 (1915) (per curiam)). But even assuming that Oklahoma courts exercised civil jurisdiction over Creek members, too, the dissent never explains why this jurisdiction implies the Creek Reservation must have been disestablished. After all, everyone agrees that the Creeks were prohibited from having their own courts at the time. So it should be no surprise that some Creek might have resorted to state courts in hope of resolving their disputes. 12 The dissent finds the statements of the Creek leadership so proba-tive that it cites them not just as evidence about the meaning of treaties the Tribe signed but even as evidence about the meaning of general purpose laws the Creek had no hand in. See post, at 26 (citing Chief Porter’sviews on the legal effects of the Oklahoma Enabling Act). That is quite a stretch from using tribal statements as “historical evidence of ‘the manner in which [treaties were] negotiated’ with the . . . Tribe.” Parker, 577 U. S., at ___ (slip op., at 9) (quoting Solem v. Bartlett, 465 U.S. 463, 471 (1984)). 13 Part of the reason for Cohen’s error might be explained by a portion of the memorandum the dissent leaves unquoted. Cohen concluded that Oklahoma was free to try Indians anywhere in the State because, among other things, the Oklahoma Enabling Act “transfer[red] . . . jurisdiction from the Federal courts to the State courts upon the establishment of the State of Oklahoma.” App. to Supp. Reply Brief for Petitioner in Carpenter v. Murphy, O. T. 2018, No. 17–1107, p. 1a (Memorandum for Commissioner of Indian Affairs (July 11, 1941)). Yet, as we explore below, the Oklahoma Enabling Act did not send cases covered by the federal MCA to state court. See Part V, infra. Other, contemporaneous Interior Department memoranda acknowledged that Oklahoma state courts had simply “assumed jurisdiction” over cases arising on restricted allotments without any clear authority in the Oklahoma Enabling Act or the MCA, and much the same appears to have occurred here. App. to Supp. Reply Brief for Respondent in Carpenter v. Murphy, O. T. 2018, No. 17–1107, p. 1a (Memorandum from N. Gray, Dept. of Interior, for Mr. Flanery (Aug. 12, 1942)). So rather than Oklahoma and the United States having a “shared understanding” that Congress had disestablished the Creek Reservation, post, at 27, it seems more accurate to say that for many years much uncertainty remained about whether the MCA applied in eastern Oklahoma. 14 The dissent asks us to examine a hodge-podge of other, but no more compelling, material. For example, the dissent points to later statutes that do no more than confirm there are former reservations in the State of Oklahoma. Post, at 30–31. It cites legislative history to show that Congress had the Creek Nation—or, at least, its neighbors—in mind when it added these in 1988. Post, at 31, n. 7. The dissent cites a Senate Report from 1989 and post-1980 statements made by representatives of other tribes. Post, at 30, 32–33. It highlights three occasions on which this Court referred to something like a “former Creek Nation,” though it neglects to add that in each the Court was referring to the loss of the Nation’s communal fee title, not its sovereignty. Grayson v. Harris, 267 U.S. 352, 357 (1925); Woodward v. DeGraffenreid, 238 U.S. 284, 289–290 (1915); Washington v. Miller, 235 U.S. 422, 423–425 (1914). The dissent points as well to a single instance in which the Creek Nation disclaimed reservation boundaries for purposes of litigation in a lower court, post, at 32, but ignores that the Creek Nation has repeatedly filed briefs in this Court to the contrary. This is thin gruel to set against treaty promises enshrined in statutes. 15 For example, Oklahoma appears to apply a general rule that “issues that were not raised previously on direct appeal, but which could have been raised, are waived for further review.” Logan v. State, 2013 OK CR 2, ¶ 1, 293 P.3d 969, 973. Indeed, Justice Thomas contends that this state-law limitation on collateral review prevents us from considering even the case now before us. Post, at 2 (dissenting opinion). But while that state-law rule may often bar our way, it doesn’t in this case. After noting a potential state-law obstacle, the Oklahoma Court of Criminal Appeals (OCCA) proceeded to address the merits of Mr. McGirt’s federal MCA claim anyway. Because the OCCA’s opinion “fairly appears to rest primarily on federal law or to be interwoven with federal law” and lacks any “plain statement” that it was relying on a state-law ground, we have jurisdiction to consider the federal-law question presented to us. See Michigan v. Long, 463 U.S. 1032, 1040–1041, 1044 (1983). 16 This sense of cooperation and a shared future is on display in this very case. The Creek Nation is supported by an array of leaders of other Tribes and the State of Oklahoma, many of whom had a role in negotiating exactly these agreements. See Brief for Tom Cole et al. as Amici Curiae 1 (“Amici are a former Governor, State Attorney General, cabinet members, and legislators of the State of Oklahoma, and two federally recognized Indian tribes, the Chickasaw Nation and Choctaw Nation of Oklahoma”) (brief authored by Robert H. Henry, also a former State Attorney General and Chief Judge of the Tenth Circuit).

589.US.2019_18-1109

An Arizona jury convicted petitioner James McKinney of two counts of first-degree murder. The trial judge found aggravating circumstances for both murders, weighed the aggravating and mitigating circumstances, and sentenced McKinney to death. Nearly 20 years later, the Ninth Circuit held on habeas review that the Arizona courts violated Eddings v. Oklahoma, 455 U.S. 104, by failing to properly consider as relevant mitigating evidence McKinney’s posttraumatic stress disorder. McKinney’s case then returned to the Arizona Supreme Court. McKinney argued that he was entitled to a jury resentencing, but the Arizona Supreme Court itself reweighed the aggravating and mitigating circumstances, as permitted by Clemons v. Mississippi, 494 U.S. 738, and upheld both death sentences. Held: A Clemons reweighing is a permissible remedy for an Eddings error, and when an Eddings error is found on collateral review, a state appellate court may conduct a Clemons reweighing on collateral review. McKinney’s argument that a jury must resentence him does not square with Clemons, where the Court held that a reweighing of the aggravating and mitigating evidence may be conducted by an appellate court. 494 U. S., at 741. Because Clemons involved an improperly considered aggravating circumstance, McKinney maintains that it is inapposite here, where the case involves an improperly ignored mitigating circumstance. Clemons, however, did not depend on any unique effect of aggravators as distinct from mitigators. For purposes of appellate reweighing, there is no meaningful difference between subtracting an aggravator from one side of the scale and adding a mitigator to the other side. McKinney also argues that Clemons is no longer good law in the wake of Ring v. Arizona, 536 U.S. 584, and Hurst v. Florida, 577 U. S. ___, where the Court held that a jury must find the aggravating circumstance that makes the defendant death eligible. But that does not mean that a jury is constitutionally required to weigh the aggravating and mitigating circumstances or to make the ultimate sentencing decision within the relevant sentencing range. See Apprendi v. New Jersey, 530 U.S. 466, 481. McKinney notes that the Arizona trial court, not the jury, made the initial aggravating circumstance finding that made him eligible for the death penalty. But McKinney’s case became final on direct review long before Ring and Hurst, which do not apply retroactively on collateral review, see Schriro v. Summerlin, 542 U.S. 348, 358, and the Arizona Supreme Court’s 2018 decision reweighing the aggravators and mitigators did not constitute a reopening of direct review. Pp. 2–7. 245 Ariz. 225, 426 P.3d 1204, affirmed. Kavanaugh, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Alito, and Gorsuch, JJ., joined. Ginsburg, J., filed a dissenting opinion, in which Breyer, Sotomayor, and Kagan, JJ., joined.

Over a 4-week span in early 1991, James McKinney and his half brother, Charles Hedlund, burglarized five residences in the Phoenix, Arizona, area. During one of the burglaries, McKinney and Hedlund beat and repeatedly stabbed Christine Mertens. McKinney then shot Mertens in the back of the head, fatally wounding her. In another burglary, McKinney and Hedlund killed Jim McClain by shooting him in the back of the head with a sawed-off rifle. In 1992, an Arizona jury convicted McKinney of two counts of first-degree murder. Under this Court’s precedents, a defendant convicted of murder is eligible for a death sentence if at least one aggravating circumstance is found. See Tuilaepa v. California, 512 U.S. 967 (1994); Zant v. Stephens, 462 U.S. 862 (1983); Gregg v. Georgia, 428 U.S. 153 (1976). McKinney’s trial judge found aggravating circumstances for both murders. For the Mertens murder, the judge found that McKinney committed the murder for pecuniary gain and that McKinney killed Mertens in an especially heinous, cruel, or depraved manner. For the McClain murder, the judge found that McKinney committed the murder for pecuniary gain and that McKinney had been convicted of another offense with a potential sentence of life imprisonment or death (the Mertens murder). The trial judge then weighed the aggravating and mitigating circumstances and sentenced McKinney to death for both murders. In 1996, the Arizona Supreme Court affirmed McKinney’s death sentences. Nearly 20 years later, on federal habeas corpus review, an en banc panel of the U. S. Court of Appeals for the Ninth Circuit decided by a 6 to 5 vote that, in sentencing McKinney, the Arizona courts had failed to properly consider McKinney’s posttraumatic stress disorder (PTSD) and had thereby run afoul of this Court’s decision in Eddings v. Oklahoma, 455 U.S. 104 (1982). In Eddings, this Court held that a capital sentencer may not refuse as a matter of law to consider relevant mitigating evidence. Id., at 113–114. McKinney’s case then returned to the Arizona Supreme Court. In that court, McKinney argued that he was entitled to resentencing by a jury. By contrast, the State asked that the Arizona Supreme Court itself conduct a reweighing of the aggravating and mitigating circumstances, as permitted by Clemons v. Mississippi, 494 U.S. 738 (1990). The Arizona Supreme Court agreed with the State. The court itself reviewed the evidence in the record and reweighed the relevant aggravating and mitigating circumstances, including McKinney’s PTSD. The court upheld both death sentences. 245 Ariz. 225, 426 P.3d 1204 (2018). McKinney petitioned for certiorari in this Court. Because of the importance of the case to capital sentencing in Arizona, we granted certiorari. 587 U. S. ___ (2019). The issue in this case is narrow. McKinney contends that after the Ninth Circuit identified an Eddings error, the Arizona Supreme Court could not itself reweigh the aggravating and mitigating circumstances. Rather, according to McKinney, a jury must resentence him. McKinney’s argument does not square with this Court’s decision in Clemons. In Clemons, a Mississippi jury sentenced the defendant to death based in part on two aggravating circumstances. After the Mississippi Supreme Court determined that one of the aggravators was unconstitutionally vague, the defendant argued that he was entitled to resentencing before a jury so that the jury could properly weigh the permissible aggravating and mitigating evidence. This Court disagreed. The Court concluded that the Mississippi Supreme Court could itself reweigh the permissible aggravating and mitigating evidence. 494 U. S., at 745–750. This Court stated that “the Federal Constitution does not prevent a state appellate court from upholding a death sentence that is based in part on an invalid or improperly defined aggravating circumstance either by reweighing of the aggravating and mitigating evidence or by harmless-error review.” Id., at 741. The Court explained that a Clemons reweighing is not a resentencing but instead is akin to harmless-error review in that both may be conducted by an appellate court. McKinney contends that Clemons does not apply here. He raises two basic arguments. First, McKinney maintains that Clemons involved an improperly considered aggravating circumstance, whereas his case involves what the Ninth Circuit said was an improperly ignored mitigating circumstance. But the Court’s analysis in Clemons hinged on its assessment of appellate courts’ ability to weigh aggravating and mitigating evidence, not on any unique effect of aggravators as distinct from mitigators. After noting that the “primary concern in the Eighth Amendment context has been that the sentencing decision be based on the facts and circumstances of the defendant, his background, and his crime,” the Court explained that nothing “inherent in the process of appellate reweighing is inconsistent” with that objective. Id., at 748. Indeed, the Court explicitly rejected the argument that “appellate courts are unable to fully consider and give effect to the mitigating evidence presented by defendants at the sentencing phase in a capital case.” Ibid. And the Court concluded that a state appellate court may uphold the death sentence after a “reweighing of the aggravating and mitigating evidence.” Id., at 741. In deciding whether a particular defendant warrants a death sentence in light of the mix of aggravating and mitigating circumstances, there is no meaningful difference for purposes of appellate reweighing between subtracting an aggravator from one side of the scale and adding a mitigator to the other side. Both involve weighing, and the Court’s decision in Clemons ruled that appellate tribunals may perform a “reweighing of the aggravating and mitigating evidence.” Ibid. In short, a Clemons reweighing is a permissible remedy for an Eddings error. Second, the Court decided Clemons back in 1990, and McKinney argues that Clemons is no longer good law in the wake of this Court’s decisions in Ring v. Arizona, 536 U.S. 584 (2002), and Hurst v. Florida, 577 U. S. ___ (2016). According to McKinney, appellate courts may no longer reweigh aggravating and mitigating circumstances in determining whether to uphold a death sentence. McKinney is incorrect. In Ring, this Court held that capital defendants “are entitled to a jury determination of any fact on which the legislature conditions an increase in their maximum punishment”—in particular, the finding of an aggravating circumstance. 536 U. S., at 589. In Hurst, the Court applied Ring and decided that Florida’s capital sentencing scheme impermissibly allowed “a sentencing judge to find an aggravating circumstance, independent of a jury’s factfinding, that is necessary for imposition of the death penalty.” 577 U. S., at ___ (slip op., at 9). Under Ring and Hurst, a jury must find the aggravating circumstance that makes the defendant death eligible. But importantly, in a capital sentencing proceeding just as in an ordinary sentencing proceeding, a jury (as opposed to a judge) is not constitutionally required to weigh the aggravating and mitigating circumstances or to make the ultimate sentencing decision within the relevant sentencing range. In Apprendi v. New Jersey, 530 U.S. 466 (2000), this Court carefully avoided any suggestion that “it is impermissible for judges to exercise discretion—taking into consideration various factors relating both to offense and offender—in imposing a judgment within the range prescribed by statute.” Id., at 481. And in the death penalty context, as Justice Scalia, joined by Justice Thomas, explained in his concurrence in Ring, the decision in Ring “has nothing to do with jury sentencing. What today’s decision says is that the jury must find the existence of the fact that an aggravating factor existed.” 536 U. S., at 612; see also Kansas v. Carr, 577 U. S. ___, ___–___ (2016) (slip op., at 9–11). Therefore, as Justice Scalia explained, the “States that leave the ultimate life-or-death decision to the judge may continue to do so.” Ring, 536 U. S., at 612. In short, Ring and Hurst did not require jury weighing of aggravating and mitigating circumstances, and Ring and Hurst did not overrule Clemons so as to prohibit appellate reweighing of aggravating and mitigating circumstances. In addition to those two arguments about Clemons, McKinney advances an additional argument based on Ring and Hurst. This argument focuses not on the weighing of aggravators and mitigators, but rather on the Arizona trial court’s initial 1993 finding of the aggravating circumstances that made McKinney eligible for the death penalty. McKinney points out that a jury did not find the aggravating circumstances, as is now required by Ring and Hurst. The hurdle is that McKinney’s case became final on direct review in 1996, long before Ring and Hurst. Ring and Hurst do not apply retroactively on collateral review. See Schriro v. Summerlin, 542 U.S. 348, 358 (2004). Because this case comes to us on state collateral review, Ring and Hurst do not apply. McKinney says, however, that this case has a twist. He asserts that the Arizona Supreme Court’s 2018 decision reweighing the aggravators and mitigators constituted a re- opening of direct review. Because this case (as McKinney sees it) is again on direct review, McKinney argues that he should receive the benefit of Ring and Hurst—namely, a jury resentencing with a jury determination of aggravating circumstances. But the premise of that argument is wrong because the Arizona Supreme Court’s reweighing of the aggravating and mitigating circumstances occurred on collateral review, not direct review. In conducting the reweighing, the Arizona Supreme Court explained that it was conducting an independent review in a collateral proceeding. The court cited its prior decision in State v. Styers, 227 Ariz. 186, 254 P.3d 1132 (2011), which concluded that Arizona could conduct such an independent review in a collateral proceeding. See also Ariz. Rev. Stat. Ann. §13–755 (2010); State v. Hedlund, 245 Ariz. 467, 470–471, 431 P.3d 181, 184–185 (2018). Under these circumstances, we may not second-guess the Arizona Supreme Court’s characterization of state law. See Mullaney v. Wilbur, 421 U.S. 684, 691 (1975); see also Jimenez v. Quarterman, 555 U.S. 113, 120, n. 4 (2009); Styers v. Ryan, 811 F.3d 292, 297, n. 5 (CA9 2015). As a matter of state law, the reweighing proceeding in McKinney’s case occurred on collateral review. McKinney responds that the state label of collateral review cannot control the finality question; that a Clemons reweighing is a sentencing proceeding; and that a Clemons reweighing therefore may occur only on direct review (or on reopening of direct review). But Clemons itself, over a vigorous dissent, stated that an appellate reweighing is not a sentencing proceeding that must be conducted by a jury. See 494 U. S., at 741, 744–755. The appellate reweighing is akin to harmless-error review. Courts routinely conduct harmless-error review in collateral proceedings. Cf., e.g., Brecht v. Abrahamson, 507 U.S. 619, 638 (1993). There is no good reason—and McKinney supplies none—why state courts may not likewise conduct a Clemons reweighing on collateral review. As relevant here, when an Eddings error is found on collateral review, a state court may conduct a Clemons reweighing on collateral review.[1]* Here, therefore, the Arizona Supreme Court permissibly conducted a Clemons reweighing on collateral review. * * * This Court’s precedents establish that state appellate courts may conduct a Clemons reweighing of aggravating and mitigating circumstances, and may do so in collateral proceedings as appropriate and provided under state law. We affirm the judgment of the Arizona Supreme Court. It is so ordered. Notes 1 *Moreover, the District Court’s conditional writ in this case merely required Arizona to correct a purported Eddings error. As we have explained, an Eddings error may be remedied on appeal or on collateral review. Our holding here does not suggest that a State, by use of a collateral label, may conduct a new trial proceeding in violation of current constitutional standards.

589.US.2019_18-935

The Hague Convention on the Civil Aspects of International Child Abduction (Hague Convention or Convention), implemented in the United States by the International Child Abduction Remedies Act, 22 U. S. C. §9001 et seq., provides that a child wrongfully removed from her country of “habitual residence” ordinarily must be returned to that country. Petitioner Monasky, a U. S. citizen, asserts that her Italian husband, respondent Taglieri, became abusive after the couple moved to Italy from the United States. Two months after the birth of the couple’s daughter, A. M. T., in Italy, Monasky fled with the infant to Ohio. Taglieri petitioned the U. S. District Court for the Northern District of Ohio for A. M. T.’s return to Italy under the Convention, pursuant to 22 U. S. C. §9003(b), on the ground that the child had been wrongfully removed from her country of “habitual residence.” The District Court granted Taglieri’s petition, concluding that the parents’ shared intent was for their daughter to live in Italy. Then two-year-old A. M. T. was returned to Italy. The en banc Sixth Circuit affirmed. Under its precedent, the court first noted, an infant’s habitual residence depends on the parents’ shared intent. It then reviewed the District Court’s habitual-residence determination for clear error and found none. In doing so, the court rejected Monasky’s argument that Italy could not qualify as A. M. T.’s “habitual residence” in the absence of an actual agreement by her parents to raise her there. Held: 1. A child’s habitual residence depends on the totality of the circumstances specific to the case, not on categorical requirements such as an actual agreement between the parents. Pp. 7–14. (a) The inquiry begins with the Convention’s text “and the context in which the written words are used.” Air France v. Saks, 470 U.S. 392, 397. The Convention does not define “habitual residence,” but, as the Convention’s text and explanatory report indicate, a child habitually resides where she is at home. This fact-driven inquiry must be “sensitive to the unique circumstances of the case and informed by common sense.” Redmond v. Redmond, 724 F.3d 729, 744. Acclimation of older children and the intentions and circumstances of caregiving parents are relevant considerations, but no single fact is dispositive across all cases. The treaty’s “negotiation and drafting history” corroborates that habitual residence depends on the specific circumstances of the particular case. Medellín v. Texas, 552 U.S. 491, 507. This interpretation also aligns with habitual-residence determinations made by other nations party to the Convention. Pp. 7–12. (b) Monasky’s arguments in favor of an actual-agreement requirement are unpersuasive. While an infant’s “mere physical presence” is not a dispositive indicator of an infant’s habitual residence, a wide range of facts other than an actual agreement, including those indicating that the parents have made their home in a particular place, can enable a trier to determine whether an infant’s residence has the quality of being “habitual.” Nor is adjudicating a dispute over whether an agreement existed a more expeditious way of promoting returns of abducted children and deterring would-be abductors than according courts leeway to consider all the circumstances. Finally, imposing a categorical actual-agreement requirement is unlikely to be an appropriate solution to the serious problem of protecting children born into domestic violence, for it would leave many infants without a habitual residence, and therefore outside the Convention’s domain. Domestic violence should be an issue fully explored in the custody adjudication upon the child’s return. The Convention also has a mechanism for guarding children from the harms of domestic violence: Article 13(b) allows a court to refrain from ordering a child’s return to her habitual residence if “there is a grave risk that [the child’s] return would expose the child to physical or psychological harm or otherwise place the child in an intolerable situation.” Pp. 12–14. 2. A first-instance habitual-residence determination is subject to deferential appellate review for clear error. A trial court’s habitual-residence determination presents a mixed question of law and fact that is heavily fact laden. The determination thus presents a task for factfinding courts and should be judged on appeal by a clear-error review standard. See U. S. Bank N. A. v. Village at Lakeridge, LLC, 583 U. S. ___, ___–___. There is no “historical tradition” indicating otherwise. Pierce v. Underwood, 487 U.S. 552, 558. Clear-error review has a particular virtue in Hague Convention cases: By speeding up appeals, it serves the Convention’s emphasis on expedition. Notably, courts of other treaty partners also review first-instance habitual-residence determinations deferentially. Pp. 14–16. 3. Given the circumstances of this case, it is unnecessary to disturb the judgment below and remand the case to give the lower courts an opportunity to apply the governing totality-of-the-circumstances standard in the first instance. Pp. 16–17. 907 F.3d 404, affirmed. Ginsburg, J., delivered the opinion of the Court, in which Roberts, C. J., and Breyer, Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., joined, and in which Thomas, J., joined as to Parts I, III, and IV. Thomas, J., and Alito, J., filed opinions concurring in part and concurring in the judgment.

Under the Hague Convention on the Civil Aspects of International Child Abduction (Hague Convention or Convention), Oct. 25, 1980, T. I. A. S. No. 11670, S. Treaty Doc. No. 99–11 (Treaty Doc.), a child wrongfully removed from her country of “habitual residence” ordinarily must be returned to that country. This case concerns the standard for determining a child’s “habitual residence” and the standard for reviewing that determination on appeal. The petitioner, Michelle Monasky, is a U. S. citizen who brought her infant daughter, A. M. T., to the United States from Italy after her Italian husband, Domenico Taglieri, became abusive to Monasky. Taglieri successfully petitioned the District Court for A. M. T.’s return to Italy under the Convention, and the Court of Appeals affirmed the District Court’s order. Monasky assails the District Court’s determination that Italy was A. M. T.’s habitual residence. First of the questions presented: Could Italy qualify as A. M. T.’s “habitual residence” in the absence of an actual agreement by her parents to raise her there? The second question: Should the Court of Appeals have reviewed the District Court’s habitual- residence determination independently rather than def- erentially? In accord with decisions of the courts of other countries party to the Convention, we hold that a child’s habitual residence depends on the totality of the circumstances specific to the case. An actual agreement between the parents is not necessary to establish an infant’s habitual residence. We further hold that a first-instance habitual-residence determination is subject to deferential appel- late review for clear error. I A The Hague Conference on Private International Law adopted the Hague Convention in 1980 “[t]o address the problem of international child abductions during domestic disputes.” Lozano v. Montoya Alvarez, 572 U.S. 1, 4 (2014) (internal quotation marks omitted). One hundred one countries, including the United States and Italy, are Convention signatories. Hague Conference on Private Int’l Law, Convention of 25 Oct. 1980 on the Civil Aspects of Int’l Child Abduction, Status Table, https://www.hcch.net/en/instruments/ conventions/status-table/?cid=24. The International Child Abduction Remedies Act (ICARA), 102Stat. 437, as amended, 22 U. S. C. §9001 et seq., implements our Nation’s obligations under the Convention. It is the Convention’s core premise that “the interests of children . . . in matters relating to their custody” are best served when custody decisions are made in the child’s country of “habitual residence.” Convention Preamble, Treaty Doc., at 7; see Abbott v. Abbott, 560 U.S. 1, 20 (2010). To that end, the Convention ordinarily requires the prompt return of a child wrongfully removed or retained away from the country in which she habitually resides. Art. 12, Treaty Doc., at 9 (cross-referencing Art. 3, id., at 7). The removal or retention is wrongful if done in violation of the custody laws of the child’s habitual residence. Art. 3, ibid. The Convention recognizes certain exceptions to the return obligation. Prime among them, a child’s return is not in order if the return would place her at a “grave risk” of harm or otherwise in “an intolerable situation.” Art. 13(b), id., at 10. The Convention’s return requirement is a “provisional” remedy that fixes the forum for custody proceedings. Silberman, Interpreting the Hague Abduction Convention: In Search of a Global Jurisprudence, 38 U. C. D. L. Rev. 1049, 1054 (2005). Upon the child’s return, the custody adjudication will proceed in that forum. See ibid. To avoid delaying the custody proceeding, the Convention instructs contracting states to “use the most expeditious procedures avail- able” to return the child to her habitual residence. Art. 2, Treaty Doc., at 7. See also Art. 11, id., at 9 (prescribing six weeks as normal time for return-order decisions). B In 2011, Monasky and Taglieri were married in the United States. Two years later, they relocated to Italy, where they both found work. Neither then had definite plans to return to the United States. During their first year in Italy, Monasky and Taglieri lived together in Milan. But the marriage soon deteriorated. Taglieri became physically abusive, Monasky asserts, and “forced himself upon [her] multiple times.” 907 F.3d 404, 406 (CA6 2018) (en banc). About a year after their move to Italy, in May 2014, Monasky became pregnant. Taglieri thereafter took up new employment in the town of Lugo, while Monasky, who did not speak Italian, remained about three hours away in Milan. The long-distance separation and a difficult pregnancy further strained their marriage. Monasky looked into returning to the United States. She applied for jobs there, asked about U. S. divorce lawyers, and obtained cost information from moving companies. At the same time, though, she and Taglieri made preparations to care for their expected child in Italy. They inquired about childcare options there, made purchases needed for their baby to live in Italy, and found a larger apartment in a Milan suburb. Their daughter, A. M. T., was born in February 2015. Shortly thereafter, Monasky told Taglieri that she wanted to divorce him, a matter they had previously broached, and that she anticipated returning to the United States. Later, however, she agreed to join Taglieri, together with A. M. T., in Lugo. The parties dispute whether they reconciled while together in that town. On March 31, 2015, after yet another heated argument, Monasky fled with her daughter to the Italian police and sought shelter in a safe house. In a written statement to the police, Monasky alleged that Taglieri had abused her and that she feared for her life. Two weeks later, in April 2015, Monasky and two-month-old A. M. T. left Italy for Ohio, where they moved in with Monasky’s parents. Taglieri sought recourse in the courts. With Monasky absent from the proceedings, an Italian court granted Taglieri’s request to terminate Monasky’s parental rights, discrediting her statement to the Italian police. App. 183. In the United States, on May 15, 2015, Taglieri petitioned the U. S. District Court for the Northern District of Ohio for the return of A. M. T. to Italy under the Hague Convention, pursuant to 22 U. S. C. §9003(b), on the ground that Italy was her habitual residence. The District Court granted Taglieri’s petition after a four-day bench trial. Sixth Circuit precedent at the time, the District Court observed, instructed courts that a child habitually resides where the child has become “acclimatiz[ed]” to her surroundings. App. to Pet. for Cert. 85a (quoting Robert v. Tesson, 507 F.3d 981, 993 (CA6 2007)). An infant, however, is “too young” to acclimate to her surroundings. App. to Pet. for Cert. 87a. The District Court therefore proceeded on the assumption that “the shared intent of the [parents] is relevant in determining the habitual residence of an infant,” though “particular facts and circumstances . . . might necessitate the consideration [of] other factors.” Id., at 97a. The shared intention of A. M. T.’s parents, the District Court found, was for their daughter to live in Italy, where the parents had established a marital home “with no definitive plan to return to the United States.” Ibid. Even if Monasky could change A. M. T.’s habitual residence unilaterally by making plans to raise A. M. T. away from Italy, the District Court added, the evidence on that score indicated that, until the day she fled her husband, Monasky had “no definitive plans” to raise A. M. T. in the United States. Id., at 98a. In line with its findings, the District Court ordered A. M. T.’s prompt return to Italy. The Sixth Circuit and this Court denied Monasky’s requests for a stay of the return order pending appeal. 907 F. 3d, at 407. In December 2016, A. M. T., nearly two years old, was returned to Italy and placed in her father’s care.[1] In the United States, Monasky’s appeal of the District Court’s return order proceeded. See Chafin v. Chafin, 568 U.S. 165, 180 (2013) (the return of a child under the Hague Convention does not moot an appeal of the return order). A divided three-judge panel of the Sixth Circuit affirmed the District Court’s order, and a divided en banc court adhered to that disposition. The en banc majority noted first that, after the District Court’s decision, a precedential Sixth Circuit opinion, Ahmed v. Ahmed, 867 F.3d 682 (2017), established that, as the District Court had assumed, an infant’s habitual residence depends on “shared parental intent.” 907 F. 3d, at 408 (quoting Ahmed, 867 F. 3d, at 690). The en banc majority then reviewed the District Court’s habitual-residence determination for clear error and found none. Sustaining the District Court’s determination that A. M. T.’s habitual residence was Italy, the majority rejected Monasky’s argument that the District Court erred because “she and Taglieri never had a ‘meeting of the minds’ about their child’s future home.” 907 F. 3d, at 410. No member of the en banc court disagreed with the majority’s rejection of Monasky’s proposed actual-agreement requirement. Nor did any judge maintain that Italy was not A. M. T.’s habitual residence. Judge Boggs wrote a concurring opinion adhering to the reasoning of his three-judge panel majority opinion: “[A]bsent unusual circumstances, where a child has resided exclusively in a single country, especially with both parents, that country is the child’s habitual residence.” Id., at 411. The dissenters urged two discrete objections. Some would have reviewed the District Court’s habitual-residence determination de novo. See id., at 419 (opinion of Moore, J.). All would have remanded for the District Court to reconsider A. M. T.’s habitual residence in light of the Sixth Circuit’s Ahmed precedent. See 907 F. 3d, at 419–420; id., at 421–422 (opinion of Gibbons, J.); id., at 423 (opinion of Stranch, J.). We granted certiorari to clarify the standard for habitual residence, an important question of federal and international law, in view of differences in emphasis among the Courts of Appeals. 587 U. S. ___ (2019). Compare, e.g., 907 F. 3d, at 407 (case below) (describing inquiry into the child’s acclimatization as the “primary” approach), with, e.g., Mozes v. Mozes, 239 F.3d 1067, 1073–1081 (CA9 2001) (placing greater weight on the shared intentions of the parents), with, e.g., Redmond v. Redmond, 724 F.3d 729, 746 (CA7 2013) (rejecting “rigid rules, formulas, or presumptions”). Certiorari was further warranted to resolve a division in Courts of Appeals over the appropriate standard of appellate review. Compare, e.g., 907 F. 3d, at 408–409 (case below) (clear error), with, e.g., Mozes, 239 F. 3d, at 1073 (de novo). II The first question presented concerns the standard for habitual residence: Is an actual agreement between the parents on where to raise their child categorically necessary to establish an infant’s habitual residence? We hold that the determination of habitual residence does not turn on the existence of an actual agreement. A We begin with “the text of the treaty and the context in which the written words are used.” Air France v. Saks, 470 U.S. 392, 397 (1985). The Hague Convention does not define the term “habitual residence.” A child “resides” where she lives. See Black’s Law Dictionary 1176 (5th ed. 1979). Her residence in a particular country can be deemed “habitual,” however, only when her residence there is more than transitory. “Habitual” implies “[c]ustomary, usual, of the nature of a habit.” Id., at 640. The Hague Convention’s text alone does not definitively tell us what makes a child’s residence sufficiently enduring to be deemed “habitual.” It surely does not say that habitual residence depends on an actual agreement between a child’s parents. But the term “habitual” does suggest a fact-sensitive inquiry, not a categorical one. The Convention’s explanatory report confirms what the Convention’s text suggests. The report informs that habitual residence is a concept “well-established . . . in the Hague Conference.” 1980 Conférence de La Haye de droit international privé, Enlèvement d’enfants, E. Pérez-Vera, Explanatory Report in 3 Actes et documents de la Quatorzième session, p. 445, ¶66 (1982) (Pérez-Vera).[2] The report refers to a child’s habitual residence in fact-focused terms: “the family and social environment in which [the child’s] life has developed.” Id., at 428, ¶11. What makes a child’s residence “habitual” is therefore “some degree of integration by the child in a social and family environment.” OL v. PQ, 2017 E. C. R. No. C–111/17, ¶42 (Judgt. of June 8); accord Office of the Children’s Lawyer v. Balev, [2018] 1 S. C. R. 398, 421, ¶43, 424 D. L. R. (4th) 391, 410, ¶43 (Can.); A v. A, [2014] A. C., ¶54 (2013) (U. K.). Accordingly, while Federal Courts of Appeals have diverged, if only in emphasis, in the standards they use to locate a child’s habitual residence, see supra, at 6, they share a “common” understanding: The place where a child is at home, at the time of removal or retention, ranks as the child’s habitual residence. Karkkainen v. Kovalchuk, 445 F.3d 280, 291 (CA3 2006). Because locating a child’s home is a fact-driven inquiry, courts must be “sensitive to the unique circumstances of the case and informed by common sense.” Redmond, 724 F. 3d, at 744. For older children capable of acclimating to their surroundings, courts have long recognized, facts indicating acclimatization will be highly relevant.[3] Because children, especially those too young or otherwise unable to acclimate, depend on their parents as caregivers, the intentions and circumstances of caregiving parents are relevant considerations. No single fact, however, is dispositive across all cases. Common sense suggests that some cases will be straightforward: Where a child has lived in one place with her family indefinitely, that place is likely to be her habitual residence. But suppose, for instance, that an infant lived in a country only because a caregiving parent had been coerced into remaining there. Those circumstances should figure in the calculus. See Karkkainen, 445 F. 3d, at 291 (“The inquiry into a child’s habitual residence is a fact-intensive determination that cannot be reduced to a predetermined formula and necessarily varies with the circumstances of each case.”). The treaty’s “negotiation and drafting history” corroborates that a child’s habitual residence depends on the specific circumstances of the particular case. Medellín v. Texas, 552 U.S. 491, 507 (2008) (noting that such history may aid treaty interpretation). The Convention’s explanatory report states that the Hague Conference regarded habitual residence as “a question of pure fact, differing in that respect from domicile.” Pérez-Vera 445, ¶66. The Conference deliberately chose “habitual residence” for its factual character, making it the foundation for the Convention’s return remedy in lieu of formal legal concepts like domicile and nationality. See Anton, The Hague Convention on International Child Abduction, 30 Int’l & Comp. L. Q. 537, 544 (1981) (history of the Convention authored by the drafting commission’s chairman). That choice is instructive. The signatory nations sought to afford courts charged with determining a child’s habitual residence “maximum flex- ibility” to respond to the particular circumstances of each case. P. Beaumont & P. McEleavy, The Hague Convention on International Child Abduction 89–90 (1999) (Beaumont & McEleavy). The aim: to ensure that custody is adjudicated in what is presumptively the most appropriate forum—the country where the child is at home. Our conclusion that a child’s habitual residence depends on the particular circumstances of each case is bolstered by the views of our treaty partners. ICARA expressly recognizes “the need for uniform international interpretation of the Convention.” 22 U. S. C. §9001(b)(3)(B). See Lozano, 572 U. S., at 13; Abbott, 560 U. S., at 16. The understanding that the opinions of our sister signatories to a treaty are due “considerable weight,” this Court has said, has “special force” in Hague Convention cases. Ibid. (quoting El Al Israel Airlines, Ltd. v. Tsui Yuan Tseng, 525 U.S. 155, 176 (1999), in turn quoting Air France, 470 U. S., at 404). The “clear trend” among our treaty partners is to treat the determination of habitual residence as a fact-driven inquiry into the particular circumstances of the case. Balev, [2018] 1 S. C. R., at 423, ¶50, 424 D. L. R. (4th), at 411, ¶50. Lady Hale wrote for the Supreme Court of the United Kingdom: A child’s habitual residence “depends on numerous factors . . . with the purposes and intentions of the parents being merely one of the relevant factors. . . . The essentially factual and individual nature of the inquiry should not be glossed with legal concepts.” A, [2014] A. C., at ¶54. The Court of Justice of the European Union, the Supreme Court of Canada, and the High Court of Australia agree. See OL, 2017 E. C. R. No. C–111/17, ¶42 (the habitual residence of a child “must be established . . . taking account of all the circumstances of fact specific to each individual case”); Balev, [2018] 1 S. C. R., at 421, 423–430, ¶¶43, 48–71, 424 D. L. R. (4th), at 410–417, ¶¶43, 48–71 (adopting an approach to habitual residence under which “[t]he judge considers all relevant links and circumstances”); LK v. Director-General, Dept. of Community Servs., [2009] 237 C. L. R. 582, 596, ¶35 (Austl.) (“to seek to identify a set list of criteria that bear upon where a child is habitually resident . . . would deny the simple observation that the question of habitual residence will fall for decision in a very wide range of circumstances”). Intermediate appellate courts in Hong Kong and New Zealand have similarly stated what “habitual residence” imports. See LCYP v. JEK, [2015] 4 H. K. L. R. D. 798, 809–810, ¶7.7 (H. K.); Punter v. Secretary for Justice, [2007] 1 N. Z. L. R. 40, 71, ¶130 (N. Z.). Tellingly, Monasky has not identified a single treaty partner that has adopted her actual-agreement proposal. See Tr. of Oral Arg. 9.[4] The bottom line: There are no categorical requirements for establishing a child’s habitual residence—least of all an actual-agreement requirement for infants. Monasky’s proposed actual-agreement requirement is not only unsupported by the Convention’s text and inconsistent with the leeway and international harmony the Convention demands; her proposal would thwart the Convention’s “objects and purposes.” Abbott, 560 U. S., at 20. An actual-agreement requirement would enable a parent, by withholding agreement, unilaterally to block any finding of habitual residence for an infant. If adopted, the requirement would undermine the Convention’s aim to stop unilateral decisions to remove children across international borders. Moreover, when parents’ relations are acrimonious, as is often the case in controversies arising under the Convention, agreement can hardly be expected. In short, as the Court of Appeals observed below, “Monasky’s approach would create a presumption of no habitual residence for infants, leaving the population most vulnerable to abduction the least protected.” 907 F. 3d, at 410. B Monasky counters that an actual-agreement requirement is necessary to ensure “that an infant’s mere physical presence in a country has a sufficiently settled quality to be deemed ‘habitual.’ ” Brief for Petitioner 32. An infant’s “mere physical presence,” we agree, is not a dispositive indicator of an infant’s habitual residence. But a wide range of facts other than an actual agreement, including facts indicating that the parents have made their home in a particular place, can enable a trier to determine whether an infant’s residence in that place has the quality of being “habitual.” Monasky also argues that a bright-line rule like her proposed actual-agreement requirement would promote prompt returns of abducted children and deter would-be abductors from “tak[ing] their chances” in the first place. Id., at 35, 38. Adjudicating a winner-takes-all evidentiary dispute over whether an agreement existed, however, is scarcely more expeditious than providing courts with leeway to make “a quick impression gained on a panoramic view of the evidence.” Beaumont & McEleavy 103 (internal quotation marks omitted). When all the circumstances are in play, would-be abductors should find it more, not less, difficult to manipulate the reality on the ground, thus impeding them from forging “artificial jurisdictional links . . . with a view to obtaining custody of a child.” Pérez-Vera 428, ¶11. Finally, Monasky and amici curiae raise a troublesome matter: An actual-agreement requirement, they say, is necessary to protect children born into domestic violence. Brief for Petitioner 42–44; Brief for Sanctuary for Families et al. as Amici Curiae 11–20. Domestic violence poses an “intractable” problem in Hague Convention cases involving caregiving parents fleeing with their children from abuse. Hale, Taking Flight—Domestic Violence and Child Abduction, 70 Current Legal Prob. 3, 11 (2017). We doubt, however, that imposing a categorical actual-agreement requirement is an appropriate solution, for it would leave many infants without a habitual residence, and therefore outside the Convention’s domain. See supra, at 11–12. Settling the forum for adjudication of a dispute over a child’s custody, of course, does not dispose of the merits of the controversy over custody. Domestic violence should be an issue fully explored in the custody adjudication upon the child’s return. The Hague Convention, we add, has a mechanism for guarding children from the harms of domestic violence: Article 13(b). See Hale, 70 Current Legal Prob., at 10–16 (on Hague Conference working group to develop a best- practices guide to the interpretation and application of Article 13(b) in cases involving domestic violence). Article 13(b), as noted supra, at 3, allows a court to refrain from ordering a child’s return to her habitual residence if “there is a grave risk that [the child’s] return would expose the child to physical or psychological harm or otherwise place the child in an intolerable situation.” Art. 13(b), Treaty Doc., at 10. Monasky raised below an Article 13(b) defense to Taglieri’s return petition. In response, the District Court credited Monasky’s “deeply troubl[ing]” allegations of her exposure to Taglieri’s physical abuse. App. to Pet. for Cert. 105a. But the District Court found “no evidence” that Taglieri ever abused A. M. T. or otherwise disregarded her well-being. Id., at 103a, 105a. That court also followed Circuit precedent disallowing consideration of psychological harm A. M. T. might experience due to separation from her mother. Id., at 102a. Monasky does not challenge those dispositions in this Court. III Turning to the second question presented: What is the appropriate standard of appellate review of an initial adjudicator’s habitual-residence determination? Neither the Convention nor ICARA prescribes modes of appellate review, other than the directive to act “expeditiously.” Art. 11, Treaty Doc., at 9; see Federal Judicial Center, J. Garbolino, The 1980 Hague Convention on the Civil Aspects of International Child Abduction: A Guide for Judges 162 (2d ed. 2015) (the Convention’s “emphasis on prompt disposition applies to appellate proceedings”).[5] Absent a treaty or statutory prescription, the appropriate level of deference to a trial court’s habitual-residence determination depends on whether that determination resolves a question of law, a question of fact, or a mixed question of law and fact. Generally, questions of law are reviewed de novo and questions of fact, for clear error, while the appropriate standard of appellate review for a mixed question “depends . . . on whether answering it entails primarily legal or factual work.” U. S. Bank N. A. v. Village at Lakeridge, LLC, 583 U. S. ___, ___–___ (2018) (slip op., at 8–9). A child’s habitual residence presents what U. S. law types a “mixed question” of law and fact—albeit barely so. Id., at ___ (slip op., at 7). The inquiry begins with a legal question: What is the appropriate standard for habitual residence? Once the trial court correctly identifies the governing totality-of-the-circumstances standard, however, what remains for the court to do in applying that standard, as we explained supra, at 7–11, is to answer a factual question: Was the child at home in the particular country at issue? The habitual-residence determination thus presents a task for factfinding courts, not appellate courts, and should be judged on appeal by a clear-error review standard deferential to the factfinding court. In selecting standards of appellate review, the Court has also asked whether there is “a long history of appellate practice” indicating the appropriate standard, for arriving at the standard from first principles can prove “uncommonly difficult.” Pierce v. Underwood, 487 U.S. 552, 558 (1988). Although some Federal Courts of Appeals have reviewed habitual-residence determinations de novo, there has been no uniform, reasoned practice in this regard, nothing resembling “a historical tradition.” Ibid. See also supra, at 6–7 (noting a Circuit split). Moreover, when a mixed question has a factual foundation as evident as the habitual-residence inquiry here does, there is scant cause to default to historical practice. Clear-error review has a particular virtue in Hague Convention cases. As a deferential standard of review, clear-error review speeds up appeals and thus serves the Convention’s premium on expedition. See Arts. 2, 11, Treaty Doc., at 7, 9. Notably, courts of our treaty partners review first-instance habitual-residence determinations deferentially. See, e.g., Balev, [2018] 1 S. C. R., at 419, ¶38, 424 D. L. R. (4th), at 408, ¶38; Punter, [2007] 1 N. Z. L. R., at 88, ¶204; AR v. RN, [2015] UKSC 35, ¶18. IV Although agreeing with the manner in which the Court has resolved the two questions presented, the United States, as an amicus curiae supporting neither party, suggests remanding to the Court of Appeals rather than affirming that court’s judgment. Brief for United States as Amicus Curiae 28. Ordinarily, we might take that course, giving the lower courts an opportunity to apply the gov- erning totality-of-the-circumstances standard in the first instance. Under the circumstances of this case, however, we decline to disturb the judgment below. True, the lower courts viewed A. M. T.’s situation through the lens of her parents’ shared intentions. But, after a four-day bench trial, the District Court had before it all the facts relevant to the dispute. Asked at oral argument to identify any additional fact the District Court did not digest, counsel for the United States offered none. Tr. of Oral Arg. 38. Monasky and Taglieri agree that their dispute “requires no ‘further factual development,’ ” and neither party asks for a remand. Reply Brief 22 (quoting Brief for Respondent 54). Monasky does urge the Court to reverse if it rests A. M. T.’s habitual residence on all relevant circumstances. She points to her “absence of settled ties to Italy” and the “unsettled and unstable conditions in which A. M. T. resided in Italy.” Reply Brief 19 (internal quotation marks and alteration omitted). The District Court considered the competing facts bearing on those assertions, however, including the fraught circumstances in which the parties’ marriage unraveled. That court nevertheless found that Monasky had sufficient ties to Italy such that “[a]rguably, [she] was a habitual resident of Italy.” App. to Pet. for Cert. 91a. And, despite the rocky state of the marriage, the District Court found beyond question that A. M. T. was born into “a marital home in Italy,” one that her parents established “with no definitive plan to return to the United States.” Id., at 97a. Nothing in the record suggests that the District Court would appraise the facts differently on remand. A remand would consume time when swift resolution is the Convention’s objective. The instant return-order proceedings began a few months after A. M. T.’s birth. She is now five years old. The more than four-and-a-half-year duration of this litigation dwarfs the six-week target time for resolving a return-order petition. See Art. 11, Treaty Doc., at 9. Taglieri represents that custody of A. M. T. has so far been resolved only “on an interim basis,” Brief for Respondent 56, n. 13, and that custody proceedings, including the matter of Monasky’s parental rights, remain pending in Italy. Tr. of Oral Arg. 60–61. Given the exhaustive record before the District Court, the absence of any reason to anticipate that the District Court’s judgment would change on a remand that neither party seeks, and the protraction of proceedings thus far, final judgment on A. M. T.’s return is in order. * * * For the reasons stated, the judgment of the Court of Appeals for the Sixth Circuit is Affirmed. Notes 1 Taglieri represents that “[a]n order issued by the Italian court in December 2018 awarded legal custody of A. M. T., on an interim basis, to the Lugo municipality . . . with placement at [Taglieri’s] residence; and provided that mother-daughter visits would continue under the plan prescribed in a court order issued earlier in 2018.” Brief for Respondent 56, n. 13. 2 According to an analysis provided by the Department of State to the Senate during the ratification process, the “explanatory report is recognized by the [Hague] Conference as the official history and commentary on the Convention and is a source of background on the meaning of the provisions of the Convention.” Hague International Child Abduction Convention; Text and Legal Analysis, 51 Fed. Reg. 10503 (1986). The explanatory report notes, however, that “it has not been approved by the Conference, and it is possible that, despite the Rapporter’s [sic] efforts to remain objective, certain passages reflect a viewpoint which is in part subjective.” Pérez-Vera 427–428, ¶8. See Abbott v. Abbott, 560 U.S. 1, 19 (2010) (“We need not decide whether this Report should be given greater weight than a scholarly commentary.”). 3 Facts courts have considered include: “a change in geography combined with the passage of an appreciable period of time,” “age of the child,” “immigration status of child and parent,” “academic activities,” “social engagements,” “participation in sports programs and excursions,” “meaningful connections with the people and places in the child’s new country,” “language proficiency,” and “location of personal belongings.” Federal Judicial Center, J. Garbolino, The 1980 Hague Convention on the Civil Aspects of International Child Abduction: A Guide for Judges 67–68 (2d ed. 2015). 4 Monasky disputes that foreign courts apply a totality-of-the-circumstances standard to infants, as opposed to older children. In this regard, she points out, the Court of Justice of the European Union instructs that, “where ‘the infant is in fact looked after by her mother,’ ‘it is necessary to assess the mother’s integration in her social and family environment’ in the relevant country.” Reply Brief 5–6 (quoting Mercredi v. Chaffe, 2010 E. C. R. I–14309, I–14379, ¶55). True, a caregiving parent’s ties to the country at issue are highly relevant. But the Court of Justice did not hold that the caregiver’s ties are the end of the inquiry. Rather, the deciding court must “tak[e] account of all the circumstances of fact specific to each individual case.” Id., ¶56 (emphasis added) (also considering, among other factors, the infant’s physical presence and duration of time in the country). 5 Monasky contends that only de novo review can satisfy “the need for uniform international interpretation of the Convention.” 22 U. S. C. §9001(b)(3)(B). See Brief for Petitioner 19–21. However, ICARA’s recognition of the need for harmonious international interpretation is hardly akin to the “clear statutory prescription” on the standard of appellate review that Congress has provided “[f]or some few trial court determinations.” Pierce v. Underwood, 487 U.S. 552, 558 (1988).

590.US.2019_18-1432

Under federal immigration law, noncitizens who commit certain crimes are removable from the United States. During removal proceedings, a noncitizen who demonstrates a likelihood of torture in the designated country of removal is entitled to relief under the international Convention Against Torture (CAT) and may not be removed to that country. If an immigration judge orders removal and denies CAT relief, the noncitizen may appeal both orders to the Board of Immigration Appeals and then to a federal court of appeals. But if the noncitizen has committed any crime specified in 8 U. S. C. §1252(a)(2)(C), the scope of judicial review of the removal order is limited to constitutional and legal challenges. See §1252(a)(2)(D). The Government sought to remove petitioner Nidal Khalid Nasrallah after he pled guilty to receiving stolen property. Nasrallah applied for CAT relief to prevent his removal to Lebanon. The Immigration Judge ordered Nasrallah removed and granted CAT relief. On appeal, the Board of Immigration Appeals vacated the CAT relief order and ordered Nasrallah removed to Lebanon. The Eleventh Circuit declined to review Nasrallah’s factual challenges to the CAT order because Nasrallah had committed a §1252(a)(2)(C) crime and Circuit precedent precluded judicial review of factual challenges to both the final order of removal and the CAT order in such cases. Held: Sections 1252(a)(2)(C) and (D) do not preclude judicial review of a noncitizen’s factual challenges to a CAT order. Pp. 5–13. (a) Three interlocking statutes establish that CAT orders may be reviewed together with final orders of removal in a court of appeals. The Illegal Immigration Reform and Immigrant Responsibility Act of 1996 authorizes noncitizens to obtain direct “review of a final order of removal” in a court of appeals, §1252(a)(1), and requires that all challenges arising from the removal proceeding be consolidated for review, §1252(b)(9). The Foreign Affairs Reform and Restructuring Act of 1998 (FARRA) implements Article 3 of CAT and provides for judicial review of CAT claims “as part of the review of a final order of removal.” §2242(d). And the REAL ID Act of 2005 clarifies that final orders of removal and CAT orders may be reviewed only in the courts of appeals. §§1252(a)(4)–(5). Pp. 5–6. (b) Sections 1252(a)(2)(C) and (D) preclude judicial review of factual challenges only to final orders of removal. A CAT order is not a final “order of removal,” which in this context is defined as an order “concluding that the alien is deportable or ordering deportation,” §1101(a)(47)(A). Nor does a CAT order merge into a final order of removal, because a CAT order does not affect the validity of a final order of removal. See INS v. Chadha, 462 U.S. 919, 938. FARRA provides that a CAT order is reviewable “as part of the review of a final order of removal,” not that it is the same as, or affects the validity of, a final order of removal. Had Congress wished to preclude judicial review of factual challenges to CAT orders, it could have easily done so. Pp. 6–9. (c) The standard of review for factual challenges to CAT orders is substantial evidence—i.e., the agency’s “findings of fact are conclusive unless any reasonable adjudicator would be compelled to conclude to the contrary.” §1252(b)(4)(B). The Government insists that the statute supplies no judicial review of factual challenges to CAT orders, but its arguments are unpersuasive. First, the holding in Foti v. INS, 375 U.S. 217, depends on an outdated interpretation of “final orders of deportation” and so does not control here. Second, the Government argues that §1252(a)(1) supplies judicial review only of final orders of removal, and if a CAT order is not merged into that final order, then no statute authorizes review of the CAT claim. But both FARRA and the REAL ID Act provide for direct review of CAT orders in the courts of appeals. Third, the Government’s assertion that Congress would not bar review of factual challenges to a removal order and allow such challenges to a CAT order ignores the importance of adherence to the statutory text as well as the good reason Congress had for distinguishing the two—the facts that rendered the noncitizen removable are often not in serious dispute, while the issues related to a CAT order will not typically have been litigated prior to the alien’s removal proceedings. Fourth, the Government’s policy argument—that judicial review of the factual components of a CAT order would unduly delay removal proceedings—has not been borne out in practice in those Circuits that have allowed factual challenges to CAT orders. Fifth, the Government fears that a decision allowing factual review of CAT orders would lead to factual challenges to other orders in the courts of appeals. But orders denying discretionary relief under §1252(a)(2)(B) are not affected by this decision, and the question whether factual challenges to statutory withholding orders under §1231(b)(3)(A) are subject to judicial review is not presented here. Pp. 9–13. 762 Fed. Appx. 638, reversed. Kavanaugh, J., delivered the opinion of the Court, in which Roberts, C. J., and Ginsburg, Breyer, Sotomayor, Kagan, and Gorsuch, JJ., joined. Thomas, J., filed a dissenting opinion, in which Alito, J., joined.

Under federal immigration law, noncitizens who commit certain crimes are removable from the United States. During removal proceedings, a noncitizen may raise claims under the international Convention Against Torture, known as CAT. If the noncitizen demonstrates that he likely would be tortured if removed to the designated country of removal, then he is entitled to CAT relief and may not be removed to that country (although he still may be removed to other countries). If the immigration judge orders removal and denies CAT relief, the noncitizen may appeal to the Board of Immigration Appeals. If the Board of Immigration Appeals orders removal and denies CAT relief, the noncitizen may obtain judicial review in a federal court of appeals of both the final order of removal and the CAT order. In the court of appeals, for cases involving noncitizens who have committed any crime specified in 8 U. S. C. §1252(a)(2)(C), federal law limits the scope of judicial review. Those noncitizens may obtain judicial review of constitutional and legal challenges to the final order of removal, but not of factual challenges to the final order of removal. Everyone agrees on all of the above. The dispute here concerns the scope of judicial review of CAT orders for those noncitizens who have committed crimes specified in §1252(a)(2)(C). The Government argues that judicial review of a CAT order is analogous to judicial review of a final order of removal. The Government contends, in other words, that the court of appeals may review the noncitizen’s constitutional and legal challenges to a CAT order, but not the noncitizen’s factual challenges to the CAT order. Nasrallah responds that the court of appeals may review the noncitizen’s constitutional, legal, and factual challenges to the CAT order, although Nasrallah acknowledges that judicial review of factual challenges to CAT orders must be highly deferential. So the narrow question before the Court is whether, in a case involving a noncitizen who committed a crime specified in §1252(a)(2)(C), the court of appeals should review the noncitizen’s factual challenges to the CAT order (i) not at all or (ii) deferentially. Based on the text of the statute, we conclude that the court of appeals should review factual challenges to the CAT order deferentially. We therefore reverse the judgment of the U. S. Court of Appeals for the Eleventh Circuit. I Nidal Khalid Nasrallah is a native and citizen of Lebanon. In 2006, when he was 17 years old, Nasrallah came to the United States on a tourist visa. In 2007, he became a lawful permanent resident. In 2013, Nasrallah pled guilty to two counts of receiving stolen property. The U. S. District Court for the Western District of North Carolina sentenced Nasrallah to 364 days in prison. Based on Nasrallah’s conviction, the Government initiated deportation proceedings. See 8 U. S. C. §1227(a)(2)(A)(i). In those proceedings, Nasrallah applied for CAT relief to prevent his removal to Lebanon. See Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, Art. 3, Dec. 10, 1984, S. Treaty Doc. No. 100–20, p. 20, 1465 U. N. T. S. 114. Nasrallah alleged that he was a member of the Druze religion, and that he had been tortured by Hezbollah before he came to the United States. Nasrallah argued that he would be tortured again if returned to Lebanon.[1] The Immigration Judge determined that Nasrallah was removable. As to the CAT claim, the Immigration Judge found that Nasrallah had previously suffered torture at the hands of Hezbollah. Based on Nasrallah’s past experience and the current political conditions in Lebanon, the Immigration Judge concluded that Nasrallah likely would be tortured again if returned to Lebanon. The Immigration Judge ordered Nasrallah removed, but also granted CAT relief and thereby blocked Nasrallah’s removal to Lebanon. On appeal, the Board of Immigration Appeals disagreed that Nasrallah likely would be tortured in Lebanon. The Board therefore vacated the order granting CAT relief and ordered Nasrallah removed to Lebanon. Nasrallah filed a petition for review in the U. S. Court of Appeals for the Eleventh Circuit, claiming (among other things) that the Board of Immigration Appeals erred in finding that he would not likely be tortured in Lebanon. Nasrallah raised factual challenges to the Board’s CAT order. Applying Circuit precedent, the Eleventh Circuit declined to review Nasrallah’s factual challenges. Nasrallah v. United States Attorney General, 762 Fed. Appx. 638 (2019). The court explained that Nasrallah had been convicted of a crime specified in 8 U. S. C. §1252(a)(2)(C). Noncitizens convicted of §1252(a)(2)(C) crimes may not obtain judicial review of factual challenges to a “final order of removal.” §§1252(a)(2)(C)–(D). Under Eleventh Circuit precedent, that statute also precludes judicial review of factual challenges to the CAT order.[2] Nasrallah contends that the Eleventh Circuit should have reviewed his factual challenges to the CAT order because the statute bars review only of factual challenges to a “final order of removal.” According to Nasrallah, a CAT order is not a “final order of removal” and does not affect the validity of a final order of removal. Therefore, Nasrallah argues, the statute by its terms does not bar judicial review of factual challenges to a CAT order. The Courts of Appeals are divided over whether §§1252(a)(2)(C) and (D) preclude judicial review of factual challenges to a CAT order. Most Courts of Appeals have sided with the Government; the Seventh and Ninth Circuits have gone the other way. Compare Gourdet v. Holder, 587 F.3d 1, 5 (CA1 2009); Ortiz-Franco v. Holder, 782 F.3d 81, 88 (CA2 2015); Pieschacon-Villegas v. Attorney General of U. S., 671 F.3d 303, 309–310 (CA3 2011); Oxygene v. Lynch, 813 F.3d 541, 545 (CA4 2016); Escudero-Arciniega v. Holder, 702 F.3d 781, 785 (CA5 2012); Tran v. Gonzales, 447 F.3d 937, 943 (CA6 2006); Lovan v. Holder, 574 F.3d 990, 998 (CA8 2009); Cole v. United States Attorney General, 712 F.3d 517, 532 (CA11 2013), with Wanjiru v. Holder, 705 F.3d 258, 264 (CA7 2013); Vinh Tan Nguyen v. Holder, 763 F.3d 1022, 1029 (CA9 2014). In light of the Circuit split on this important question of federal law, we granted certiorari. 589 U. S. ___ (2019).[3] II When a noncitizen is removable because he committed a crime specified in §1252(a)(2)(C), immigration law bars judicial review of the noncitizen’s factual challenges to his final order of removal. In the Government’s view, the law also bars judicial review of the noncitizen’s factual challenges to a CAT order. Nasrallah disagrees. We conclude that Nasrallah has the better of the statutory argument. A We begin by describing the three interlocking statutes that provide for judicial review of final orders of removal and CAT orders. The first relevant statute is the Illegal Immigration Reform and Immigrant Responsibility Act of 1996. That Act authorizes noncitizens to obtain direct “review of a final order of removal” in a court of appeals. 110Stat. 3009–607, 8 U. S. C. §1252(a)(1). As the parties agree, in the deportation context, a “final order of removal” is a final order “concluding that the alien is deportable or ordering deportation.” §1101(a)(47)(A); see §309(d)(2), 110Stat. 3009–627; Calcano-Martinez v. INS, 533 U.S. 348, 350, n. 1 (2001). The Act also states that judicial review “of all questions of law and fact . . . arising from any action taken or proceeding brought to remove an alien from the United States under this subchapter shall be available only in judicial review of a final order under this section.” 8 U. S. C. §1252(b)(9); see 110Stat. 3009–610. In other words, a noncitizen’s various challenges arising from the removal proceeding must be “consolidated in a petition for review and considered by the courts of appeals.” INS v. St. Cyr, 533 U.S. 289, 313, and n. 37 (2001). By consolidating the issues arising from a final order of removal, eliminating review in the district courts, and supplying direct review in the courts of appeals, the Act expedites judicial review of final orders of removal. The second relevant statute is the Foreign Affairs Reform and Restructuring Act of 1998, known as FARRA. FARRA implements Article 3 of the international Convention Against Torture, known as CAT. As relevant here, CAT prohibits removal of a noncitizen to a country where the noncitizen likely would be tortured. Importantly for present purposes, §2242(d) of FARRA provides for judicial review of CAT claims “as part of the review of a final order of removal pursuant to section 242 of the Immigration and Nationality Act (8 U. S. C. 1252).” 112Stat. 2681–822, note following 8 U. S. C. §1231. The third relevant statute is the REAL ID Act of 2005. As relevant here, that Act responded to this Court’s 2001 decision in St. Cyr. In St. Cyr, this Court ruled that the 1996 Act, although purporting to eliminate district court review of final orders of removal, did not eliminate district court review via habeas corpus of constitutional or legal challenges to final orders of removal. 533 U. S., at 312–313. The REAL ID Act clarified that final orders of removal may not be reviewed in district courts, even via habeas corpus, and may be reviewed only in the courts of appeals. See 119Stat. 310, 8 U. S. C. §1252(a)(5). The REAL ID Act also provided that CAT orders likewise may not be reviewed in district courts, even via habeas corpus, and may be reviewed only in the courts of appeals. See 119Stat. 310, 8 U. S. C. §1252(a)(4). B Those three Acts establish that CAT orders may be reviewed together with final orders of removal in a court of appeals. But judicial review of final orders of removal is somewhat limited in cases (such as Nasrallah’s) involving noncitizens convicted of crimes specified in §1252(a)(2)(C). In those cases, a court of appeals may review constitutional or legal challenges to a final order of removal, but the court of appeals may not review factual challenges to a final order of removal. §§1252(a)(2)(C)–(D); see Guerrero-Lasprilla v. Barr, 589 U. S. ___, ___–___ (2020) (slip op., at 11–13). The question in this case is the following: By precluding judicial review of factual challenges to final orders of removal, does the law also preclude judicial review of factual challenges to CAT orders? We conclude that it does not. The relevant statutory text precludes judicial review of factual challenges to final orders of removal—and only to final orders of removal. In the deportation context, a final “order of removal” is a final order “concluding that the alien is deportable or ordering deportation.” §1101(a)(47)(A).[4] A CAT order is not itself a final order of removal because it is not an order “concluding that the alien is deportable or ordering deportation.” As the Government acknowledges, a CAT order does not disturb the final order of removal. Brief for Respondent 26. An order granting CAT relief means only that, notwithstanding the order of removal, the noncitizen may not be removed to the designated country of removal, at least until conditions change in that country. But the noncitizen still “may be removed at any time to another country where he or she is not likely to be tortured.” 8 CFR §§1208.17(b)(2), 1208.16(f ). Even though CAT orders are not the same as final orders of removal, a question remains: Do CAT orders merge into final orders of removal in the same way as, say, an immigration judge’s evidentiary rulings merge into final orders of removal? The answer is no. For purposes of this statute, final orders of removal encompass only the rulings made by the immigration judge or Board of Immigration Appeals that affect the validity of the final order of removal. As this Court phrased it in INS v. Chadha, review of a final order of removal “includes all matters on which the validity of the final order is contingent.” 462 U.S. 919, 938 (1983) (internal quotation marks omitted). The rulings that affect the validity of the final order of removal merge into the final order of removal for purposes of judicial review. But the immigration judge’s or the Board’s ruling on a CAT claim does not affect the validity of the final order of removal and therefore does not merge into the final order of removal. To be sure, as noted above, FARRA provides that a CAT order is reviewable “as part of the review of a final order of removal” under 8 U. S. C. §1252. §2242(d), 112Stat. 2681–822; see also 8 U. S. C. §1252(a)(4). Likewise, §1252(b)(9) provides that “[j]udicial review of all questions of law and fact . . . arising from any action taken or proceeding brought to remove an alien from the United States under this subchapter shall be available only in judicial review of a final order under this section.” §1252(b)(9). But FARRA and §1252(b)(9) simply establish that a CAT order may be reviewed together with the final order of removal, not that a CAT order is the same as, or affects the validity of, a final order of removal. Consider an analogy. Suppose a statute furnishes appellate review of convictions and sentences in a single appellate proceeding. Suppose that the statute also precludes appellate review of certain factual challenges to the sentence. Would that statute bar appellate review of factual challenges to the conviction, just because the conviction and sentence are reviewed together? No. The same is true here. A CAT order may be reviewed together with the final order of removal. But a CAT order is distinct from a final order of removal and does not affect the validity of the final order of removal. The CAT order therefore does not merge into the final order of removal for purposes of §§1252(a)(2)(C)–(D)’s limitation on the scope of judicial review. In short, as a matter of straightforward statutory interpretation, Congress’s decision to bar judicial review of factual challenges to final orders of removal does not bar judicial review of factual challenges to CAT orders. It would be easy enough for Congress to preclude judicial review of factual challenges to CAT orders, just as Congress has precluded judicial review of factual challenges to certain final orders of removal. But Congress has not done so, and it is not the proper role of the courts to rewrite the laws passed by Congress and signed by the President. C Although a noncitizen may obtain judicial review of factual challenges to CAT orders, that review is highly deferential, as Nasrallah acknowledges. See Reply Brief 19–20; Tr. of Oral Arg. 5. The standard of review is the substantial-evidence standard: The agency’s “findings of fact are conclusive unless any reasonable adjudicator would be compelled to conclude to the contrary.” §1252(b)(4)(B); see Kenyeres v. Ashcroft, 538 U.S. 1301, 1306 (2003) (Kennedy, J., in chambers); INS v. Elias-Zacarias, 502 U.S. 478, 481, n. 1, 483–484 (1992). But the Government still insists that the statute supplies no judicial review of factual challenges to CAT orders. The Government advances a slew of arguments, but none persuades us. First, the Government raises an argument based on precedent. In Foti v. INS, 375 U.S. 217 (1963), this Court interpreted the statutory term “final orders of deportation” in the Immigration and Nationality Act of 1952, as amended in 1961, to encompass “all determinations made during and incident to the administrative proceeding” on removability. Id., at 229. The Government points out (correctly) that the Foti definition of a final order—if it still applied here—would cover CAT orders and therefore would bar judicial review of factual challenges to CAT orders. But Foti’s interpretation of the INA as it existed as of 1963 no longer applies. Since 1996, the INA has defined final “order of deportation” more narrowly than this Court interpreted the term in Foti. A final order of deportation is now defined as a final order “concluding that the alien is deportable or ordering deportation.” 8 U. S. C. §1101(a)(47)(A); Antiterrorism and Effective Death Penalty Act of 1996, 110Stat. 1277; see §309(d)(2) of the Illegal Immigration Reform and Immigrant Responsibility Act of 1996, 110Stat. 3009–627. And as we have explained, an order denying CAT relief does not fall within the statutory definition of an “order of deportation” because it is not an order “concluding that the alien is deportable or ordering deportation.” Therefore, Foti does not control here. Second, the Government puts forward a structural argument. As the Government sees it, if a CAT order is not merged into a final order of removal, then no statute would authorize a court of appeals to review a CAT order in the first place. That is because, in the Government’s view, the only statute that supplies judicial review of CAT claims is the statute that provides for judicial review of final orders of removal. See §1252(a)(1). The premise of that argument is incorrect. Section 2242(d) of FARRA, enacted in 1998, expressly provides for judicial review of CAT claims together with the review of final orders of removal. Moreover, as a result of the 2005 REAL ID Act, §1252(a)(4) now provides for direct review of CAT orders in the courts of appeals. See also 8 U. S. C. §1252(b)(9). In short, our decision does not affect the authority of the courts of appeals to review CAT orders. Third, the Government asserts a congressional intent argument: Why would Congress bar review of factual challenges to a removal order, but allow factual challenges to a CAT order? To begin with, we must adhere to the statutory text, which differentiates between the two kinds of orders for those purposes. In any event, Congress had good reason to distinguish the two. For noncitizens who have committed crimes that subject them to removal, the facts that rendered the noncitizen removable are often not in serious dispute. The relevant facts will usually just be the existence of the noncitizen’s prior criminal convictions. By barring review of factual challenges to final orders of removal, Congress prevented further relitigation of the underlying factual bases for those criminal convictions—a point that Senator Abraham, a key proponent of the statutory bar to judicial review, stressed back in 1996. See 142 Cong. Rec. 7348–7350 (1996). By contrast, the issues related to a CAT order will not typically have been litigated prior to the alien’s removal proceedings. Those factual issues may range from the noncitizen’s past experiences in the designated country of removal, to the noncitizen’s credibility, to the political or other current conditions in that country. Because the factual components of CAT orders will not previously have been litigated in court and because those factual issues may be critical to determining whether the noncitizen is likely to be tortured if returned, it makes some sense that Congress would provide an opportunity for judicial review, albeit deferential judicial review, of the factual components of a CAT order. Fourth, the Government advances a policy argument—that judicial review of the factual components of a CAT order would unduly delay removal proceedings. But today’s decision does not affect whether the noncitizen is entitled to judicial review of a CAT order and does not add a new layer of judicial review. All agree that a noncitizen facing removal under these provisions may already seek judicial review in a court of appeals of constitutional and legal claims relating to both the final order of removal and the CAT order. Our holding today means only that, in that same case in the court of appeals, the court may also review the noncitizen’s factual challenges to the CAT order under the deferential substantial-evidence standard. For many years, the Seventh and Ninth Circuits have allowed factual challenges to CAT orders, and the Government has not informed this Court of any significant problems stemming from review in those Circuits. Fifth, what about the slippery slope? If factual challenges to CAT orders may be reviewed, what other orders will now be subject to factual challenges in the courts of appeals? Importantly, another jurisdiction-stripping provision, §1252(a)(2)(B), states that a noncitizen may not bring a factual challenge to orders denying discretionary relief, including cancellation of removal, voluntary departure, adjustment of status, certain inadmissibility waivers, and other determinations “made discretionary by statute.” Kucana v. Holder, 558 U.S. 233, 248 (2010). Our decision today therefore has no effect on judicial review of those discretionary determinations.[5] The Government suggests that our decision here might lead to judicial review of factual challenges to statutory withholding orders. A statutory withholding order prevents the removal of a noncitizen to a country where the noncitizen’s “life or freedom would be threatened” because of the noncitizen’s “race, religion, nationality, membership in a particular social group, or political opinion.” 8 U. S. C. §1231(b)(3)(A). That question is not presented in this case, and we therefore leave its resolution for another day. * * * In cases where a noncitizen has committed a crime specified in 8 U. S. C. §1252(a)(2)(C), §§1252(a)(2)(C) and (D) preclude judicial review of the noncitizen’s factual challenges to a final order of removal. A CAT order is distinct from a final order of removal and does not affect the validity of a final order of removal. Therefore, §§1252(a)(2)(C) and (D) do not preclude judicial review of a noncitizen’s factual challenges to a CAT order. We reverse the judgment of the U. S. Court of Appeals for the Eleventh Circuit. It is so ordered. Notes 1 To qualify as torture, actions must be “inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity.” 8 CFR §1208.18(a)(1) (2019). 2 This opinion uses the term “noncitizen” as equivalent to the statutory term “alien.” See 8 U. S. C. §1101(a)(3). 3 This case comes to us on the premise that Nasrallah committed a crime specified in §1252(a)(2)(C). That said, courts are divided on the question whether §§1252(a)(2)(C)–(D)’s limitation on judicial review applies when a noncitizen has committed only a single crime of moral turpitude. But that issue is not the question presented in this Court, and we do not address it. Compare Keungne v. United States Attorney General, 561 F.3d 1281, 1283 (CA11 2009), with Yeremin v. Holder, 738 F.3d 708, 713 (CA6 2013); Wanjiru v. Holder, 705 F.3d 258, 262–263 (CA7 2013); Lee v. Gonzales, 410 F.3d 778, 781–782 (CA5 2005). 4 Title 8 U. S. C. §1252(a)(2)(C) provides: “Notwithstanding any other provision of law (statutory or nonstatutory), including section 2241 of title 28, or any other habeas corpus provision, and sections 1361 and 1651 of such title, and except as provided in subparagraph (D), no court shall have jurisdiction to review any final order of removal against an alien who is removable by reason of having committed a criminal offense covered in section 1182(a)(2) or 1227(a)(2)(A)(iii), (B), (C), or (D) of this title, or any offense covered by section 1227(a)(2)(A)(ii) of this title for which both predicate offenses are, without regard to their date of commission, otherwise covered by section 1227(a)(2)(A)(i) of this title.” (Emphasis added.) Section 1252(a)(2)(D) provides: “Nothing in subparagraph (B) or (C), or in any other provision of this chapter (other than this section) which limits or eliminates judicialreview, shall be construed as precluding review of constitutional claims or questions of law raised upon a petition for review filed with anappropriate court of appeals in accordance with this section.” (Emphasis added.) 5 In expedited removal proceedings, the immigration laws do not provide for any judicial review of CAT claims. See 8 U. S. C. §§1225(b)(1)(B)(iii), 1252(a)(2)(A), and 1252(e). Our ruling today does not affect that law.

591.US.2019_19-267

The First Amendment protects the right of religious institutions “to decide for themselves, free from state interference, matters of church government as well as those of faith and doctrine.” Kedroff v. Saint Nicholas Cathedral of Russian Orthodox Church in North America, 344 U.S. 94, 116. Applying this principle, this Court held in Hosanna-Tabor Evangelical Lutheran Church and School v. EEOC, 565 U.S. 171, that the First Amendment barred a court from entertaining an employment discrimination claim brought by an elementary school teacher, Cheryl Perich, against the religious school where she taught. Adopting the so-called “ministerial exception” to laws governing the employment relationship between a religious institution and certain key employees, the Court found relevant Perich’s title as a “Minister of Religion, Commissioned,” her educational training, and her responsibility to teach religion and participate with students in religious activities. Id., at 190–191. In these cases, two elementary school teachers at Roman Catholic schools in the Archdiocese of Los Angeles had teaching responsibilities similar to Perich’s. Agnes Morrissey-Berru taught at Our Lady of Guadalupe School (OLG), and Kristen Biel taught at St. James School. Both were employed under nearly identical agreements that set out the schools’ mission to develop and promote a Catholic School faith community; imposed commitments regarding religious instruction, worship, and personal modeling of the faith; and explained that teachers’ performance would be reviewed on those bases. Each was also required to comply with her school’s faculty handbook, which set out similar expectations. Each taught religion in the classroom, worshipped with her students, prayed with her students, and had her performance measured on religious bases. Both teachers sued their schools after their employment was terminated. Morrissey-Berru claimed that OLG had demoted her and had failed to renew her contract in order to replace her with a younger teacher in violation of the Age Discrimination in Employment Act of 1967. OLG invoked Hosanna-Tabor’s “ministerial exception” and successfully moved for summary judgment, but the Ninth Circuit reversed, holding that Morrissey-Berru did not fall within the exception because she did not have the formal title of “minister,” had limited formal religious training, and did not hold herself out publicly as a religious leader. Biel alleged that St. James discharged her because she had requested a leave of absence to obtain breast cancer treatment. Like OLG, St. James obtained summary judgment under the “ministerial exception.” But the Ninth Circuit reversed, reasoning that Biel lacked Perich’s credentials, religious training, and ministerial background. Held: The First Amendment’s Religion Clauses foreclose the adjudication of Morrissey-Berru’s and Biel’s employment-discrimination claims. Pp. 10–27. (a) The independence of religious institutions in matters of “faith and doctrine” is closely linked to independence in what the Court has termed “ ‘matters of church government.’ ” Hosanna-Tabor, 565 U. S., at 186. For this reason, courts are bound to stay out of employment disputes involving those holding certain important positions with churches and other religious institutions. Pp. 10–11. (b) When the “ministerial exception” reached this Court in Hosanna-Tabor, the Court looked to precedent and the “background” against which “the First Amendment was adopted,” 565 U. S., at 183, and unanimously recognized that the Religion Clauses foreclose certain employment-discrimination claims brought against religious organizations, id., at 188. Pp. 11–14. (c) In Hosanna-Tabor, the Court applied the “ministerial exception” but declined “to adopt a rigid formula for deciding when an employee qualifies as a minister.” 565 U. S., at 190. Instead, the Court identified four relevant circumstances of Perich’s employment at an Evangelical Lutheran school. First, Perich’s church had given her the title of “minister, with a role distinct from that of most of its members.” Id., at 191. Second, her position “reflected a significant degree of religious training followed by a formal process of commissioning.” Ibid. Third, she “held herself out as a minister of the Church” and claimed certain tax benefits. Id., at 191–192. Fourth, her “job duties reflected a role in conveying the Church’s message and carrying out its mission.” Id., at 192. Pp. 14–16. (d) A variety of factors may be important in determining whether a particular position falls within the ministerial exception. The circumstances that informed the Court’s decision in Hosanna-Tabor were relevant because of their relationship to Perich’s “role in conveying the Church’s message and carrying out its mission.” 565 U. S., at 192. But the recognition of the significance of those factors in Perich’s case did not mean that they must be met in all other cases. What matters is what an employee does. Implicit in the Hosanna-Tabor decision was a recognition that educating young people in their faith, inculcating its teachings, and training them to live their faith are responsibilities that lie at the very core of a private religious school’s mission. Pp. 16–21. (e) Applying this understanding of the Religion Clauses here, it is apparent that Morrissey-Berru and Biel qualify for the exception recognized in Hosanna-Tabor. There is abundant record evidence that they both performed vital religious duties, such as educating their students in the Catholic faith and guiding their students to live their lives in accordance with that faith. Their titles did not include the term “minister” and they had less formal religious training than Perich, but their core responsibilities were essentially the same. And their schools expressly saw them as playing a vital role in carrying out the church’s mission. A religious institution’s explanation of the role of its employees in the life of the religion in question is important. Pp. 21–22. (f) The Ninth Circuit mistakenly treated the circumstances the Court found relevant in Hosanna-Tabor as a checklist of items to be assessed and weighed against each other. That rigid test produced a distorted analysis. First, it invested undue significance in the fact that Morrissey-Berru and Biel did not have clerical titles. Second, it assigned too much weight to the fact that Morrissey-Berru and Biel had less formal religious schooling that Perich. Third, the St. James panel inappropriately diminished the significance of Biel’s duties. Respondents would make Hosanna-Tabor’s governing test even more rigid. And they go further astray in suggesting that an employee can never come within the Hosanna-Tabor exception unless the employee is a “practicing” member of the religion with which the employer is associated. Deciding such questions risks judicial entanglement in religious issues. Pp. 22–27. No. 19–267, 769 Fed. Appx. 460; No. 19–348, 911 F.3d 603, reversed and remanded. Alito, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Breyer, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., filed a concurring opinion, in which Gorsuch, J., joined. Sotomayor, J., filed a dissenting opinion, in which Ginsburg, J., joined. Notes 1 Together with No. 19–348, St. James School v. Biel, as Personal Representative of the Estate of Biel, on certiorari to the same Court.

These cases require us to decide whether the First Amendment permits courts to intervene in employment disputes involving teachers at religious schools who are entrusted with the responsibility of instructing their students in the faith. The First Amendment protects the right of religious institutions “to decide for themselves, free from state interference, matters of church government as well as those of faith and doctrine.” Kedroff v. Saint Nicholas Cathedral of Russian Orthodox Church in North America, 344 U.S. 94, 116 (1952). Applying this principle, we held in Hosanna-Tabor Evangelical Lutheran Church and School v. EEOC, 565 U.S. 171 (2012), that the First Amendment barred a court from entertaining an employment discrimination claim brought by an elementary school teacher, Cheryl Perich, against the religious school where she taught. Our decision built on a line of lower court cases adopting what was dubbed the “ministerial exception” to laws governing the employment relationship between a religious institution and certain key employees. We did not announce “a rigid formula” for determining whether an employee falls within this exception, but we identified circumstances that we found relevant in that case, including Perich’s title as a “Minister of Religion, Commissioned,” her educational training, and her responsibility to teach religion and participate with students in religious activities. Id., at 190–191. In the cases now before us, we consider employment discrimination claims brought by two elementary school teachers at Catholic schools whose teaching responsibilities are similar to Perich’s. Although these teachers were not given the title of “minister” and have less religious training than Perich, we hold that their cases fall within the same rule that dictated our decision in Hosanna-Tabor. The religious education and formation of students is the very reason for the existence of most private religious schools, and therefore the selection and supervision of the teachers upon whom the schools rely to do this work lie at the core of their mission. Judicial review of the way in which religious schools discharge those responsibilities would undermine the independence of religious institutions in a way that the First Amendment does not tolerate. I A 1 The first of the two cases we now decide involves Agnes Morrissey-Berru, who was employed at Our Lady of Guadalupe School (OLG), a Roman Catholic primary school in the Archdiocese of Los Angeles. Excerpts of Record (ER) 58 in No. 17–56624 (CA9) (OLG).[1] For many years, Morrissey-Berru was employed at OLG as a lay fifth or sixth grade teacher. Like most elementary school teachers, she taught all subjects, and since OLG is a Catholic school, the curriculum included religion. App. 23, 75. As a result, she was her students’ religion teacher. Morrissey-Berru earned a B. A. in English Language Arts, with a minor in secondary education, and she holds a California teaching credential. Id., at 21–22. While on the faculty at OLG, she took religious education courses at the school’s request, ER 41–ER 42, ER 44–ER 45, ER 276, and was expected to attend faculty prayer services, App. to Pet. for Cert. in No. 19–267, p. 87a.[2] Each year, Morrissey-Berru and OLG entered into an employment agreement, App. 21,[3] that set out the school’s “mission” and Morrissey-Berru’s duties. See, e.g., id., at 154–164.[4] The agreement stated that the school’s mission was “to develop and promote a Catholic School Faith Community,” id., at 154, and it informed Morrissey-Berru that “[a]ll [her] duties and responsibilities as a Teache[r were to] be performed within this overriding commitment.” Ibid. The agreement explained that the school’s hiring and retention decisions would be guided by its Catholic mission, and the agreement made clear that teachers were expected to “model and promote” Catholic “faith and morals.” Id., at 155. Under the agreement, Morrissey-Berru was required to participate in “[s]chool liturgical activities, as requested,” ibid., and the agreement specified that she could be terminated “for ‘cause’ ” for failing to carry out these duties or for “conduct that brings discredit upon the School or the Roman Catholic Church.” Id., at 155–157. The agreement required compliance with the faculty handbook, which sets out similar expectations. Id., at 156; App. to Pet. for Cert. in No. 19–267, at 52a–55a. The pastor of the parish, a Catholic priest, had to approve Morrissey-Berru’s hiring each year. Id., at 14a; see also App. 164. Like all teachers in the Archdiocese of Los Angeles, Morrissey-Berru was “considered a catechist,” i.e., “a teacher of religio[n].” App. to Pet. for Cert. in No. 19–267, at 56a, 60a. Catechists are “responsible for the faith formation of the students in their charge each day.” Id., at 56a. Morrissey-Berru provided religious instruction every day using a textbook designed for use in teaching religion to young Catholic students. Id., at 45a–51a, 90a–92a; see App. 79–80. Under the prescribed curriculum, she was expected to teach students, among other things, “to learn and express belief that Jesus is the son of God and the Word made flesh”; to “identify the ways” the church “carries on the mission of Jesus”; to “locate, read and understand stories from the Bible”; to “know the names, meanings, signs and symbols of each of the seven sacraments”; and to be able to “explain the communion of saints.” App. to Pet. for Cert. in No. 19–267, at 91a–92a. She tested her students on that curriculum in a yearly exam. Id., at 87a. She also directed and produced an annual passion play. Id., at 26a. Morrissey-Berru prepared her students for participation in the Mass and for communion and confession. Id., at 68a, 81a, 88a–89a. She also occasionally selected and prepared students to read at Mass. Id., at 83a, 89a. And she was expected to take her students to Mass once a week and on certain feast days (such as the Feast Day of St. Juan Diego, All Saints Day, and the Feast of Our Lady), and to take them to confession and to pray the Stations of the Cross. Id., at 68a–69a, 83a, 88a. Each year, she brought them to the Catholic Cathedral in Los Angeles, where they participated as altar servers. Id., at 95a–96a. This visit, she explained, was “an important experience” because “[i]t is a big honor” for children to “serve the altar” at the cathedral. Id., at 96a. Morrissey-Berru also prayed with her students. Her class began or ended every day with a Hail Mary. Id., at 87a. She led the students in prayer at other times, such as when a family member was ill. Id., at 21a, 81a, 86a–87a. And she taught them to recite the Apostle’s Creed and the Nicene Creed, as well as prayers for specific purposes, such as in connection with the sacrament of confession. Id., at 20a–21a, 92a. The school reviewed Morrissey-Berru’s performance under religious standards. The “ ‘Classroom Observation Report’ ” evaluated whether Catholic values were “infused through all subject areas” and whether there were religious signs and displays in the classroom. Id., at 94a, 95a; App. 59. Morrissey-Berru testified that she tried to instruct her students “in a manner consistent with the teachings of the Church,” App. to Pet. for Cert. in No. 19–267, at 96a, and she said that she was “committed to teaching children Catholic values” and providing a “faith-based education.” Id., at 82a. And the school principal confirmed that Morrissey-Berru was expected to do these things.[5] 2 In 2014, OLG asked Morrissey-Berru to move from a full-time to a part-time position, and the next year, the school declined to renew her contract. She filed a claim with the Equal Employment Opportunity Commission (EEOC), received a right-to-sue letter, App. 169, and then filed suit under the Age Discrimination in Employment Act of 1967, 81Stat. 602, as amended, 29 U. S. C. §621 et seq., claiming that the school had demoted her and had failed to renew her contract so that it could replace her with a younger teacher. App. 168–169. The school maintains that it based its decisions on classroom performance—specifically, Morrissey-Berru’s difficulty in administering a new reading and writing program, which had been introduced by the school’s new principal as part of an effort to maintain accreditation and improve the school’s academic program. App. to Pet. for Cert. in No. 19–267, at 66a–67a, 70a, 73a. Invoking the “ministerial exception” that we recognized in Hosanna-Tabor, OLG successfully moved for summary judgment, but the Ninth Circuit reversed in a brief opinion. 769 Fed. Appx. 460, 461 (2019). The court acknowledged that Morrissey-Berru had “significant religious responsibilities” but reasoned that “an employee’s duties alone are not dispositive under Hosanna-Tabor’s framework.” Ibid. Unlike Perich, the court noted, Morrissey-Berru did not have the formal title of “minister,” had limited formal religious training, and “did not hold herself out to the public as a religious leader or minister.” Ibid. In the court’s view, these “factors” outweighed the fact that she was invested with significant religious responsibilities. Ibid. The court therefore held that Morrissey-Berru did not fall within the “ministerial exception.” OLG filed a petition for certiorari, and we granted review. B 1 The second case concerns the late Kristen Biel, who worked for about a year and a half as a lay teacher at St. James School, another Catholic primary school in Los Angeles. For part of one academic year, Biel served as a long-term substitute teacher for a first grade class, and for one full year she was a full-time fifth grade teacher. App. 336–337. Like Morrissey-Berru, she taught all subjects, including religion. Id., at 288; ER 588 in No. 17–55180 (CA9) (St. James).[6] Biel had a B. A. in liberal studies and a teaching credential. App. 244. During her time at St. James, she attended a religious conference that imparted “[d]ifferent techniques on teaching and incorporating God” into the classroom. Id., at 260–262. Biel was Catholic.[7] Biel’s employment agreement was in pertinent part nearly identical to Morrissey-Berru’s. Compare id., at 154–164, with id., at 320–329. The agreement set out the same religious mission; required teachers to serve that mission; imposed commitments regarding religious instruction, worship, and personal modeling of the faith; and explained that teachers’ performance would be reviewed on those bases. Biel’s agreement also required compliance with the St. James faculty handbook, which resembles the OLG handbook. Id., at 322. Compare ER 641–ER 651 (OLG) with ER 565–ER 597 (St. James). The St. James handbook defines “religious development” as the school’s first goal and provides that teachers must “mode[l] the faith life,” “exemplif[y] the teachings of Jesus Christ,” “integrat[e] Catholic thought and principles into secular subjects,” and “prepar[e] students to receive the sacraments.” Id., at ER 570–ER 572. The school principal confirmed these expectations.[8] Like Morrissey-Berru, Biel instructed her students in the tenets of Catholicism. She was required to teach religion for 200 minutes each week, App. 257–258, and administered a test on religion every week, id., at 256–257. She used a religion textbook selected by the school’s principal, a Catholic nun. Id., at 255; ER 37 (St. James). The religious curriculum covered “the norms and doctrines of the Catholic Faith, including . . . the sacraments of the Catholic Church, social teachings according to the Catholic Church, morality, the history of Catholic saints, [and] Catholic prayers.” App. to Pet. for Cert. in No. 19–348, p. 83a. Biel worshipped with her students. At St. James, teachers are responsible for “prepar[ing] their students to be active participants at Mass, with particular emphasis on Mass responses,” ER 587, and Biel taught her students about “Catholic practices like the Eucharist and confession,” id., at ER 226–ER 227. At monthly Masses, she prayed with her students. App. to Pet. for Cert. in No. 19–348, at 82a, 94a–96a. Her students participated in the liturgy on some occasions by presenting the gifts (bringing bread and wine to the priest). Ibid. Teachers at St. James were “required to pray with their students every day,” id., at 80a–81a, 110a, and Biel observed this requirement by opening and closing each school day with prayer, including the Lord’s Prayer or a Hail Mary, id., at 81a–82a, 93a, 110a. As at OLG, teachers at St. James are evaluated on their fulfillment of the school’s religious mission. Id., at 83a–84a. St. James used the same classroom observation standards as OLG and thus examined whether teachers “infus[ed]” Catholic values in all their teaching and included religious displays in their classrooms. Id., at 83a–84a, 92a. The school’s principal, a Catholic nun, evaluated Biel on these measures. Id., at 106a. 2 St. James declined to renew Biel’s contract after one full year at the school. She filed charges with the EEOC, and after receiving a right-to-sue letter, brought this suit, alleging that she was discharged because she had requested a leave of absence to obtain treatment for breast cancer. App. 337–338. The school maintains that the decision was based on poor performance—namely, a failure to observe the planned curriculum and keep an orderly classroom. See id., at 303; App. to Pet. for Cert. in No. 19–348, at 85a–89a, 114a–115a, 120a–121a. Like OLG, St. James obtained summary judgment under the ministerial exception, id., at 74a, but a divided panel of the Ninth Circuit reversed, reasoning that Biel lacked Perich’s “credentials, training, [and] ministerial background,” 911 F.3d 603, 608 (2018). Judge D. Michael Fisher, sitting by designation, dissented. Considering the totality of the circumstances, he would have held that the ministerial exception applied “because of the substance reflected in [Biel’s] title and the important religious functions she performed” as a “stewar[d] of the Catholic faith to the children in her class.” Id., at 621, 622. An unsuccessful petition for rehearing en banc ensued. Judge Ryan D. Nelson, joined by eight other judges, dissented. 926 F.3d 1238, 1239 (2019). Judge Nelson faulted the panel majority for “embrac[ing] the narrowest construction” of the ministerial exception, departing from “the consensus of our sister circuits that the employee’s ministerial function should be the key focus,” and demanding nothing less than a “carbon copy” of the specific facts in Hosanna-Tabor. Ibid. We granted review and consolidated the case with OLG’s. 589 U. S. ___ (2019). II A The First Amendment provides that “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof.” Among other things, the Religion Clauses protect the right of churches and other religious institutions to decide matters “ ‘of faith and doctrine’ ” without government intrusion. Hosanna-Tabor, 565 U. S., at 186 (quoting Kedroff, 344 U. S., at 116). State interference in that sphere would obviously violate the free exercise of religion, and any attempt by government to dictate or even to influence such matters would constitute one of the central attributes of an establishment of religion. The First Amendment outlaws such intrusion. The independence of religious institutions in matters of “faith and doctrine” is closely linked to independence in what we have termed “ ‘matters of church government.’ ” 565 U. S., at 186. This does not mean that religious institutions enjoy a general immunity from secular laws, but it does protect their autonomy with respect to internal management decisions that are essential to the institution’s central mission. And a component of this autonomy is the selection of the individuals who play certain key roles. The “ministerial exception” was based on this insight. Under this rule, courts are bound to stay out of employment disputes involving those holding certain important positions with churches and other religious institutions. The rule appears to have acquired the label “ministerial exception” because the individuals involved in pioneering cases were described as “ministers.” See McClure v. Salvation Army, 460 F.2d 553, 558–559 (CA5 1972); Rayburn v. General Conference of Seventh-day Adventists, 772 F.2d 1164, 1168 (CA4 1985). Not all pre-Hosanna-Tabor decisions applying the exception involved “ministers” or even members of the clergy. See, e.g., EEOC v. Southwestern Baptist Theological Seminary, 651 F.2d 277, 283–284 (CA5 1981); EEOC v. Roman Catholic Diocese of Raleigh, N. C., 213 F.3d 795, 800–801 (CA4 2000). But it is instructive to consider why a church’s independence on matters “of faith and doctrine” requires the authority to select, supervise, and if necessary, remove a minister without interference by secular authorities. Without that power, a wayward minister’s preaching, teaching, and counseling could contradict the church’s tenets and lead the congregation away from the faith.[9] The ministerial exception was recognized to preserve a church’s independent authority in such matters. B When the so-called ministerial exception finally reached this Court in Hosanna-Tabor, we unanimously recognized that the Religion Clauses foreclose certain employment discrimination claims brought against religious organizations. 565 U. S., at 188. The constitutional foundation for our holding was the general principle of church autonomy to which we have already referred: independence in matters of faith and doctrine and in closely linked matters of internal government. The three prior decisions on which we primarily relied drew on this broad principle, and none was exclusively concerned with the selection or supervision of clergy. Watson v. Jones, 13 Wall. 679 (1872), involved a dispute about the control of church property, and both Kedroff, 344 U.S. 94, and Serbian Eastern Orthodox Diocese for United States and Canada v. Milivojevich, 426 U.S. 696 (1976), also concerned the control of property, as well as the appointment and authority of bishops. In addition to these precedents, we looked to the “background” against which “the First Amendment was adopted.” Hosanna-Tabor, 565 U. S., at 183. We noted that 16th-century British statutes had given the Crown the power to fill high “religious offices” and to control the exercise of religion in other ways, and we explained that the founding generation sought to prevent a repetition of these practices in our country. Ibid. Because Cheryl Perich, the teacher in Hosanna-Tabor, had a title that included the word “minister,” we naturally concentrated on historical events involving clerical offices, but the abuses we identified were not limited to the control of appointments. We pointed to the various Acts of Uniformity, id., at 182, which dictated what ministers could preach and imposed penalties for non-compliance. Under the 1549 Act, a minister who “preach[ed,] declare[d,] or [spoke] any thing” in derogation of any part of the Book of Common Prayer could be sentenced to six months in jail for a first offense and life imprisonment for a third violation. Act of Uniformity, 2 & 3 Edw. 6, ch. 1. In addition, all other English subjects were forbidden to say anything against the Book of Common Prayer in “[i]nterludes[,] play[s,] song[s,] r[h]ymes, or by other open [w]ord[s].” Ibid. A 1559 law contained similar prohibitions. See Act of Uniformity, 1 Eliz., ch. 2. After the Restoration, Parliament enacted a new law with a similar aim. Ministers and “Lecturer[s]” were required to pledge “unfeigned assent and consent” to the Book of Common Prayer, and all schoolmasters, private tutors, and university professors were required to “conforme to the Liturgy of the Church of England” and not “to endeavour any change or alteration” of the church. Act of Uniformity, 1662, 14 Car. 2, ch. 4. British law continued to impose religious restrictions on education in the 18th century and past the time of the adoption of the First Amendment. The Schism or Established Church Act of 1714, 13 Ann., ch. 7, required that schoolmasters and tutors be licensed by a bishop. Non-conforming Protestants, as well as Catholics and Jews, could not teach at or attend the two universities, and as Blackstone wrote, “[p]ersons professing the popish religion [could] not keep or teach any school under pain of perpetual imprisonment.” 4 W. Blackstone, Commentaries on the Laws of England 55 (8th ed. 1778). The law also imposed penalties on “any person [who] sen[t] another abroad to be educated in the popish religion . . . or [who] contribute[d] to their maintenance when there.” Id., at 55–56. British colonies in North America similarly controlled both the appointment of clergy, see Hosanna-Tabor, 565 U. S., at 183, and the teaching of students. A Maryland law “prohibited any Catholic priest or lay person from keeping school, or taking upon himself the education of youth.” 2 T. Hughes, History of the Society of Jesus in North America: Colonial and Federal 443–444 (1917). In 1771, the Governor of New York was instructed to require that all schoolmasters arriving from England obtain a license from the Bishop of London. 3 C. Lincoln, The Constitutional History of New York 485, 745 (1906). New York law also required an oath and license for any “ ‘vagrant Preacher, Moravian, or disguised Papist’ ” to “ ‘Preach or Teach, Either in Public or Private.’ ” S. Cobb, The Rise of Religious Liberty in America 358 (1902). C In Hosanna-Tabor, Cheryl Perich, a kindergarten and fourth grade teacher at an Evangelical Lutheran school, filed suit in federal court, claiming that she had been discharged because of a disability, in violation of the Americans with Disabilities Act of 1990 (ADA), 42 U. S. C. §12112(a). The school responded that the real reason for her dismissal was her violation of the Lutheran doctrine that disputes should be resolved internally and not by going to outside authorities. We held that her suit was barred by the “ministerial exception” and noted that it “concern[ed] government interference with an internal church decision that affects the faith and mission of the church.” 565 U. S., at 190. We declined “to adopt a rigid formula for deciding when an employee qualifies as a minister,” and we added that it was “enough for us to conclude, in this our first case involving the ministerial exception, that the exception covers Perich, given all the circumstances of her employment.” Id., at 190–191. We identified four relevant circumstances but did not highlight any as essential. First, we noted that her church had given Perich the title of “minister, with a role distinct from that of most of its members.” Id., at 191. Although she was not a minister in the usual sense of the term—she was not a pastor or deacon, did not lead a congregation, and did not regularly conduct religious services—she was classified as a “called” teacher, as opposed to a lay teacher, and after completing certain academic requirements, was given the formal title “ ‘Minister of Religion, Commissioned.’ ” Id., at 177–178, 191. Second, Perich’s position “reflected a significant degree of religious training followed by a formal process of commissioning.” Id., at 191. Third, “Perich held herself out as a minister of the Church by accepting the formal call to religious service, according to its terms,” and by claiming certain tax benefits. Id., at 191–192. Fourth, “Perich’s job duties reflected a role in conveying the Church’s message and carrying out its mission.” Id., at 192. The church charged her with “ ‘lead[ing] others toward Christian maturity’ ” and “ ‘teach[ing] faithfully the Word of God, the Sacred Scriptures, in its truth and purity and as set forth in all the symbolical books of the Evangelical Lutheran Church.’ ” Ibid. Although Perich also provided instruction in secular subjects, she taught religion four days a week, led her students in prayer three times a day, took her students to a chapel service once a week, and participated in the liturgy twice a year. “As a source of religious instruction,” we explained, “Perich performed an important role in transmitting the Lutheran faith to the next generation.” Ibid. The case featured two concurrences. In the first, Justice Thomas stressed that courts should “defer to a religious organization’s good-faith understanding of who qualifies as its minister.” Id., at 196. That is so, Justice Thomas explained, because “[a] religious organization’s right to choose its ministers would be hollow . . . if secular courts could second-guess” the group’s sincere application of its religious tenets. Id., at 197. The second concurrence argued that application of the “ministerial exception” should “focus on the function performed by persons who work for religious bodies” rather than labels or designations that may vary across faiths. Id., at 198 (opinion of Alito, J., joined by Kagan, J.). This opinion viewed the title of “minister” as “relevant” but “neither necessary nor sufficient.” Id., at 202. It noted that “most faiths do not employ the term ‘minister’ ” and that some “consider the ministry to consist of all or a very large percentage of their members.” Ibid. The opinion concluded that the “ ‘ministerial’ exception” “should apply to any ‘employee’ who leads a religious organization, conducts worship services or important religious ceremonies or rituals, or serves as a messenger or teacher of its faith.” Id., at 199. D 1 In determining whether a particular position falls within the Hosanna-Tabor exception, a variety of factors may be important.[10] The circumstances that informed our decision in Hosanna-Tabor were relevant because of their relationship to Perich’s “role in conveying the Church’s message and carrying out its mission,” id., at 192, but the other noted circumstances also shed light on that connection. In a denomination that uses the term “minister,” conferring that title naturally suggests that the recipient has been given an important position of trust. In Perich’s case, the title that she was awarded and used demanded satisfaction of significant academic requirements and was conferred only after a formal approval process, id., at 191, and those circumstances also evidenced the importance attached to her role, ibid. But our recognition of the significance of those factors in Perich’s case did not mean that they must be met—or even that they are necessarily important—in all other cases. Take the question of the title “minister.” Simply giving an employee the title of “minister” is not enough to justify the exception. And by the same token, since many religious traditions do not use the title “minister,” it cannot be a necessary requirement. Requiring the use of the title would constitute impermissible discrimination, and this problem cannot be solved simply by including positions that are thought to be the counterparts of a “minister,” such as priests, nuns, rabbis, and imams. See Brief for Respondents 21. Nuns are not the same as Protestant ministers. A brief submitted by Jewish organizations makes the point that “Judaism has many ‘ministers,’ ” that is, “the term ‘minister’ encompasses an extensive breadth of religious functionaries in Judaism.”[11] For Muslims, “an inquiry into whether imams or other leaders bear a title equivalent to ‘minister’ can present a troubling choice between denying a central pillar of Islam—i.e., the equality of all believers—and risking loss of ministerial exception protections.”[12] If titles were all-important, courts would have to decide which titles count and which do not, and it is hard to see how that could be done without looking behind the titles to what the positions actually entail. Moreover, attaching too much significance to titles would risk privileging religious traditions with formal organizational structures over those that are less formal. For related reasons, the academic requirements of a position may show that the church in question regards the position as having an important responsibility in elucidating or teaching the tenets of the faith. Presumably the purpose of such requirements is to make sure that the person holding the position understands the faith and can explain it accurately and effectively. But insisting in every case on rigid academic requirements could have a distorting effect. This is certainly true with respect to teachers. Teaching children in an elementary school does not demand the same formal religious education as teaching theology to divinity students. Elementary school teachers often teach secular subjects in which they have little if any special training. In addition, religious traditions may differ in the degree of formal religious training thought to be needed in order to teach. See, e.g., Brief for Ethics and Religious Liberty Commission of the Southern Baptist Convention et al. as Amici Curiae 12 (“many Protestant groups have historically rejected any requirement of formal theological training”). In short, these circumstances, while instructive in Hosanna-Tabor, are not inflexible requirements and may have far less significance in some cases. What matters, at bottom, is what an employee does. And implicit in our decision in Hosanna-Tabor was a recognition that educating young people in their faith, inculcating its teachings, and training them to live their faith are responsibilities that lie at the very core of the mission of a private religious school. As we put it, Perich had been entrusted with the responsibility of “transmitting the Lutheran faith to the next generation.” 565 U. S., at 192. One of the concurrences made the same point, concluding that the exception should include “any ‘employee’ who leads a religious organization, conducts worship services or important religious ceremonies or rituals, or serves as a messenger or teacher of its faith.” Id., at 199 (opinion of Alito, J.) (emphasis added). Religious education is vital to many faiths practiced in the United States. This point is stressed by briefs filed in support of OLG and St. James by groups affiliated with a wide array of faith traditions. In the Catholic tradition, religious education is “ ‘intimately bound up with the whole of the Church’s life.’ ” Catechism of the Catholic Church 8 (2d ed. 2016). Under canon law, local bishops must satisfy themselves that “those who are designated teachers of religious instruction in schools . . . are outstanding in correct doctrine, the witness of a Christian life, and teaching skill.” Code of Canon Law, Canon 804, §2 (Eng. transl. 1998). Similarly, Protestant churches, from the earliest settlements in this country, viewed education as a religious obligation. A core belief of the Puritans was that education was essential to thwart the “chief project of that old deluder, Satan, to keep men from the knowledge of the Scriptures.”[13] Thus, in 1647, the Massachusetts General Court passed what has been called the Old Deluder Satan Act requiring every sizable town to establish a school.[14] Most of the oldest educational institutions in this country were originally established by or affiliated with churches, and in recent years, non-denominational Christian schools have proliferated with the aim of inculcating Biblical values in their students.[15] Many such schools expressly set themselves apart from public schools that they believe do not reflect their values.[16] Religious education is a matter of central importance in Judaism. As explained in briefs submitted by Jewish organizations, the Torah is understood to require Jewish parents to ensure that their children are instructed in the faith.[17] One brief quotes Maimonides’s statement that religious instruction “is an obligation of the highest order, entrusted only to a schoolteacher possessing ‘fear of Heaven.’ ”[18] “The contemporary American Jewish community continues to place the education of children in its faith and rites at the center of its communal efforts.”[19] Religious education is also important in Islam. “[T]he acquisition of at least rudimentary knowledge of religion and its duties [is] mandatory for the Muslim individual.”[20] This precept is traced to the Prophet Muhammad, who proclaimed that “ ‘[t]he pursuit of knowledge is incumbent on every Muslim.’ ”[21] “[T]he development of independent private Islamic schools ha[s] become an important part of the picture of Muslim education in America.”[22] The Church of Jesus Christ of Latter-day Saints has a long tradition of religious education, with roots in revelations given to Joseph Smith. See Doctrine and Covenants of the Church of Jesus Christ of Latter-day Saints §93:36 (2013). “The Church Board of Education has established elementary, middle, or secondary schools in which both secular and religious instruction is offered.”[23] Seventh-day Adventists “trace the importance of education back to the Garden of Eden.”[24] Seventh-day Adventist formation “restore[s] human beings into the image of God as revealed by the life of Jesus Christ” and focuses on the development of “knowledge, skills, and understandings to serve God and humanity.”[25] This brief survey does not do justice to the rich diversity of religious education in this country, but it shows the close connection that religious institutions draw between their central purpose and educating the young in the faith. 2 When we apply this understanding of the Religion Clauses to the cases now before us, it is apparent that Morrissey-Berru and Biel qualify for the exemption we recognized in Hosanna-Tabor. There is abundant record evidence that they both performed vital religious duties. Educating and forming students in the Catholic faith lay at the core of the mission of the schools where they taught, and their employment agreements and faculty handbooks specified in no uncertain terms that they were expected to help the schools carry out this mission and that their work would be evaluated to ensure that they were fulfilling that responsibility. As elementary school teachers responsible for providing instruction in all subjects, including religion, they were the members of the school staff who were entrusted most directly with the responsibility of educating their students in the faith. And not only were they obligated to provide instruction about the Catholic faith, but they were also expected to guide their students, by word and deed, toward the goal of living their lives in accordance with the faith. They prayed with their students, attended Mass with the students, and prepared the children for their participation in other religious activities. Their positions did not have all the attributes of Perich’s. Their titles did not include the term “minister,” and they had less formal religious training, but their core responsibilities as teachers of religion were essentially the same. And both their schools expressly saw them as playing a vital part in carrying out the mission of the church, and the schools’ definition and explanation of their roles is important. In a country with the religious diversity of the United States, judges cannot be expected to have a complete understanding and appreciation of the role played by every person who performs a particular role in every religious tradition. A religious institution’s explanation of the role of such employees in the life of the religion in question is important. III In holding that Morrissey-Berru and Biel did not fall within the Hosanna-Tabor exception, the Ninth Circuit misunderstood our decision. Both panels treated the circumstances that we found relevant in that case as checklist items to be assessed and weighed against each other in every case, and the dissent does much the same. That approach is contrary to our admonition that we were not imposing any “rigid formula.” 565 U. S., at 190. Instead, we called on courts to take all relevant circumstances into account and to determine whether each particular position implicated the fundamental purpose of the exception.[26] The Ninth Circuit’s rigid test produced a distorted analysis. First, it invested undue significance in the fact that Morrissey-Berru and Biel did not have clerical titles. 769 Fed. Appx., at 460; 911 F. 3d, at 608–609; Post, at 15–16. It is true that Perich’s title included the term “minister,” but we never said that her title (or her reference to herself as a “minister”) was necessary to trigger the Hosanna-Tabor exception. Instead, “those considerations . . . merely made Perich’s case an especially easy one.” Brief for United States as Amicus Curiae 19. Moreover, both Morrissey-Berru and Biel had titles. They were Catholic elementary school teachers, which meant that they were their students’ primary teachers of religion. The concept of a teacher of religion is loaded with religious significance. The term “rabbi” means teacher, and Jesus was frequently called rabbi.[27] And if a more esoteric title is needed, they were both regarded as “catechists.”[28] Second, the Ninth Circuit assigned too much weight to the fact that Morrissey-Berru and Biel had less formal religious schooling than Perich. 769 Fed. Appx., at 460–461; 911 F. 3d, at 608; post, at 16–17. The significance of formal training must be evaluated in light of the age of the students taught and the judgment of a religious institution regarding the need for formal training. The schools in question here thought that Morrissey-Berru and Biel had a sufficient understanding of Catholicism to teach their students,[29] and judges have no warrant to second-guess that judgment or to impose their own credentialing requirements. Third, the St. James panel inappropriately diminished the significance of Biel’s duties because they did not evince “close guidance and involvement” in “students’ spiritual lives.” 911 F. 3d, at 609; post, at 12, 17–18. Specifically, the panel majority suggested that Biel merely taught “religion from a book required by the school,” “joined” students in prayer, and accompanied students to Mass in order to keep them “ ‘quiet and in their seats.’ ” 911 F. 3d, at 609. This misrepresents the record and its significance. For better or worse, many primary school teachers tie their instruction closely to textbooks, and many faith traditions prioritize teaching from authoritative texts. See Brief for InterVarsity Christian Fellowship USA et al. as Amici Curiae 26; Brief for Senator Mike Lee et al. as Amici Curiae 24–27. As for prayer, Biel prayed with her students, taught them prayers, and supervised the prayers led by students. She prepared them for Mass, accompanied them to Mass, and prayed with them there. See supra, at 8–9. In Biel’s appeal, the Ninth Circuit suggested that the Hosanna-Tabor exception should be interpreted narrowly because the ADA, 42 U. S. C. §12101 et seq., and Title VII, §2000e–2, contain provisions allowing religious employers to give preference to members of a particular faith in employing individuals to do work connected with their activities. 911 F. 3d, at 611, n. 5; post, at 2–3. But the Hosanna-Tabor exception serves an entirely different purpose. Think of the quintessential case where a church wants to dismiss its minister for poor performance. The church’s objection in that situation is not that the minister has gone over to some other faith but simply that the minister is failing to perform essential functions in a satisfactory manner. While the Ninth Circuit treated the circumstances that we cited in Hosanna-Tabor as factors to be assessed and weighed in every case, respondents would make the governing test even more rigid. In their view, courts should begin by deciding whether the first three circumstances—a ministerial title, formal religious education, and the employee’s self-description as a minister—are met and then, in order to check the conclusion suggested by those factors, ask whether the employee performed a religious function. Brief for Respondents 20–24. For reasons already explained, there is no basis for treating the circumstances we found relevant in Hosanna-Tabor in such a rigid manner. Respondents go further astray in suggesting that an employee can never come within the Hosanna-Tabor exception unless the employee is a “practicing” member of the religion with which the employer is associated. Brief for Respondents 12–13, 21. In hiring a teacher to provide religious instruction, a religious school is very likely to try to select a person who meets this requirement, but insisting on this as a necessary condition would create a host of problems. As pointed out by petitioners, determining whether a person is a “co-religionist” will not always be easy. See Reply Brief 14 (“Are Orthodox Jews and non-Orthodox Jews co- religionists? . . . Would Presbyterians and Baptists be similar enough? Southern Baptists and Primitive Baptists?”). Deciding such questions would risk judicial entanglement in religious issues. Expanding the “co-religionist” requirement, Brief for Respondents 28–29, 44, to exclude those who no longer practice the faith would be even worse, post, at 13. Would the test depend on whether the person in question no longer considered himself or herself to be a member of a particular faith? Or would the test turn on whether the faith tradition in question still regarded the person as a member in some sense? Respondents argue that Morrissey-Berru cannot fall within the Hosanna-Tabor exception because she said in connection with her lawsuit that she was not “a practicing Catholic,” but acceptance of that argument would require courts to delve into the sensitive question of what it means to be a “practicing” member of a faith, and religious employers would be put in an impossible position. Morrissey-Berru’s employment agreements required her to attest to “good standing” with the church. See App. 91, 144, 154. Beyond insisting on such an attestation, it is not clear how religious groups could monitor whether an employee is abiding by all religious obligations when away from the job. Was OLG supposed to interrogate Morrissey-Berru to confirm that she attended Mass every Sunday? Respondents argue that the Hosanna-Tabor exception is not workable unless it is given a rigid structure, but we declined to adopt a “rigid formula” in Hosanna-Tabor, and the lower courts have been applying the exception for many years without such a formula. Here, as in Hosanna-Tabor, it is sufficient to decide the cases before us. When a school with a religious mission entrusts a teacher with the responsibility of educating and forming students in the faith, judicial intervention into disputes between the school and the teacher threatens the school’s independence in a way that the First Amendment does not allow. * * * For these reasons, the judgment of the Court of Appeals in each case is reversed, and the cases are remanded for proceedings consistent with this opinion. It is so ordered. Notes 1 A major theme of the dissent is that we do not heed the rule that, in deciding whether summary judgment is proper, a court must view the facts in the light most favorable to the party against whom summary judgment is sought. See post, at 1–2, 8, 10–11, 14 (opinion of Sotomayor, J.). But the dissent, which approves of the Ninth Circuit’s reasoning, seems to forget that the Ninth Circuit in effect granted summary judgment in favor of the teachers on the issue of the applicability of the so-called ministerial exception. It did not remand for a trial on that issue but instead held that the exception did not apply. 769 Fed. Appx. 460, 460–461 (2019); 911 F.3d 603, 605, 611, n. 6 (2018). Therefore, if any material facts were genuinely in dispute, the relevant parts of the record would have to be viewed in the light most favorable to the schools. The dissent, however, does exactly the opposite. In any event, the dissent’s comments about summary judgment are so much smoke. It does not identify any disputed fact that is essential to our holding, and, although there are differences of opinion on certain facts, neither party takes the position that any material fact is genuinely in dispute. 2 After bringing suit, Morrissey-Berru filed a declaration stating that she is “not currently a practicing Catholic.” ER 248. It is unclear what Morrissey-Berru means by “practicing.” There is, however, no hint in the record that Morrissey-Berru considered herself a non-practicing Catholic during her employment at OLG. See infra, at 5 (describing religious observation). 3 This appears to have been a standard contract used within the Archdiocese of Los Angeles. See App. 154; cf. id., at 230. 4 It is not entirely clear from the record whether teachers at OLG must be Catholic. Id., at 113 (“ [Q.] ‘Is it a requirement that a teacher be Catholic in order to teach at OLG School? Yes or no?’ [A.] Yes”); but see ibid. (“Exceptions can be made”); id., at 154 (“If you are Roman Catholic[,] you must be in good standing with the Church” (emphasis added)). But it is clearly preferred. Id., at 110. 5 Record in No. 2:16–CV–09353 (CD Cal.), Doc. 33, ¶9. 6 Biel died during the pendency of this suit, which has subsequently been litigated by her husband as representative of her estate. Record in No. 17–55180 (CA9), Docs. 112, 113. 7 The school principal stated that she prefers that teachers at the school be Catholic. ER 32 (St. James). 8 Record in No. 2:15–CV–04248 (CD Cal.), Doc. 67–1, ¶¶4–7. 9 Cf. McConnell, Establishment and Disestablishment at the Founding, Part I: Establishment of Religion, 44 Wm. & Mary L. Rev. 2105, 2141 (2003) (politically appointed ministers in colonial Virginia were, in the view of the faithful, often “less than zealous in their spiritual responsibilities and less than irreproachable in their personal morals”). 10 In considering the circumstances of any given case, courts must take care to avoid “resolving underlying controversies over religious doctrine.” Presbyterian Church in U. S. v. Mary Elizabeth Blue Hull Memorial Presbyterian Church, 393 U.S. 440, 449 (1969); ibid. (“ First Amendment values are plainly jeopardized when . . . litigation is made to turn on the resolution by civil courts of controversies over religious doctrine and practice”); see also Serbian Eastern Orthodox Diocese for United States and Canada v. Milivojevich, 426 U.S. 696, 715, n. 8 (1976) (“ ‘It is not to be supposed that the judges of the civil courts can be as competent in the ecclesiastical law and religious faith of all these bodies as the ablest men in each are in reference to their own’ ” (quoting Watson v. Jones, 13 Wall. 679, 729 (1872))); cf. Thomas v. Review Bd. of Ind. Employment Security Div., 450 U.S. 707, 714–716 (1981). 11 Brief for Colpa et al. as Amici Curiae i, 3 (quotation modified). 12 Brief for Asma T. Uddin as Amicus Curiae 2. 13 Old Deluder Satan Act of 1647, in The Laws and Liberties of Massachusetts 47 (M. Farrand ed. 1929). 14 Ibid. 15 See P. Parsons, Inside America’s Christian Schools (1987); see also Association of Christian Schools International, Why Christian Schooling?, https://www.acsi.org/membership/why-christian-schooling; Association of Classical Christian Schools, What is CCE?, https://classicalchristian.org/what-is-cce/?v=a44707111a05. 16 R. Dreher, The Benedict Option 146, 155, 160 (2017); see, e.g., J. Ekeland & B. Walton, Discover Christian Schools: Ten Differences, https : / / discoverchristianschools.com/wp-content/uploads/2019/02/DCS_TenDifferences.pdf. 17 See Deuteronomy 6:7, 11:19. 18 Brief for General Conference of Seventh-day Adventists et al. as Amici Curiae 7–8 (quoting Maimonides, Mishne Torah, Hilkhot Talmud Torah 1:2; 2:1, 3). 19 Brief for Church of God in Christ, Inc., et al. as Amici Curiae 15. 20 Afsaruddin, Muslim Views on Education: Parameters, Purview, and Possibilities, 44 J. Cath. Legal Studies 143, 143–144 (2005). 21 Id., at 143. 22 Haddad & Smith, Introduction: The Challenge of Islamic Education in North America, in Educating the Muslims of America 3, 6, 11 (Y. Haddad, F. Senzai, & J. Smith eds. 2009). 23 Berrett, Church Educational System (CES) in 1 Encyclopedia of Mormonism 274, 275 (D. Ludlow ed. 1992). 24 Brief for General Conference of Seventh-day Adventists et al. as Amici Curiae 9. 25 Seventh-day Adventist Church, About Us, https://adventisteducation.org/abt.html. 26 The dissent charges that we transform the holding in Hosanna-Tabor, but that is what the dissent does. Post, at 8. According to the dissent: “Hosanna-Tabor charted a way to separate leaders who ‘personify’ a church’s ‘beliefs’ [and] ‘minister to the faithful’ from individuals who may simply relay religious tenets.” Post, at 7 (quoting 565 U. S., at 188, 195). The dissent cobbles together this new test by taking phrases out of context from separate passages and inserting a proposition never suggested in Hosanna-Tabor, namely, that an individual cannot qualify for the exception if he or she “simply relay[s] religious tenets” without “ ‘minister[ing] to the faithful.’ ” Post, at 7. Hosanna-Tabor never adopted this unworkable test. It did not suggest that the exception it recognized applied only to “leaders.” Post, at 4–5, and n. 1. The term is never used in the opinion of the Court. Insisting on leadership as a qualification would shrink the exception even more than respondents advocate. For example, they agree that it should apply to nuns, see Brief for Respondents 21, but, under the dissent’s test, is every cloistered nun—or every cloistered monk—disqualified? And even if leadership were a requirement, why couldn’t a religious teacher be regarded as a leader of the students in the class? Nor did our opinion in Hosanna-Tabor draw a critical distinction between a person who “simply relay[s] religious tenets” and one who relays such tenets while also “ ‘minister[ing] to the faithful.’ ” Post, at 7. A teacher, such as an instructor in a class on world religions, who merely provides a description of the beliefs and practices of a religion without making any effort to inculcate those beliefs could not qualify for the exception, but otherwise the distinction makes no sense. If a member of the Christian clergy or a rabbi spends almost all of his or her time studying Scripture or theology and writing instead of ministering to a congregation, would that individual fall outside the exception as understood by the dissent? 27 See, e.g., Mark 9:5, 11:21; John 1:38, 3:26, 4:31, 6:25, 9:2. 28 See App. to Pet. for Cert. in No. 19–267, at 56a, 60a; ER 593 (St. James) (“teachers are expected to . . . engage in catechetical . . . development”); Record in No. 2:15–CV–04248 (CD Cal.), Doc. 67–1, ¶10 (“requir[ing]” attendance at “Catholic education conference” to “prepare teachers as religious educators”). 29 The record also makes clear (contrary to the Ninth Circuit’s and dissent’s conclusion, post, at 17) that Morrissey-Berru and Biel “held themselves out” as authorities on religion to their students, and, by extension, their families. See supra, at 2–9.

589.US.2019_18-801

The Patent Act provides two mutually exclusive methods for challenging an adverse decision by the Patent and Trademark Office (PTO). A dissatisfied applicant may appeal directly to the Federal Circuit, 35 U. S. C. §141, or, as relevant here, may file a new civil action against the PTO Director in the United States District Court for the Eastern District of Virginia, §145. Under this second proceeding, the applicant must pay “[a]ll the expenses of the proceedings.” Ibid. Respondent NantKwest, Inc., filed a §145 civil action after its patent application was denied. The District Court granted summary judgment to the PTO, and the Federal Circuit affirmed. The PTO moved for reimbursement of expenses, including the pro rata salaries of PTO attorneys and a paralegal who worked on the case. The District Court denied the motion, concluding that the statutory language referencing expenses was not sufficient to rebut the “American Rule” presumption that parties are responsible for their own attorney’s fees. The en banc Federal Circuit affirmed. Held: The PTO cannot recover the salaries of its legal personnel under §145. Pp. 3–10. (a) The “American Rule”—the bedrock principle that “[e]ach litigant pays his own attorney’s fees, win or lose, unless a statute or contract provides otherwise,” Hardt v. Reliance Standard Life Ins. Co., 560 U.S. 242, 253—provides the starting point for assessing whether §145 authorizes payment of the PTO’s legal fees. Contrary to the Government’s view, this Court has never suggested that any statute is exempt from the presumption against fee shifting or limited its American Rule inquiries to prevailing party statutes. Rather, it has developed a line of precedents addressing statutory deviations from the American Rule that do not limit attorney’s fees awards to prevailing parties. See, e.g., id., at 254. The presumption against fee shifting is particularly important here because reading §145 to permit an unsuccessful government agency to recover attorney’s fees from a prevailing party “would be a radical departure from longstanding fee-shifting principles adhered to in a wide range of contexts.” Ruckelshaus v. Sierra Club, 463 U.S. 680, 683. Pp. 3–6. (b) Section 145’s plain text does not overcome the American Rule’s presumption against fee shifting. Definitions of “expenses,” while capacious enough to include attorney’s fees, provide scant guidance. The mere failure to foreclose a fee award “neither specifically nor explicitly authorizes courts to shift [fees].” Baker Botts L. L. P. v. ASARCO LLC, 576 U.S. 121, ___. The complete phrase “expenses of the proceeding” would not have been commonly understood to include attorney’s fees at the time §145 was enacted. Finally, the modifier “all” does not transform “expenses” to reach an outlay it would not otherwise include. In common statutory usage, the term “expenses” alone has never been considered to authorize an award of attorney’s fees with sufficient clarity to overcome the American Rule presumption. The appearance of “expenses” and “attorney’s fees” together across various statutes indicates that Congress understands the terms to be distinct and not inclusive of each other. See, e.g., 11 U. S. C. §363(n). Other statutes that refer to attorney’s fees as a subset of expenses show only that “expenses” can include attorney’s fees when so defined. See, e.g., 28 U. S. C. §361. Nor do this Court’s cases further the Government’s position that the Court has used “expenses” to mean “attorney’s fees.” See, e.g., Taniguchi v. Kan Pacific Saipan, Ltd., 566 U.S. 560, 573. The Patent Act’s history reinforces that Congress did not intend to shift attorney’s fees in §145 actions. There is no evidence that the original Patent Office ever paid its personnel from sums collected from adverse parties. Neither has the PTO, until this litigation, sought its attorney’s fees under §145. When Congress intended to provide for attorney’s fees in the Patent Act, it has stated so explicitly. See, e.g., 35 U. S. C. §285. Pp. 6–10. 898 F.3d 1177, affirmed. Sotomayor, J., delivered the opinion for a unanimous Court.

Section 145 of the Patent Act affords applicants “dissatisfied with the decision of the Patent Trial and Appeal Board” an opportunity to file a civil action in the United States District Court for the Eastern District of Virginia. 35 U. S. C. §145. The statute specifies that “[a]ll the expenses of the proceedings shall be paid by the applicant.” Ibid. The question presented in this case is whether such “expenses” include the salaries of attorney and paralegal employees of the United States Patent and Trademark Office (PTO). We hold that they do not. I A The Patent Act creates two mutually exclusive pathways to challenge an adverse decision by the PTO. The first permits judicial review by direct appeal to the United States Court of Appeals for the Federal Circuit. §141. There is “no opportunity for the applicant to offer new evidence” in a §141 proceeding, and the Federal Circuit “must review the PTO’s decision on the same administrative record that was before the [agency].” Kappos v. Hyatt, 566 U.S. 431, 434 (2012); 35 U. S. C. §144. The second pathway allows applicants to file a new civil action against the Director of the PTO in federal district court. §145. Unlike §141, §145 “permits the applicant to present new evidence . . . not presented to the PTO.” Kappos, 566 U. S., at 435. The district court “acts as a factfinder when new evidence is introduced in a §145 proceeding” and must make de novo determinations that take into account “both the new evidence and the administrative record before the PTO.” Id., at 444, 446. The parties may appeal the district court’s final decision to the Federal Circuit. 28 U. S. C. §1295(a)(4)(C). Because §145 does not limit an applicant’s ability to introduce new evidence to challenge the denial of a patent, Kappos, 566 U. S., at 439, it can result in protracted litigation. As a condition for permitting such extensive review, the Patent Act requires applicants who avail themselves of §145 to pay “[a]ll the expenses of the proceedings.” 35 U. S. C. §145. B After the PTO denied respondent NantKwest, Inc.’s patent application directed to a method for treating cancer, NantKwest filed a complaint against the PTO Director in the Eastern District of Virginia under §145. The District Court granted summary judgment to the PTO, and the Federal Circuit affirmed. NantKwest, Inc. v. Lee, 686 Fed. Appx. 864 (2017). The PTO moved for reimbursement of expenses that included—for the first time in the 170-year history of §145—the pro rata salaries of PTO attorneys and a paralegal who worked on the case. The District Court denied the PTO’s motion to recover its pro rata legal fees as “expenses” of the §145 proceeding. The court concluded that the statutory language referencing expenses was not clear enough to rebut the “American Rule”—the background principle that parties are responsible for their own attorney’s fees. NantKwest, Inc. v. Lee, 162 F. Supp. 3d 540, 542 (ED Va. 2016). A divided Federal Circuit panel reversed, with Judge Stoll dissenting. NantKwest, Inc. v. Matal, 860 F.3d 1352 (2017). The majority expressed “substantial doub[t ]” that §145 even implicated the American Rule’s presumption against fee shifting in a case in which the payment was not made to a prevailing party. Id., at 1355. The majority concluded that, even assuming the American Rule presumption applied, the term “expenses” in §145 “specific[ally]” and “explicit[ly]” authorized an award of fees. Id., at 1356. The en banc Federal Circuit voted sua sponte to rehear the case and reversed the panel over a dissent. NantKwest, Inc. v. Iancu, 898 F.3d 1177, 1184 (2018). The majority opinion—now authored by Judge Stoll—held that the American Rule presumption applied to §145 because it is “the starting point whenever a party seeks to shift fees from one side to the other in adversarial litigation.” Id., at 1184 (citing Baker Botts L. L. P. v. ASARCO LLC, 576 U.S. 121, ___ (2015)). After examining the plain text and statutory history of §145, the judicial and congressional understanding of similar language, and overarching policy considerations, the majority concluded that “[a]warding ‘[a]ll the expenses’ simply cannot supply the ‘specific and explicit’ directive from Congress to shift attorneys’ fees, and nothing else in the statute evinces congressional intent to make them available.” 898 F. 3d, at 1196 (quoting Alyeska Pipeline Service Co. v. Wilderness Society, 421 U.S. 240, 260 (1975)). We granted certiorari, 586 U. S. ___ (2019), and now affirm. II This Court’s “ ‘basic point of reference’ when considering the award of attorney’s fees is the bedrock principle known as the ‘ “American Rule” ’: Each litigant pays his own attorney’s fees, win or lose, unless a statute or contract provides otherwise.” Hardt v. Reliance Standard Life Ins. Co., 560 U.S. 242, 252–253 (2010) (quoting Ruckelshaus v. Sierra Club, 463 U.S. 680, 683 (1983)). The American Rule has “roots in our common law reaching back to at least the 18th century.” Baker Botts, 576 U. S., at ___ (slip op., at 3) (citing Arcambel v. Wiseman, 3 Dall. 306 (1796)); see also Summit Valley Industries, Inc. v. Carpenters, 456 U.S. 717, 721 (1982) (observing that the American Rule “has been consistently followed for almost 200 years”); Alyeska Pipeline, 421 U. S., at 257 (referring to the presumption against shifting attorney’s fees as a “general” rule). The Government does not dispute this principle or its pedigree, but argues instead that it does not apply at all. Because the American Rule presumption is most often overcome when a statute awards fees to a “prevailing party,” the Government maintains, the presumption applies only to prevailing-party statutes. And because §145 requires one party to pay all expenses regardless of outcome, the argument goes, it is not a statute subject to the presumption. That view is incorrect. This Court has never suggested that any statute is exempt from the presumption against fee shifting. Nor has it limited its American Rule inquiries to prevailing-party statutes. Indeed, the Court has developed a “line of precedents” “addressing statutory deviations from the American Rule that do not limit attorney’s fees awards to the ‘prevailing party.’ ” Hardt, 560 U. S., at 254; see also Baker Botts, 576 U. S., at ___–___ (slip op., at 5–7) (analyzing a bankruptcy provision that did not mention prevailing parties under the American Rule’s presumption against fee shifting). Sebelius v. Cloer, 569 U.S. 369 (2013), confirms that the presumption against fee shifting applies to all statutes—even those like §145 that do not explicitly award attorney’s fees to “prevailing parties.” In Cloer, the Court interpreted a provision of the National Childhood Vaccine Injury Act that permitted courts to “award attorney’s fees . . . ‘incur- red [by a claimant] in any proceeding on’ an unsuccessful vaccine-injury ‘petition . . . brought in good faith [with] a reasonable basis for the claim.’ ” 569 U. S., at 371 (quoting 42 U. S. C. §300aa–15(e)(1)). The Court held that the provision’s clear language authorized attorney’s fees, even though the statute exclusively applied to unsuccessful litigants. 569 U. S., at 372. Cloer establishes two points: First, contrary to the Government’s suggestion, Congress has indeed enacted fee-shifting statutes that apply to nonprevailing parties. Second, and again contrary to the Government’s view, the American Rule applies to such statutes. The Government itself argued in Cloer that the presumption against fee shifting applied by default, but maintained that the statute “depart[ed] so far from background principles about who pays a litigant’s attorney’s fees that it [could not] be justified without a clearer statement than the Act can supply.’ ” Brief for Petitioner in Sebelius v. Cloer, O. T. 2012, No. 12–236, p. 32. The Court acknowledged the Government’s position but concluded that the “rul[e ] of thumb” against fee shifting gave way because the “words of [the] statute [were] unambiguous.” Cloer, 569 U. S., at 380–381 (citing the Government’s brief ). The dissenting en banc Federal Circuit Judges also doubted that the American Rule could apply to a §145 action. They characterized the proceeding as an intermediate step in obtaining a patent and the payment of legal fees as a portion of the application costs. 898 F. 3d, at 1200 (opinion of Prost, J.). Yet §145 has all the marks of the kind of adversarial litigation in which fee shifting, and the presumption against it, is common; the statute authorizes filing a separate civil action where new evidence can be introduced for de novo review by a district judge. Thus, the presumption against fee shifting not only applies, but is particularly important because §145 permits an unsuccessful government agency to recover its expenses from a prevailing party. Reading §145 to award attorney’s fees in that circumstance “would be a radical departure from longstanding fee-shifting principles adhered to in a wide range of contexts.” Ruckelshaus, 463 U. S., at 683. The American Rule thus provides the starting point for assessing whether §145 authorizes payment of the PTO’s legal fees. III To determine whether Congress intended to depart from the American Rule presumption, the Court first “look[s] to the language of the section” at issue. Hardt, 560 U. S., at 254 (internal quotation marks omitted). While “[t]he absence of [a] specific reference to attorney’s fees is not dispositive,” Key Tronic Corp. v. United States, 511 U.S. 809, 815 (1994), Congress must provide a sufficiently “specific and explicit” indication of its intent to overcome the American Rule’s presumption against fee shifting. Alyeska Pipeline, 421 U. S., at 260. A The reference to “expenses” in §145 does not invoke attorney’s fees with the kind of “clarity we have required to deviate from the American Rule.” Baker Botts, 576 U. S., at ___ (slip op., at 4). Definitions of “expenses” provide scant guidance. The term, standing alone, encompasses wide-ranging “expenditure[s] of money, time, labor, or resources to accomplish a result,” Black’s Law Dictionary 698 (10th ed. 2014), “charges or costs met with in . . . doing one’s work,” Webster’s New World College Dictionary 511 (5th ed. 2014), and “outlay[s]” for labor, Merriam-Webster’s Dictionary of Law 180 (1996); see also N. Webster, An American Dictionary of the English Language 319 (3d ed. 1830) (defining the term broadly to include “the employment and consumption, as of time or labor,” or the “disbursing of money”). Though these definitions are capacious enough to include attorney’s fees, the mere failure to foreclose a fee award “neither specifically nor explicitly authorizes courts to shift [fees].” Baker Botts, 576 U. S., at ___ (slip op., at 6). Reading the term “expenses” alongside neighboring words in the statute, however, supports a conclusion excluding legal fees from the scope of §145. The complete phrase “expenses of the proceeding” is similar to the Latin expensæ litis, or “expenses of the litigation.” This term has long referred to a class of expenses commonly recovered in litigation to which attorney’s fees did not traditionally belong. See Black’s Law Dictionary 461 (1891) (defining “expensæ litis” to mean “generally allowed” costs); 1 J. Bouvier, Law Dictionary 392 (1839) (defining the term to mean the “costs which are generally allowed to the successful party”); id., at 244 (excluding from the definition of “costs” the “extraordinary fees [a party] may have paid counsel”). These definitions suggest that the use of “expenses” in §145 would not have been commonly understood to include attorney’s fees at its enactment. Finally, the modifier “all” does not expand §145’s reach to include attorney’s fees. Although the word conveys breadth, it cannot transform “expenses” to reach an outlay it would not otherwise include. Cf. Rimini Street, Inc. v. Oracle USA, Inc., 586 U. S. ___, ___–___ (2019) (slip op., at 6–7) (“The adjective ‘full’ in §505 therefore does not alter the meaning of the word ‘costs.’ Rather, ‘full costs’ are all the ‘costs’ otherwise available under law”). Section 145’s plain text thus does not overcome the American Rule’s presumption against fee shifting to permit the PTO to recoup its legal personnel salaries as “expenses of the proceedings.” B “The record of statutory usage” also illustrates how the term “expenses” alone does not authorize recovery of attorney’s fees. See West Virginia Univ. Hospitals, Inc. v. Casey, 499 U.S. 83, 88 (1991) (looking to statutory usage to determine whether attorney’s fees and expert fees were distinct expenses in the fee-shifting context). That “expenses” and “attorney’s fees” appear in tandem across various statutes shifting litigation costs indicates that Congress understands the two terms to be distinct and not inclusive of each other. See, e.g., 898 F. 3d, at 1188 (quoting 11 U. S. C. §363(n) (allowing trustee to recover “any costs, attorneys’ fees, or expenses incurred”); 12 U. S. C. §1786(p) (permitting courts to “allow to any such party such reasonable expenses and attorneys’ fees as it deems just and proper”); 25 U. S. C. §1401(a) (allowing distribution of funds after payment of “attorney fees and litigation expenses”); 26 U. S. C. §6673(a)(2)(A) (authorizing recovery of “costs, expenses, and attorneys’ fees” against an attorney who “unreasonably and vexatiously” multiplies proceedings); 31 U. S. C. §3730(d)(1) (permitting recovery of “reasonable expenses . . . plus reasonable attorneys’ fees and costs”); 38 U. S. C. §4323(h)(2) (allowing courts to award “reasonable attorney fees, expert witness fees, and other litigation expenses”) (all internal quotation marks omitted)). While some other statutes refer to attorney’s fees as a subset of expenses, they show only that “expenses” can include attorney’s fees when so defined. See, e.g., 28 U. S. C. §361 (authorizing “reasonable expenses, including attorneys’ fees”); §1447(c) (“An order remanding the case may require payment of just costs and any actual expenses, including attorney fees, incurred as a result of the removal”); 29 U. S. C. §1370(e)(1) (“[T]he court in its discretion may award all or a portion of the costs and expenses incurred in connection with such action including reasonable attorney’s fees”); 42 U. S. C. §247d–6d(e)(9) (allowing a party to recover “reasonable expenses incurred . . . , including a reasonable attorney’s fee”). The Government cites several decisions to argue how, on occasion, this Court has used the term “expenses” to mean “attorney’s fees.” None of the cases furthers its position. See, e.g., Rimini Street, 586 U. S., at ___, ___ (slip op., at 4, 11) (reasoning that the term “costs” in the general federal costs statutes does not include attorney’s fees); Taniguchi v. Kan Pacific Saipan, Ltd., 566 U.S. 560, 573 (2012) (mentioning that a party may bear “expenses” related to attorneys, without specifying whether these “expenses” include attorney’s fees); Arlington Central School Dist. Bd. of Ed. v. Murphy, 548 U.S. 291, 297–303 (2006) (distinguishing “attorney’s fees” from “costs” and “costs” from “expenses,” without indicating whether “expenses” encompasses attorney’s fees); Casey, 499 U. S., at 99 (suggesting that an explicit reference to “expert witness fees” or “litigation expenses” could shift expert fees in addition to attorney’s fees—not that the term “litigation expenses” alone could shift attorney’s fees). Simply put, in common statutory usage, the term “expenses” alone has never been considered to authorize an award of attorney’s fees with sufficient clarity to overcome the American Rule presumption. C In fact, the Patent Act’s history reinforces that Congress did not intend to shift fees in §145 actions. There is no evidence that the Patent Office, the PTO’s predecessor, originally paid its personnel from sums collected from adverse parties in litigation, or that the Office initially even employed attorneys. See Act of July 4, 1836, §9, 5Stat. 121 (“[T]he moneys received into the Treasury under this act shall constitute a fund for the payment of the salaries of the officers and clerks herein provided for, and all other expenses of the Patent Office, and to be called the patent fund”). That salaries of PTO employees might have qualified as an “expense” of the agency, however, does not mean that they are an “expense” of a §145 proceeding. Neither has the PTO, until this litigation, sought its attorney’s fees under §145. That the agency has managed to pay its attorneys consistently suggests that financial necessity does not require reading §145 to shift fees, either. In later years, when Congress intended to provide for attorney’s fees in the Patent Act, it stated so explicitly. See, e.g., 35 U. S. C. §285 (“The court in exceptional cases may award reasonable attorney fees to the prevailing party”); §271(e)(4) (“[A] court may award attorney fees under section 285”); §273(f ) (same); §296(b) (same); §297(b)(1) (“Any customer . . . who is found by a court to have been injured by any material false or fraudulent statement . . . may recover . . . reasonable costs and attorneys’ fees”). Because Congress failed to make its intention similarly clear in §145, the Court will not read the statute to “contravene fundamental precepts of the common law.” United States v. Rodgers, 461 U.S. 677, 716 (1983). The history of the Patent Act thus reaffirms the Court’s view that the statute does not specifically or explicitly authorize the PTO to recoup its lawyers’ or paralegals’ pro rata salaries in §145 civil actions. * * * For the foregoing reasons, we conclude that the PTO cannot recover the pro rata salaries of its legal personnel under §145 and therefore affirm the judgment of the Court of Appeals for the Federal Circuit. It is so ordered.

589.US.2019_18-938

An appeal of right lies from “final judgments, orders, and decrees” entered by bankruptcy courts “in cases and proceedings.” 28 U. S. C. §158(a). Bankruptcy court orders are considered final and immediately appealable if they “dispose of discrete disputes within the larger [bankruptcy] case.” Bullard v. Blue Hills, 575 U.S. 496, 501. Ritzen Group, Inc. (Ritzen) sued Jackson Masonry, LLC (Jackson) in Tennessee state court for breach of a land-sale contract. Jackson filed for bankruptcy under Chapter 11 of the Bankruptcy Code. The state-court litigation was put on hold by operation of 11 U. S. C. §362(a), which provides that filing a bankruptcy petition automatically “operates as a stay” of creditors’ debt-collection efforts outside the umbrella of the bankruptcy case. The Bankruptcy Court denied Ritzen’s motion for relief from the automatic stay filed pursuant to §362(d). Ritzen did not appeal that disposition. Instead, its next step was to file a proof of claim against the bankruptcy estate. The Bankruptcy Court subsequently disallowed Ritzen’s claim and confirmed Jackson’s plan of reorganization. Ritzen then filed a notice of appeal in the District Court, challenging the Bankruptcy Court’s order denying relief from the automatic stay. The District Court rejected Ritzen’s appeal as untimely under 28 U. S. C. §158(c)(2) and Federal Rule of Bankruptcy Procedure 8002(a), which require appeals from a bankruptcy court order to be filed “within 14 days after entry of [that] order.” The Sixth Circuit affirmed, concluding that the order denying Ritzen’s motion to lift the stay was final under §158(a), and that the 14-day appeal clock therefore ran from entry of that order. Held: A bankruptcy court’s order unreservedly denying relief from the automatic stay constitutes a final, immediately appealable order under §158(a). Pp. 6–12. (a) This Court’s application of §158(a)’s finality requirement is guided by the opinion in Bullard v. Blue Hills Bank, 575 U.S. 496. Addressing repayment plan confirmations under Chapter 13, the Court held in Bullard that a bankruptcy court’s order rejecting a proposed plan was not final because it did not conclusively resolve the relevant “proceeding.” Rather, the proceeding would continue until approval of a plan. Id., at 502. P. 6. (b) In applying Bullard’s analysis here, the key inquiry is “how to define the immediately appealable ‘proceeding’ in the context of [stay-relief motions].” 575 U. S., at 502. Adjudication of a creditor’s motion for relief from the stay is properly considered a discrete “proceeding.” A bankruptcy court’s order ruling on a stay-relief motion disposes of a procedural unit anterior to, and separate from, claim-resolution proceedings. It occurs before and apart from proceedings on the merits of creditors’ claims. And its resolution forms no part of the adversary claims-adjudication process, proceedings typically governed by state substantive law. Relief from bankruptcy’s automatic stay thus presents a discrete dispute qualifying as an independent “proceeding” within the meaning of §158(a). Bullard, 575 U. S., at 502–505. Pp. 6–8. (c) Ritzen incorrectly characterizes denial of stay relief as determining nothing more than the forum for claim adjudication and thus a preliminary step in the claims-adjudication process. Resolution of a stay-relief motion can have large practical consequences, however, including whether a creditor can isolate its claim from those of other creditors and go it alone outside bankruptcy or the manner in which adversary claims will be adjudicated. Moreover, bankruptcy’s automatic stay stops even nonjudicial efforts to obtain or control the debtor’s assets, matters that often do not concern the forum for, and cannot be considered part of, any subsequent claim adjudication. Ritzen errs in arguing that the order should nonetheless rank as nonfinal where, as here, the bankruptcy court’s decision turns on a substantive issue that may be raised later in the litigation. Section 158(a) asks whether the order in question terminates a procedural unit separate from the remaining case, not whether the bankruptcy court has preclusively resolved a substantive issue. Finally, rather than disrupting the efficiency of the bankruptcy process, immediate appeal may permit creditors to establish their rights expeditiously outside the bankruptcy process, affecting the relief sought and awarded later in the bankruptcy case. Pp. 8–11. 906 F.3d 494, affirmed. Ginsburg, J., delivered the opinion for a unanimous Court.

Under the Bankruptcy Code, filing a petition for bankruptcy automatically “operates as a stay” of creditors’ debt-collection efforts outside the umbrella of the bankruptcy case. 11 U. S. C. §362(a). The question this case presents concerns the finality of, and therefore the time allowed for appeal from, a bankruptcy court’s order denying a creditor’s request for relief from the automatic stay. In civil litigation generally, a court’s decision ordinarily becomes “final,” for purposes of appeal, only upon completion of the entire case, i.e., when the decision “terminate[s the] action” or “ends the litigation on the merits and leaves nothing for the court to do but execute the judgment.” Gelboim v. Bank of America Corp., 574 U.S. 405, 409 (2015) (internal quotation marks omitted). The regime in bankruptcy is different. A bankruptcy case embraces “an aggregation of individual controversies.” 1 Collier on Bankruptcy ¶5.08[1][b], p. 5–43 (16th ed. 2019). Orders in bankruptcy cases qualify as “final” when they definitively dispose of discrete disputes within the overarching bankruptcy case. Bullard v. Blue Hills Bank, 575 U.S. 496, 501 (2015). The precise issue the Court today decides: Does a creditor’s motion for relief from the automatic stay initiate a distinct proceeding terminating in a final, appealable order when the bankruptcy court rules dispositively on the motion? In agreement with the courts below, our answer is “yes.” We hold that the adjudication of a motion for relief from the automatic stay forms a discrete procedural unit within the embracive bankruptcy case. That unit yields a final, appealable order when the bankruptcy court unreservedly grants or denies relief. I In civil litigation generally, 28 U. S. C. §1291 governs appeals from “final decisions.” Under that provision, a party may appeal to a court of appeals as of right from “final decisions of the district courts.” Ibid. A “final decision” within the meaning of §1291 is normally limited to an order that resolves the entire case. Accordingly, the appellant must raise all claims of error in a single appeal. See In re Saco Local Development Corp., 711 F.2d 441, 443 (CA1 1983) (Breyer, J.) (“Traditionally, every civil action in a federal court has been viewed as a ‘single judicial unit,’ from which only one appeal would lie.”). This understanding of the term “final decision” precludes “piecemeal, prejudgment appeals” that would “undermin[e] efficient judicial administration and encroac[h] upon the prerogatives of district court judges.” Bullard, 575 U. S., at 501 (quoting Mohawk Industries, Inc. v. Carpenter, 558 U.S. 100, 106 (2009); internal quotation marks omitted). The ordinary understanding of “final decision” is not attuned to the distinctive character of bankruptcy litigation. A bankruptcy case encompasses numerous “individual controversies, many of which would exist as stand-alone lawsuits but for the bankrupt status of the debtor.” Bullard, 575 U. S., at 501 (internal quotation marks omitted). It is thus common for bankruptcy courts to resolve discrete controversies definitively while the umbrella bankruptcy case re- mains pending. Delaying appeals from discrete, controversy- resolving decisions in bankruptcy cases would long postpone appellate review of fully adjudicated disputes. Moreover, controversies adjudicated during the life of a bankruptcy case may be linked, one dependent on the outcome of another. Delaying appeal until the termination of the entire bankruptcy case, therefore, could have this untoward consequence: Reversal of a decision made early on could require the bankruptcy court to unravel later adjudications rendered in reliance on an earlier decision. The provision on appeals to U. S. district courts from decisions of bankruptcy courts is 28 U. S. C. §158(a). Under that provision, an appeal of right lies from “final judgments, orders, and decrees” entered by bankruptcy courts “in cases and proceedings.” Ibid. By providing for appeals from final decisions in bankruptcy “proceedings,” as distinguished from bankruptcy “cases,” Congress made “orders in bankruptcy cases . . . immediately appeal[able] if they finally dispose of discrete disputes within the larger [bankruptcy] case.” Bullard, 575 U. S., at 501 (quoting Howard Delivery Service, Inc. v. Zurich American Ins. Co., 547 U.S. 651, 657, n. 3 (2006)); see In re Saco Local Development Corp., 711 F. 2d, at 444–447. In short, “the usual judicial unit for analyzing finality in ordinary civil litigation is the case, [but] in bankruptcy[,] it is [often] the proceeding.” Brief for United States as Amicus Curiae 10. Correct delineation of the dimensions of a bankruptcy “proceeding” is a matter of considerable importance. An erroneous identification of an interlocutory order as a final decision may yield an appeal over which the appellate forum lacks jurisdiction. Conversely, an erroneous identification of a final order as interlocutory may cause a party to miss the appellate deadline. II The dispute at hand involves a contract in which Ritzen Group, Inc. (Ritzen) agreed to buy land in Nashville, Tennessee from Jackson Masonry, LLC (Jackson). The land sale was never effected. Blaming Jackson for the deal’s unraveling, Ritzen sued for breach of contract in Tennessee state court. After over a year of litigation, just days before trial was to begin, Jackson filed for bankruptcy under Chapter 11 of the Bankruptcy Code. By operation of the Bankruptcy Code’s automatic stay provision, 11 U. S. C. §362(a), the state-court litigation was put on hold. Ritzen filed a motion in the Federal Bankruptcy Court for relief from the automatic stay, seeking an order allowing the trial to proceed in state court. Ritzen argued that relief would promote judicial economy and that Jackson had filed for bankruptcy in bad faith. After a hearing, the Bankruptcy Court denied the motion. The Bankruptcy Code and Federal Rules of Bankruptcy Procedure require parties to appeal from a final order “within 14 days after entry of the . . . order . . . being appealed.” 28 U. S. C. §158(c)(2); Fed. Rule Bkrtcy. Proc. 8002(a). Ritzen did not appeal from the order refusing to lift the stay within the prescribed period. In pursuit of the breach-of-contract claim initially commenced in state court, Ritzen filed a proof of claim against the bankruptcy estate. Following an adversary proceeding, the Bankruptcy Court found that Ritzen, not Jackson, was the party in breach of the land-sale contract because Ritzen failed to secure financing by the closing date. The court therefore disallowed Ritzen’s claim against the bankruptcy estate. Without objection from Ritzen, the court confirmed Jackson’s plan of reorganization. The plan permanently enjoined all creditors from the “commencement or continuation of any . . . proceeding against [d]ebtor . . . on account of [c]laims against [d]ebtor.” Debtor’s Plan of Reorganization in No. 3:16–bk–02065 (MD Tenn.), p. 15. Thereafter, Ritzen filed two separate notices of appeal in the District Court for the Middle District of Tennessee. First, Ritzen challenged the Bankruptcy Court’s order denying relief from the automatic stay. Second, Ritzen challenged the court’s resolution of its breach-of-contract claim. The District Court rejected the first of Ritzen’s appeals as untimely, holding that under §158(c)(2) and Federal Rule of Bankruptcy Procedure 8002(a), time to appeal expired 14 days after the Bankruptcy Court’s entry of the order denying relief from the automatic stay. Turning to the appeal from the Bankruptcy Court’s rejection of Ritzen’s breach-of-contract claim, the District Court ruled against Ritzen on the merits. On further appeal, the Court of Appeals for the Sixth Circuit affirmed the District Court’s dispositions. As to the timeliness of the first notice of appeal, the Court of Appeals rendered this determination: Adjudication of Ritzen’s motion for relief from the automatic stay qualified as a discrete “proceeding,” commencing with the filing of the motion, followed by procedural steps, and culminating in a “[dispositive] decision based on the application of a legal standard.” In re Jackson Masonry, LLC, 906 F.3d 494, 499–500 (2018).[1] The 14-day appeal clock, the Court of Appeals therefore concluded, ran from the order denying the motion to lift the stay, a disposition “(1) entered in a proceeding and (2) final[ ly] terminating that proceeding.” Id., at 499 (alterations omitted). We granted certiorari to resolve whether orders denying relief from bankruptcy’s automatic stay are final, therefore immediately appealable under §158(a)(1). 587 U. S. ___ (2019). III A This Court’s opinion in Bullard v. Blue Hills Bank, 575 U.S. 496, guides our application of §158(a)’s finality requirement. Addressing repayment plan confirmations under Chapter 13, we held in Bullard that a bankruptcy court’s order rejecting a proposed plan was not “final” under §158(a) because it did not conclusively resolve the relevant “proceeding.” Id., at 499, 502–503. The plan-confirmation process, the Bullard opinion explains, involves back and forth negotiations. See id., at 502. Plan proposal rejections may be followed by amended or new proposals. Only plan approval, we observed, “alters the status quo and fixes the rights and obligations of the parties.” Ibid. “Denial of confirmation with leave to amend,” by contrast, leaves the “parties’ rights and obligations . . . unsettled,” and therefore cannot be typed “final.” Id., at 503. The appropriate procedural unit for determining finality, we concluded, is not a plan proposal, it is “the process of attempting to arrive at an approved plan.” Id., at 502. B We take up next the application of Bullard’s analysis to a bankruptcy court’s order denying relief from the automatic stay. As earlier stated, see supra, at 1, under the Bankruptcy Code, the filing of a bankruptcy petition automatically halts efforts to collect prepetition debts from the bankrupt debtor outside the bankruptcy forum. 11 U. S. C. §362(a). The stay serves to “maintai[n] the status quo and preven[t] dismemberment of the estate” during the pendency of the bankruptcy case. 1 Collier ¶1.05[1], p. 1–19; 3 id., ¶362.03, p. 362–23. Among other things, the stay bars commencement or continuation of lawsuits to recover from the debtor, enforcement of liens or judgments against the debtor, and exercise of control over the debtor’s property. §362(a). A creditor may seek relief from the stay by filing in the bankruptcy court a motion for an order “terminating, annulling, modifying, or conditioning” the stay, asserting in support of the motion either “cause” or the presence of specified conditions. §362(d). A majority of circuits and the leading treatises regard orders denying such motions as final, immediately appealable decisions.[2] We reach the same conclusion. Bullard instructs that we inquire “how to define the immediately appealable ‘proceeding’ in the context of [stay- relief motions].” 575 U. S., at 502. Jackson urges that, as the Court of Appeals held, adjudication of a stay-relief motion is a discrete “proceeding.” Ritzen urges that stay-relief adjudication is properly considered a first step in the process of adjudicating a creditor’s claim against the estate. We agree with the Court of Appeals and Jackson that the appropriate “proceeding” is the stay-relief adjudication. A bankruptcy court’s order ruling on a stay-relief motion disposes of a procedural unit anterior to, and separate from, claim-resolution proceedings. Adjudication of a stay-relief motion, as just observed, occurs before and apart from proceedings on the merits of creditors’ claims: The motion initiates a discrete procedural sequence, including notice and a hearing, and the creditor’s qualification for relief turns on the statutory standard, i.e., “cause” or the presence of specified conditions. §362(d), (e); Fed. Rules Bkrtcy. Proc. 4001(a)(1) and (2), 9014 (describing procedure for adjudicating motions for relief from automatic stay). Resolution of stay-relief motions does not occur as part of the adversary claims-adjudication process, proceedings typically governed by state substantive law. See Butner v. United States, 440 U.S. 48, 54–55 (1979). Under Bullard, a discrete dispute of this kind constitutes an independent “proceeding” within the meaning of 28 U. S. C. §158(a). 575 U. S., at 502–505. Our conclusion that the relevant “proceeding” is the stay-relief adjudication is consistent with statutory text. See id., at 503. A provision neighboring §158(a), §157(b)(2)(G), types motions to terminate, annul, or modify the automatic stay as “core proceedings” arising in a bankruptcy case. Section 157(b)(2) lists those motions separately from the “allowance or disallowance of claims against the estate.” §157(b)(2)(B), (G). Although the discrete “core proceedings” listings “hardly clinc[h] the matter,” as the “provision’s purpose is not to explain appealability,” they are a “textual clue” that Congress viewed adjudication of stay-relief motions as “proceedings” distinct from claim adjudication. Bullard, 575 U. S., at 503. C In Ritzen’s view, the position Jackson advances and we adopt “slic[es] the case too thin.” Id., at 502. Ritzen asserts that an order denying stay relief simply decides the forum for adjudication of adversary claims—bankruptcy court or state court—and therefore should be treated as merely a preliminary step in the claims-adjudication process. Brief for Petitioner 19–21, 26–28. Courts, we agree, should not define “proceeding” to include disputes over minor details about how a bankruptcy case will unfold. As we put it in Bullard, “[t]he concept of finality cannot stretch to cover, for example, an order resolving a disputed request for an extension of time.” 575 U. S., at 505. But Ritzen incorrectly characterizes denial of stay relief as determining nothing more than the forum for claim adjudication. Resolution of a motion for stay relief can have large practical consequences. See 3 Collier ¶362.03, pp. 362–23 to 362–24. Disposition of the motion determines whether a creditor can isolate its claim from those of other creditors and go it alone outside bankruptcy. It can also affect the manner in which adversary claims will be adjudicated. See 11 U. S. C. §502 (permitting summary adjudication or estimation of amounts due in bankruptcy claims adjudication). These are not matters of minor detail; they can significantly increase creditors’ costs. Leaving the stay in place may, inter alia, delay collection of a debt or cause collateral to decline in value. See Brief for United States as Amicus Curiae 14. Ruling on a motion for stay relief, it is true, will determine where the adjudication of an adversary claim will take place—in the bankruptcy forum or state court. But that effect does not render a ruling nonfinal. Orders denying a plaintiff the opportunity to seek relief in its preferred forum often qualify as final and immediately appealable, though they leave the plaintiff free to sue elsewhere. Notably, dismissal for want of personal jurisdiction ranks as a final decision. See Daimler AG v. Bauman, 571 U.S. 117, 124–125 (2014). So too, dismissal for improper venue, or under the doctrine of forum non conveniens. See United States v. Wallace & Tiernan Co., 336 U.S. 793, 794–795, n. 1 (1949); 15A C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §§3914.6, 3914.12 (2d ed. 1992 and Supp. 2019) (collecting cases on appealability of dismissal without prejudice to filing in another forum).[3] Ritzen’s position encounters a further shoal: Many motions to lift the automatic stay do not involve adversary claims against the debtor that would be pursued in another forum but for bankruptcy. Bankruptcy’s embracive automatic stay stops even nonjudicial efforts to obtain or control the debtor’s assets. See §362(a). Motions for stay relief may, for example, seek permission to repossess or liquidate collateral, to terminate a lease, or to set off debts. Ibid. These matters do not concern the forum for, and cannot be considered part of, any subsequent claim adjudication. See Brief for National Association of Consumer Bankruptcy Attorneys as Amicus Curiae 23–24. We see no good reason to treat stay adjudication as the relevant “proceeding” in only a subset of cases. As we have held in another context, “the issue of appealability” should “be determined for the entire category to which a claim belongs.” Digital Equipment Corp. v. Desktop Direct, Inc., 511 U.S. 863, 868 (1994) (addressing collateral order doctrine). Ritzen alternatively argues that, even if an order denying stay relief is not part of the claims-adjudication process, the order should nonetheless rank as nonfinal where, as here, the bankruptcy court’s decision turns on a substantive issue that may be raised later in the litigation. Brief for Petitioner 45. Specifically, Ritzen stresses that it based its stay-relief motion largely on an argument that Jackson filed for bankruptcy in bad faith, an issue that could have been urged again later in the bankruptcy case. Ibid. That argument is misaddressed. Section 158(a) asks whether the order in question terminates a procedural unit separate from the remaining case, not whether the bankruptcy court has preclusively resolved a substantive issue. It does not matter whether the court rested its decision on a determination potentially pertinent to other disputes in the bankruptcy case, so long as the order conclusively resolved the movant’s entitlement to the requested relief. Finally, Ritzen protests that the rule we adopt will encourage piecemeal appeals and unduly disrupt the efficiency of the bankruptcy process. Id., at 48–52. As we see it, classifying as final all orders conclusively resolving stay-relief motions will avoid, rather than cause, “delays and inefficiencies.” Bullard, 575 U. S., at 504. Immediate appeal, if successful, will permit creditors to establish their rights expeditiously outside the bankruptcy process, affecting the relief sought and awarded later in the bankruptcy case. The rule Ritzen urges “would force creditors who lose stay-relief motions to fully litigate their claims in bankruptcy court and then, after the bankruptcy case is over, appeal and seek to redo the litigation all over again in the original court.” 906 F. 3d, at 503. This case is illustrative. After the Bankruptcy Court denied Ritzen’s motion for relief from the automatic stay, Ritzen filed a claim against Jackson in the Bankruptcy Court. The parties and court expended substantial resources definitively litigating the dueling breach-of- contract allegations, and Ritzen lost. The Bankruptcy Court thereafter considered and confirmed Jackson’s reorganization plan. By endeavoring now to appeal the stay-relief order, after forgoing an appeal directly after the denial, Ritzen seeks to return to square one. Its aim, to relitigate the opposing contract claims in state court. Nevermind that the Bankruptcy Court has fully adjudicated the contract claims and has, without objection from Ritzen, approved Jackson’s reorganization plan. The second bite Ritzen seeks scarcely advances the finality principle. IV Because the appropriate “proceeding” in this case is the adjudication of the motion for relief from the automatic stay, the Bankruptcy Court’s order conclusively denying that motion is “final.” The court’s order ended the stay- relief adjudication and left nothing more for the Bankruptcy Court to do in that proceeding.[4] The Court of Appeals therefore correctly ranked the order as final and immediately appealable, and correctly affirmed the District Court’s dismissal of Ritzen’s appeal as untimely. * * * For the reasons stated, the judgment of the Court of Appeals is Affirmed. Notes 1 The “procedural steps” included Ritzen’s provision of notice to Jackson and the Bankruptcy Court’s conduct of a hearing at which the parties presented witness testimony and other evidence. App. to Pet. for Cert. 48a. The question under the “applicable legal standard”: Did Ritzen establish “cause” to permit the state-court litigation to proceed. See id., at 52a–67a; 11 U. S. C. §362(d)(1). 2 See, e.g., Rajala v. Gardner, 709 F.3d 1031, 1034 (CA10 2013); In re Excel Innovations, Inc., 502 F.3d 1086, 1092 (CA9 2007); In re James Wilson Assocs., 965 F.2d 160, 166 (CA7 1992); In re Sonnax Industries, Inc., 907 F.2d 1280, 1284–1285 (CA2 1990); In re Lieb, 915 F.2d 180, 185, n. 3 (CA5 1990); Grundy Nat. Bank v. Tandem Mining Corp., 754 F.2d 1436, 1439 (CA4 1985), overruled in part on other grounds by United Sav. Assn. of Tex. v. Timbers of Inwood Forest Associates, Ltd., 484 U.S. 365 (1988); In re American Mariner Industries, Inc., 734 F.2d 426, 429 (CA9 1984), overruled in part on other grounds by Timbers of Inwood Forest Associates, Ltd., 484 U.S. 365; In re Leimer, 724 F.2d 744, 745 (CA8 1984); 16 C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §3926.2, p. 352, nn. 39–40 (3d ed. 2012 and Supp. 2019) (“Automatic-stay rulings by a bankruptcy judge or appellate panel should be appealable as final decisions.”). See also 1 Collier on Bankruptcy ¶5.09, pp. 5–55 to 5–57 (16th ed. 2019). 3 We note, however, that within the federal court system, when venue is laid in the wrong district, or when the plaintiff chooses an inconvenient forum, transfer rather than dismissal is ordinarily ordered if “in the interest of justice.” 28 U. S. C. §§1404(a), 1406. 4 We do not decide whether finality would attach to an order denying stay relief if the bankruptcy court enters it “without prejudice” because further developments might change the stay calculus. Nothing in the record before us suggests that this is such an order.

589.US.2019_18-1269

The Internal Revenue Service (IRS) allows an affiliated group of corporations to file a consolidated federal return. See 26 U. S. C. §1501. The IRS issues any refund as a single payment to the group’s designated agent. The tax regulations say very little about how the group members should then distribute that refund among themselves. If a dispute arises and the members have no tax allocation agreement in place, federal courts normally turn to state law to resolve the distribution question. Some courts, however, have crafted their own federal common law rule, known as the Bob Richards rule. See In re Bob Richards Chrysler-Plymouth Corp., 473 F.2d 262. The rule initially provided that, in the absence of an agreement, a refund belongs to the group member responsible for the losses that led to it. But it has since evolved, in some jurisdictions, into a general rule that is always followed unless an agreement unambiguously specifies a different result. Soon after United Western Bank suffered huge losses, its parent, United Western Bancorp, Inc., was forced into bankruptcy. When the IRS issued the group a $4 million tax refund, the bank’s receiver, respondent Federal Deposit Insurance Corporation (FDIC), and the parent corporation’s bankruptcy trustee, petitioner Simon Rodriguez, each sought to claim it. The dispute wound its way through a bankruptcy court and a federal district court before the Tenth Circuit examined the parties’ tax allocation agreement, applied the more expansive version of Bob Richards, and ruled for the FDIC. Held: The Bob Richards rule is not a legitimate exercise of federal common lawmaking. Federal judges may appropriately craft the rule of decision in only limited areas, Sosa v. Alvarez-Machain, 542 U.S. 692, 729, and claiming a new area is subject to strict conditions. One of the most basic is that federal common lawmaking must be “ ‘necessary to protect uniquely federal interests.’ ” Texas Industries, Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 640. The Bob Richards rule has not satisfied this condition. The federal courts applying and extending Bob Richards have not pointed to any significant federal interest sufficient to support the Bob Richards rule. Nor have the parties in this case. State law is well-equipped to handle disputes involving corporate property rights, even in cases, like this one, that involve federal bankruptcy and a tax dispute. Whether this case might yield the same or a different result without Bob Richards is a matter the court of appeals may take up on remand. Pp. 4–6. 914 F.3d 1262, vacated and remanded. Gorsuch, J., delivered the opinion for a unanimous Court.

This case grows from a fight over a tax refund. But the question we face isn’t who gets the money, only how to decide the dispute. Should federal courts rely on state law, together with any applicable federal rules, or should they devise their own federal common law test? To ask the question is nearly to answer it. The cases in which federal courts may engage in common lawmaking are few and far between. This is one of the cases that lie between. The trouble here started when the United Western Bank hit hard times, entered receivership, and the Federal Deposit Insurance Corporation took the reins. Not long after that, the bank’s parent, United Western Bancorp, Inc., faced its own problems and was forced into bankruptcy, led now by a trustee, Simon Rodriguez. When the Internal Revenue Service issued a $4 million tax refund, each of these newly assigned caretakers understandably sought to claim the money. Unable to resolve their differences, they took the matter to court. The case wound its way through a bankruptcy court and a federal district court before eventually landing in the Tenth Circuit. At the end of it all, the court of appeals ruled for the FDIC, as receiver for the subsidiary bank, rather than for Mr. Rodriguez, as trustee for the corporate parent. How could two separate corporate entities both claim entitlement to a single tax refund? For many years, the IRS has allowed an affiliated group of corporations to file a consolidated federal return. See 26 U. S. C. §1501. This serves as a convenience for the government and taxpayers alike. Unsurprisingly, though, a corporate group seeking to file a single return must comply with a host of regulations. See 26 U. S. C. §1502; 26 CFR §1.1502–0 et seq. (2019). These regulations are pretty punctilious about ensuring the government gets all the taxes due from corporate group members. See, e.g., §1.1502–6. But when it comes to the distribution of refunds, the regulations say considerably less. They describe how the IRS will pay the group’s designated agent a single refund. See §1.1502–77(d)(5). And they warn that the IRS’s payment discharges the government’s refund liability to all group members. Ibid. But how should the members distribute the money among themselves once the government sends it to their designated agent? On that, federal law says little. To fill the gap, many corporate groups have developed “tax allocation agreements.” These agreements usually specify what share of a group’s tax liability each member will pay, along with the share of any tax refund each member will receive. But what if there is no tax allocation agreement? Or what if the group members dispute the meaning of the terms found in their agreement? Normally, courts would turn to state law to resolve questions like these. State law is replete with rules readymade for such tasks—rules for interpreting contracts, creating equitable trusts, avoiding unjust enrichment, and much more. Some federal courts, however, have charted a different course. They have crafted their own federal common law rule—one known to those who practice in the area as the Bob Richards rule, so named for the Ninth Circuit case from which it grew: In re Bob Richards Chrysler-Plymouth Corp., 473 F.2d 262 (1973). As initially conceived, the Bob Richards rule provided that, in the absence of a tax allocation agreement, a refund belongs to the group member responsible for the losses that led to it. See id., at 265. With the passage of time, though, Bob Richards evolved. Now, in some jurisdictions, Bob Richards doesn’t just supply a stopgap rule for situations when group members lack an allocation agreement. It represents a general rule always to be followed unless the parties’ tax allocation agreement unambiguously specifies a different result. At the urging of the FDIC and consistent with circuit precedent, the Tenth Circuit employed this more expansive version of Bob Richards in the case now before us. Because the parties did have a tax allocation agreement, the court of appeals explained, the question it faced was whether the agreement unambiguously deviated from Bob Richards’s default rule. In re United Western Bancorp, Inc., 914 F.3d 1262, 1269–1270 (2019). After laying out this “analytical framework” for decision, id., at 1269 (emphasis deleted), the court proceeded to hold that the FDIC, as receiver for the bank, owned the tax refund. Not all circuits, however, follow Bob Richards. The Sixth Circuit, for example, has observed that “federal common law constitutes an unusual exercise of lawmaking which should be indulged . . . only when there is a significant conflict between some federal policy or interest and the use of state law.” FDIC v. AmFin Financial Corp., 757 F.3d 530, 535 (2014) (internal quotation marks omitted). In the Sixth Circuit’s view, courts employing Bob Richards have simply “bypassed th[is] threshold question.” 757 F. 3d, at 536. And any fair examination of it, the Sixth Circuit has submitted, reveals no conflict that might justify resort to federal common law. Ibid. We took this case to decide Bob Richards’s fate. 588 U. S. ___ (2019) Judicial lawmaking in the form of federal common law plays a necessarily modest role under a Constitution that vests the federal government’s “legislative Powers” in Congress and reserves most other regulatory authority to the States. See Art. I, §1; Amdt. 10. As this Court has put it, there is “no federal general common law.” Erie R. Co. v. Tompkins, 304 U.S. 64, 78 (1938). Instead, only limited areas exist in which federal judges may appropriately craft the rule of decision. Sosa v. Alvarez-Machain, 542 U.S. 692, 729 (2004). These areas have included admiralty disputes and certain controversies between States. See, e.g., Norfolk Southern R. Co. v. James N. Kirby, Pty Ltd., 543 U.S. 14, 23 (2004); Hinderlider v. La Plata River & Cherry Creek Ditch Co., 304 U.S. 92, 110 (1938). In contexts like these, federal common law often plays an important role. But before federal judges may claim a new area for common lawmaking, strict conditions must be satisfied. The Sixth Circuit correctly identified one of the most basic: In the absence of congressional authorization, common lawmaking must be “ ‘necessary to protect uniquely federal interests.’ ” Texas Industries, Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 640 (1981) (quoting Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 426 (1964)). Nothing like that exists here. The federal government may have an interest in regulating how it receives taxes from corporate groups. See, e.g., 26 CFR §§1.1502–6, –12, –13. The government also may have an interest in regulating the delivery of any tax refund due a corporate group. For example and as we’ve seen, the government may wish to ensure that others in the group have no recourse against federal coffers once it pays the group’s designated agent. See §1.1502–77(d)(5). But what unique interest could the federal government have in determining how a consolidated corporate tax refund, once paid to a designated agent, is distributed among group members? The Sixth Circuit correctly observed that Bob Richards offered no answer—it just bypassed the question. Nor have the courts applying and extending Bob Richards provided satisfactory answers of their own. Even the FDIC, which advocated for the Bob Richards rule in the Tenth Circuit, failed to point that court to any unique federal interest the rule might protect. In this Court, the FDIC, now represented by the Solicitor General, has gone a step further, expressly conceding that federal courts “should not apply a federal common law rule to . . . put a thumb on . . . the scale” when deciding which corporate group member owns some or all of a consolidated refund. Tr. of Oral Arg. 40; see also id., at 32–36. Understandably too. Corporations are generally “creatures of state law,” Cort v. Ash, 422 U.S. 66, 84 (1975), and state law is well equipped to handle disputes involving corporate property rights. That cases like the one now before us happen to involve corporate property rights in the context of a federal bankruptcy and a tax dispute doesn’t change much. As this Court has long recognized, “Congress has generally left the determination of property rights in the assets of a bankrupt’s estate to state law.” Butner v. United States, 440 U.S. 48, 54 (1979). So too with the Internal Revenue Code—it generally “ ‘creates no property rights.’ ” United States v. National Bank of Commerce, 472 U.S. 713, 722 (1985) (quoting United States v. Bess, 357 U.S. 51, 55 (1958)). If special exceptions to these usual rules sometimes might be warranted, no one has explained why the distribution of a consolidated corporate tax refund should be among them. Even if the Tenth Circuit’s reliance on Bob Richards’s analytical framework was mistaken, the FDIC suggests we might affirm the court’s judgment in this case anyway. The FDIC points out that the court of appeals proceeded to consult applicable state law—and the FDIC assures us its result follows naturally from state law. The FDIC also suggests that the IRS regulations concerning the appointment and duties of a corporate group’s agent found in 26 CFR §§1.1502–77(a) and (d) tend to support the court of appeals’s judgment. Unsurprisingly, Mr. Rodriguez disagrees with these assessments and contends that, absent Bob Richards, the Tenth Circuit would have reached a different outcome. Who is right about all this we do not decide. Some, maybe many, cases will come out the same way under state law or Bob Richards. But we did not take this case to decide how this case should be resolved under state law or to determine how IRS regulations might interact with state law. We took this case only to underscore the care federal courts should exercise before taking up an invitation to try their hand at common lawmaking. Bob Richards made the mistake of moving too quickly past important threshold questions at the heart of our separation of powers. It supplies no rule of decision, only a cautionary tale. Whether this case might yield the same or a different result without Bob Richards is a matter the court of appeals may consider on remand. See, e.g., Conkright v. Frommert, 559 U.S. 506, 521–522 (2010); Travelers Casualty & Surety Co. of America v. Pacific Gas & Elec. Co., 549 U.S. 443, 455–456 (2007); Gonzales v. Duenas-Alvarez, 549 U.S. 183, 194 (2007). The judgment of the court of appeals is vacated, and the case is remanded for further proceedings consistent with this opinion. It is so ordered.

590.US.2019_18-1233

Romag Fasteners, Inc., and Fossil, Inc., signed an agreement to use Romag’s fasteners in Fossil’s leather goods. Romag eventually discovered that factories in China making Fossil products were using counterfeit Romag fasteners. Romag sued Fossil and certain retailers of Fossil products (collectively, Fossil) for trademark infringement pursuant to 15 U. S. C. §1125(a). Relying on Second Circuit precedent, the district court rejected Romag’s request for an award of profits, because the jury, while finding that Fossil had acted callously, rejected Romag’s accusation that Fossil had acted willfully. Held: A plaintiff in a trademark infringement suit is not required to show that a defendant willfully infringed the plaintiff’s trademark as a precondition to a profits award. The Lanham Act provision governing remedies for trademark violations, §1117(a), makes a showing of willfulness a precondition to a profits award in a suit under §1125(c) for trademark dilution, but §1125(a) has never required such a showing. Reading words into a statute should be avoided, especially when they are included elsewhere in the very same statute. That absence seems all the more telling here, where the Act speaks often, expressly, and with considerable care about mental states. See, e.g., §§1117(b), (c), 1118. Pointing to §1117(a)’s language indicating that a violation under §1125(a) can trigger an award of the defendant’s profits “subject to the principles of equity,” Fossil argues that equity courts historically required a showing of willfulness before authorizing a profits remedy in trademark disputes. But this suggestion relies on the curious assumption that Congress intended to incorporate a willfulness requirement here obliquely while it prescribed mens rea conditions expressly elsewhere throughout the Act. Nor is it likely that Congress meant to direct “principles of equity”—a term more naturally suggesting fundamental rules that apply more systematically across claims and practice areas—to a narrow rule about a profits remedy within trademark law. Even crediting Fossil’s assumption, all that can be said with certainty is that Pre-Lanham Act case law supports the ordinary principle that a defendant’s mental state is relevant to assigning an appropriate remedy. The place for reconciling the competing and incommensurable policy goals advanced by the parties is before policymakers. Pp. 2–7. Vacated and remanded. Gorsuch, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Ginsburg, Breyer, Alito, Kagan, and Kavanaugh, JJ., joined. Alito, J., filed a concurring opinion, in which Breyer and Kagan, JJ., joined. Sotomayor, J., filed an opinion concurring in the judgment.

When it comes to remedies for trademark infringement, the Lanham Act authorizes many. A district court may award a winning plaintiff injunctive relief, damages, or the defendant’s ill-gotten profits. Without question, a defendant’s state of mind may have a bearing on what relief a plaintiff should receive. An innocent trademark violator often stands in very different shoes than an intentional one. But some circuits have gone further. These courts hold a plaintiff can win a profits remedy, in particular, only after showing the defendant willfully infringed its trademark. The question before us is whether that categorical rule can be reconciled with the statute’s plain language. The question comes to us in a case involving handbag fasteners. Romag sells magnetic snap fasteners for use in leather goods. Fossil designs, markets, and distributes a wide range of fashion accessories. Years ago, the pair signed an agreement allowing Fossil to use Romag’s fasteners in Fossil’s handbags and other products. Initially, both sides seemed content with the arrangement. But in time Romag discovered that the factories Fossil hired in China to make its products were using counterfeit Romag fasteners—and that Fossil was doing little to guard against the practice. Unable to resolve its concerns amicably, Romag sued. The company alleged that Fossil had infringed its trademark and falsely represented that its fasteners came from Romag. After trial, a jury agreed with Romag, and found that Fossil had acted “in callous disregard” of Romag’s rights. At the same time, however, the jury rejected Romag’s accusation that Fossil had acted willfully, as that term was defined by the district court. For our purposes, the last finding is the important one. By way of relief for Fossil’s trademark violation, Romag sought (among other things) an order requiring Fossil to hand over the profits it had earned thanks to its trademark violation. But the district court refused this request. The court pointed out that controlling Second Circuit precedent requires a plaintiff seeking a profits award to prove that the defendant’s violation was willful. Not all circuits, however, agree with the Second Circuit’s rule. We took this case to resolve that dispute over the law’s demands. 588 U. S. ___ (2019). Where does Fossil’s proposed willfulness rule come from? The relevant section of the Lanham Act governing remedies for trademark violations, §35, 60Stat. 439–440, as amended, 15 U. S. C. §1117(a), says this: “When a violation of any right of the registrant of a mark registered in the Patent and Trademark Office, a violation under section 1125(a) or (d) of this title, or a willful violation under section 1125(c) of this title, shall have been established . . . , the plaintiff shall be entitled, subject to the provisions of sections 1111 and 1114 of this title, and subject to the principles of equity, to recover (1) defendant’s profits, (2) any damages sustained by the plaintiff, and (3) the costs of the action.” Immediately, this language spells trouble for Fossil and the circuit precedent on which it relies. The statute does make a showing of willfulness a precondition to a profits award when the plaintiff proceeds under §1125(c). That section, added to the Lanham Act some years after its initial adoption, creates a cause of action for trademark dilution—conduct that lessens the association consumers have with a trademark. But Romag alleged and proved a violation of §1125(a), a provision establishing a cause of action for the false or misleading use of trademarks. And in cases like that, the statutory language has never required a showing of willfulness to win a defendant’s profits. Yes, the law tells us that a profits award is subject to limitations found in §§1111 and 1114. But no one suggests those cross-referenced sections contain the rule Fossil seeks. Nor does this Court usually read into statutes words that aren’t there. It’s a temptation we are doubly careful to avoid when Congress has (as here) included the term in question elsewhere in the very same statutory provision. A wider look at the statute’s structure gives us even more reason for pause. The Lanham Act speaks often and expressly about mental states. Section 1117(b) requires courts to treble profits or damages and award attorney’s fees when a defendant engages in certain acts intentionally and with specified knowledge. Section 1117(c) increases the cap on statutory damages from $200,000 to $2,000,000 for certain willful violations. Section 1118 permits courts to order the infringing items be destroyed if a plaintiff proves any violation of §1125(a) or a willful violation of §1125(c). Section 1114 makes certain innocent infringers subject only to injunctions. Elsewhere, the statute specifies certain mens rea standards needed to establish liability, before even getting to the question of remedies. See, e.g., §§1125(d)(1)(A)(i), (B)(i) (prohibiting certain conduct only if undertaken with “bad faith intent” and listing nine factors relevant to ascertaining bad faith intent). Without doubt, the Lanham Act exhibits considerable care with mens rea standards. The absence of any such standard in the provision before us, thus, seems all the more telling. So how exactly does Fossil seek to conjure a willfulness requirement out of §1117(a)? Lacking any more obvious statutory hook, the company points to the language indicating that a violation under §1125(a) can trigger an award of the defendant’s profits “subject to the principles of equity.” In Fossil’s telling, equity courts historically required a showing of willfulness before authorizing a profits remedy in trademark disputes. Admittedly, equity courts didn’t require so much in patent infringement cases and other arguably analogous suits. See, e.g., Dowagiac Mfg. Co. v. Minnesota Moline Plow Co., 235 U.S. 641, 644, 650–651 (1915). But, Fossil says, trademark is different. There alone, a willfulness requirement was so long and universally recognized that today it rises to the level of a “principle of equity” the Lanham Act carries forward. It’s a curious suggestion. Fossil’s contention that the term “principles of equity” includes a willfulness requirement would not directly contradict the statute’s other, express mens rea provisions or render them wholly superfluous. But it would require us to assume that Congress intended to incorporate a willfulness requirement here obliquely while it prescribed mens rea conditions expressly elsewhere throughout the Lanham Act. That might be possible, but on first blush it isn’t exactly an obvious construction of the statute. Nor do matters improve with a second look. The phrase “principles of equity” doesn’t readily bring to mind a substantive rule about mens rea from a discrete domain like trademark law. In the context of this statute, it more naturally suggests fundamental rules that apply more systematically across claims and practice areas. A principle is a “fundamental truth or doctrine, as of law; a comprehensive rule or doctrine which furnishes a basis or origin for others.” Black’s Law Dictionary 1417 (3d ed. 1933); Black’s Law Dictionary 1357 (4th ed. 1951). And treatises and handbooks on the “principles of equity” generally contain transsubstantive guidance on broad and fundamental questions about matters like parties, modes of proof, defenses, and remedies. See, e.g., E. Merwin, Principles of Equity and Equity Pleading (1895); J. Indermaur & C. Thwaites, Manual of the Principles of Equity (7th ed. 1913); H. Smith, Practical Exposition of the Principles of Equity (5th ed. 1914); R. Megarry, Snell’s Principles of Equity (23d ed. 1947). Our precedent, too, has used the term “principles of equity” to refer to just such transsubstantive topics. See, e.g., eBay Inc. v. MercExchange, L. L. C., 547 U.S. 388, 391, 393 (2006); Holmberg v. Armbrecht, 327 U.S. 392, 395 (1946). Congress itself has elsewhere used “equitable principles” in just this way: An amendment to a different section of the Lanham Act lists “laches, estoppel, and acquiescence” as examples of “equitable principles.” 15 U. S. C. §1069. Given all this, it seems a little unlikely Congress meant “principles of equity” to direct us to a narrow rule about a profits remedy within trademark law. But even if we were to spot Fossil that first essential premise of its argument, the next has problems too. From the record the parties have put before us, it’s far from clear whether trademark law historically required a showing of willfulness before allowing a profits remedy. The Trademark Act of 1905—the Lanham Act’s statutory predecessor which many earlier cases interpreted and applied—did not mention such a requirement. It’s true, as Fossil notes, that some courts proceeding before the 1905 Act, and even some later cases following that Act, did treat willfulness or something like it as a prerequisite for a profits award and rarely authorized profits for purely good-faith infringement. See, e.g., Horlick’s Malted Milk Corp. v. Horluck’s, Inc., 51 F.2d 357, 359 (WD Wash. 1931) (explaining that the plaintiff “cannot recover defendant’s profits unless it has been shown beyond a reasonable doubt that defendant was guilty of willful fraud in the use of the enjoined trade-name”); see also Saxlehner v. Siegel-Cooper Co., 179 U.S. 42, 42–43 (1900) (holding that one defendant “should not be required to account for gains and profits” when it “appear[ed] to have acted in good faith”). But Romag cites other cases that expressly rejected any such rule. See, e.g., Oakes v. Tonsmierre, 49 F. 447, 453 (CC SD Ala. 1883); see also Stonebraker v. Stonebraker, 33 Md. 252, 268 (1870); Lawrence-Williams Co. v. Societe Enfants Gombault et Cie, 52 F.2d 774, 778 (CA6 1931). The confusion doesn’t end there. Other authorities advanced still different understandings about the relationship between mens rea and profits awards in trademark cases. See, e.g., H. Nims, Law of Unfair Competition and Trade-Marks §424 (2d ed. 1917) (“An accounting will not be ordered where the infringing party acted innocently and in ignorance of the plaintiff’s rights”); N. Hesseltine, Digest of the Law of Trade-Marks and Unfair Trade 305 (1906) (contrasting a case holding “[n]o account as to profits allowed except as to user after knowledge of plaintiff’s right to trademark” and one permitting profits “although defendant did not know of infringement” (emphasis added)). And the vast majority of the cases both Romag and Fossil cite simply failed to speak clearly to the issue one way or another. See, e.g., Hostetter v. Vowinkle, 12 F. Cas. 546, 547 (No. 6,714) (CC Neb. 1871); Graham v. Plate, 40 Cal. 593, 597–599 (1871); Hemmeter Cigar Co. v. Congress Cigar Co., 118 F.2d 64, 71–72 (CA6 1941). At the end of it all, the most we can say with certainty is this. Mens rea figured as an important consideration in awarding profits in pre-Lanham Act cases. This reflects the ordinary, transsubstantive principle that a defendant’s mental state is relevant to assigning an appropriate remedy. That principle arises not only in equity, but across many legal contexts. See, e.g., Smith v. Wade, 461 U.S. 30, 38–51 (1983) ( 42 U. S. C. §1983); Morissette v. United States, 342 U.S. 246, 250–263 (1952) (criminal law); Wooden-Ware Co. v. United States, 106 U.S. 432, 434–435 (1882) (common law trespass). It’s a principle reflected in the Lanham Act’s text, too, which permits greater statutory damages for certain willful violations than for other violations. 15 U. S. C. §1117(c). And it is a principle long reflected in equity practice where district courts have often considered a defendant’s mental state, among other factors, when exercising their discretion in choosing a fitting remedy. See, e.g., L. P. Larson, Jr., Co. v. Wm. Wrigley, Jr., Co., 277 U.S. 97, 99–100 (1928); Lander v. Lujan, 888 F.2d 153, 155–156 (CADC 1989); United States v. Klimek, 952 F. Supp. 1100, 1117 (ED Pa. 1997). Given these traditional principles, we do not doubt that a trademark defendant’s mental state is a highly important consideration in determining whether an award of profits is appropriate. But acknowledging that much is a far cry from insisting on the inflexible precondition to recovery Fossil advances. With little to work with in the statute’s language, structure, and history, Fossil ultimately rests on an appeal to policy. The company tells us that stouter restraints on profits awards are needed to deter “baseless” trademark suits. Meanwhile, Romag insists that its reading of the statute will promote greater respect for trademarks in the “modern global economy.” As these things go, amici amplify both sides’ policy arguments. Maybe, too, each side has a point. But the place for reconciling competing and incommensurable policy goals like these is before policymakers. This Court’s limited role is to read and apply the law those policymakers have ordained, and here our task is clear. The judgment of the court of appeals is vacated, and the case is remanded for further proceedings consistent with this opinion. It is so ordered.

591.US.2019_19-7

In the wake of the 2008 financial crisis, Congress established the Consumer Financial Protection Bureau (CFPB), an independent regulatory agency tasked with ensuring that consumer debt products are safe and transparent. See Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank), 124Stat. 1376. Congress transferred the administration of 18 existing federal statutes to the CFPB, including the Fair Credit Reporting Act, the Fair Debt Collection Practices Act, and the Truth in Lending Act; and Congress enacted a new prohibition on unfair and deceptive practices in the consumer-finance sector. 12 U. S. C. §5536(a)(1)(B). In doing so, Congress gave the CFPB extensive rulemaking, enforcement, and adjudicatory powers, including the authority to conduct investigations, issue subpoenas and civil investigative demands, initiate administrative adjudications, prosecute civil actions in federal court, and issue binding decisions in administrative proceedings. The CFPB may seek restitution, disgorgement, injunctive relief, and significant civil penalties for violations of the 19 federal statutes under its purview. So far, the agency has obtained over $11 billion in relief for more than 25 million consumers. Unlike traditional independent agencies headed by multimember boards or commissions, the CFPB is led by a single Director, §5491(b)(1), who is appointed by the President with the advice and consent of the Senate, §5491(b)(2), for a five-year term, during which the President may remove the Director only for “inefficiency, neglect of duty, or malfeasance in office,” §§5491(c)(1), (3). The CFPB receives its funding outside the annual appropriations process from the Federal Reserve, which is itself funded outside the appropriations process through bank assessments. In 2017, the CFPB issued a civil investigative demand to Seila Law LLC, a California-based law firm that provides debt-related legal services to clients. The civil investigative demand (essentially a subpoena) sought information and documents related to the firm’s business practices. Seila Law asked the CFPB to set aside the demand on the ground that the agency’s leadership by a single Director removable only for cause violated the separation of powers. When the CFPB declined, Seila Law refused to comply with the demand, and the CFPB filed a petition to enforce the demand in District Court. Seila Law renewed its claim that the CFPB’s structure violated the separation of powers, but the District Court disagreed and ordered Seila Law to comply with the demand. The Ninth Circuit affirmed, concluding that Seila Law’s challenge was foreclosed by Humphrey’s Executor v. United States, 295 U.S. 602, and Morrison v. Olson, 487 U.S. 654. Held: The judgment is vacated and remanded. 923 F.3d 680, vacated and remanded. The Chief Justice delivered the opinion of the Court with respect to Parts I, II, and III, concluding: 1. Appointed amicus raises three threshold arguments for why this Court may not or should not reach the merits of petitioner’s constitutional challenge, but they are unavailing. Pp. 8–11. 2. The CFPB’s leadership by a single individual removable only for inefficiency, neglect, or malfeasance violates the separation of powers. Pp. 11–30. (a) Article II vests the entire “executive Power” in the President alone, but the Constitution presumes that lesser executive officers will assist the President in discharging his duties. The President’s executive power generally includes the power to supervise—and, if necessary, remove—those who exercise the President’s authority on his behalf. The President’s removal power has long been confirmed by history and precedent. It was recognized by the First Congress in 1789, confirmed by this Court in Myers v. United States, 272 U.S. 52, and reiterated in Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U.S. 477. In Free Enterprise Fund, the Court recognized that it had previously upheld certain congressional limits on the President’s removal power. But the Court declined to extend those limits to “a new situation not yet encountered by the Court.” 561 U. S., at 483. Free Enterprise Fund left in place only two exceptions to the President’s unrestricted removal power. First, Humphrey’s Executor permitted Congress to give for-cause removal protection to a multimember body of experts who were balanced along partisan lines, appointed to staggered terms, performed only “quasi-legislative” and “quasi-judicial functions,” and were said not to exercise any executive power. Second, Morrison approved for-cause removal protection for an inferior officer—the independent counsel—who had limited duties and no policymaking or administrative authority. Pp. 11–16. (b) Neither Humphrey’s Executor nor Morrison resolves whether the CFPB Director’s insulation from removal is constitutional. The New Deal-era FTC upheld in Humphrey’s Executor bears little resemblance to the CFPB. Unlike the multiple Commissioners of the FTC, who were balanced along partisan lines and served staggered terms to ensure the accumulation of institutional knowledge, the CFPB Director serves a five-year term that guarantees abrupt shifts in leadership and the loss of agency expertise. In addition, the Director cannot be dismissed as a mere legislative or judicial aid. Rather, the Director possesses significant administrative and enforcement authority, including the power to seek daunting monetary penalties against private parties in federal court—a quintessentially executive power not considered in Humphrey’s Executor. The logic of Morrison also does not apply. The independent counsel approved in Morrison was an inferior officer who lacked policymaking or administrative authority and exercised narrow authority to initiate criminal investigations and prosecutions of Governmental actors identified by others. By contrast, the CFPB Director is a principal officer whose duties are far from limited. The Director promulgates binding rules fleshing out 19 consumer-protection statutes that cover everything from credit cards and car payments to mortgages and student loans. And the Director brings the coercive power of the state to bear on millions of private citizens and businesses, imposing potentially billion-dollar penalties through administrative adjudications and civil actions. The question here is therefore whether to extend the Humphrey’s Executor and Morrison exceptions to a “new situation.” Free Enterprise Fund, 561 U. S., at 433. Pp. 16–18. (c) The Court declines to extend these precedents to an independent agency led by a single Director and vested with significant executive power. Pp. 18–30. (1) The CFPB’s structure has no foothold in history or tradition. Congress has provided removal protection to principal officers who alone wield power in only four isolated instances: the Comptroller of the Currency (for a one-year period during the Civil War); the Office of Special Counsel; the Administrator of the Social Security Administration; and the Director of the Federal Housing Finance Agency. Aside from the one-year blip for the Comptroller of the Currency, these examples are modern and contested; and they do not involve regulatory or enforcement authority comparable to that exercised by the CFPB. Pp. 18–21. (2) The CFPB’s single-Director configuration is also incompatible with the structure of the Constitution, which—with the sole exception of the Presidency—scrupulously avoids concentrating power in the hands of any single individual. The Framers’ constitutional strategy is straightforward: divide power everywhere except for the Presidency, and render the President directly accountable to the people through regular elections. In that scheme, individual executive officials may wield significant authority, but that authority remains subject to the ongoing supervision and control of the elected President. The CFPB’s single-Director structure contravenes this carefully calibrated system by vesting significant governmental power in the hands of a single individual who is neither elected by the people nor meaningfully controlled (through the threat of removal) by someone who is. The Director may unilaterally, without meaningful supervision, issue final regulations, oversee adjudications, set enforcement priorities, initiate prosecutions, and determine what penalties to impose on private parties. And the Director may do so without even having to rely on Congress for appropriations. While the CFPB’s independent, single-Director structure is sufficient to render the agency unconstitutional, the Director’s five-year term and receipt of funds outside the appropriations process heighten the concern that the agency will “slip from the Executive’s control, and thus from that of the people.” Free Enterprise Fund, 561 U. S., at 499. Pp. 21–25. (3) Amicus raises three principal arguments in the agency’s defense. First, amicus challenges the textual basis for the President’s removal power and highlights statements from individual Framers expressing divergent views on the subject. This Court’s precedents, however, make clear that the President’s removal power derives from the “executive Power” vested exclusively in the President by Article II. And this Court has already discounted the founding-era statements cited by amicus in light of their context. Second, amicus claims that Humphrey’s Executor and Morrison establish a general rule that Congress may freely constrain the President’s removal power, with only two limited exceptions not applicable here. But text, first principles, the First Congress’s decision in 1789, Myers, and Free Enterprise Fund all establish that the President’s removal power is the rule, not the exception. Finally, amicus submits that this Court can cure any constitutional defect in the CFPB’s structure by interpreting the language “inefficiency, neglect of duty, or malfeasance in office,” 12 U. S. C. §5491(c)(3), to reserve substantial discretion to the President. But Humphrey’s Executor implicitly rejected this position, and the CFPB’s defenders have not advanced any workable standard derived from the statutory text. Nor have they explained how a lenient removal standard can be squared with the Dodd-Frank Act as a whole, which makes plain that the CFPB is an “independent bureau.” §5491(a). The dissent advances several additional arguments in the agency’s defense, but they have already been expressly considered and rejected by the Court in Free Enterprise Fund. Pp. 25–30. The Chief Justice, joined by Justice Alito and Justice Kav-anaugh, concluded in Part IV that the Director’s removal protection is severable from the other provisions of the Dodd-Frank Act that establish the CFPB and define its authority. Pp. 30–37. Roberts, C. J., delivered the opinion of the Court with respect to Parts I, II, and III, in which Thomas, Alito, Gorsuch, and Kavanaugh, JJ., joined, and an opinion with respect to Part IV, in which Alito and Kav-anaugh, JJ., joined. Thomas, J., filed an opinion concurring in part and dissenting in part, in which Gorsuch, J., joined. Kagan, J., filed an opinion concurring in the judgment with respect to severability and dissenting in part, in which Ginsburg, Breyer, and Sotomayor, JJ., joined.

with respect to Parts I, II, and III. In the wake of the 2008 financial crisis, Congress established the Consumer Financial Protection Bureau (CFPB), an independent regulatory agency tasked with ensuring that consumer debt products are safe and transparent. In organizing the CFPB, Congress deviated from the structure of nearly every other independent administrative agency in our history. Instead of placing the agency under the leadership of a board with multiple members, Congress provided that the CFPB would be led by a single Director, who serves for a longer term than the President and cannot be removed by the President except for inefficiency, neglect, or malfeasance. The CFPB Director has no boss, peers, or voters to report to. Yet the Director wields vast rulemaking, enforcement, and adjudicatory authority over a significant portion of the U. S. economy. The question before us is whether this arrangement violates the Constitution’s separation of powers. Under our Constitution, the “executive Power”—all of it—is “vested in a President,” who must “take Care that the Laws be faithfully executed.” Art. II, §1, cl. 1; id., §3. Because no single person could fulfill that responsibility alone, the Framers expected that the President would rely on subordinate officers for assistance. Ten years ago, in Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U.S. 477 (2010), we reiterated that, “as a general matter,” the Constitution gives the President “the authority to remove those who assist him in carrying out his duties,” id., at 513–514. “Without such power, the President could not be held fully accountable for discharging his own responsibilities; the buck would stop somewhere else.” Id., at 514. The President’s power to remove—and thus supervise—those who wield executive power on his behalf follows from the text of Article II, was settled by the First Congress, and was confirmed in the landmark decision Myers v. United States, 272 U.S. 52 (1926). Our precedents have recognized only two exceptions to the President’s unrestricted removal power. In Humphrey’s Executor v. United States, 295 U.S. 602 (1935), we held that Congress could create expert agencies led by a group of principal officers removable by the President only for good cause. And in United States v. Perkins, 116 U.S. 483 (1886), and Morrison v. Olson, 487 U.S. 654 (1988), we held that Congress could provide tenure protections to certain inferior officers with narrowly defined duties. We are now asked to extend these precedents to a new configuration: an independent agency that wields significant executive power and is run by a single individual who cannot be removed by the President unless certain statutory criteria are met. We decline to take that step. While we need not and do not revisit our prior decisions allowing certain limitations on the President’s removal power, there are compelling reasons not to extend those precedents to the novel context of an independent agency led by a single Director. Such an agency lacks a foundation in historical practice and clashes with constitutional structure by concentrating power in a unilateral actor insulated from Presidential control. We therefore hold that the structure of the CFPB violates the separation of powers. We go on to hold that the CFPB Director’s removal protection is severable from the other statutory provisions bearing on the CFPB’s authority. The agency may therefore continue to operate, but its Director, in light of our decision, must be removable by the President at will. I A In the summer of 2007, then-Professor Elizabeth Warren called for the creation of a new, independent federal agency focused on regulating consumer financial products. Warren, Unsafe at Any Rate, Democracy (Summer 2007). Professor Warren believed the financial products marketed to ordinary American households—credit cards, student loans, mortgages, and the like—had grown increasingly unsafe due to a “regulatory jumble” that paid too much attention to banks and too little to consumers. Ibid. To remedy the lack of “coherent, consumer-oriented” financial regulation, she proposed “concentrat[ing] the review of financial products in a single location”—an independent agency modeled after the multimember Consumer Product Safety Commission. Ibid. That proposal soon met its moment. Within months of Professor Warren’s writing, the subprime mortgage market collapsed, precipitating a financial crisis that wiped out over $10 trillion in American household wealth and cost millions of Americans their jobs, their retirements, and their homes. In the aftermath, the Obama administration embraced Professor Warren’s recommendation. Through the Treasury Department, the administration encouraged Congress to establish an agency with a mandate to ensure that “consumer protection regulations” in the financial sector “are written fairly and enforced vigorously.” Dept. of Treasury, Financial Regulatory Reform: A New Foundation 55 (2009). Like Professor Warren, the administration envisioned a traditional independent agency, run by a multimember board with a “diverse set of viewpoints and experiences.” Id., at 58. In 2010, Congress acted on these proposals and created the Consumer Financial Protection Bureau (CFPB) as an independent financial regulator within the Federal Reserve System. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank), 124Stat. 1376. Congress tasked the CFPB with “implement[ing]” and “enforc[ing]” a large body of financial consumer protection laws to “ensur[e] that all consumers have access to markets for consumer financial products and services and that markets for consumer financial products and services are fair, transparent, and competitive.” 12 U. S. C. §5511(a). Congress transferred the administration of 18 existing federal statutes to the CFPB, including the Fair Credit Reporting Act, the Fair Debt Collection Practices Act, and the Truth in Lending Act. See §§5512(a), 5481(12), (14). In addition, Congress enacted a new prohibition on “any unfair, deceptive, or abusive act or practice” by certain participants in the consumer-finance sector. §5536(a)(1)(B). Congress authorized the CFPB to implement that broad standard (and the 18 pre-existing statutes placed under the agency’s purview) through binding regulations. §§5531(a)–(b), 5581(a)(1)(A), (b). Congress also vested the CFPB with potent enforcement powers. The agency has the authority to conduct investigations, issue subpoenas and civil investigative demands, initiate administrative adjudications, and prosecute civil actions in federal court. §§5562, 5564(a), (f ). To remedy violations of federal consumer financial law, the CFPB may seek restitution, disgorgement, and injunctive relief, as well as civil penalties of up to $1,000,000 (inflation adjusted) for each day that a violation occurs. §§5565(a), (c)(2); 12 CFR §1083.1(a), Table (2019). Since its inception, the CFPB has obtained over $11 billion in relief for over 25 million consumers, including a $1 billion penalty against a single bank in 2018. See CFPB, Financial Report of the Consumer Financial Protection Bureau, Fiscal Year 2015, p. 3; CFPB, Bureau of Consumer Financial Protection Announces Settlement With Wells Fargo for Auto-Loan Administration and Mortgage Practices (Apr. 20, 2018). The CFPB’s rulemaking and enforcement powers are coupled with extensive adjudicatory authority. The agency may conduct administrative proceedings to “ensure or enforce compliance with” the statutes and regulations it administers. 12 U. S. C. §5563(a). When the CFPB acts as an adjudicator, it has “jurisdiction to grant any appropriate legal or equitable relief.” §5565(a)(1). The “hearing officer” who presides over the proceedings may issue subpoenas, order depositions, and resolve any motions filed by the parties. 12 CFR §1081.104(b). At the close of the proceedings, the hearing officer issues a “recommended decision,” and the CFPB Director considers that recommendation and “issue[s] a final decision and order.” §§1081.400(d), 1081.402(b); see also §1081.405. Congress’s design for the CFPB differed from the proposals of Professor Warren and the Obama administration in one critical respect. Rather than create a traditional independent agency headed by a multimember board or commission, Congress elected to place the CFPB under the leadership of a single Director. 12 U. S. C. §5491(b)(1). The CFPB Director is appointed by the President with the advice and consent of the Senate. §5491(b)(2). The Director serves for a term of five years, during which the President may remove the Director from office only for “inefficiency, neglect of duty, or malfeasance in office.” §§5491(c)(1), (3). Unlike most other agencies, the CFPB does not rely on the annual appropriations process for funding. Instead, the CFPB receives funding directly from the Federal Reserve, which is itself funded outside the appropriations process through bank assessments. Each year, the CFPB requests an amount that the Director deems “reasonably necessary to carry out” the agency’s duties, and the Federal Reserve grants that request so long as it does not exceed 12% of the total operating expenses of the Federal Reserve (inflation adjusted). §§5497(a)(1), (2)(A)(iii), 2(B). In recent years, the CFPB’s annual budget has exceeded half a billion dollars. See CFPB, Fiscal Year 2019: Ann. Performance Plan and Rep., p. 7. B Seila Law LLC is a California-based law firm that provides debt-related legal services to clients. In 2017, the CFPB issued a civil investigative demand to Seila Law to determine whether the firm had “engag[ed] in unlawful acts or practices in the advertising, marketing, or sale of debt relief services.” 2017 WL 6536586, *1 (CD Cal., Aug. 25, 2017). See also 12 U. S. C. §5562(c)(1) (authorizing the agency to issue such demands to persons who “may have any information[ ] relevant to a violation” of one of the laws enforced by the CFPB). The demand (essentially a subpoena) directed Seila Law to produce information and documents related to its business practices. Seila Law asked the CFPB to set aside the demand, objecting that the agency’s leadership by a single Director removable only for cause violated the separation of powers. The CFPB declined to address that claim and directed Seila Law to comply with the demand. When Seila Law refused, the CFPB filed a petition to enforce the demand in the District Court. See §5562(e)(1) (creating cause of action for that purpose). In response, Seila Law renewed its defense that the demand was invalid and must be set aside because the CFPB’s structure violated the Constitution. The District Court disagreed and ordered Seila Law to comply with the demand (with one modification not relevant here). The Court of Appeals affirmed. 923 F.3d 680 (CA9 2019). The Court observed that the “arguments for and against” the constitutionality of the CFPB’s structure had already been “thoroughly canvassed” in majority, concurring, and dissenting opinions by the en banc Court of Appeals for the District of Columbia Circuit in PHH Corp. v. CFPB, 881 F.3d 75 (2018), which had rejected a challenge similar to the one presented here. 923 F. 3d, at 682. The Court saw “no need to re-plow the same ground.” Ibid. Instead, it provided a brief explanation for why it agreed with the PHH Court’s core holding. The Court took as its starting point Humphrey’s Executor, which had approved for-cause removal protection for the Commissioners of the Federal Trade Commission (FTC). In applying that precedent, the Court recognized that the CFPB wields “substantially more executive power than the FTC did back in 1935” and that the CFPB’s leadership by a single Director (as opposed to a multimember commission) presented a “structural difference” that some jurists had found “dispositive.” 923 F. 3d, at 683–684. But the Court felt bound to disregard those differences in light of our decision in Morrison, which permitted a single individual (an independent counsel) to exercise a core executive power (prosecuting criminal offenses) despite being insulated from removal except for cause. Because the Court found Humphrey’s Executor and Morrison “controlling,” it affirmed the District Court’s order requiring compliance with the demand. 923 F. 3d, at 684. We granted certiorari to address the constitutionality of the CFPB’s structure. 589 U. S. ___ (2019). We also requested argument on an additional question: whether, if the CFPB’s structure violates the separation of powers, the CFPB Director’s removal protection can be severed from the rest of the Dodd-Frank Act. Because the Government agrees with petitioner on the merits of the constitutional question, we appointed Paul Clement to defend the judgment below as amicus curiae. He has ably discharged his responsibilities. II We first consider three threshold arguments raised by the appointed amicus for why we may not or should not reach the merits. Each is unavailing. First, amicus argues that the demand issued to petitioner is not “traceable” to the alleged constitutional defect because two of the three Directors who have in turn played a role in enforcing the demand were (or now consider themselves to be) removable by the President at will. Brief for Court-Appointed Amicus Curiae 21–24. Amicus highlights the Government’s argument below that the demand, originally issued by former Director Richard Cordray, had been ratified by an acting CFPB Director who, according to the Office of Legal Counsel (OLC), was removable by the President at will. See Brief for Appellee in No. 17–56324 (CA9), pp. 1, 10, 13–19 (citing Designating an Acting Director of the Bureau of Consumer Financial Protection, 41 Op. OLC ___, ___ (Nov. 25, 2017)). Amicus further observes that current CFPB Director Kathleen Kraninger, now responsible for enforcing the demand, agrees with the Solicitor General’s position in this case that her for-cause removal protection is unconstitutional. See Brief for Respondent on Pet. for Cert. 20; Letter from K. Kraninger, CFPB Director, to M. McConnell, Majority Leader, U. S. Senate, p. 2 (Sept. 17, 2019); Letter from K. Kraninger, CFPB Director, to N. Pelosi, Speaker, U. S. House of Representatives, p. 2 (Sept. 17, 2019).[1] In amicus’ view, these developments reveal that the demand would have been issued—and would continue to be enforced—even in the absence of the CFPB Director’s removal protection, making the asserted separation of powers dispute “artificial.” Brief for Court-Appointed Amicus Curiae 22. Even if that were true, it would not deprive us of jurisdiction. Amicus’ traceability argument appears to challenge petitioner’s Article III standing. See Lujan v. Defenders of Wildlife, 504 U.S. 555, 560 (1992) (explaining that the plaintiff ’s injury must be “fairly traceable to the challenged action of the defendant” (internal quotation marks and alterations omitted)). But amicus’ argument does not cast any doubt on the jurisdiction of the District Court because petitioner is the defendant and did not invoke the Court’s jurisdiction. See Bond v. United States, 564 U.S. 211, 217 (2011) (When the plaintiff has standing, “Article III does not restrict the opposing party’s ability to object to relief being sought at its expense.”). It is true that “standing must be met by persons seeking appellate review, just as it must be met by persons appearing in courts of first instance.” Hollingsworth v. Perry, 570 U.S. 693, 705 (2013) (internal quotation marks omitted). But petitioner’s appellate standing is beyond dispute. Petitioner is compelled to comply with the civil investigative demand and to provide documents it would prefer to withhold, a concrete injury. That injury is traceable to the decision below and would be fully redressed if we were to reverse the judgment of the Court of Appeals and remand with instructions to deny the Government’s petition to enforce the demand. Without engaging with these principles, amicus contends that a litigant wishing to challenge an executive act on the basis of the President’s removal power must show that the challenged act would not have been taken if the responsible official had been subject to the President’s control. See Brief for Court-Appointed Amicus Curiae 21–24. Our precedents say otherwise. We have held that a litigant challenging governmental action as void on the basis of the separation of powers is not required to prove that the Government’s course of conduct would have been different in a “counterfactual world” in which the Government had acted with constitutional authority. Free Enterprise Fund, 561 U. S., at 512, n. 12. In the specific context of the President’s removal power, we have found it sufficient that the challenger “sustain[s] injury” from an executive act that allegedly exceeds the official’s authority. Bowsher v. Synar, 478 U.S. 714, 721 (1986). Second, amicus contends that the proper context for assessing the constitutionality of an officer’s removal restriction is a contested removal. See Brief for Court-Appointed Amicus Curiae 24–27. While that is certainly one way to review a removal restriction, it is not the only way. Our precedents have long permitted private parties aggrieved by an official’s exercise of executive power to challenge the official’s authority to wield that power while insulated from removal by the President. See Bowsher, 478 U. S., at 721 (lawsuit filed by aggrieved third party in the absence of contested removal); Free Enterprise Fund, 561 U. S., at 487 (same); Morrison, 487 U. S., at 668–669 (defense to subpoena asserted by third party in the absence of contested removal). Indeed, we have expressly “reject[ed]” the “argument that consideration of the effect of a removal provision is not ‘ripe’ until that provision is actually used,” because when such a provision violates the separation of powers it inflicts a “here-and-now” injury on affected third parties that can be remedied by a court. Bowsher, 478 U. S., at 727, n. 5 (internal quotation marks omitted). The Court of Appeals therefore correctly entertained petitioner’s constitutional defense on the merits. Lastly, amicus contends that we should dismiss the case because the parties agree on the merits of the constitutional question and the case therefore lacks “adverseness.” Tr. of Oral Arg. 42–43, 45–46. That contention, however, is foreclosed by United States v. Windsor, 570 U.S. 744 (2013). There, we explained that a lower court order that presents real-world consequences for the Government and its adversary suffices to support Article III jurisdiction—even if “the Executive may welcome” an adverse order that “is accompanied by the constitutional ruling it wants.” Id., at 758. Here, petitioner and the Government disagree about whether petitioner must comply with the civil investigative demand. The lower courts sided with the Government, and the Government has not volunteered to relinquish that victory and withdraw the demand. To the contrary, while the Government agrees that the agency is unconstitutionally structured, it believes it may nevertheless enforce the demand on remand. See infra, at 30. Accordingly, our “decision will have real meaning” for the parties. INS v. Chadha, 462 U.S. 919, 939 (1983). And, as in Windsor, any prudential concerns with deciding an important legal question in this posture can be addressed by “the practice of entertaining arguments made by an amicus when the Solicitor General confesses error with respect to a judgment below,” which we have done. 570 U. S., at 760. We therefore turn to the merits of petitioner’s constitutional challenge. III We hold that the CFPB’s leadership by a single individual removable only for inefficiency, neglect, or malfeasance violates the separation of powers. A Article II provides that “[t]he executive Power shall be vested in a President,” who must “take Care that the Laws be faithfully executed.” Art. II, §1, cl. 1; id., §3. The entire “executive Power” belongs to the President alone. But because it would be “impossib[le]” for “one man” to “perform all the great business of the State,” the Constitution assumes that lesser executive officers will “assist the supreme Magistrate in discharging the duties of his trust.” 30 Writings of George Washington 334 (J. Fitzpatrick ed. 1939). These lesser officers must remain accountable to the President, whose authority they wield. As Madison explained, “[I]f any power whatsoever is in its nature Executive, it is the power of appointing, overseeing, and controlling those who execute the laws.” 1 Annals of Cong. 463 (1789). That power, in turn, generally includes the ability to remove executive officials, for it is “only the authority that can remove” such officials that they “must fear and, in the performance of [their] functions, obey.” Bowsher, 478 U. S., at 726 (internal quotation marks omitted). The President’s removal power has long been confirmed by history and precedent. It “was discussed extensively in Congress when the first executive departments were created” in 1789. Free Enterprise Fund, 561 U. S., at 492. “The view that ‘prevailed, as most consonant to the text of the Constitution’ and ‘to the requisite responsibility and harmony in the Executive Department,’ was that the executive power included a power to oversee executive officers through removal.” Ibid. (quoting Letter from James Madison to Thomas Jefferson (June 30, 1789), 16 Documentary History of the First Federal Congress 893 (2004)). The First Congress’s recognition of the President’s removal power in 1789 “provides contemporaneous and weighty evidence of the Constitution’s meaning,” Bowsher, 478 U. S., at 723 (internal quotation marks omitted), and has long been the “settled and well understood construction of the Constitution,” Ex parte Hennen, 13 Pet. 230, 259 (1839). The Court recognized the President’s prerogative to remove executive officials in Myers v. United States, 272 U.S. 52. Chief Justice Taft, writing for the Court, conducted an exhaustive examination of the First Congress’s determination in 1789, the views of the Framers and their contemporaries, historical practice, and our precedents up until that point. He concluded that Article II “grants to the President” the “general administrative control of those executing the laws, including the power of appointment and removal of executive officers.” Id., at 163–164 (emphasis added). Just as the President’s “selection of administrative officers is essential to the execution of the laws by him, so must be his power of removing those for whom he cannot continue to be responsible.” Id., at 117. “[T]o hold otherwise,” the Court reasoned, “would make it impossible for the President . . . to take care that the laws be faithfully executed.” Id., at 164. We recently reiterated the President’s general removal power in Free Enterprise Fund. “Since 1789,” we recapped, “the Constitution has been understood to empower the President to keep these officers accountable—by removing them from office, if necessary.” 561 U. S., at 483. Although we had previously sustained congressional limits on that power in certain circumstances, we declined to extend those limits to “a new situation not yet encountered by the Court”—an official insulated by two layers of for-cause removal protection. Id., at 483, 514. In the face of that novel impediment to the President’s oversight of the Executive Branch, we adhered to the general rule that the President possesses “the authority to remove those who assist him in carrying out his duties.” Id., at 513–514. Free Enterprise Fund left in place two exceptions to the President’s unrestricted removal power. First, in Humphrey’s Executor, decided less than a decade after Myers, the Court upheld a statute that protected the Commissioners of the FTC from removal except for “inefficiency, neglect of duty, or malfeasance in office.” 295 U. S., at 620 (quoting 15 U. S. C. §41). In reaching that conclusion, the Court stressed that Congress’s ability to impose such removal restrictions “will depend upon the character of the office.” 295 U. S., at 631. Because the Court limited its holding “to officers of the kind here under consideration,” id., at 632, the contours of the Humphrey’s Executor exception depend upon the characteristics of the agency before the Court. Rightly or wrongly, the Court viewed the FTC (as it existed in 1935) as exercising “no part of the executive power.” Id., at 628. Instead, it was “an administrative body” that performed “specified duties as a legislative or as a judicial aid.” Ibid. It acted “as a legislative agency” in “making investigations and reports” to Congress and “as an agency of the judiciary” in making recommendations to courts as a master in chancery. Ibid. “To the extent that [the FTC] exercise[d] any executive function[,] as distinguished from executive power in the constitutional sense,” it did so only in the discharge of its “quasi-legislative or quasi-judicial powers.” Ibid. (emphasis added).[2] The Court identified several organizational features that helped explain its characterization of the FTC as non-executive. Composed of five members—no more than three from the same political party—the Board was designed to be “non-partisan” and to “act with entire impartiality.” Id., at 624; see id., at 619–620. The FTC’s duties were “neither political nor executive,” but instead called for “the trained judgment of a body of experts” “informed by experience.” Id., at 624 (internal quotation marks omitted). And the Commissioners’ staggered, seven-year terms enabled the agency to accumulate technical expertise and avoid a “complete change” in leadership “at any one time.” Ibid. In short, Humphrey’s Executor permitted Congress to give for-cause removal protections to a multimember body of experts, balanced along partisan lines, that performed legislative and judicial functions and was said not to exercise any executive power. Consistent with that understanding, the Court later applied “[t]he philosophy of Humphrey’s Executor” to uphold for-cause removal protections for the members of the War Claims Commission—a three-member “adjudicatory body” tasked with resolving claims for compensation arising from World War II. Wiener v. United States, 357 U.S. 349, 356 (1958). While recognizing an exception for multimember bodies with “quasi-judicial” or “quasi-legislative” functions, Humphrey’s Executor reaffirmed the core holding of Myers that the President has “unrestrictable power . . . to remove purely executive officers.” 295 U. S., at 632. The Court acknowledged that between purely executive officers on the one hand, and officers that closely resembled the FTC Commissioners on the other, there existed “a field of doubt” that the Court left “for future consideration.” Ibid. We have recognized a second exception for inferior officers in two cases, United States v. Perkins and Morrison v. Olson.[3] In Perkins, we upheld tenure protections for a naval cadet-engineer. 116 U. S., at 485. And, in Morrison, we upheld a provision granting good-cause tenure protection to an independent counsel appointed to investigate and prosecute particular alleged crimes by high-ranking Government officials. 487 U. S., at 662–663, 696–697. Backing away from the reliance in Humphrey’s Executor on the concepts of “quasi-legislative” and “quasi-judicial” power, we viewed the ultimate question as whether a removal restriction is of “such a nature that [it] impede[s] the President’s ability to perform his constitutional duty.” 487 U. S., at 691. Although the independent counsel was a single person and performed “law enforcement functions that typically have been undertaken by officials within the Executive Branch,” we concluded that the removal protections did not unduly interfere with the functioning of the Executive Branch because “the independent counsel [was] an inferior officer under the Appointments Clause, with limited jurisdiction and tenure and lacking policymaking or significant administrative authority.” Ibid. These two exceptions—one for multimember expert agencies that do not wield substantial executive power, and one for inferior officers with limited duties and no policymaking or administrative authority—“represent what up to now have been the outermost constitutional limits of permissible congressional restrictions on the President’s removal power.” PHH, 881 F. 3d, at 196 (Kavanaugh, J., dissenting) (internal quotation marks omitted). B Neither Humphrey’s Executor nor Morrison resolves whether the CFPB Director’s insulation from removal is constitutional. Start with Humphrey’s Executor. Unlike the New Deal-era FTC upheld there, the CFPB is led by a single Director who cannot be described as a “body of experts” and cannot be considered “non-partisan” in the same sense as a group of officials drawn from both sides of the aisle. 295 U. S., at 624. Moreover, while the staggered terms of the FTC Commissioners prevented complete turnovers in agency leadership and guaranteed that there would always be some Commissioners who had accrued significant expertise, the CFPB’s single-Director structure and five-year term guarantee abrupt shifts in agency leadership and with it the loss of accumulated expertise. In addition, the CFPB Director is hardly a mere legislative or judicial aid. Instead of making reports and recommendations to Congress, as the 1935 FTC did, the Director possesses the authority to promulgate binding rules fleshing out 19 federal statutes, including a broad prohibition on unfair and deceptive practices in a major segment of the U. S. economy. And instead of submitting recommended dispositions to an Article III court, the Director may unilaterally issue final decisions awarding legal and equitable relief in administrative adjudications. Finally, the Director’s enforcement authority includes the power to seek daunting monetary penalties against private parties on behalf of the United States in federal court—a quintessentially executive power not considered in Humphrey’s Executor.[4] The logic of Morrison also does not apply. Everyone agrees the CFPB Director is not an inferior officer, and her duties are far from limited. Unlike the independent counsel, who lacked policymaking or administrative authority, the Director has the sole responsibility to administer 19 separate consumer-protection statutes that cover everything from credit cards and car payments to mortgages and student loans. It is true that the independent counsel in Morrison was empowered to initiate criminal investigations and prosecutions, and in that respect wielded core executive power. But that power, while significant, was trained inward to high-ranking Governmental actors identified by others, and was confined to a specified matter in which the Department of Justice had a potential conflict of interest. By contrast, the CFPB Director has the authority to bring the coercive power of the state to bear on millions of private citizens and businesses, imposing even billion-dollar penalties through administrative adjudications and civil actions. In light of these differences, the constitutionality of the CFPB Director’s insulation from removal cannot be settled by Humphrey’s Executor or Morrison alone. C The question instead is whether to extend those precedents to the “new situation” before us, namely an independent agency led by a single Director and vested with significant executive power. Free Enterprise Fund, 561 U. S., at 483. We decline to do so. Such an agency has no basis in history and no place in our constitutional structure. 1 “Perhaps the most telling indication of [a] severe constitutional problem” with an executive entity “is [a] lack of historical precedent” to support it. Id., at 505 (internal quotation marks omitted). An agency with a structure like that of the CFPB is almost wholly unprecedented. After years of litigating the agency’s constitutionality, the Courts of Appeals, parties, and amici have identified “only a handful of isolated” incidents in which Congress has provided good-cause tenure to principal officers who wield power alone rather than as members of a board or commission. Ibid. “[T]hese few scattered examples”—four to be exact—shed little light. NLRB v. Noel Canning, 573 U.S. 513, 538 (2014). First, the CFPB’s defenders point to the Comptroller of the Currency, who enjoyed removal protection for one year during the Civil War. That example has rightly been dismissed as an aberration. It was “adopted without discussion” during the heat of the Civil War and abandoned before it could be “tested by executive or judicial inquiry.” Myers, 272 U. S., at 165. (At the time, the Comptroller may also have been an inferior officer, given that he labored “under the general direction of the Secretary of the Treasury.” Ch. 58, 12Stat. 665.)[5] Second, the supporters of the CFPB point to the Office of the Special Counsel (OSC), which has been headed by a single officer since 1978.[6] But this first enduring single-leader office, created nearly 200 years after the Constitution was ratified, drew a contemporaneous constitutional objection from the Office of Legal Counsel under President Carter and a subsequent veto on constitutional grounds by President Reagan. See Memorandum Opinion for the General Counsel, Civil Service Commission, 2 Op. OLC 120, 122 (1978); Public Papers of the Presidents, Ronald Reagan, Vol. II, Oct. 26, 1988, pp. 1391–1392 (1991).[7] In any event, the OSC exercises only limited jurisdiction to enforce certain rules governing Federal Government employers and employees. See 5 U. S. C. §1212. It does not bind private parties at all or wield regulatory authority comparable to the CFPB. Third, the CFPB’s defenders note that the Social Security Administration (SSA) has been run by a single Administrator since 1994. That example, too, is comparatively recent and controversial. President Clinton questioned the constitutionality of the SSA’s new single-Director structure upon signing it into law. See Public Papers of the Presidents, William J. Clinton, Vol. II, Aug. 15, 1994, pp. 1471–1472 (1995) (inviting a “corrective amendment” from Congress). In addition, unlike the CFPB, the SSA lacks the authority to bring enforcement actions against private parties. Its role is largely limited to adjudicating claims for Social Security benefits. The only remaining example is the Federal Housing Finance Agency (FHFA), created in 2008 to assume responsibility for Fannie Mae and Freddie Mac. That agency is essentially a companion of the CFPB, established in response to the same financial crisis. See Housing and Economic Recovery Act of 2008, 122Stat. 2654. It regulates primarily Government-sponsored enterprises, not purely private actors. And its single-Director structure is a source of ongoing controversy. Indeed, it was recently held unconstitutional by the Fifth Circuit, sitting en banc. See Collins v. Mnuchin, 938 F.3d 553, 587–588 (2019). With the exception of the one-year blip for the Comptroller of the Currency, these isolated examples are modern and contested. And they do not involve regulatory or enforcement authority remotely comparable to that exercised by the CFPB. The CFPB’s single-Director structure is an innovation with no foothold in history or tradition.[8] 2 In addition to being a historical anomaly, the CFPB’s single-Director configuration is incompatible with our constitutional structure. Aside from the sole exception of the Presidency, that structure scrupulously avoids concentrating power in the hands of any single individual. “The Framers recognized that, in the long term, structural protections against abuse of power were critical to preserving liberty.” Bowsher, 478 U. S., at 730. Their solution to governmental power and its perils was simple: divide it. To prevent the “gradual concentration” of power in the same hands, they enabled “[a]mbition . . . to counteract ambition” at every turn. The Federalist No. 51, p. 349 (J. Cooke ed. 1961) (J. Madison). At the highest level, they “split the atom of sovereignty” itself into one Federal Government and the States. Gamble v. United States, 587 U. S. ___, ___ (2019) (slip op., at 9) (internal quotation marks omitted). They then divided the “powers of the new Federal Government into three defined categories, Legislative, Executive, and Judicial.” Chadha, 462 U. S., at 951. They did not stop there. Most prominently, the Framers bifurcated the federal legislative power into two Chambers: the House of Representatives and the Senate, each composed of multiple Members and Senators. Art. I, §§2, 3. The Executive Branch is a stark departure from all this division. The Framers viewed the legislative power as a special threat to individual liberty, so they divided that power to ensure that “differences of opinion” and the “jarrings of parties” would “promote deliberation and circumspection” and “check excesses in the majority.” See The Federalist No. 70, at 475 (A. Hamilton); see also id., No. 51, at 350. By contrast, the Framers thought it necessary to secure the authority of the Executive so that he could carry out his unique responsibilities. See id., No. 70, at 475–478. As Madison put it, while “the weight of the legislative authority requires that it should be . . . divided, the weakness of the executive may require, on the other hand, that it should be fortified.” Id., No. 51, at 350. The Framers deemed an energetic executive essential to “the protection of the community against foreign attacks,” “the steady administration of the laws,” “the protection of property,” and “the security of liberty.” Id., No. 70, at 471. Accordingly, they chose not to bog the Executive down with the “habitual feebleness and dilatoriness” that comes with a “diversity of views and opinions.” Id., at 476. Instead, they gave the Executive the “[d]ecision, activity, secrecy, and dispatch” that “characterise the proceedings of one man.” Id., at 472. To justify and check that authority—unique in our constitutional structure—the Framers made the President the most democratic and politically accountable official in Government. Only the President (along with the Vice President) is elected by the entire Nation. And the President’s political accountability is enhanced by the solitary nature of the Executive Branch, which provides “a single object for the jealousy and watchfulness of the people.” Id., at 479. The President “cannot delegate ultimate responsibility or the active obligation to supervise that goes with it,” because Article II “makes a single President responsible for the actions of the Executive Branch.” Free Enterprise Fund, 561 U. S., at 496–497 (quoting Clinton v. Jones, 520 U.S. 681, 712–713 (1997) (Breyer, J., concurring in judgment)). The resulting constitutional strategy is straightforward: divide power everywhere except for the Presidency, and render the President directly accountable to the people through regular elections. In that scheme, individual executive officials will still wield significant authority, but that authority remains subject to the ongoing supervision and control of the elected President. Through the President’s oversight, “the chain of dependence [is] preserved,” so that “the lowest officers, the middle grade, and the highest” all “depend, as they ought, on the President, and the President on the community.” 1 Annals of Cong. 499 (J. Madison). The CFPB’s single-Director structure contravenes this carefully calibrated system by vesting significant governmental power in the hands of a single individual accountable to no one. The Director is neither elected by the people nor meaningfully controlled (through the threat of removal) by someone who is. The Director does not even depend on Congress for annual appropriations. See The Federalist No. 58, at 394 (J. Madison) (describing the “power over the purse” as the “most compleat and effectual weapon” in representing the interests of the people). Yet the Director may unilaterally, without meaningful supervision, issue final regulations, oversee adjudications, set enforcement priorities, initiate prosecutions, and determine what penalties to impose on private parties. With no colleagues to persuade, and no boss or electorate looking over her shoulder, the Director may dictate and enforce policy for a vital segment of the economy affecting millions of Americans. The CFPB Director’s insulation from removal by an accountable President is enough to render the agency’s structure unconstitutional. But several other features of the CFPB combine to make the Director’s removal protection even more problematic. In addition to lacking the most direct method of presidential control—removal at will—the agency’s unique structure also forecloses certain indirect methods of Presidential control. Because the CFPB is headed by a single Director with a five-year term, some Presidents may not have any opportunity to shape its leadership and thereby influence its activities. A President elected in 2020 would likely not appoint a CFPB Director until 2023, and a President elected in 2028 may never appoint one. That means an unlucky President might get elected on a consumer-protection platform and enter office only to find herself saddled with a holdover Director from a competing political party who is dead set against that agenda. To make matters worse, the agency’s single-Director structure means the President will not have the opportunity to appoint any other leaders—such as a chair or fellow members of a Commission or Board—who can serve as a check on the Director’s authority and help bring the agency in line with the President’s preferred policies. The CFPB’s receipt of funds outside the appropriations process further aggravates the agency’s threat to Presidential control. The President normally has the opportunity to recommend or veto spending bills that affect the operation of administrative agencies. See Art. I, §7, cl. 2; Art. II, §3. And, for the past century, the President has annually submitted a proposed budget to Congress for approval. See Budget and Accounting Act, 1921, ch. 18, §201, 42Stat. 20. Presidents frequently use these budgetary tools “to influence the policies of independent agencies.” PHH, 881 F. 3d, at 147 (Henderson, J., dissenting) (citing Pasachoff, The President’s Budget as a Source of Agency Policy Control, 125 Yale L. J. 2182, 2191, 2203–2204 (2016)). But no similar opportunity exists for the President to influence the CFPB Director. Instead, the Director receives over $500 million per year to fund the agency’s chosen priorities. And the Director receives that money from the Federal Reserve, which is itself funded outside of the annual appropriations process. This financial freedom makes it even more likely that the agency will “slip from the Executive’s control, and thus from that of the people.” Free Enterprise Fund, 561 U. S., at 499.[9] 3 Amicus raises three principal arguments in the agency’s defense. At the outset, amicus questions the textual basis for the removal power and highlights statements from Madison, Hamilton, and Chief Justice Marshall expressing “heterodox” views on the subject. Brief for Court-Appointed Amicus Curiae 4–5, 28–29. But those concerns are misplaced. It is true that “there is no ‘removal clause’ in the Constitution,” id., at 1, but neither is there a “separation of powers clause” or a “federalism clause.” These foundational doctrines are instead evident from the Constitution’s vesting of certain powers in certain bodies. As we have explained many times before, the President’s removal power stems from Article II’s vesting of the “executive Power” in the President. Free Enterprise Fund, 561 U. S., at 483 (quoting Art. II, §1, cl. 1). As for the opinions of Madison, Hamilton, and Chief Justice Marshall, we have already considered the statements cited by amicus and discounted them in light of their context (Madison), the fact they reflect initial impressions later abandoned by the speaker (Hamilton), or their subsequent rejection as ill-considered dicta (Chief Justice Marshall). See Free Enterprise Fund, 561 U. S., at 500, n. 6 (Madison); Myers, 272 U. S., at 136–139, 142–144 (Hamilton and Chief Justice Marshall).[10] Next, amicus offers a grand theory of our removal precedents that, if accepted, could leave room for an agency like the CFPB—and many other innovative intrusions on Article II. According to amicus, Humphrey’s Executor and Morrison establish a general rule that Congress may impose “modest” restrictions on the President’s removal power, with only two limited exceptions. Brief for Court-Appointed Amicus Curiae 33–37. Congress may not reserve a role for itself in individual removal decisions (as it attempted to do in Myers and Bowsher). And it may not eliminate the President’s removal power altogether (as it effectively did in Free Enterprise Fund). Outside those two situations, amicus argues, Congress is generally free to constrain the President’s removal power. See also post, at 16–22 (Kagan, J., concurring in judgment with respect to severability and dissenting in part) (hereinafter dissent) (expressing similar view). But text, first principles, the First Congress’s decision in 1789, Myers, and Free Enterprise Fund all establish that the President’s removal power is the rule, not the exception. While we do not revisit Humphrey’s Executor or any other precedent today, we decline to elevate it into a freestanding invitation for Congress to impose additional restrictions on the President’s removal authority.[11] Finally, amicus contends that if we identify a constitutional problem with the CFPB’s structure, we should avoid it by broadly construing the statutory grounds for removing the CFPB Director from office. See Brief for Court-Appointed Amicus Curiae 50–53; Tr. of Oral Arg. 57–62. The Dodd-Frank Act provides that the Director may be removed for “inefficiency, neglect of duty, or malfeasance in office.” 12 U. S. C. §5491(c)(3). In amicus’ view, that language could be interpreted to reserve substantial discretion to the President. Brief for Court-Appointed Amicus Curiae 51. We are not persuaded. For one, Humphrey’s Executor implicitly rejected an interpretation that would leave the President free to remove an officer based on disagreements about agency policy. See 295 U. S., at 619, 625–626. In addition, while both amicus and the House of Representatives invite us to adopt whatever construction would cure the constitutional problem, they have not advanced any workable standard derived from the statutory language. Amicus suggests that the proper standard might permit removals based on general policy disagreements, but not specific ones; the House suggests that the permissible bases for removal might vary depending on the context and the Presidential power involved. See Tr. of Oral Arg. 58–60, 76–77. They do not attempt to root either of those standards in the statutory text. Further, although nearly identical language governs the removal of some two-dozen multimember independent agencies, amicus suggests that the standard should vary from agency to agency, morphing as necessary to avoid constitutional doubt. Tr. of Oral Arg. 55–56. We decline to embrace such an uncertain and elastic approach to the text. Amicus and the House also fail to engage with the Dodd-Frank Act as a whole, which makes plain that the CFPB is an “independent bureau.” 12 U. S. C. §5491(a); see also 44 U. S. C. §3502(5) (listing the CFPB as an “independent regulatory agency”). Neither amicus nor the House explains how the CFPB would be “independent” if its head were required to implement the President’s policies upon pain of removal. See Black’s Law Dictionary 838 (9th ed. 2009) (defining “independent” as “[n]ot subject to the control or influence of another”). The Constitution might of course compel the agency to be dependent on the President notwithstanding Congress’s contrary intent, but that result cannot fairly be inferred from the statute Congress enacted. Constitutional avoidance is not a license to rewrite Congress’s work to say whatever the Constitution needs it to say in a given situation. Without a proffered interpretation that is rooted in the statutory text and structure, and would avoid the constitutional violation we have identified, we take Congress at its word that it meant to impose a meaningful restriction on the President’s removal authority. The dissent, for its part, largely reprises points that the Court has already considered and rejected: It notes the lack of an express removal provision, invokes Congress’s general power to create and define executive offices, highlights isolated statements from individual Framers, downplays the decision of 1789, minimizes Myers, brainstorms methods of Presidential control short of removal, touts the need for creative congressional responses to technological and economic change, and celebrates a pragmatic, flexible approach to American governance. See post, at 1–25, 32–33, 38. If these arguments sound familiar, it’s because they are. They were raised by the dissent in Free Enterprise Fund. Compare post, at 1–25, 32–33, 38, with Free Enterprise Fund, 561 U. S., at 515–524, 530 (Breyer, J., dissenting). The answers to these repeated concerns (beyond those we have already covered) are the same today as they were ten years ago. Today, as then, Congress’s “plenary control over the salary, duties, and even existence of executive offices” makes “Presidential oversight” more critical—not less—as the “[o]nly” tool to “counter [Congress’s] influence.” Id., at 500 (opinion of the Court). Today, as then, the various “bureaucratic minutiae” a President might use to corral agency personnel is no substitute for at will removal. Ibid. And today, as always, the urge to meet new technological and societal problems with novel governmental structures must be tempered by constitutional restraints that are not known—and were not chosen—for their efficiency or flexibility. Id., at 499. As we explained in Free Enterprise Fund, “One can have a government that functions without being ruled by functionaries, and a government that benefits from expertise without being ruled by experts.” Ibid. While “[n]o one doubts Congress’s power to create a vast and varied federal bureaucracy,” the expansion of that bureaucracy into new territories the Framers could scarcely have imagined only sharpens our duty to ensure that the Executive Branch is overseen by a President accountable to the people. Ibid. IV Having concluded that the CFPB’s leadership by a single independent Director violates the separation of powers, we now turn to the appropriate remedy. We directed the parties to brief and argue whether the Director’s removal protection was severable from the other provisions of the Dodd-Frank Act that establish the CFPB. If so, then the CFPB may continue to exist and operate notwithstanding Congress’s unconstitutional attempt to insulate the agency’s Director from removal by the President. There is a live controversy between the parties on that question, and resolving it is a necessary step in determining petitioner’s entitlement to its requested relief. As the defendant in this action, petitioner seeks a straightforward remedy. It asks us to deny the Government’s petition to enforce the civil investigative demand and dismiss the case. The Government counters that the demand, though initially issued by a Director unconstitutionally insulated from removal, can still be enforced on remand because it has since been ratified by an Acting Director accountable to the President. The parties dispute whether this alleged ratification in fact occurred and whether, if so, it is legally sufficient to cure the constitutional defect in the original demand. That debate turns on case-specific factual and legal questions not addressed below and not briefed here. A remand for the lower Courts to consider those questions in the first instance is therefore the appropriate course—unless such a remand would be futile. In petitioner’s view, it would be. Before the Court of Appeals, petitioner contended that, regardless of any ratification, the demand is unenforceable because the statutory provision insulating the CFPB Director from removal cannot be severed from the other statutory provisions that define the CFPB’s authority. See Brief for Appellant in No. 17–56324 (CA9), pp. 27–28, 30–32. If petitioner is correct, and the offending removal provision means the entire agency is unconstitutional and powerless to act, then a remand would be pointless. With no agency left with statutory authority to maintain this suit or otherwise enforce the demand, the appropriate disposition would be to reverse with instructions to deny the Government’s petition to enforce the agency’s demand for documents and dismiss the case, as petitioner requests. Accordingly, there is a live controversy over the question of severability. And that controversy is essential to our ability to provide petitioner the relief it seeks: If the removal restriction is not severable, then we must grant the relief requested, promptly rejecting the demand outright. If, on the other hand, the removal restriction is severable, we must instead remand for the Government to press its ratification arguments in further proceedings. Unlike the lingering ratification issue, severability presents a pure question of law that has been fully briefed and argued by the parties. We therefore proceed to address it.[12] It has long been settled that “one section of a statute may be repugnant to the Constitution without rendering the whole act void.” Loeb v. Columbia Township Trustees, 179 U.S. 472, 490 (1900) (quoting Treasurer of Fayette Cty. v. People’s & Drovers’ Bank, 47 Ohio St. 503, 523, 25 N.E. 697, 702 (1890)). Because a “statute bad in part is not necessarily void in its entirety,” “[p]rovisions within the legislative power may stand if separable from the bad.” Dorchy v. Kansas, 264 U.S. 286, 289–290 (1924). “Generally speaking, when confronting a constitutional flaw in a statute, we try to limit the solution to the problem, severing any problematic portions while leaving the remainder intact.” Free Enterprise Fund, 561 U. S., at 508 (internal quotation marks omitted). Even in the absence of a severability clause, the “traditional” rule is that “the unconstitutional provision must be severed unless the statute created in its absence is legislation that Congress would not have enacted.” Alaska Airlines, Inc. v. Brock, 480 U.S. 678, 685 (1987). When Congress has expressly provided a severability clause, our task is simplified. We will presume “that Congress did not intend the validity of the statute in question to depend on the validity of the constitutionally offensive provision . . . unless there is strong evidence that Congress intended otherwise.” Id., at 686. The only constitutional defect we have identified in the CFPB’s structure is the Director’s insulation from removal. If the Director were removable at will by the President, the constitutional violation would disappear. We must therefore decide whether the removal provision can be severed from the other statutory provisions relating to the CFPB’s powers and responsibilities. In Free Enterprise Fund, we found a set of unconstitutional removal provisions severable even in the absence of an express severability clause because the surviving provisions were capable of “functioning independently” and “nothing in the statute’s text or historical context [made] it evident that Congress, faced with the limitations imposed by the Constitution, would have preferred no Board at all to a Board whose members are removable at will.” 561 U. S., at 509 (internal quotation marks omitted). So too here. The provisions of the Dodd-Frank Act bearing on the CFPB’s structure and duties remain fully operative without the offending tenure restriction. Those provisions are capable of functioning independently, and there is nothing in the text or history of the Dodd-Frank Act that demonstrates Congress would have preferred no CFPB to a CFPB supervised by the President. Quite the opposite. Unlike the Sarbanes-Oxley Act at issue in Free Enterprise Fund, the Dodd-Frank Act contains an express severability clause. There is no need to wonder what Congress would have wanted if “any provision of this Act” is “held to be unconstitutional” because it has told us: “the remainder of this Act” should “not be affected.” 12 U. S. C. §5302. Petitioner urges us to disregard this plain language for three reasons. None is persuasive. First, petitioner dismisses the clause as non-probative “boilerplate” because it applies “to the entire, 848-page Dodd-Frank Act” and “appears almost 600 pages before the removal provision at issue.” Brief for Petitioner 45. In petitioner’s view, that means we cannot be certain that Congress really meant to apply the clause to each of the Act’s provisions. But boilerplate is boilerplate for a reason—because it offers tried-and-true language to ensure a precise and predictable result. That is the case here. The language unmistakably references “any provision of this Act.” 12 U. S. C. §5302 (emphasis added). And it appears in a logical and prominent place, immediately following the Act’s title and definitions sections, reinforcing the conclusion that it applies to the entirety of the Act. Congress was not required to laboriously insert duplicative severability clauses, provision by provision, to accomplish its stated objective. Second, petitioner points to an additional severability clause in the Act that applies only to one of the Act’s subtitles. See 15 U. S. C. §8232. In petitioner’s view, that clause would be superfluous if Congress meant the general severability clause to apply across the Act. But “our preference for avoiding surplusage constructions is not absolute.” Lamie v. United States Trustee, 540 U.S. 526, 536 (2004). In this instance, the redundant language appears to reflect the fact that the subtitle to which it refers originated as a standalone bill that was later incorporated into Dodd-Frank. Compare 15 U. S. C. §8232 with H. R. 2571, 111th Cong., 1st Sess., §302 (2009). And petitioner does not offer any construction that would give effect to both provisions, making the redundancy both inescapable and unilluminating. See Microsoft Corp. v. i4i L. P., 564 U.S. 91, 106 (2011) (“The canon against superfluity assists only where a competing interpretation gives effect to every clause and word of a statute.” (internal quotation marks omitted)). Finally, petitioner argues more broadly that Congress would not have wanted to give the President unbridled control over the CFPB’s vast authority. Petitioner highlights the references to the CFPB’s independence in the statutory text and legislative history, as well as in Professor Warren’s and the Obama administration’s original proposals. See Brief for Petitioner 43–44 (collecting examples). And petitioner submits that Congress might not have exempted the CFPB from congressional oversight via the appropriations process if it had known that the CFPB would come under executive control. These observations certainly confirm that Congress preferred an independent CFPB to a dependent one; but they shed little light on the critical question whether Congress would have preferred a dependent CFPB to no agency at all. That is the only question we have the authority to decide, and the answer seems clear. Petitioner assumes that, if we eliminate the CFPB, regulatory and enforcement authority over the statutes it administers would simply revert back to the handful of independent agencies previously responsible for them. See id., at 46. But, as the Solicitor General and House of Representatives explain, that shift would trigger a major regulatory disruption and would leave appreciable damage to Congress’s work in the consumer-finance arena. See Reply Brief for Respondent 21–22; Tr. of Oral Arg. 67–68. One of the agencies whose regulatory authority was transferred to the CFPB no longer exists. See 12 U. S. C. §§5412–5413 (Office of Thrift Supervision). The others do not have the staff or appropriations to absorb the CFPB’s 1,500-employee, 500-million-dollar operations. And none has the authority to administer the Dodd-Frank Act’s new prohibition on unfair and deceptive practices in the consumer-finance sector. Given these consequences, it is far from evident that Congress would have preferred no CFPB to a CFPB led by a Director removable at will by the President. Justice Thomas would have us junk our settled severability doctrine and start afresh, even though no party has asked us to do so. See post, at 15–16, 21–24 (opinion concurring in part and dissenting in part). Among other things, he objects that it is sheer “speculation” that Congress would prefer that its consumer protection laws be enforced by a Director accountable to the President rather than not at all. Post, at 23–24. We think it clear that Congress would prefer that we use a scalpel rather than a bulldozer in curing the constitutional defect we identify today. And such an approach by this Court can come as no surprise to Congress, which was on notice of constitutional objections to single-Director agencies by multiple past Presidents from both political parties, supra, at 19–20, and enacted Dodd-Frank against the background of our established severability doctrine. As in every severability case, there may be means of remedying the defect in the CFPB’s structure that the Court lacks the authority to provide. Our severability analysis does not foreclose Congress from pursuing alternative responses to the problem—for example, converting the CFPB into a multimember agency. The Court’s only instrument, however, is a blunt one. We have “the negative power to disregard an unconstitutional enactment,” Massachusetts v. Mellon, 262 U.S. 447, 488 (1923); see Marbury v. Madison, 1 Cranch 137, 178 (1803), but we cannot re-write Congress’s work by creating offices, terms, and the like. “[S]uch editorial freedom . . . belongs to the Legislature, not the Judiciary.” Free Enterprise Fund, 561 U. S., at 510. Because we find the Director’s removal protection severable from the other provisions of Dodd-Frank that establish the CFPB, we remand for the Court of Appeals to consider whether the civil investigative demand was validly ratified. * * * A decade ago, we declined to extend Congress’s authority to limit the President’s removal power to a new situation, never before confronted by the Court. We do the same today. In our constitutional system, the executive power belongs to the President, and that power generally includes the ability to supervise and remove the agents who wield executive power in his stead. While we have previously upheld limits on the President’s removal authority in certain contexts, we decline to do so when it comes to principal officers who, acting alone, wield significant executive power. The Constitution requires that such officials remain dependent on the President, who in turn is accountable to the people. The judgment of the United States Court of Appeals for the Ninth Circuit is vacated, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 Director Kraninger did not indicate whether she would disregard her statutory removal protection if the President attempted to remove her without cause. 2 The Court’s conclusion that the FTC did not exercise executive power has not withstood the test of time. As we observed in Morrison v. Olson, 487 U.S. 654 (1988), “[I]t is hard to dispute that the powers of the FTC at the time of Humphrey’s Executor would at the present time be considered ‘executive,’ at least to some degree.” Id., at 690, n. 28. See also Arlington v. FCC, 569 U.S. 290, 305, n. 4 (2013) (even though the activities of administrative agencies “take ‘legislative’ and ‘judicial’ forms,” “they are exercises of—indeed, under our constitutional structure they must be exercises of—the ‘executive Power’ ” (quoting Art. II, §1, cl. 1)). 3 Article II distinguishes between two kinds of officers—principal officers (who must be appointed by the President with the advice and consent of the Senate) and inferior officers (whose appointment Congress may vest in the President, courts, or heads of Departments). §2, cl. 2. While “[o]ur cases have not set forth an exclusive criterion for distinguishing between principal and inferior officers,” we have in the past examined factors such as the nature, scope, and duration of an officer’s duties. Edmond v. United States, 520 U.S. 651, 661 (1997). More recently, we have focused on whether the officer’s work is “directed and supervised” by a principal officer. Id., at 663. 4 The dissent would have us ignore the reasoning of Humphrey’s Executor and instead apply the decision only as part of a reimagined Humphrey’s-through-Morrison framework. See post, at 18, n. 7, 19–22 (Kagan, J., concurring in judgment with respect to severability and dissenting in part) (hereinafter dissent). But we take the decision on its own terms, not through gloss added by a later Court in dicta. The dissent also criticizes us for suggesting that the 1935 FTC may have had lesser responsibilities than the present FTC. See post, at 27, n. 10. Perhaps the FTC possessed broader rulemaking, enforcement, and adjudicatory powers than the Humphrey’s Court appreciated. Perhaps not. Either way, what matters is the set of powers the Court considered as the basis for its decision, not any latent powers that the agency may have had not alluded to by the Court. 5 The dissent suggests that the Comptroller still enjoyed some degree of insulation after his removal protection was repealed because the President faced a new requirement to “communicate[ ]” his “reasons” for terminating the Comptroller to the Senate. Post, at 15 (quoting Act of June 3, 1864, ch. 106, §1, 13Stat. 100). But the President could still remove the Comptroller for any reason so long as the President was, in the dissent’s phrase, “in a firing mood.” Post, at 15. 6 The OSC should not be confused with the independent counsel in Morrison or the special counsel recently appointed to investigate allegations related to the 2016 Presidential election. Despite sharing similar titles, those individuals have no relationship to the OSC. 7 An Act similar to the one vetoed by President Reagan was eventually signed by President George H. W. Bush after extensive negotiations and compromises with Congress. See Public Papers of the Presidents, George H. W. Bush, Vol. I, Apr. 10, 1989, p. 391 (1990). 8 The dissent categorizes the CFPB as one of many “financial regulators” that have historically enjoyed some insulation from the President. See post, at 11–16. But even assuming financial institutions like the Second Bank and the Federal Reserve can claim a special historical status, the CFPB is in an entirely different league. It acts as a mini legislature, prosecutor, and court, responsible for creating substantive rules for a wide swath of industries, prosecuting violations, and levying knee-buckling penalties against private citizens. See supra, at 4–5. And, of course, it is the only agency of its kind run by a single Director. 9 Amicus and the dissent try to diminish the CFPB’s insulation from Presidential control by observing that the CFPB’s final rules can be set aside by a super majority of the Financial Stability and Oversight Council (FSOC). See Brief for Court-Appointed Amicus Curiae 40; post, at 33, n. 13, 36. But the FSOC’s veto power is statutorily reserved for extreme situations, when two-thirds of the Council concludes that a CFPB regulation would “put the safety and soundness of the United States banking system or the stability of the financial system of the United States at risk.” 12 U. S. C. §§5513(a), (c)(3). That narrow escape hatch has no impact on the CFPB’s enforcement or adjudicatory authority and has never been used in the ten years since the agency’s creation. It certainlydoes not render the CFPB’s independent, single-Director structureconstitutional. 10 The dissent likewise points to Madison’s statement in The Federalist No. 39 that the “tenure” of “ministerial offices generally will be a subject of legal regulation.” Post, at 10 (quoting The Federalist No. 39, p. 253 (J. Cooke ed. 1961)). But whatever Madison may have meant by that statement, he later led the charge in contending, on the floor of the First Congress, that “inasmuch as the power of removal is of an Executive nature . . . it is beyond the reach of the Legislative body.” 1 Annals of Cong. 464 (1789); see also id., at 462–464, 495–496. Like the dissent in Free Enterprise Fund, the dissent goes on to “attribute[ ] to Madison a belief that . . . the Comptroller[ ] could be made independent of the President. But Madison’s actual proposal, consistent with his view of the Constitution, was that the Comptroller hold office for a term of ‘years, unless sooner removed by the President’; he would thus be ‘dependent upon the President, because he can be removed by him,’ and also ‘dependent upon the Senate, because they must consent to his [reappointment] for every term of years.’ ” Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U.S. 477, 499, 500 n. 6 (2010) (citation omitted) (quoting 1 Annals of Cong. 612). See post, at 10, n. 4. The dissent further notes that, at the time of the founding, some States placed limitations on their Governors’ removal power. See post, at 7. But the Framers hardly viewed State Governors as a reliable guide in fashioning the Federal Executive. Indeed, they expressly rejected the “executive council” structure favored by most States, fearing that subjecting the President to oversight, as the States had, would “distract and . . . enervate the whole system of administration” and inject it with “habitual feebleness and dilatoriness.” The Federalist No. 70, at 473, 476 (A. Hamilton). 11 Building on amicus’ proposal, the dissent would endorse whatever “the times demand, so long as the President retains the ability to carry out his constitutional functions.” Post, at 4. But that amorphous test provides no real limiting principle. The “clearest” (and only) “example” the dissent can muster for what may be prohibited is a for-cause removal restriction placed on the President’s “close military or diplomatic advisers.” Post, at 17. But that carveout makes no logical or constitutional sense. In the dissent’s view, for-cause removal restrictions are permissible because they guarantee the President “meaningful control” over his subordinates. Post, at 28 (internal quotation marks and alterations omitted); see also post, at 8, 20, 26, 36. If that is the theory, then what is the harm in giving the President the same “meaningful control” over his close advisers? The dissent claims to see a constitutional distinction between the President’s “own constitutional duties in foreign relations and war” and his duty to execute laws passed by Congress. Post, at 13. But the same Article that establishes the President’s foreign relations and war duties expressly entrusts him to take care that the laws be faithfully executed. And, from the perspective of the governed, it is far from clear that the President’s core and traditional powers present greater cause for concern than peripheral and modern ones. If anything, “[t]he growth of the Executive Branch, which now wields vast power and touches almost every aspect of daily life, heightens the concern that it may slip from the Executive’s control, and thus from that of the people.” Free Enterprise Fund, 561 U. S., at 499 (emphasis added). 12 Justice Thomas believes that any ratification is irrelevant. In his view, even if the issuance of the demand and initiation of this suit have been validly ratified, Director Kraninger’s activities in litigating the case—after inheriting it from an Acting Director, but before becoming removable at will herself in light of our decision—present a distinct constitutional injury requiring immediate dismissal. See post, at 17–19 (opinion concurring in part and dissenting in part). But whether and when the temporary involvement of an unconstitutionally insulated officer in an otherwise valid prosecution requires dismissal falls outside the questions presented, has not been fully briefed, and is best resolved by the lower courts in the first instance.

589.US.2019_18-6662

The Armed Career Criminal Act (ACCA) mandates a 15-year minimum sentence for a defendant convicted of being a felon in possession of a firearm who has at least three convictions for “serious drug offense[s].” 18 U. S. C. §924(e)(1). A state offense ranks as a “serious drug offense” only if it “involv[es] manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance.” §924(e)(2)(A)(ii). To determine whether an offender’s prior convictions qualify for ACCA enhancement, this Court has used a “categorical approach,” looking “only to the statutory definitions of the prior offenses.” Taylor v. United States, 495 U.S. 575, 600. Under some statutes, a court employing a categorical approach must come up with a “generic” version of a crime—that is, the elements of the offense as commonly understood. The court then determines whether the elements of the offense of conviction match those of the generic crime. Other statutes, which ask the court to determine whether the conviction meets some other criterion, require no such generic-offense analysis. Shular pleaded guilty to being a felon in possession of a firearm and received a 15-year sentence, the mandatory minimum under ACCA. In imposing this sentence, the District Court held that Shular’s six prior cocaine-related convictions under Florida law qualified as “serious drug offense[s]” triggering ACCA enhancement. The Eleventh Circuit affirmed, concluding that §924(e)(2)(A)(ii)’s “serious drug offense” definition does not require a comparison to a generic offense. Held: Section 924(e)(2)(A)(ii)’s “serious drug offense” definition requires only that the state offense involve the conduct specified in the statute; it does not require that the state offense match certain generic offenses. Pp. 5–11. (a) The parties agree that §924(e)(2)(A)(ii) requires a categorical approach. They differ, however, on what comparison the statute requires. In the Government’s view, §924(e)(2)(A)(ii) identifies conduct a court should compare directly against the state crime’s elements. In Shular’s view, §924(e)(2)(A)(ii) identifies generic offenses whose elements a court must first expound, then compare against the state crime’s elements. Pp. 5–6. (b) The statutory text and context show that §924(e)(2)(A)(ii) refers to conduct, not offenses. In two respects, §924(e)(2)(A)(ii) contrasts with neighboring §924(e)(2)(B)(ii), which refers to a crime that “is burglary, arson, or extortion” and calls for the generic-offense analysis that Shular urges. First, the terms in §924(e)(2)(A)(ii)—“manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance”—can be used to describe conduct. Unlike “burglary,” “arson,” and “extortion,” those terms do not unambiguously name offenses. Second, by speaking of activities a state-law drug offense “involv[es],” §924(e)(2)(A)(ii) suggests that the descriptive terms immediately following the word “involving” identify conduct. To refer to offenses, it would have been far more natural for the drafter to follow §924(e)(2)(B)(ii) in using “is.” Pp. 6–7. (c) Shular argues that Congress meant to capture the drug offenses generally existing in state laws at the time of §924(e)(2)(A)(ii)’s enactment. But he admits that those state laws lacked common nomenclature. The evident solution was for Congress to identify offenses by the conduct involved, not by the name of the offenses. Shular offers no persuasive explanation for why Congress would have chosen “involving” over “is” to refer to offenses. Nor do the other ACCA provisions on which Shular relies shed light on whether §924(e)(2)(A)(ii) refers to conduct or offenses. Pp. 7–9. (d) Rejecting a generic-offense approach, Shular contends, would subject defendants to ACCA enhancement based on outlier state laws. He emphasizes that the Florida drug offenses of which he was convicted do not require, as an element, knowledge of the illicit nature of the controlled substance. But Shular overstates the extent to which Florida law is idiosyncratic, for if a defendant asserts that he was unaware of the substance’s illicit nature, the jury must find knowledge beyond a reasonable doubt. In any event, Shular’s interpretation is scarcely the only one that promotes consistency. Congress intended consistent application of ACCA to all offenders who engaged—according to the elements of their prior convictions—in certain conduct. Pp. 9–10. (e) The rule of lenity has no application here, for after consulting traditional canons of interpretation there remains no ambiguity for the rule of lenity to resolve. Pp. 10–11. 736 Fed. Appx. 876, affirmed. Ginsburg, J., delivered the opinion for a unanimous Court. Kav- anaugh, J., filed a concurring opinion.

The Armed Career Criminal Act (ACCA), 18 U. S. C. §924(e), mandates a 15-year minimum sentence of imprisonment for certain defendants with prior convictions for a “serious drug offense.” A state offense ranks as a “serious drug offense” only if it “involv[es] manufacturing, distributing, or possessing with intent to manufacture or dis- tribute, a controlled substance.” §924(e)(2)(A)(ii). This case concerns the methodology courts use to apply that definition. While the parties agree that a court should look to the state offense’s elements, they disagree over what the court should measure those elements against. In the Government’s view, the court should ask whether those elements involve the conduct identified in §924(e)(2)(A)(ii)—namely, “manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance.” Petitioner Eddie Lee Shular, however, contends that the terms employed in the statute identify not conduct, but offenses. In his view, those terms are shorthand for the elements of the offenses as commonly understood. According to Shular, the court must first identify the elements of the “generic” offense, then ask whether the elements of the state offense match those of the generic crime. Under the approach he advances, Shular argues, his sentence is not subject to ACCA enhancement. The generic offenses named in §924(e)(2)(A)(ii), as Shular understands them, include a mens rea element of knowledge that the substance is illicit. He emphasizes that his prior convictions were for state offenses that do not make knowledge of the substance’s illegality an element of the offense; the state offenses, he therefore maintains, do not match the generic offenses in §924(e)(2)(A)(ii). The question presented: Does §924(e)(2)(A)(ii)’s “serious drug offense” definition call for a comparison to a generic offense? We hold it does not. The “serious drug offense” definition requires only that the state offense involve the conduct specified in the federal statute; it does not require that the state offense match certain generic offenses. I Ordinarily, a defendant convicted of being a felon in possession of a firearm, in violation of §922(g)(1), faces a maximum sentence of ten years. §924(a)(2). If the offender’s prior criminal record includes at least three convictions for “serious drug offense[s]” or “violent felon[ies],” however, ACCA mandates a minimum sentence of 15 years. §924(e)(1). To determine whether an offender’s prior convictions qualify for ACCA enhancement, we have used a “categorical approach,” under which we look “only to the statutory definitions of the prior offenses.” Taylor v. United States, 495 U.S. 575, 600 (1990). Under this approach, we consider neither “the particular facts underlying the prior convictions” nor “the label a State assigns to [the] crime[s].” Mathis v. United States, 579 U. S. ___, ___ (2016) (slip op., at 8) (internal quotation marks and alterations omitted). So, for example, to apply ACCA’s provision defining “violent felony” to include “burglary,” §924(e)(2)(B)(ii), we ask only whether the elements of the prior conviction constitute burglary; we do not ask what the person did or whether the offense of conviction was named “burglary.” Under some statutes, using a categorical approach requires the court to come up with a “generic” version of a crime—that is, the elements of “the offense as commonly understood,” id., at ___ (slip op., at 1).[1] We have required that step when the statute refers generally to an offense without specifying its elements. In that situation, the court must define the offense so that it can compare elements, not labels. For example, in Taylor, confronted with ACCA’s unadorned reference to “burglary,” we identified the elements of “generic burglary” based on the “sense in which the term is now used in the criminal codes of most States.” 495 U. S., at 598–599; §924(e)(2)(B)(ii). We then inquired whether the elements of the offense of conviction matched those of the generic crime. Id., at 602. See also, e.g., Esquivel-Quintana v. Sessions, 581 U. S. ___, ___ (2017) (slip op., at 4) (“generic federal definition of sexual abuse of a minor” for purposes of 8 U. S. C. §1101(a)(43)(A)). In contrast, other statutes calling for a categorical approach ask the court to determine not whether the prior conviction was for a certain offense, but whether the conviction meets some other criterion. For example, in Kawashima v. Holder, 565 U.S. 478 (2012), we applied a categorical approach to a statute assigning immigration consequences to prior convictions for “an offense that . . . involves fraud or deceit” with a loss exceeding $10,000. §1101(a)(43)(M)(i). The quoted language, we held, “mean[s] offenses with elements that necessarily entail fraudulent or deceitful conduct.” Id., at 484 (emphasis added). Consequently, no identification of generic offense elements was necessary; we simply asked whether the prior convictions before us met that measure. Id., at 483–485. See also, e.g., Stokeling v. United States, 586 U. S. ___, ___–___ (2019) (slip op., at 12–13) (determining whether an offense “has as an element the use, attempted use, or threatened use of physical force against the person of another,” 18 U. S. C. §924(e)(2)(B)(i)). This case invites us to decide which of the two categorical methodologies just described applies in determining whether a state offense is a “serious drug offense” under ACCA. ACCA defines that term to include: “an offense under State law, involving manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance (as defined in section 102 of the Controlled Substances Act (21 U. S. C. [§]802)), for which a maximum term of imprisonment of ten years or more is prescribed by law.” 18 U. S. C. §924(e)(2)(A)(ii). II Shular pleaded guilty in the United States District Court for the Northern District of Florida to possessing a firearm after having been convicted of a felony, in violation of §922(g)(1), and possessing with intent to distribute cocaine and cocaine base, in violation of 21 U. S. C. §841(a)(1) and (b)(1)(C). The District Court sentenced Shular to imprisonment for 15 years, the mandatory minimum under ACCA, to be followed by three years of supervised release. In imposing that enhanced sentence, the District Court took account of Shular’s prior convictions under Florida law. In 2012, Shular pleaded guilty to five counts of selling cocaine and one count of possessing cocaine with intent to sell, all in violation of Fla. Stat. §893.13(1)(a). That law makes it a crime to “sell, manufacture, or deliver, or possess with intent to sell, manufacture, or deliver, a controlled substance.” Ibid. For those offenses, “knowledge of the illicit nature of a controlled substance is not an element,” but lack of such knowledge “is an affirmative defense.” §893.101(2). Shular’s six convictions under that Florida law, the District Court concluded, qualified as “serious drug offense[s]” triggering ACCA enhancement under 18 U. S. C. §924(e)(2)(A)(ii). The United States Court of Appeals for the Eleventh Circuit affirmed the sentence. 736 Fed. Appx. 876 (2018). It relied on Circuit precedent holding that a court applying §924(e)(2)(A)(ii) “need not search for the elements of ‘generic’ definitions” of any offense, because the statute “require[s] only that the predicate offense ‘involv[e]’ . . . certain activities.” United States v. Smith, 775 F.3d 1262, 1267 (2014). Courts of Appeals have divided on whether §924(e)(2)(A)(ii)’s “serious drug offense” definition requires a comparison to a generic offense. Compare, e.g., id., at 1267 (no generic-offense comparison), with United States v. Franklin, 904 F.3d 793, 800 (CA9 2018) (court must define a generic crime). We granted certiorari to resolve this conflict, 588 U. S. ___ (2019), and now affirm the Eleventh Circuit’s judgment. III A The parties here agree that §924(e)(2)(A)(ii) requires a categorical approach. A court must look only to the state offense’s elements, not the facts of the case or labels pinned to the state conviction. They differ, however, on what comparison §924(e)(2)(A)(ii) requires. Shular would require “a generic-offense matching exercise”: A court should define the elements of the generic offenses identified in §924(e)(2)(A)(ii), then compare those elements to the elements of the state offense. Brief for Petitioner 13–14. In the Government’s view, a court should apply “the Kawashima categorical approach”: It should ask whether the state offense’s elements “necessarily entail one of the types of conduct” identified in §924(e)(2)(A)(ii). Brief for United States 13, 20 (emphasis added). This methodological dispute is occasioned by an interpretive disagreement over §924(e)(2)(A)(ii)’s reference to “manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance.” Those terms, in the Government’s view, describe conduct a court can compare directly against the state crime’s elements. Shular sees them instead as offenses whose elements a court must first expound. B The Government’s reading, we are convinced, correctly interprets the statutory text and context. Two features of §924(e)(2)(A)(ii), compared against a neighboring provision referring to offenses, §924(e)(2)(B)(ii), show that §924(e)(2)(A)(ii) refers to conduct. First, the terms in §924(e)(2)(A)(ii)—“manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance”—are unlikely names for generic offenses. Those words undoubtedly can be used to describe conduct. But as Shular acknowledges, they are not universal names of offenses; instead, States define “core drug offenses with all manner of terminology, including: trafficking, selling, giving, dispensing, distributing, delivering, promoting, and producing.” Reply Brief 7. Contrast §924(e)(2)(A)(ii) with §924(e)(2)(B)(ii), the enumerated-offense clause of ACCA’s “violent felony” definition, appearing in the same section of the Career Criminals Amendment Act of 1986, 100Stat. 3207–39 to 3207–40. That provision, which refers to a crime that “is burglary, arson, or extortion,” requires a generic-offense analysis. See Mathis, 579 U. S., at ___ (slip op., at 2). The terms “burglary,” “arson,” and “extortion”—given their common-law history and widespread usage—un- ambiguously name offenses. Cf., e.g., Taylor, 495 U. S., at 590–599 (discussing “burglary”). Drug offenses, Shular admits, lack “the same heritage and the same established lexicon.” Brief for Petitioner 14. Second, by speaking of activities a state-law drug offense “involv[es],” §924(e)(2)(A)(ii) suggests that the descriptive terms immediately following the word “involving” identify conduct. The parties agree that “involve” means “necessarily requir[e].” Brief for Petitioner 14 (citing Random House Dictionary of the English Language 1005 (2d ed. 1987) (“to include as a necessary circumstance, condition, or consequence”)); Brief for United States 21 (same). It is natural to say that an offense “involves” or “requires” certain conduct. E.g., §924(e)(2)(B)(ii) (addressing a crime “involv[ing] conduct that presents a serious potential risk of physical injury to another”); Mathis, 579 U. S., at ___ (slip op., at 5) (“The generic offense [of burglary] requires unlawful entry into a building or other structure.” (internal quotation marks omitted)). To refer to offenses as Shular urges, it would have been far more natural for the drafter to follow the enumerated-offense clause in using “is,” not “involving.” See §924(e)(2)(B)(ii) (crime that “is burglary, arson, or extortion”). There, the word “is” indicates a congruence between “crime” and the terms that follow, terms that are also crimes. See American Heritage Dictionary 114 (def. 7a) (1981) (“To equal in meaning or identity”). Yet Congress did not adopt that formulation in §924(e)(2)(A)(ii), opting instead for language suited to conduct. C Shular principally urges that at the time of §924(e)(2)(A)(ii)’s enactment, federal and state criminal laws widely prohibited the “core conduct” of manufacturing, distributing, and possessing with intent to manufacture or distribute drugs. Brief for Petitioner 10–12. Some laws, Shular observes, used those very terms. See, e.g., 21 U. S. C. §841(a)(1) (1982 ed.). But even if the substance of state drug laws was well established—rather than their nomenclature, which Shular concedes was not—Congress could capture that substance by reference to conduct, rather than offenses. Shular points out that the word “involving” can accommodate a generic-offense approach. Cf. Scheidler v. National Organization for Women, Inc., 537 U.S. 393, 409 (2003) (“act or threat involving . . . extortion,” 18 U. S. C. §1961(1), contemplates “ ‘generic’ extortion” (some internal quotation marks omitted)). But we have no reason to think Congress intended that approach for §924(e)(2)(A)(ii)—which uses no deeply rooted offense name like “extortion” and contrasts with the offense-oriented language of a neighboring provision. Endeavoring to explain why Congress might have chosen “involving” over “is” in §924(e)(2)(A)(ii), Shular suggests that variation in state drug-offense terminology required a word more approximate than “is.” But if Congress was concerned that state drug offenses lacked clear, universally employed names, the evident solution was to identify them instead by conduct. Using “involving” rather than “is” does not clarify that the terms are names of offenses; quite the opposite. See supra, at 7. Shular asserts that to describe conduct rather than offenses, Congress would have used the language of the elements clause of the “violent felony” definition, which captures a crime that “has as an element the use, attempted use, or threatened use of physical force against the person of another.” §924(e)(2)(B)(i) (emphasis added). It would have been awkward, however, to describe “possessing with intent to manufacture or distribute”—requiring both possession and intent—as “an element.” Congress may also have wanted to clarify that the state offense need not include the identified conduct as a formal element. Cf. Kawashima, 565 U. S., at 483–484 (the statutory phrase “an offense that . . . involves fraud or deceit” “is not limited to offenses that include fraud or deceit as formal elements” but extends to offenses “that necessarily entail fraudulent or deceitful conduct”). Whatever the reason, Congress’ choice not to describe each term in §924(e)(2)(A)(ii) as “an element” neither refutes that those terms refer to conduct nor shows that they refer to offenses. Nor does the other clause of the “serious drug offense” definition shed light on the question before us. Section 924(e)(2)(A)(i) includes as “serious drug offenses” “offense[s] under” specific portions of the U. S. Code.[2] That provision, Shular observes, refers to fully defined crimes. But “the divergent text of the two provisions” of the serious-drug-offense definition, as the Government explains, “makes any divergence in their application unremarkable.” Brief for United States 22. Congress’ decision to identify federal offenses by reference to the U. S. Code does not speak to whether it identified state offenses by reference to named offenses or conduct. D Shular expresses concern that rejecting a generic-offense approach would yield an anomalous result. Unlike other drug laws, Shular contends, the Florida law under which he was previously convicted does not require that the defendant know the substance is illicit. Unless §924(e)(2)(A)(ii) takes into account all the elements of the offense as commonly understood, Shular maintains, defendants would face ACCA enhancement based on outlier state laws. As an initial matter, Shular overstates Florida’s disregard for mens rea. Charged under Fla. Stat. §893.13(1)(a), a defendant unaware of the substance’s illicit nature can raise that unawareness as an affirmative defense, in which case the standard jury instructions require a finding of knowledge beyond a reasonable doubt. §893.101(2); Fla. Crim. Jury Instr. §25.2 (2020), https:// www . floridasupremecourt . org / content / download / 568865/ 6425767/file/EntireDocument.rtf. In any event, both parties’ interpretations of 18 U. S. C. §924(e)(2)(A)(ii) achieve a measure of consistency. Resolving this case requires us to determine which form of consistency Congress intended: application of ACCA to all offenders who engaged in certain conduct or to all who committed certain generic offenses (in either reading, judging only by the elements of their prior convictions). For the reasons explained, we are persuaded that Congress chose the former. E Shular urges us to apply the rule of lenity in determining whether §924(e)(2)(A)(ii) requires a generic-offense-matching analysis. The rule “applies only when, after consulting traditional canons of statutory construction, we are left with an ambiguous statute.” United States v. Shabani, 513 U.S. 10, 17 (1994). Here, we are left with no ambiguity for the rule of lenity to resolve. Section 924(e)(2)(A)(ii)’s text and context leave no doubt that it refers to an offense involving the conduct of “manufacturing, distributing, or possessing with intent to manufacture or distribute, a controlled substance.” Because those terms describe conduct and do not name offenses, a court applying §924(e)(2)(A)(ii) need not delineate the elements of generic offenses.[3] * * * For the reasons stated, the judgment of the Court of Appeals for the Eleventh Circuit is Affirmed. Notes 1 We have also used the term “generic crime” to mean the crime “in general” as opposed to “the specific acts in which an offender engaged on a specific occasion.” Nijhawan v. Holder, 557 U.S. 29, 33–34 (2009). That is not the sense in which we use “generic” in this opinion. 2 Section 924(e)(2)(A)(i) provides that the term “serious drug offense” includes “an offense under the Controlled Substances Act (21 U. S. C. [§]801 et seq.), the Controlled Substances Import and Export Act (21 U. S. C. [§]951 et seq.), or chapter 705 of title 46 for which a maximum term of imprisonment of ten years or more is prescribed by law.” 3 Shular argues in the alternative that even if §924(e)(2)(A)(ii) does not call for a generic-offense-matching analysis, it requires knowledge of the substance’s illicit nature. See Brief for Petitioner 23; Reply Brief 8–10. We do not address that argument. Not only does it fall outside the question presented, Pet. for Cert. i, Shular disclaimed it at the certiorari stage, Supp. Brief for Petitioner 3.

590.US.2019_17-1712

Plaintiffs James Thole and Sherry Smith are retired participants in U. S. Bank’s defined-benefit retirement plan, which guarantees them a fixed payment each month regardless of the plan’s value or its fiduciaries’ good or bad investment decisions. Both have been paid all of their monthly pension benefits so far and are legally and contractually entitled to those payments for the rest of their lives. Nevertheless, they filed a putative class-action suit against U. S. Bank and others (collectively, U. S. Bank) under the Employee Retirement Income Security Act of 1974 (ERISA), alleging that the defendants violated ERISA’s duties of loyalty and prudence by poorly investing the plan’s assets. They request the repayment of approximately $750 million to the plan in losses suffered due to mismanagement; injunctive relief, including replacement of the plan’s fiduciaries; and attorney’s fees. The District Court dismissed the case, and the Eighth Circuit affirmed on the ground that the plaintiffs lack statutory standing. Held: Because Thole and Smith have no concrete stake in the lawsuit, they lack Article III standing. See Lujan v. Defenders of Wildlife, 504 U.S. 555, 560–561. Win or lose, they would still receive the exact same monthly benefits they are already entitled to receive. None of the plaintiffs’ arguments suffices to establish Article III standing. First, the plaintiffs rely on a trust analogy in arguing that an ERISA participant has an equitable or property interest in the plan and that injuries to the plan are therefore injuries to the participants. But participants in a defined-benefit plan are not similarly situated to the beneficiaries of a private trust or to participants in a defined- contribution plan, and they possess no equitable or property interest in the plan, see Hughes Aircraft Co. v. Jacobson, 525 U.S. 432, 439–441. Second, the plaintiffs cannot assert representative standing based on injuries to the plan where they themselves have not “suffered an injury in fact,” Hollingsworth v. Perry, 570 U.S. 693, 708, or been legally or contractually appointed to represent the plan. Third, the fact that ERISA affords all participants—including defined-benefit plan participants—a cause of action to sue does not satisfy the injury-in-fact requirement here. “Article III standing requires a concrete injury even in the context of a statutory violation.” Spokeo, Inc. v. Robins, 578 U. S. ___, ___. Fourth, the plaintiffs contend that meaningful regulation of plan fiduciaries is possible only if they may sue to target perceived fiduciary misconduct. But this Court has long rejected that argument for Article III standing, see Valley Forge Christian College v. Americans United for Separation of Church and State, Inc., 454 U.S. 464, 489, and defined-benefit plans are regulated and monitored in multiple ways. The plaintiffs’ amici assert that defined-benefit plan participants have standing to sue if the plan’s mismanagement was so egregious that it substantially increased the risk that the plan and the employer would fail and be unable to pay the participants’ future benefits. The plaintiffs do not assert that theory of standing here, nor did their complaint allege that level of mismanagement. Pp. 2–8. 873 F.3d 617, affirmed. Kavanaugh, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Alito, and Gorsuch, JJ., joined. Thomas, J., filed a concurring opinion, in which Gorsuch, J., joined. Sotomayor, J., filed a dissenting opinion, in which Ginsburg, Breyer, and Kagan, JJ., joined.

To establish standing under Article III of the Constitution, a plaintiff must demonstrate (1) that he or she suffered an injury in fact that is concrete, particularized, and actual or imminent, (2) that the injury was caused by the defendant, and (3) that the injury would likely be redressed by the requested judicial relief. See Lujan v. Defenders of Wildlife, 504 U.S. 555, 560–561 (1992). Plaintiffs James Thole and Sherry Smith are two retired participants in U. S. Bank’s retirement plan. Of decisive importance to this case, the plaintiffs’ retirement plan is a defined-benefit plan, not a defined-contribution plan. In a defined-benefit plan, retirees receive a fixed payment each month, and the payments do not fluctuate with the value of the plan or because of the plan fiduciaries’ good or bad investment decisions. By contrast, in a defined-contribution plan, such as a 401(k) plan, the retirees’ benefits are typically tied to the value of their accounts, and the benefits can turn on the plan fiduciaries’ particular investment decisions. See Beck v. PACE Int’l Union, 551 U.S. 96, 98 (2007); Hughes Aircraft Co. v. Jacobson, 525 U.S. 432, 439–440 (1999). As retirees and vested participants in U. S. Bank’s defined-benefit plan, Thole receives $2,198.38 per month, and Smith receives $42.26 per month, regardless of the plan’s value at any one moment and regardless of the investment decisions of the plan’s fiduciaries. Thole and Smith have been paid all of their monthly pension benefits so far, and they are legally and contractually entitled to receive those same monthly payments for the rest of their lives. Even though the plaintiffs have not sustained any monetary injury, they filed a putative class-action suit against U. S. Bank and others (collectively, U. S. Bank) for alleged mismanagement of the defined-benefit plan. The alleged mismanagement occurred more than a decade ago, from 2007 to 2010. The plaintiffs sued under ERISA, the aptly named Employee Retirement Income Security Act of 1974, 88Stat. 829, as amended, 29 U. S. C. §1001 et seq. The plaintiffs claimed that the defendants violated ERISA’s duties of loyalty and prudence by poorly investing the assets of the plan. The plaintiffs requested that U. S. Bank repay the plan approximately $750 million in losses that the plan allegedly suffered. The plaintiffs also asked for injunctive relief, including replacement of the plan’s fiduciaries. See ERISA §§502(a)(2), (3), 29 U. S. C. §§1132(a)(2), (3). No small thing, the plaintiffs also sought attorney’s fees. In the District Court, the plaintiffs’ attorneys requested at least $31 million in attorney’s fees. The U. S. District Court for the District of Minnesota dismissed the case, and the U. S. Court of Appeals for the Eighth Circuit affirmed on the ground that the plaintiffs lack statutory standing. 873 F.3d 617 (2017). We granted certiorari. 588 U. S. ___ (2019). We affirm the judgment of the U. S. Court of Appeals for the Eighth Circuit on the ground that the plaintiffs lack Article III standing. Thole and Smith have received all of their monthly benefit payments so far, and the outcome of this suit would not affect their future benefit payments. If Thole and Smith were to lose this lawsuit, they would still receive the exact same monthly benefits that they are already slated to receive, not a penny less. If Thole and Smith were to win this lawsuit, they would still receive the exact same monthly benefits that they are already slated to receive, not a penny more. The plaintiffs therefore have no concrete stake in this lawsuit. To be sure, their attorneys have a stake in the lawsuit, but an “interest in attorney’s fees is, of course, insufficient to create an Article III case or controversy where none exists on the merits of the underlying claim.” Lewis v. Continental Bank Corp., 494 U.S. 472, 480 (1990); see Steel Co. v. Citizens for Better Environment, 523 U.S. 83, 107 (1998) (same). Because the plaintiffs themselves have no concrete stake in the lawsuit, they lack Article III standing. * * * If Thole and Smith had not received their vested pension benefits, they would of course have Article III standing to sue and a cause of action under ERISA §502(a)(1)(B) to recover the benefits due to them. See 29 U. S. C. §1132(a)(1)(B). But Thole and Smith have received all of their monthly pension benefits so far, and they will receive those same monthly payments for the rest of their lives. To nonetheless try to demonstrate their standing to challenge alleged plan mismanagement, the plaintiffs have advanced four alternative arguments. First, analogizing to trust law, Thole and Smith contend that an ERISA defined-benefit plan participant possesses an equitable or property interest in the plan, meaning in essence that injuries to the plan are by definition injuries to the plan participants. Thole and Smith contend, in other words, that a plan fiduciary’s breach of a trust-law duty of prudence or duty of loyalty itself harms ERISA defined-benefit plan participants, even if the participants themselves have not suffered (and will not suffer) any monetary losses. The basic flaw in the plaintiffs’ trust-based theory of standing is that the participants in a defined-benefit plan are not similarly situated to the beneficiaries of a private trust or to the participants in a defined-contribution plan. See Varity Corp. v. Howe, 516 U.S. 489, 497 (1996) (trust law informs but does not control interpretation of ERISA). In the private trust context, the value of the trust property and the ultimate amount of money received by the beneficiaries will typically depend on how well the trust is managed, so every penny of gain or loss is at the beneficiaries’ risk. By contrast, a defined-benefit plan is more in the nature of a contract. The plan participants’ benefits are fixed and will not change, regardless of how well or poorly the plan is managed. The benefits paid to the participants in a defined-benefit plan are not tied to the value of the plan. Moreover, the employer, not plan participants, receives any surplus left over after all of the benefits are paid; the employer, not plan participants, is on the hook for plan shortfalls. See Beck, 551 U. S., at 98–99. As this Court has stated before, plan participants possess no equitable or property interest in the plan. See Hughes Aircraft Co., 525 U. S., at 439–441; see also LaRue v. DeWolff, Boberg & Associates, Inc., 552 U.S. 248, 254–256 (2008). The trust-law analogy therefore does not fit this case and does not support Article III standing for plaintiffs who allege mismanagement of a defined-benefit plan. Second, Thole and Smith assert standing as representatives of the plan itself. But in order to claim “the interests of others, the litigants themselves still must have suffered an injury in fact, thus giving” them “a sufficiently concrete interest in the outcome of the issue in dispute.” Hollingsworth v. Perry, 570 U.S. 693, 708 (2013) (internal quotation marks omitted); cf. Gollust v. Mendell, 501 U.S. 115, 125–126 (1991) (suggesting that shareholder must “maintain some continuing financial stake in the litigation” in order to have Article III standing to bring an insider trading suit on behalf of the corporation); Craig v. Boren, 429 U.S. 190, 194–195 (1976) (vendor who “independently” suffered an Article III injury in fact could then assert the rights of her customers). The plaintiffs themselves do not have a concrete stake in this suit. The plaintiffs point to the Court’s decisions upholding the Article III standing of assignees—that is, where a party’s right to sue has been legally or contractually assigned to another party. But here, the plan’s claims have not been legally or contractually assigned to Thole or Smith. Cf. Sprint Communications Co. v. APCC Services, Inc., 554 U.S. 269, 290 (2008); Vermont Agency of Natural Resources v. United States ex rel. Stevens, 529 U.S. 765, 771–774 (2000) (qui tam statute makes a relator a partial assignee and “gives the relator himself an interest in the lawsuit”) (emphasis deleted). The plaintiffs’ invocation of cases involving guardians, receivers, and executors falls short for basically the same reason. The plaintiffs have not been legally or contractually appointed to represent the plan. Third, in arguing for standing, Thole and Smith stress that ERISA affords the Secretary of Labor, fiduciaries, beneficiaries, and participants—including participants in a defined-benefit plan—a general cause of action to sue for restoration of plan losses and other equitable relief. See ERISA §§502(a)(2), (3), 29 U. S. C. §§1132(a)(2), (3). But the cause of action does not affect the Article III standing analysis. This Court has rejected the argument that “a plaintiff automatically satisfies the injury-in-fact requirement whenever a statute grants a person a statutory right and purports to authorize that person to sue to vindicate that right.” Spokeo, Inc. v. Robins, 578 U. S. ___, ___ (2016) (slip op., at 9); see Raines v. Byrd, 521 U.S. 811, 820, n. 3 (1997). The Court has emphasized that “Article III standing requires a concrete injury even in the context of a statutory violation.” Spokeo, 578 U. S., at ___ (slip op., at 9). Here, the plaintiffs have failed to plausibly and clearly allege a concrete injury.[1] Fourth, Thole and Smith contend that if defined-benefit plan participants may not sue to target perceived fiduciary misconduct, no one will meaningfully regulate plan fiduciaries. For that reason, the plaintiffs suggest that defined-benefit plan participants must have standing to sue. But this Court has long rejected that kind of argument for Article III standing. See Valley Forge Christian College v. Americans United for Separation of Church and State, Inc., 454 U.S. 464, 489 (1982) (the “ ‘assumption that if respondents have no standing to sue, no one would have standing, is not a reason to find standing’ ”) (quoting Schlesinger v. Reservists Comm. to Stop the War, 418 U.S. 208, 227 (1974)). In any event, the argument rests on a faulty premise in this case because defined-benefit plans are regulated and monitored in multiple ways. To begin with, employers and their shareholders often possess strong incentives to root out fiduciary misconduct because the employers are entitled to the plan surplus and are often on the hook for plan shortfalls. Therefore, about the last thing a rational employer wants or needs is a mismanaged retirement plan. Cf. ERISA §4062(a), 29 U. S. C. §1362(a). Moreover, ERISA expressly authorizes the Department of Labor to enforce ERISA’s fiduciary obligations. See ERISA §502(a)(2), 29 U. S. C. §1132(a)(2). And the Department of Labor has a substantial motive to aggressively pursue fiduciary misconduct, particularly to avoid the financial burden of failed defined-benefit plans being backloaded onto the Federal Government. When a defined-benefit plan fails and is unable to pay benefits to retirees, the federal Pension Benefit Guaranty Corporation is required by law to pay the vested pension benefits of the retirees, often in full. The Department of Labor is well positioned to understand the relationship between plan failure and the PBGC because, by law, the PBGC operates within the Department of Labor, and the Secretary of Labor chairs the Board of the PBGC. See ERISA §§4002(a), (d), 29 U. S. C. §§1302(a), (d). On top of all that, fiduciaries (including trustees who are fiduciaries) can sue other fiduciaries—and they would have good reason to sue if, as Thole and Smith posit, one fiduciary were using the plan’s assets as a “personal piggybank.” Brief for Petitioners 2. In addition, depending on the nature of the fiduciary misconduct, state and federal criminal laws may apply. See, e.g., 18 U. S. C. §§664, 1954; ERISA §514(b)(4), 29 U. S. C. §1144(b)(4). In short, under ERISA, fiduciaries who manage defined-benefit plans face a regulatory phalanx. In sum, none of the plaintiffs’ four theories supports their Article III standing in this case. One last wrinkle remains. According to the plaintiffs’ amici, plan participants in a defined-benefit plan have standing to sue if the mismanagement of the plan was so egregious that it substantially increased the risk that the plan and the employer would fail and be unable to pay the participants’ future pension benefits. Cf. Clapper v. Amnesty Int’l USA, 568 U.S. 398, 414, n. 5 (2013); Lee v. Verizon Communications, Inc., 837 F.3d 523, 545–546 (CA5 2016); David v. Alphin, 704 F.3d 327, 336–338 (CA4 2013). But the plaintiffs do not assert that theory of standing in this Court. In any event, the plaintiffs’ complaint did not plausibly and clearly claim that the alleged mismanagement of the plan substantially increased the risk that the plan and the employer would fail and be unable to pay the plaintiffs’ future pension benefits. It is true that the plaintiffs’ complaint alleged that the plan was underfunded for a period of time. But a bare allegation of plan underfunding does not itself demonstrate a substantially increased risk that the plan and the employer would both fail. Cf. LaRue, 552 U. S., at 255 (“Misconduct by the administrators of a defined benefit plan will not affect an individual’s entitlement to a defined benefit unless it creates or enhances the risk of default by the entire plan”).[2] * * * Courts sometimes make standing law more complicated than it needs to be. There is no ERISA exception to Article III. And under ordinary Article III standing analysis, the plaintiffs lack Article III standing for a simple, commonsense reason: They have received all of their vested pension benefits so far, and they are legally entitled to receive the same monthly payments for the rest of their lives. Winning or losing this suit would not change the plaintiffs’ monthly pension benefits. The plaintiffs have no concrete stake in this dispute and therefore lack Article III standing. We affirm the judgment of the U. S. Court of Appeals for the Eighth Circuit. It is so ordered. Notes 1 To be clear, our decision today does not concern suits to obtain plan information. See, e.g., ERISA §502(a)(1)(A), 29 U. S. C. §1132(a)(1)(A). 2 Even if a defined-benefit plan is mismanaged into plan termination, the federal PBGC by law acts as a backstop and covers the vested pension benefits up to a certain amount and often in full. For example, if the plan and the employer in this case were to fail, the PBGC would be required to pay these two plaintiffs all of their vested pension benefits in full. See ERISA §§4022(a), (b), 29 U. S. C. §§1322(a), (b); Tr. of Oral Arg. 18–19; see also Congressional Research Service, Pension Benefit Guaranty Corporation (PBGC): A Primer 1 (2019); PBGC, General FAQs About PBGC, https://www.pbgc.gov/about/faq/general-faqs-about-pbgc. Any increased-risk-of-harm theory of standing therefore might not be available for plan participants whose benefits are guaranteed in full by the PBGC. But we need not decide that question in this case.

590.US.2019_18-916

Inter partes review is an administrative process that permits a patent challenger to ask the U. S. Patent and Trademark Office to reconsider the validity of earlier granted patent claims. For inter partes review to proceed, the agency must agree to institute review. See 35 U. S. C. §314. Among other conditions set by statute, if a request comes more than a year after suit against the requesting party for patent infringement, “[a]n inter partes review may not be instituted.” §315(b). The agency’s “determination . . . whether to institute an inter partes review under this section shall be final and nonappealable.” §314(d). Entities associated with petitioner Thryv, Inc. sought inter partes review of a patent owned by respondent Click-to-Call Technologies, LP. Click-to-Call countered that the petition was untimely under §315(b). The Patent Trial and Appeal Board (Board) disagreed and instituted review. After proceedings on the merits, the Board issued a final written decision reiterating its §315(b) decision and canceling 13 of the patent’s claims as obvious or lacking novelty. Click-to-Call appealed the Board’s §315(b) determination. Treating the Board’s application of §315(b) as judicially reviewable, the Court of Appeals concluded that the petition was untimely, vacated the Board’s decision, and remanded with instructions to dismiss. Held: Section 314(d) precludes judicial review of the agency’s application of §315(b)’s time prescription. Pp. 6–14. (a) A party generally cannot contend on appeal that the agency should have refused “to institute an inter partes review.” §314(d). That follows from §314(d)’s text and Cuozzo Speed Technologies, LLC v. Lee, 579 U. S. ___. In Cuozzo, this Court explained that §314(d) “preclud[es] review of the Patent Office’s institution decisions”—at least “where the grounds for attacking the decision to institute inter partes review consist of questions that are closely tied to the application and interpretation of statutes related to the Patent Office’s decision to initiate inter partes review.” Id., at ___. Pp. 6–7. (b) The question here is whether a challenge based on §315(b) ranks as an appeal of the agency’s decision “to institute an inter partes review.” §314(d). There is no need to venture beyond Cuozzo’s holding that §314(d) bars review at least of matters “closely tied to the application and interpretation of statutes related to” the institution decision, 579 U. S., at ___. A §315(b) challenge easily meets that measurement. Section 315(b), setting forth a circumstance in which “[a]n inter partes review may not be instituted,” expressly governs institution and nothing more. Pp. 7–8. (c) This conclusion is strongly reinforced by the statute’s purpose and design. Congress designed inter partes review to weed out bad patent claims efficiently. Allowing §315(b) appeals, however, would unwind agency proceedings determining patentability and leave bad patents enforceable. Pp. 8–10. (d) In Click-to-Call’s view, §314(d)’s bar on judicial review is limited to the agency’s threshold determination under §314(a) of the question whether the petitioner has a reasonable likelihood of prevailing. Cuozzo is fatal to that interpretation, for the Court in that case held unreviewable the agency’s application of a provision other than §314(a). Contrary to Click-to-Call’s contention, §314(d)’s text does not limit the review bar to §314(a). Rather than borrowing language from related provisions that would have achieved Click-to-Call’s preferred meaning, Congress used broader language in §314(d). Click-to-Call also insists that Congress intended judicial supervision of the agency’s application of §315(b), but the statute instead reflects a choice to entrust that issue to the agency. Finally, SAS Institute Inc. v. Iancu, 584 U. S. ___, offers Click-to-Call no assistance. Unlike the appeal held reviewable in SAS Institute, Click-to-Call’s appeal challenges not the manner in which the agency’s review proceeds once instituted, but whether the agency should have instituted review at all. Pp. 10–13. (e) Click-to-Call argues in the alternative that its §315(b) objection is authorized as an appeal from the Board’s final written decision, which addressed the §315(b) issue. Even labeled that way, Click-to-Call’s appeal is still barred by §314(d) because Click-to-Call’s contention remains, essentially, that the agency should have refused to institute inter partes review. P. 14. 899 F.3d 1321, vacated and remanded. Ginsburg, J., delivered the opinion of the Court, in which Roberts, C. J., and Breyer, Kagan, and Kavanaugh, JJ., joined, and in which Thomas and Alito, JJ., joined except as to Part III–C. Gorsuch, J., filed a dissenting opinion, in which Sotomayor, J., joined as to Parts I, II, III, and IV.

of questions that are closely tied to the application and interpretation of statutes related to” the institution decision. 579 U. S., at ___, ___ (slip op., at 7, 11). The appeal bar, we therefore reiterate, is not limited to the agency’s application of §314(a).

591.US.2019_19-715

In April 2019, three committees of the U. S. House of Representatives issued four subpoenas seeking information about the finances of President Donald J. Trump, his children, and affiliated businesses. The House Committee on Financial Services issued a subpoena to Deutsche Bank seeking any document related to account activity, due diligence, foreign transactions, business statements, debt schedules, statements of net worth, tax returns, and suspicious activity identified by Deutsche Bank. It issued a second subpoena to Capital One for similar information. The Permanent Select Committee on Intelligence issued a subpoena to Deutsche Bank that mirrored the subpoena issued by the Financial Services Committee. And the House Committee on Oversight and Reform issued a subpoena to the President’s personal accounting firm, Mazars USA, LLP, demanding information related to the President and several affiliated businesses. Although each of the committees sought overlapping sets of financial documents, each supplied different justifications for the requests, explaining that the information would help guide legislative reform in areas ranging from money laundering and terrorism to foreign involvement in U. S. elections. Petitioners—the President in his personal capacity, along with his children and affiliated businesses—contested the subpoena issued by the Oversight Committee in the District Court for the District of Columbia (Mazars, No. 19–715) and the subpoenas issued by the Financial Services and Intelligence Committees in the Southern District of New York (Deutsche Bank, No. 19–760). In both cases, petitioners contended that the subpoenas lacked a legitimate legislative purpose and violated the separation of powers. The President did not, however, argue that any of the requested records were protected by executive privilege. In Mazars, the District Court granted judgment for the House and the D. C. Circuit affirmed, finding that the subpoena issued by the Oversight Committee served a valid legislative purpose because the requested information was relevant to reforming financial disclosure requirements for Presidents and presidential candidates. In Deutsche Bank, the District Court denied a preliminary injunction and the Second Circuit affirmed in substantial part, holding that the Intelligence Committee properly issued its subpoena to Deutsche Bank as part of an investigation into alleged foreign influence in the U. S. political process, which could inform legislation to strengthen national security and combat foreign meddling. The court also concluded that the subpoenas issued by the Financial Services Committee to Deutsche Bank and Capital One were adequately related to potential legislation on money laundering, terrorist financing, and the global movement of illicit funds through the real estate market. Held: The courts below did not take adequate account of the significant separation of powers concerns implicated by congressional subpoenas for the President’s information. Pp. 7–20. (a) Historically, disputes over congressional demands for presidential documents have been resolved by the political branches through negotiation and compromise without involving this Court. The Court recognizes that this dispute is the first of its kind to reach the Court; that such disputes can raise important issues concerning relations between the branches; that similar disputes recur on a regular basis, including in the context of deeply partisan controversy; and that Congress and the Executive have nonetheless managed for over two centuries to resolve these disputes among themselves without Supreme Court guidance. Such longstanding practice “ ‘is a consideration of great weight’ ” in cases concerning “the allocation of power between [the] two elected branches of Government,” and it imposes on the Court a duty of care to ensure that it does not needlessly disturb “the compromises and working arrangements” reached by those branches. NLRB v. Noel Canning, 573 U.S. 513, 524–526 (quoting The Pocket Veto Case, 279 U.S. 655, 689). Pp. 7–11. (b) Each House of Congress has the power “to secure needed information” in order to legislate. McGrain v. Daugherty, 273 U.S. 135, 161. This power is “indispensable” because, without information, Congress would be unable to legislate wisely or effectively. Watkins v. United States, 354 U.S. 178, 215. Because this power is “justified solely as an adjunct to the legislative process,” it is subject to several limitations. Id., at 197. Most importantly, a congressional subpoena is valid only if it is “related to, and in furtherance of, a legitimate task of the Congress.” Id., at 187. The subpoena must serve a “valid legislative purpose.” Quinn v. United States, 349 U.S. 155, 161. Furthermore, Congress may not issue a subpoena for the purpose of “law enforcement,” because that power is assigned to the Executive and the Judiciary. Ibid. Finally, recipients of congressional subpoenas retain their constitutional rights and various privileges throughout the course of an investigation. Pp. 11–12. (c) The President contends, as does the Solicitor General on behalf of the United States, that congressional subpoenas for the President’s information should be evaluated under the standards set forth in United States v. Nixon, 418 U.S. 683, and Senate Select Committee on Presidential Campaign Activities v. Nixon, 498 F.2d 725, which would require the House to show that the requested information satisfies a “demonstrated, specific need,” 418 U. S., at 713, and is “demonstrably critical” to a legislative purpose, 498 F. 2d, at 731. Nixon and Senate Select Committee, however, involved subpoenas for communications between the President and his close advisers, over which the President asserted executive privilege. Because executive privilege safeguards the public interest in candid, confidential deliberations within the Executive Branch, information subject to the privilege deserves “the greatest protection consistent with the fair administration of justice.” 418 U. S., at 715. That protection should not be transplanted root and branch to cases involving nonprivileged, private information, which by definition does not implicate sensitive Executive Branch deliberations. The standards proposed by the President and the Solicitor General—if applied outside the context of privileged information—would risk seriously impeding Congress in carrying out its responsibilities, giving short shrift to its important interests in conducting inquiries to obtain information needed to legislate effectively. Pp. 12–14. (d) The approach proposed by the House, which relies on precedents that did not involve the President’s papers, fails to take adequate account of the significant separation of powers issues raised by congressional subpoenas for the President’s information. The House’s approach would leave essentially no limits on the congressional power to subpoena the President’s personal records. A limitless subpoena power could transform the established practice of the political branches and allow Congress to aggrandize itself at the President’s expense. These separation of powers concerns are unmistakably implicated by the subpoenas here, which represent not a run-of-the-mill legislative effort but rather a clash between rival branches of government over records of intense political interest for all involved. The interbranch conflict does not vanish simply because the subpoenas seek personal papers or because the President sued in his personal capacity. Nor are separation of powers concerns less palpable because the subpoenas were issued to third parties. Pp. 14–18. (e) Neither side identifies an approach that adequately accounts for these weighty separation of powers concerns. A balanced approach is necessary, one that takes a “considerable impression” from “the practice of the government,” McCulloch v. Maryland, 4 Wheat. 316, 401, and “resist[s]” the “pressure inherent within each of the separate Branches to exceed the outer limits of its power,” INS v. Chadha, 462 U.S. 919, 951. In assessing whether a subpoena directed at the President’s personal information is “related to, and in furtherance of, a legitimate task of the Congress,” Watkins, 354 U. S., at 187, courts must take adequate account of the separation of powers principles at stake, including both the significant legislative interests of Congress and the unique position of the President. Several special considerations inform this analysis. First, courts should carefully assess whether the asserted legislative purpose warrants the significant step of involving the President and his papers. “ ‘[O]ccasion[s] for constitutional confrontation between the two branches’ should be avoided whenever possible.” Cheney v. United States Dist. Court for D. C., 542 U.S. 367, 389–390 (quoting Nixon, 418 U. S., at 692). Congress may not rely on the President’s information if other sources could reasonably provide Congress the information it needs in light of its particular legislative objective. Second, to narrow the scope of possible conflict between the branches, courts should insist on a subpoena no broader than reasonably necessary to support Congress’s legislative objective. The specificity of the subpoena’s request “serves as an important safeguard against unnecessary intrusion into the operation of the Office of the President.” Cheney, 542 U. S., at 387. Third, courts should be attentive to the nature of the evidence offered by Congress to establish that a subpoena advances a valid legislative purpose. The more detailed and substantial, the better. That is particularly true when Congress contemplates legislation that raises sensitive constitutional issues, such as legislation concerning the Presidency. Fourth, courts should assess the burdens imposed on the President by a subpoena, particularly because they stem from a rival political branch that has an ongoing relationship with the President and incentives to use subpoenas for institutional advantage. Other considerations may be pertinent as well; one case every two centuries does not afford enough experience for an exhaustive list. Pp. 18–20. No. 19–715, 940 F.3d 710; No. 19–760, 943 F.3d 627, vacated and remanded. Roberts, C. J., delivered the opinion of the Court, in which Ginsburg, Breyer, Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Thomas, J., and Alito, J., filed dissenting opinions. Notes 1 Together with 19–760, Trump et al. v. Deutsche Bank AG et al., on certiorari to the United States Court of Appeals for the Second Circuit.

Over the course of five days in April 2019, three committees of the U. S. House of Representatives issued four subpoenas seeking information about the finances of President Donald J. Trump, his children, and affiliated businesses. We have held that the House has authority under the Constitution to issue subpoenas to assist it in carrying out its legislative responsibilities. The House asserts that the financial information sought here—encompassing a decade’s worth of transactions by the President and his family—will help guide legislative reform in areas ranging from money laundering and terrorism to foreign involvement in U. S. elections. The President contends that the House lacked a valid legislative aim and instead sought these records to harass him, expose personal matters, and conduct law enforcement activities beyond its authority. The question presented is whether the subpoenas exceed the authority of the House under the Constitution. We have never addressed a congressional subpoena for the President’s information. Two hundred years ago, it was established that Presidents may be subpoenaed during a federal criminal proceeding, United States v. Burr, 25 F. Cas. 30 (No. 14,692d) (CC Va. 1807) (Marshall, Cir. J.), and earlier today we extended that ruling to state criminal proceedings, Trump v. Vance, ante, p. ___. Nearly fifty years ago, we held that a federal prosecutor could obtain information from a President despite assertions of executive privilege, United States v. Nixon, 418 U.S. 683 (1974), and more recently we ruled that a private litigant could subject a President to a damages suit and appropriate discovery obligations in federal court, Clinton v. Jones, 520 U.S. 681 (1997). This case is different. Here the President’s information is sought not by prosecutors or private parties in connection with a particular judicial proceeding, but by committees of Congress that have set forth broad legislative objectives. Congress and the President—the two political branches established by the Constitution—have an ongoing relationship that the Framers intended to feature both rivalry and reciprocity. See The Federalist No. 51, p. 349 (J. Cooke ed. 1961) (J. Madison); Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 635 (1952) (Jackson, J., concurring). That distinctive aspect necessarily informs our analysis of the question before us. I A Each of the three committees sought overlapping sets of financial documents, but each supplied different justifications for the requests. The House Committee on Financial Services issued two subpoenas, both on April 11, 2019. App. 128, 154, 226. The first, issued to Deutsche Bank, seeks the financial information of the President, his children, their immediate family members, and several affiliated business entities. Specifically, the subpoena seeks any document related to account activity, due diligence, foreign transactions, business statements, debt schedules, statements of net worth, tax returns, and suspicious activity identified by Deutsche Bank. The second, issued to Capital One, demands similar financial information with respect to more than a dozen business entities associated with the President. The Deutsche Bank subpoena requests materials from “2010 through the present,” and the Capital One subpoena covers “2016 through the present,” but both subpoenas impose no time limitations for certain documents, such as those connected to account openings and due diligence. Id., at 128, 155. According to the House, the Financial Services Committee issued these subpoenas pursuant to House Resolution 206, which called for “efforts to close loopholes that allow corruption, terrorism, and money laundering to infiltrate our country’s financial system.” H. Res. 206, 116th Cong., 1st Sess., 5 (Mar. 13, 2019). Such loopholes, the resolution explained, had allowed “illicit money, including from Russian oligarchs,” to flow into the United States through “anonymous shell companies” using investments such as “luxury high-end real estate.” Id., at 3. The House also invokes the oversight plan of the Financial Services Committee, which stated that the Committee intends to review banking regulation and “examine the implementation, effectiveness, and enforcement” of laws designed to prevent money laundering and the financing of terrorism. H. R. Rep. No. 116–40, p. 84 (2019). The plan further provided that the Committee would “consider proposals to prevent the abuse of the financial system” and “address any vulnerabilities identified” in the real estate market. Id., at 85. On the same day as the Financial Services Committee, the Permanent Select Committee on Intelligence issued an identical subpoena to Deutsche Bank—albeit for different reasons. According to the House, the Intelligence Committee subpoenaed Deutsche Bank as part of an investigation into foreign efforts to undermine the U. S. political process. Committee Chairman Adam Schiff had described that investigation in a previous statement, explaining that the Committee was examining alleged attempts by Russia to influence the 2016 election; potential links between Russia and the President’s campaign; and whether the President and his associates had been compromised by foreign actors or interests. Press Release, House Permanent Select Committee on Intelligence, Chairman Schiff Statement on House Intelligence Committee Investigation (Feb. 6, 2019). Chairman Schiff added that the Committee planned “to develop legislation and policy reforms to ensure the U. S. government is better positioned to counter future efforts to undermine our political process and national security.” Ibid. Four days after the Financial Services and Intelligence Committees, the House Committee on Oversight and Reform issued another subpoena, this time to the President’s personal accounting firm, Mazars USA, LLP. The subpoena demanded information related to the President and several affiliated business entities from 2011 through 2018, including statements of financial condition, independent auditors’ reports, financial reports, underlying source documents, and communications between Mazars and the President or his businesses. The subpoena also requested all engagement agreements and contracts “[w]ithout regard to time.” App. to Pet. for Cert. in 19–715, p. 230. Chairman Elijah Cummings explained the basis for the subpoena in a memorandum to the Oversight Committee. According to the chairman, recent testimony by the President’s former personal attorney Michael Cohen, along with several documents prepared by Mazars and supplied by Cohen, raised questions about whether the President had accurately represented his financial affairs. Chairman Cummings asserted that the Committee had “full authority to investigate” whether the President: (1) “may have engaged in illegal conduct before and during his tenure in office,” (2) “has undisclosed conflicts of interest that may impair his ability to make impartial policy decisions,” (3) “is complying with the Emoluments Clauses of the Constitution,” and (4) “has accurately reported his finances to the Office of Government Ethics and other federal entities.” App. in No. 19–5142 (CADC), p. 107. “The Committee’s interest in these matters,” Chairman Cummings concluded, “informs its review of multiple laws and legislative proposals under our jurisdiction.” Ibid. B Petitioners—the President in his personal capacity, along with his children and affiliated businesses—filed two suits challenging the subpoenas. They contested the subpoena issued by the Oversight Committee in the District Court for the District of Columbia (Mazars, No. 19–715), and the subpoenas issued by the Financial Services and Intelligence Committees in the Southern District of New York (Deutsche Bank, No. 19–760). In both cases, petitioners contended that the subpoenas lacked a legitimate legislative purpose and violated the separation of powers. The President did not, however, resist the subpoenas by arguing that any of the requested records were protected by executive privilege. For relief, petitioners asked for declaratory judgments and injunctions preventing Mazars and the banks from complying with the subpoenas. Although named as defendants, Mazars and the banks took no positions on the legal issues in these cases, and the House committees intervened to defend the subpoenas. Petitioners’ challenges failed. In Mazars, the District Court granted judgment for the House, 380 F. Supp. 3d 76 (DC 2019), and the D. C. Circuit affirmed, 940 F.3d 710 (2019). In upholding the subpoena issued by the Oversight Committee to Mazars, the Court of Appeals found that the subpoena served a “valid legislative purpose” because the requested information was relevant to reforming financial disclosure requirements for Presidents and presidential candidates. Id., at 726–742 (internal quotation marks omitted). Judge Rao dissented. As she saw it, the “gravamen” of the subpoena was investigating alleged illegal conduct by the President, and the House must pursue such wrongdoing through its impeachment powers, not its legislative powers. Id., at 773–774. Otherwise, the House could become a “roving inquisition over a co-equal branch of government.” Id., at 748. The D. C. Circuit denied rehearing en banc over several more dissents. 941 F.3d 1180, 1180–1182 (2019). In Deutsche Bank, the District Court denied a preliminary injunction, 2019 WL 2204898 (SDNY, May 22, 2019), and the Second Circuit affirmed “in substantial part,” 943 F.3d 627, 676 (2019). While acknowledging that the subpoenas are “surely broad in scope,” the Court of Appeals held that the Intelligence Committee properly issued its subpoena to Deutsche Bank as part of an investigation into alleged foreign influence over petitioners and Russian interference with the U. S. political process. Id., at 650, 658–659. That investigation, the court concluded, could inform legislation to combat foreign meddling and strengthen national security. Id., at 658–659, and n. 59. As to the subpoenas issued by the Financial Services Committee to Deutsche Bank and Capital One, the Court of Appeals concluded that they were adequately related to potential legislation on money laundering, terrorist financing, and the global movement of illicit funds through the real estate market. Id., at 656–659. Rejecting the contention that the subpoenas improperly targeted the President, the court explained in part that the President’s financial dealings with Deutsche Bank made it “appropriate” for the House to use him as a “case study” to determine “whether new legislation is needed.” Id., at 662–663, n. 67.[1] Judge Livingston dissented, seeing no “clear reason why a congressional investigation aimed generally at closing regulatory loopholes in the banking system need focus on over a decade of financial information regarding this President, his family, and his business affairs.” Id., at 687. We granted certiorari in both cases and stayed the judgments below pending our decision. 589 U. S. ___ (2019). II A The question presented is whether the subpoenas exceed the authority of the House under the Constitution. Historically, disputes over congressional demands for presidential documents have not ended up in court. Instead, they have been hashed out in the “hurly-burly, the give-and-take of the political process between the legislative and the executive.” Hearings on S. 2170 et al. before the Subcommittee on Intergovernmental Relations of the Senate Committee on Government Operations, 94th Cong., 1st Sess., 87 (1975) (A. Scalia, Assistant Attorney General, Office of Legal Counsel). That practice began with George Washington and the early Congress. In 1792, a House committee requested Executive Branch documents pertaining to General St. Clair’s campaign against the Indians in the Northwest Territory, which had concluded in an utter rout of federal forces when they were caught by surprise near the present-day border between Ohio and Indiana. See T. Taylor, Grand Inquest: The Story of Congressional Investigations 19–23 (1955). Since this was the first such request from Congress, President Washington called a Cabinet meeting, wishing to take care that his response “be rightly conducted” because it could “become a precedent.” 1 Writings of Thomas Jefferson 189 (P. Ford ed. 1892). The meeting, attended by the likes of Alexander Hamilton, Thomas Jefferson, Edmund Randolph, and Henry Knox, ended with the Cabinet of “one mind”: The House had authority to “institute inquiries” and “call for papers” but the President could “exercise a discretion” over disclosures, “communicat[ing] such papers as the public good would permit” and “refus[ing]” the rest. Id., at 189–190. President Washington then dispatched Jefferson to speak to individual congressmen and “bring them by persuasion into the right channel.” Id., at 190. The discussions were apparently fruitful, as the House later narrowed its request and the documents were supplied without recourse to the courts. See 3 Annals of Cong. 536 (1792); Taylor, supra, at 24. Jefferson, once he became President, followed Washington’s precedent. In early 1807, after Jefferson had disclosed that “sundry persons” were conspiring to invade Spanish territory in North America with a private army, 16 Annals of Cong. 686–687, the House requested that the President produce any information in his possession touching on the conspiracy (except for information that would harm the public interest), id., at 336, 345, 359. Jefferson chose not to divulge the entire “voluminous” correspondence on the subject, explaining that much of it was “private” or mere “rumors” and “neither safety nor justice” permitted him to “expos[e] names” apart from identifying the conspiracy’s “principal actor”: Aaron Burr. Id., at 39–40. Instead of the entire correspondence, Jefferson sent Congress particular documents and a special message summarizing the conspiracy. Id., at 39–43; see generally Vance, ante, at 3–4. Neither Congress nor the President asked the Judiciary to intervene.[2] Ever since, congressional demands for the President’s information have been resolved by the political branches without involving this Court. The Reagan and Clinton presidencies provide two modern examples: During the Reagan administration, a House subcommittee subpoenaed all documents related to the Department of the Interior’s decision whether to designate Canada a reciprocal country for purposes of the Mineral Lands Leasing Act. President Reagan directed that certain documents be withheld because they implicated his confidential relationship with subordinates. While withholding those documents, the administration made “repeated efforts” at accommodation through limited disclosures and testimony over a period of several months. 6 Op. of Office of Legal Counsel 751, 780 (1982). Unsatisfied, the subcommittee and its parent committee eventually voted to hold the Secretary of the Interior in contempt, and an innovative compromise soon followed: All documents were made available, but only for one day with no photocopying, minimal notetaking, and no participation by non-Members of Congress. Id., at 780–781; see H. R. Rep. No. 97–898, pp. 3–8 (1982). In 1995, a Senate committee subpoenaed notes taken by a White House attorney at a meeting with President Clinton’s personal lawyers concerning the Whitewater controversy. The President resisted the subpoena on the ground that the notes were protected by attorney-client privilege, leading to “long and protracted” negotiations and a Senate threat to seek judicial enforcement of the subpoena. S. Rep. No. 104–204, pp. 16–17 (1996). Eventually the parties reached an agreement, whereby President Clinton avoided the threatened suit, agreed to turn over the notes, and obtained the Senate’s concession that he had not waived any privileges. Ibid.; see L. Fisher, Congressional Research Service, Congressional Investigations: Subpoenas and Contempt Power 16–18 (2003). Congress and the President maintained this tradition of negotiation and compromise—without the involvement of this Court—until the present dispute. Indeed, from President Washington until now, we have never considered a dispute over a congressional subpoena for the President’s records. And, according to the parties, the appellate courts have addressed such a subpoena only once, when a Senate committee subpoenaed President Nixon during the Watergate scandal. See infra, at 13 (discussing Senate Select Committee on Presidential Campaign Activities v. Nixon, 498 F.2d 725 (CADC 1974) (en banc)). In that case, the court refused to enforce the subpoena, and the Senate did not seek review by this Court. This dispute therefore represents a significant departure from historical practice. Although the parties agree that this particular controversy is justiciable, we recognize that it is the first of its kind to reach this Court; that disputes of this sort can raise important issues concerning relations between the branches; that related disputes involving congressional efforts to seek official Executive Branch information recur on a regular basis, including in the context of deeply partisan controversy; and that Congress and the Executive have nonetheless managed for over two centuries to resolve such disputes among themselves without the benefit of guidance from us. Such longstanding practice “ ‘is a consideration of great weight’ ” in cases concerning “the allocation of power between [the] two elected branches of Government,” and it imposes on us a duty of care to ensure that we not needlessly disturb “the compromises and working arrangements that [those] branches . . . themselves have reached.” NLRB v. Noel Canning, 573 U.S. 513, 524–526 (2014) (quoting The Pocket Veto Case, 279 U.S. 655, 689 (1929)). With that in mind, we turn to the question presented. B Congress has no enumerated constitutional power to conduct investigations or issue subpoenas, but we have held that each House has power “to secure needed information” in order to legislate. McGrain v. Daugherty, 273 U.S. 135, 161 (1927). This “power of inquiry—with process to enforce it—is an essential and appropriate auxiliary to the legislative function.” Id., at 174. Without information, Congress would be shooting in the dark, unable to legislate “wisely or effectively.” Id., at 175. The congressional power to obtain information is “broad” and “indispensable.” Watkins v. United States, 354 U.S. 178, 187, 215 (1957). It encompasses inquiries into the administration of existing laws, studies of proposed laws, and “surveys of defects in our social, economic or political system for the purpose of enabling the Congress to remedy them.” Id., at 187. Because this power is “justified solely as an adjunct to the legislative process,” it is subject to several limitations. Id., at 197. Most importantly, a congressional subpoena is valid only if it is “related to, and in furtherance of, a legitimate task of the Congress.” Id., at 187. The subpoena must serve a “valid legislative purpose,” Quinn v. United States, 349 U.S. 155, 161 (1955); it must “concern[ ] a subject on which legislation ‘could be had,’ ” Eastland v. United States Servicemen’s Fund, 421 U.S. 491, 506 (1975) (quoting McGrain, 273 U. S., at 177). Furthermore, Congress may not issue a subpoena for the purpose of “law enforcement,” because “those powers are assigned under our Constitution to the Executive and the Judiciary.” Quinn, 349 U. S., at 161. Thus Congress may not use subpoenas to “try” someone “before [a] committee for any crime or wrongdoing.” McGrain, 273 U. S., at 179. Congress has no “ ‘general’ power to inquire into private affairs and compel disclosures,” id., at 173–174, and “there is no congressional power to expose for the sake of exposure,” Watkins, 354 U. S., at 200. “Investigations conducted solely for the personal aggrandizement of the investigators or to ‘punish’ those investigated are indefensible.” Id., at 187. Finally, recipients of legislative subpoenas retain their constitutional rights throughout the course of an investigation. See id., at 188, 198. And recipients have long been understood to retain common law and constitutional privileges with respect to certain materials, such as attorney-client communications and governmental communications protected by executive privilege. See, e.g., Congressional Research Service, supra, at 16–18 (attorney-client privilege); Senate Select Committee, 498 F. 2d, at 727, 730–731 (executive privilege). C The President contends, as does the Solicitor General appearing on behalf of the United States, that the usual rules for congressional subpoenas do not govern here because the President’s papers are at issue. They argue for a more demanding standard based in large part on cases involving the Nixon tapes—recordings of conversations between President Nixon and close advisers discussing the break-in at the Democratic National Committee’s headquarters at the Watergate complex. The tapes were subpoenaed by a Senate committee and the Special Prosecutor investigating the break-in, prompting President Nixon to invoke executive privilege and leading to two cases addressing the showing necessary to require the President to comply with the subpoenas. See Nixon, 418 U.S. 683; Senate Select Committee, 498 F.2d 725. Those cases, the President and the Solicitor General now contend, establish the standard that should govern the House subpoenas here. Quoting Nixon, the President asserts that the House must establish a “demonstrated, specific need” for the financial information, just as the Watergate special prosecutor was required to do in order to obtain the tapes. 418 U. S., at 713. And drawing on Senate Select Committee—the D. C. Circuit case refusing to enforce the Senate subpoena for the tapes—the President and the Solicitor General argue that the House must show that the financial information is “demonstrably critical” to its legislative purpose. 498 F. 2d, at 731. We disagree that these demanding standards apply here. Unlike the cases before us, Nixon and Senate Select Committee involved Oval Office communications over which the President asserted executive privilege. That privilege safeguards the public interest in candid, confidential deliberations within the Executive Branch; it is “fundamental to the operation of Government.” Nixon, 418 U. S., at 708. As a result, information subject to executive privilege deserves “the greatest protection consistent with the fair administration of justice.” Id., at 715. We decline to transplant that protection root and branch to cases involving nonprivileged, private information, which by definition does not implicate sensitive Executive Branch deliberations. The standards proposed by the President and the Solicitor General—if applied outside the context of privileged information—would risk seriously impeding Congress in carrying out its responsibilities. The President and the Solicitor General would apply the same exacting standards to all subpoenas for the President’s information, without recognizing distinctions between privileged and nonprivileged information, between official and personal information, or between various legislative objectives. Such a categorical approach would represent a significant departure from the longstanding way of doing business between the branches, giving short shrift to Congress’s important interests in conducting inquiries to obtain the information it needs to legislate effectively. Confounding the legislature in that effort would be contrary to the principle that: “It is the proper duty of a representative body to look diligently into every affair of government and to talk much about what it sees. It is meant to be the eyes and the voice, and to embody the wisdom and will of its constituents. Unless Congress have and use every means of acquainting itself with the acts and the disposition of the administrative agents of the government, the country must be helpless to learn how it is being served.” United States v. Rumely, 345 U.S. 41, 43 (1953) (internal quotation marks omitted). Legislative inquiries might involve the President in appropriate cases; as noted, Congress’s responsibilities extend to “every affair of government.” Ibid. (internal quotation marks omitted). Because the President’s approach does not take adequate account of these significant congressional interests, we do not adopt it. D The House meanwhile would have us ignore that these suits involve the President. Invoking our precedents concerning investigations that did not target the President’s papers, the House urges us to uphold its subpoenas because they “relate[ ] to a valid legislative purpose” or “concern[ ] a subject on which legislation could be had.” Brief for Respondent 46 (quoting Barenblatt v. United States, 360 U.S. 109, 127 (1959), and Eastland, 421 U. S., at 506). That approach is appropriate, the House argues, because the cases before us are not “momentous separation-of-powers disputes.” Brief for Respondent 1. Largely following the House’s lead, the courts below treated these cases much like any other, applying precedents that do not involve the President’s papers. See 943 F. 3d, at 656–670; 940 F. 3d, at 724–742. The Second Circuit concluded that “this case does not concern separation of powers” because the House seeks personal documents and the President sued in his personal capacity. 943 F. 3d, at 669. The D. C. Circuit, for its part, recognized that “separation-of-powers concerns still linger in the air,” and therefore it did not afford deference to the House. 940 F. 3d, at 725–726. But, because the House sought only personal documents, the court concluded that the case “present[ed] no direct interbranch dispute.” Ibid. The House’s approach fails to take adequate account of the significant separation of powers issues raised by congressional subpoenas for the President’s information. Congress and the President have an ongoing institutional relationship as the “opposite and rival” political branches established by the Constitution. The Federalist No. 51, at 349. As a result, congressional subpoenas directed at the President differ markedly from congressional subpoenas we have previously reviewed, e.g., Barenblatt, 360 U. S., at 127; Eastland, 421 U. S., at 506, and they bear little resemblance to criminal subpoenas issued to the President in the course of a specific investigation, see Vance, ante, p. ___; Nixon, 418 U.S. 683. Unlike those subpoenas, congressional subpoenas for the President’s information unavoidably pit the political branches against one another. Cf. In re Sealed Case, 121 F.3d 729, 753 (CADC 1997) (“The President’s ability to withhold information from Congress implicates different constitutional considerations than the President’s ability to withhold evidence in judicial proceedings.”). Far from accounting for separation of powers concerns, the House’s approach aggravates them by leaving essentially no limits on the congressional power to subpoena the President’s personal records. Any personal paper possessed by a President could potentially “relate to” a conceivable subject of legislation, for Congress has broad legislative powers that touch a vast number of subjects. Brief for Respondent 46. The President’s financial records could relate to economic reform, medical records to health reform, school transcripts to education reform, and so on. Indeed, at argument, the House was unable to identify any type of information that lacks some relation to potential legislation. See Tr. of Oral Arg. 52–53, 62–65. Without limits on its subpoena powers, Congress could “exert an imperious controul” over the Executive Branch and aggrandize itself at the President’s expense, just as the Framers feared. The Federalist No. 71, at 484 (A. Hamilton); see id., No. 48, at 332–333 (J. Madison); Bowsher v. Synar, 478 U.S. 714, 721–722, 727 (1986). And a limitless subpoena power would transform the “established practice” of the political branches. Noel Canning, 573 U. S., at 524 (internal quotation marks omitted). Instead of negotiating over information requests, Congress could simply walk away from the bargaining table and compel compliance in court. The House and the courts below suggest that these separation of powers concerns are not fully implicated by the particular subpoenas here, but we disagree. We would have to be “blind” not to see what “[a]ll others can see and understand”: that the subpoenas do not represent a run-of-the-mill legislative effort but rather a clash between rival branches of government over records of intense political interest for all involved. Rumely, 345 U. S., at 44 (quoting Child Labor Tax Case, 259 U.S. 20, 37 (1922) (Taft, C. J.)). The interbranch conflict here does not vanish simply because the subpoenas seek personal papers or because the President sued in his personal capacity. The President is the only person who alone composes a branch of government. As a result, there is not always a clear line between his personal and official affairs. “The interest of the man” is often “connected with the constitutional rights of the place.” The Federalist No. 51, at 349. Given the close connection between the Office of the President and its occupant, congressional demands for the President’s papers can implicate the relationship between the branches regardless whether those papers are personal or official. Either way, a demand may aim to harass the President or render him “complaisan[t] to the humors of the Legislature.” Id., No. 71, at 483. In fact, a subpoena for personal papers may pose a heightened risk of such impermissible purposes, precisely because of the documents’ personal nature and their less evident connection to a legislative task. No one can say that the controversy here is less significant to the relationship between the branches simply because it involves personal papers. Quite the opposite. That appears to be what makes the matter of such great consequence to the President and Congress. In addition, separation of powers concerns are no less palpable here simply because the subpoenas were issued to third parties. Congressional demands for the President’s information present an interbranch conflict no matter where the information is held—it is, after all, the President’s information. Were it otherwise, Congress could sidestep constitutional requirements any time a President’s information is entrusted to a third party—as occurs with rapidly increasing frequency. Cf. Carpenter v. United States, 585 U. S. ___, ___, ___ (2018) (slip op., at 15, 17). Indeed, Congress could declare open season on the President’s information held by schools, archives, internet service providers, e-mail clients, and financial institutions. The Constitution does not tolerate such ready evasion; it “deals with substance, not shadows.” Cummings v. Missouri, 4 Wall. 277, 325 (1867). E Congressional subpoenas for the President’s personal information implicate weighty concerns regarding the separation of powers. Neither side, however, identifies an approach that accounts for these concerns. For more than two centuries, the political branches have resolved information disputes using the wide variety of means that the Constitution puts at their disposal. The nature of such interactions would be transformed by judicial enforcement of either of the approaches suggested by the parties, eroding a “[d]eeply embedded traditional way[ ] of conducting government.” Youngstown Sheet & Tube Co., 343 U. S., at 610 (Frankfurter, J., concurring). A balanced approach is necessary, one that takes a “considerable impression” from “the practice of the government,” McCulloch v. Maryland, 4 Wheat. 316, 401 (1819); see Noel Canning, 573 U. S., at 524–526, and “resist[s]” the “pressure inherent within each of the separate Branches to exceed the outer limits of its power,” INS v. Chadha, 462 U.S. 919, 951 (1983). We therefore conclude that, in assessing whether a subpoena directed at the President’s personal information is “related to, and in furtherance of, a legitimate task of the Congress,” Watkins, 354 U. S., at 187, courts must perform a careful analysis that takes adequate account of the separation of powers principles at stake, including both the significant legislative interests of Congress and the “unique position” of the President, Clinton, 520 U. S., at 698 (internal quotation marks omitted). Several special considerations inform this analysis. First, courts should carefully assess whether the asserted legislative purpose warrants the significant step of involving the President and his papers. “ ‘[O]ccasion[s] for constitutional confrontation between the two branches’ should be avoided whenever possible.” Cheney v. United States Dist. Court for D. C., 542 U.S. 367, 389–390 (2004) (quoting Nixon, 418 U. S., at 692). Congress may not rely on the President’s information if other sources could reasonably provide Congress the information it needs in light of its particular legislative objective. The President’s unique constitutional position means that Congress may not look to him as a “case study” for general legislation. Cf. 943 F. 3d, at 662–663, n. 67. Unlike in criminal proceedings, where “[t]he very integrity of the judicial system” would be undermined without “full disclosure of all the facts,” Nixon, 418 U. S., at 709, efforts to craft legislation involve predictive policy judgments that are “not hamper[ed] . . . in quite the same way” when every scrap of potentially relevant evidence is not available, Cheney, 542 U. S., at 384; see Senate Select Committee, 498 F. 2d, at 732. While we certainly recognize Congress’s important interests in obtaining information through appropriate inquiries, those interests are not sufficiently powerful to justify access to the President’s personal papers when other sources could provide Congress the information it needs. Second, to narrow the scope of possible conflict between the branches, courts should insist on a subpoena no broader than reasonably necessary to support Congress’s legislative objective. The specificity of the subpoena’s request “serves as an important safeguard against unnecessary intrusion into the operation of the Office of the President.” Cheney, 542 U. S., at 387. Third, courts should be attentive to the nature of the evidence offered by Congress to establish that a subpoena advances a valid legislative purpose. The more detailed and substantial the evidence of Congress’s legislative purpose, the better. See Watkins, 354 U. S., at 201, 205 (preferring such evidence over “vague” and “loosely worded” evidence of Congress’s purpose). That is particularly true when Congress contemplates legislation that raises sensitive constitutional issues, such as legislation concerning the Presidency. In such cases, it is “impossible” to conclude that a subpoena is designed to advance a valid legislative purpose unless Congress adequately identifies its aims and explains why the President’s information will advance its consideration of the possible legislation. Id., at 205–206, 214–215. Fourth, courts should be careful to assess the burdens imposed on the President by a subpoena. We have held that burdens on the President’s time and attention stemming from judicial process and litigation, without more, generally do not cross constitutional lines. See Vance, ante, at 12–14; Clinton, 520 U. S., at 704–705. But burdens imposed by a congressional subpoena should be carefully scrutinized, for they stem from a rival political branch that has an ongoing relationship with the President and incentives to use subpoenas for institutional advantage. Other considerations may be pertinent as well; one case every two centuries does not afford enough experience for an exhaustive list. When Congress seeks information “needed for intelligent legislative action,” it “unquestionably” remains “the duty of all citizens to cooperate.” Watkins, 354 U. S., at 187 (emphasis added). Congressional subpoenas for information from the President, however, implicate special concerns regarding the separation of powers. The courts below did not take adequate account of those concerns. The judgments of the Courts of Appeals for the D. C. Circuit and the Second Circuit are vacated, and the cases are remanded for further proceedings consistent with this opinion. It is so ordered. Notes 1 The Court of Appeals directed a “limited” remand for the District Court to consider whether it was necessary to disclose certain “sensitive personal details” (such as documents reflecting medical services received by employees of the Trump business entities) and a “few” documents that might not relate to the committees’ legislative purposes. 943 F.3d 627, 667–668, 675 (2019). The Court of Appeals ordered that all other documents be “promptly transmitted” to the committees. Id., at 669. 2 By contrast, later that summer, the Judiciary was called on to resolve whether President Jefferson could be issued a subpoena duces tecum arising from Burr’s criminal trial. See United States v. Burr, 25 F. Cas. 30 (No. 14,692d) (CC Va. 1807); see also Trump v. Vance, ante, at 5–7.

591.US.2019_19-635

In 2019, the New York County District Attorney’s Office—acting on behalf of a grand jury—served a subpoena duces tecum on Mazars USA, LLP, the personal accounting firm of President Donald J. Trump, for financial records relating to the President and his businesses. The President, acting in his personal capacity, sued the district attorney and Mazars in Federal District Court to enjoin enforcement of the subpoena, arguing that a sitting President enjoys absolute immunity from state criminal process under Article II and the Supremacy Clause. The District Court dismissed the case under the abstention doctrine of Younger v. Harris, 401 U.S. 37, and, in the alternative, held that the President was not entitled to injunctive relief. The Second Circuit rejected the District Court’s dismissal under Younger but agreed with the court’s denial of injunctive relief, concluding that presidential immunity did not bar enforcement of the subpoena and rejecting the argument of the United States as amicus curiae that a state grand jury subpoena seeking the President’s documents must satisfy a heightened showing of need. Held: Article II and the Supremacy Clause do not categorically preclude, or require a heightened standard for, the issuance of a state criminal subpoena to a sitting President. Pp. 3–22. (a) In 1807, John Marshall, presiding as Circuit Justice for Virginia over the treason trial of Aaron Burr, granted Burr’s motion for a subpoena duces tecum directed at President Jefferson. In rejecting the prosecution’s argument that a President was not subject to such a subpoena, Marshall held that a President does not “stand exempt” from the Sixth Amendment’s guarantee that the accused have compulsory process for obtaining witnesses for their defense. United States v. Burr, 25 F. Cas. 30, 33–34. The sole argument for an exemption was that a President’s “duties as chief magistrate demand his whole time for national objects.” Ibid. But, in Marshall’s assessment, those duties were “not unremitting,” ibid., and any conflict could be addressed by the court upon return of the subpoena. Marshall also concluded that the Sixth Amendment’s guarantee extended to the production of papers. “[T]he propriety of introducing any papers,” he explained, would “depend on the character of the paper, not the character of the person who holds it,” and would have “due consideration” upon the return of the subpoena. Id., at 34, 37. Jefferson agreed to furnish whatever justice required, subject to the prerogative to decide whether particular executive communications should be withheld. In the two centuries since Burr, successive Presidents from Monroe to Clinton have accepted Marshall’s ruling that the Chief Executive is subject to subpoena and have uniformly agreed to testify when called in criminal proceedings. In 1974, the question whether to compel the disclosure of official communications over the President’s objection came to a head when the Watergate Special Prosecutor secured a subpoena duces tecum directing President Nixon to produce, among other things, tape recordings of Oval Office meetings. This Court rejected Nixon’s claim of an absolute privilege of confidentiality for all presidential communications. Recognizing that “compulsory process” was imperative for both the prosecution and the defense, the Court held that the President’s “generalized assertion of privilege must yield to the demonstrated, specific need for evidence in a pending criminal trial.” United States v. Nixon, 418 U.S. 683, 713. President Nixon dutifully released the tapes. Pp. 3–10. (b) This history all involved federal criminal proceedings. Here, the President claims that the Supremacy Clause gives a sitting President absolute immunity from state criminal subpoenas because compliance with such subpoenas would categorically impair the performance of his Article II functions. The Solicitor General, arguing on behalf of the United States, claims that a state grand jury subpoena for a sitting President’s personal records must, at the very least, meet a heightened standard of need. Pp. 10–22. (1) The President’s unique duties as head of the Executive Branch come with protections that safeguard his ability to perform his vital functions. The Constitution also guarantees “the entire independence of the General Government from any control by the respective States.” Farmers and Mechanics Sav. Bank of Minneapolis v. Minnesota, 232 U.S. 516, 521. Marshall’s ruling in Burr, entrenched by 200 years of practice and this Court’s decision in Nixon, confirms that federal criminal subpoenas do not “rise to the level of constitutionally forbidden impairment of the Executive’s ability to perform its constitutionally mandated functions.” Clinton v. Jones, 520 U.S. 681, 702–703. But the President claims that state criminal subpoenas necessarily pose a unique threat of impairment and thus require absolute immunity. His categorical argument focuses on three burdens: diversion, stigma, and harassment. Pp. 10–17. (i) The President contends that complying with state criminal subpoenas would necessarily distract the Chief Executive from his duties. He grounds that concern on Nixon v. Fitzgerald, which recognized a President’s “absolute immunity from damages liability predicated on his official acts.” 457 U.S. 731, 749. But, contrary to the President’s suggestion, that case did not hold that distraction was sufficient to confer absolute immunity. Indeed, the Court expressly rejected immunity based on distraction alone 15 years later in Clinton v. Jones, when President Clinton sought absolute immunity from civil liability for private acts. As the Court explained, Fitzgerald’s “dominant concern” was not mere distraction but the distortion of the Executive’s “decisionmaking process.” 520 U. S., at 694, n. 19. The prospect that a President may become “preoccupied by pending litigation” did not ordinarily implicate constitutional concerns. Id., at 705, n. 40. Two centuries of experience likewise confirm that a properly tailored criminal subpoena will not normally hamper the performance of a President’s constitutional duties. The President claims this case is different. He believes that he is under investigation and argues that the toll will necessarily be heavier in that circumstance. But the President is not seeking immunity from the diversion occasioned by the prospect of future criminal liability. He concedes that he may be investigated while in office. His objection is instead limited to the additional distraction caused by the subpoena itself. That argument, however, runs up against the 200 years of precedent establishing that Presidents, and their official communications, are subject to judicial process, see Burr, 25 F. Cas., at 34, even when the President is under investigation, see Nixon, 418 U. S., at 706. Pp. 12–14. (ii) The President next claims that the stigma of being subpoenaed will undermine his leadership at home and abroad. But even if a tarnished reputation were a cognizable impairment, there is nothing inherently stigmatizing about a President performing “the citizen’s normal duty of . . . furnishing information relevant” to a criminal investigation. Branzburg v. Hayes, 408 U.S. 665, 691. Nor can the risk of association with persons or activities under criminal investigation absolve a President of such an important public duty. The consequences for a President’s public standing will likely increase if he is the one under investigation, but the President concedes that such investigations are permitted under Article II and the Supremacy Clause. And the receipt of a subpoena would not seem to categorically magnify the harm to the President’s reputation. Additionally, in the grand jury context longstanding secrecy rules aim to prevent the very stigma the President anticipates. Pp. 14–15. (iii) Finally, the President argues that subjecting Presidents to state criminal subpoenas will make them “easily identifiable target[s]” for harassment. Fitzgerald, 457 U. S., at 753. The Court rejected a nearly identical argument in Clinton, concluding that the risk posed by harassing civil litigation was not “serious” because federal courts have the tools to deter and dismiss vexatious lawsuits. 520 U. S., at 708. Harassing state criminal subpoenas could, under certain circumstances, threaten the independence or effectiveness of the Executive. But here again the law already seeks to protect against such abuse. First, grand juries are prohibited from engaging in “arbitrary fishing expeditions” or initiating investigations “out of malice or an intent to harass,” United States v. R. Enterprises, Inc., 498 U.S. 292, 299, and federal courts may intervene in state proceedings that are motivated by or conducted in bad faith. Second, because the Supremacy Clause prohibits state judges and prosecutors from interfering with a President’s official duties, any effort to manipulate a President’s policy decisions or to retaliate against a President for official acts through issuance of a subpoena would be an unconstitutional attempt to “influence” a superior sovereign “exempt” from such obstacles, see McCulloch v. Maryland, 4 Wheat. 316, 417. And federal law allows a President to challenge any such allegedly unconstitutional influence in a federal forum. Pp. 15–17. (2) A state grand jury subpoena seeking a President’s private papers need not satisfy a heightened need standard, for three reasons. First, although a President cannot be treated as an “ordinary individual” when executive communications are sought, Burr teaches that, with regard to private papers, a President stands in “nearly the same situation with any other individual.” 25 F. Cas., at 191–192. Second, there has been no showing here that heightened protection against state subpoenas is necessary for the Executive to fulfill his Article II functions. Finally, absent a need to protect the Executive, the public interest in fair and effective law enforcement cuts in favor of comprehensive access to evidence. Rejecting a heightened need standard does not leave Presidents without recourse. A President may avail himself of the same protections available to every other citizen, including the right to challenge the subpoena on any grounds permitted by state law, which usually include bad faith and undue burden or breadth. When the President invokes such protections, “[t]he high respect that is owed to the office of the Chief Executive . . . should inform the conduct of the entire proceeding, including the timing and scope of discovery.” Clinton, 520 U. S., at 707. In addition, a President can raise subpoena-specific constitutional challenges in either a state or a federal forum. As noted above, he can challenge the subpoena as an attempt to influence the performance of his official duties, in violation of the Supremacy Clause. And he can argue that compliance with a particular subpoena would impede his constitutional duties. Pp. 17–21. 941 F.3d 631, affirmed and remanded. Roberts, C. J., delivered the opinion of the Court, in which Ginsburg, Breyer, Sotomayor, and Kagan, JJ., joined. Kavanaugh, J., filed an opinion concurring in the judgment, in which Gorsuch, J., joined. Thomas, J., and Alito, J., filed dissenting opinions.

In our judicial system, “the public has a right to every man’s evidence.”[1] Since the earliest days of the Republic, “every man” has included the President of the United States. Beginning with Jefferson and carrying on through Clinton, Presidents have uniformly testified or produced documents in criminal proceedings when called upon by federal courts. This case involves—so far as we and the parties can tell—the first state criminal subpoena directed to a President. The President contends that the subpoena is unenforceable. We granted certiorari to decide whether Article II and the Supremacy Clause categorically preclude, or require a heightened standard for, the issuance of a state criminal subpoena to a sitting President. I In the summer of 2018, the New York County District Attorney’s Office opened an investigation into what it opaquely describes as “business transactions involving multiple individuals whose conduct may have violated state law.” Brief for Respondent Vance 2. A year later, the office—acting on behalf of a grand jury—served a subpoena duces tecum (essentially a request to produce evidence) on Mazars USA, LLP, the personal accounting firm of President Donald J. Trump. The subpoena directed Mazars to produce financial records relating to the President and business organizations affiliated with him, including “[t]ax returns and related schedules,” from “2011 to the present.” App. to Pet. for Cert. 119a.[2] The President, acting in his personal capacity, sued the district attorney and Mazars in Federal District Court to enjoin enforcement of the subpoena. He argued that, under Article II and the Supremacy Clause, a sitting President enjoys absolute immunity from state criminal process. He asked the court to issue a “declaratory judgment that the subpoena is invalid and unenforceable while the President is in office” and to permanently enjoin the district attorney “from taking any action to enforce the subpoena.” Amended Complaint in No. 1:19–cv–8694 (SDNY, Sept. 25, 2019), p. 19. Mazars, concluding that the dispute was between the President and the district attorney, took no position on the legal issues raised by the President. The District Court abstained from exercising jurisdiction and dismissed the case based on Younger v. Harris, 401 U.S. 37 (1971), which generally precludes federal courts from intervening in ongoing state criminal prosecutions. 395 F. Supp. 3d 283, 290 (SDNY 2019). In an alternative holding, the court ruled that the President was not entitled to injunctive relief. Ibid. The Second Circuit met the District Court halfway. As to the dismissal, the Court of Appeals held that Younger abstention was inappropriate because that doctrine’s core justification—“preventing friction” between States and the Federal Government—is diminished when state and federal actors are already in conflict, as the district attorney and the President were. 941 F.3d 631, 637, 639 (2019). On the merits, the Court of Appeals agreed with the District Court’s denial of a preliminary injunction. Drawing on the 200-year history of Presidents being subject to federal judicial process, the Court of Appeals concluded that “presidential immunity does not bar the enforcement of a state grand jury subpoena directing a third party to produce non-privileged material, even when the subject matter under investigation pertains to the President.” Id., at 640. It also rejected the argument raised by the United States as amicus curiae that a state grand jury subpoena must satisfy a heightened showing of need. The court reasoned that the proposed test, derived from cases addressing privileged Executive Branch communications, “ha[d] little bearing on a subpoena” seeking “information relating solely to the President in his private capacity and disconnected from the discharge of his constitutional obligations.” Id., at 645–646. We granted certiorari. 589 U. S. ___ (2019). II In the summer of 1807, all eyes were on Richmond, Virginia. Aaron Burr, the former Vice President, was on trial for treason.[3] Fallen from political grace after his fatal duel with Alexander Hamilton, and with a murder charge pending in New Jersey, Burr followed the path of many down-and-out Americans of his day—he headed West in search of new opportunity. But Burr was a man with outsized ambitions. Together with General James Wilkinson, the Governor of the Louisiana Territory, he hatched a plan to establish a new territory in Mexico, then controlled by Spain.[4] Both men anticipated that war between the United States and Spain was imminent, and when it broke out they intended to invade Spanish territory at the head of a private army. But while Burr was rallying allies to his cause, tensions with Spain eased and rumors began to swirl that Burr was conspiring to detach States by the Allegheny Mountains from the Union. Wary of being exposed as the principal co-conspirator, Wilkinson took steps to ensure that any blame would fall on Burr. He sent a series of letters to President Jefferson accusing Burr of plotting to attack New Orleans and revolutionize the Louisiana Territory. Jefferson, who despised his former running mate Burr for trying to steal the 1800 presidential election from him, was predisposed to credit Wilkinson’s version of events. The President sent a special message to Congress identifying Burr as the “prime mover” in a plot “against the peace and safety of the Union.” 16 Annals of Cong. 39–40 (1807). According to Jefferson, Burr contemplated either the “severance of the Union” or an attack on Spanish territory. Id., at 41. Jefferson acknowledged that his sources contained a “mixture of rumors, conjectures, and suspicions” but, citing Wilkinson’s letters, he assured Congress that Burr’s guilt was “beyond question.” Id., at 39–40. The trial that followed was “the greatest spectacle in the short history of the republic,” complete with a Founder-studded cast. N. Isenberg, Fallen Founder: The Life of Aaron Burr 351 (2007). People flocked to Richmond to watch, massing in tents and covered wagons along the banks of the James River, nearly doubling the town’s population of 5,000. Burr’s defense team included Edmund Randolph and Luther Martin, both former delegates at the Constitutional Convention and renowned advocates. Chief Justice John Marshall, who had recently squared off with the Jefferson administration in Marbury v. Madison, 1 Cranch 137 (1803), presided as Circuit Justice for Virginia. Meanwhile Jefferson, intent on conviction, orchestrated the prosecution from afar, dedicating Cabinet meetings to the case, peppering the prosecutors with directions, and spending nearly $100,000 from the Treasury on the five-month proceedings. In the lead-up to trial, Burr, taking aim at his accusers, moved for a subpoena duces tecum directed at Jefferson. The draft subpoena required the President to produce an October 21, 1806 letter from Wilkinson and accompanying documents, which Jefferson had referenced in his message to Congress. The prosecution opposed the request, arguing that a President could not be subjected to such a subpoena and that the letter might contain state secrets. Following four days of argument, Marshall announced his ruling to a packed chamber. The President, Marshall declared, does not “stand exempt from the general provisions of the constitution” or, in particular, the Sixth Amendment’s guarantee that those accused have compulsory process for obtaining witnesses for their defense. United States v. Burr, 25 F. Cas. 30, 33–34 (No. 14,692d) (CC Va. 1807). At common law the “single reservation” to the duty to testify in response to a subpoena was “the case of the king,” whose “dignity” was seen as “incompatible” with appearing “under the process of the court.” Id., at 34. But, as Marshall explained, a king is born to power and can “do no wrong.” Ibid. The President, by contrast, is “of the people” and subject to the law. Ibid. According to Marshall, the sole argument for exempting the President from testimonial obligations was that his “duties as chief magistrate demand his whole time for national objects.” Ibid. But, in Marshall’s assessment, those demands were “not unremitting.” Ibid. And should the President’s duties preclude his attendance at a particular time and place, a court could work that out upon return of the subpoena. Ibid. Marshall also rejected the prosecution’s argument that the President was immune from a subpoena duces tecum because executive papers might contain state secrets. “A subpoena duces tecum,” he said, “may issue to any person to whom an ordinary subpoena may issue.” Ibid. As he explained, no “fair construction” of the Constitution supported the conclusion that the right “to compel the attendance of witnesses[ ] does not extend” to requiring those witnesses to “bring[ ] with them such papers as may be material in the defence.” Id., at 35. And, as a matter of basic fairness, permitting such information to be withheld would “tarnish the reputation of the court.” Id., at 37. As for “the propriety of introducing any papers,” that would “depend on the character of the paper, not on the character of the person who holds it.” Id., at 34. Marshall acknowledged that the papers sought by Burr could contain information “the disclosure of which would endanger the public safety,” but stated that, again, such concerns would have “due consideration” upon the return of the subpoena. Id., at 37. While the arguments unfolded, Jefferson, who had received word of the motion, wrote to the prosecutor indicating that he would—subject to the prerogative to decide which executive communications should be withheld—“furnish on all occasions, whatever the purposes of justice may require.” Letter from T. Jefferson to G. Hay (June 12, 1807), in 10 Works of Thomas Jefferson 398, n. (P. Ford ed. 1905). His “personal attendance,” however, was out of the question, for it “would leave the nation without” the “sole branch which the constitution requires to be always in function.” Letter from T. Jefferson to G. Hay (June 17, 1807), in id., at 400–401, n. Before Burr received the subpoenaed documents, Marshall rejected the prosecution’s core legal theory for treason and Burr was accordingly acquitted. Jefferson, however, was not done. Committed to salvaging a conviction, he directed the prosecutors to proceed with a misdemeanor (yes, misdemeanor) charge for inciting war against Spain. Burr then renewed his request for Wilkinson’s October 21 letter, which he later received a copy of, and subpoenaed a second letter, dated November 12, 1806, which the prosecutor claimed was privileged. Acknowledging that the President may withhold information to protect public safety, Marshall instructed that Jefferson should “state the particular reasons” for withholding the letter. United States v. Burr, 25 F. Cas. 187, 192 (No. 14,694) (CC Va. 1807). The court, paying “all proper respect” to those reasons, would then decide whether to compel disclosure. Ibid. But that decision was averted when the misdemeanor trial was cut short after it became clear that the prosecution lacked the evidence to convict. In the two centuries since the Burr trial, successive Presidents have accepted Marshall’s ruling that the Chief Executive is subject to subpoena. In 1818, President Monroe received a subpoena to testify in a court-martial against one of his appointees. See Rotunda, Presidents and Ex-Presidents as Witnesses: A Brief Historical Footnote, 1975 U. Ill. L. Forum 1, 5. His Attorney General, William Wirt—who had served as a prosecutor during Burr’s trial—advised Monroe that, per Marshall’s ruling, a subpoena to testify may “be properly awarded to the President.” Id., at 5–6. Monroe offered to sit for a deposition and ultimately submitted answers to written interrogatories. Following Monroe’s lead, his successors have uniformly agreed to testify when called in criminal proceedings, provided they could do so at a time and place of their choosing. In 1875, President Grant submitted to a three-hour deposition in the criminal prosecution of a political appointee embroiled in a network of tax-evading whiskey distillers. See 1 R. Rotunda & J. Nowak, Constitutional Law §7.1(b)(ii), p. 996 (5th ed. 2012) (Rotunda & Nowak). A century later, President Ford’s attempted assassin subpoenaed him to testify in her defense. See United States v. Fromme, 405 F. Supp. 578 (ED Cal. 1975). Ford obliged—from a safe distance—in the first videotaped deposition of a President. President Carter testified via the same means in the trial of two local officials who, while Carter was Governor of Georgia, had offered to contribute to his campaign in exchange for advance warning of any state gambling raids. See Carter’s Testimony, on Videotape, Is Given to Georgia Gambling Trial, N. Y. Times, Apr. 20, 1978, p. A20 (Carter recounted that he “rejected the proposition instantly.”). Two years later, Carter gave videotaped testimony to a federal grand jury investigating whether a fugitive financier had entreated the White House to quash his extradition proceedings. See Rotunda & Nowak §7.1(b)(vi), at 997. President Clinton testified three times, twice via deposition pursuant to subpoenas in federal criminal trials of associates implicated during the Whitewater investigation, and once by video for a grand jury investigating possible perjury. See id., §7.1(c)(viii), at 1007–1008. The bookend to Marshall’s ruling came in 1974 when the question he never had to decide—whether to compel the disclosure of official communications over the objection of the President—came to a head. That spring, the Special Prosecutor appointed to investigate the break-in of the Democratic National Committee Headquarters at the Watergate complex filed an indictment charging seven defendants associated with President Nixon and naming Nixon as an unindicted co-conspirator. As the case moved toward trial, the Special Prosecutor secured a subpoena duces tecum directing Nixon to produce, among other things, tape recordings of Oval Office meetings. Nixon moved to quash the subpoena, claiming that the Constitution provides an absolute privilege of confidentiality to all presidential communications. This Court rejected that argument in United States v. Nixon, 418 U.S. 683 (1974), a decision we later described as “unequivocally and emphatically endors[ing] Marshall’s” holding that Presidents are subject to subpoena. Clinton v. Jones, 520 U.S. 681, 704 (1997). The Nixon Court readily acknowledged the importance of preserving the confidentiality of communications “between high Government officials and those who advise and assist them.” 418 U. S., at 705. “Human experience,” the Court explained, “teaches that those who expect public dissemination of their remarks may well temper candor with a concern for appearances and for their own interests to the detriment of the decisionmaking process.” Ibid. Confidentiality thus promoted the “public interest in candid, objective, and even blunt or harsh opinions in Presidential decisionmaking.” Id., at 708. But, like Marshall two centuries prior, the Court recognized the countervailing interests at stake. Invoking the common law maxim that “the public has a right to every man’s evidence,” the Court observed that the public interest in fair and accurate judicial proceedings is at its height in the criminal setting, where our common commitment to justice demands that “guilt shall not escape” nor “innocence suffer.” Id., at 709 (internal quotation marks and alteration omitted). Because these dual aims would be “defeated if judgments” were “founded on a partial or speculative presentation of the facts,” the Nixon Court recognized that it was “imperative” that “compulsory process be available for the production of evidence needed either by the prosecution or the defense.” Ibid. The Court thus concluded that the President’s “generalized assertion of privilege must yield to the demonstrated, specific need for evidence in a pending criminal trial.” Id., at 713. Two weeks later, President Nixon dutifully released the tapes. III The history surveyed above all involved federal criminal proceedings. Here we are confronted for the first time with a subpoena issued to the President by a local grand jury operating under the supervision of a state court.[5] In the President’s view, that distinction makes all the difference. He argues that the Supremacy Clause gives a sitting President absolute immunity from state criminal subpoenas because compliance with those subpoenas would categorically impair a President’s performance of his Article II functions. The Solicitor General, arguing on behalf of the United States, agrees with much of the President’s reasoning but does not commit to his bottom line. Instead, the Solicitor General urges us to resolve this case by holding that a state grand jury subpoena for a sitting President’s personal records must, at the very least, “satisfy a heightened standard of need,” which the Solicitor General contends was not met here. Brief for United States as Amicus Curiae 26, 29. A We begin with the question of absolute immunity. No one doubts that Article II guarantees the independence of the Executive Branch. As the head of that branch, the President “occupies a unique position in the constitutional scheme.” Nixon v. Fitzgerald, 457 U.S. 731, 749 (1982). His duties, which range from faithfully executing the laws to commanding the Armed Forces, are of unrivaled gravity and breadth. Quite appropriately, those duties come with protections that safeguard the President’s ability to perform his vital functions. See, e.g., ibid. (concluding that the President enjoys “absolute immunity from damages liability predicated on his official acts”); Nixon, 418 U. S., at 708 (recognizing that presidential communications are presumptively privileged). In addition, the Constitution guarantees “the entire independence of the General Government from any control by the respective States.” Farmers and Mechanics Sav. Bank of Minneapolis v. Minnesota, 232 U.S. 516, 521 (1914). As we have often repeated, “States have no power . . . to retard, impede, burden, or in any manner control the operations of the constitutional laws enacted by Congress.” McCulloch v. Maryland, 4 Wheat. 316, 436 (1819). It follows that States also lack the power to impede the President’s execution of those laws. Marshall’s ruling in Burr, entrenched by 200 years of practice and our decision in Nixon, confirms that federal criminal subpoenas do not “rise to the level of constitutionally forbidden impairment of the Executive’s ability to perform its constitutionally mandated functions.” Clinton, 520 U. S., at 702–703. But the President, joined in part by the Solicitor General, argues that state criminal subpoenas pose a unique threat of impairment and thus demand greater protection. To be clear, the President does not contend here that this subpoena, in particular, is impermissibly burdensome. Instead he makes a categorical argument about the burdens generally associated with state criminal subpoenas, focusing on three: diversion, stigma, and harassment. We address each in turn. 1 The President’s primary contention, which the Solicitor General supports, is that complying with state criminal subpoenas would necessarily divert the Chief Executive from his duties. He grounds that concern in Nixon v. Fitzgerald, which recognized a President’s “absolute immunity from damages liability predicated on his official acts.” 457 U. S., at 749. In explaining the basis for that immunity, this Court observed that the prospect of such liability could “distract a President from his public duties, to the detriment of not only the President and his office but also the Nation that the Presidency was designed to serve.” Id., at 753. The President contends that the diversion occasioned by a state criminal subpoena imposes an equally intolerable burden on a President’s ability to perform his Article II functions. But Fitzgerald did not hold that distraction was sufficient to confer absolute immunity. We instead drew a careful analogy to the common law absolute immunity of judges and prosecutors, concluding that a President, like those officials, must “deal fearlessly and impartially with the duties of his office”—not be made “unduly cautious in the discharge of [those] duties” by the prospect of civil liability for official acts. Id., at 751–752, and n. 32 (internal quotation marks omitted). Indeed, we expressly rejected immunity based on distraction alone 15 years later in Clinton v. Jones. There, President Clinton argued that the risk of being “distracted by the need to participate in litigation” entitled a sitting President to absolute immunity from civil liability, not just for official acts, as in Fitzgerald, but for private conduct as well. 520 U. S., at 694, n. 19. We disagreed with that rationale, explaining that the “dominant concern” in Fitzgerald was not mere distraction but the distortion of the Executive’s “decisionmaking process” with respect to official acts that would stem from “worry as to the possibility of damages.” 520 U. S., at 694, n. 19. The Court recognized that Presidents constantly face myriad demands on their attention, “some private, some political, and some as a result of official duty.” Id., at 705, n. 40. But, the Court concluded, “[w]hile such distractions may be vexing to those subjected to them, they do not ordinarily implicate constitutional . . . concerns.” Ibid. The same is true of criminal subpoenas. Just as a “properly managed” civil suit is generally “unlikely to occupy any substantial amount of ” a President’s time or attention, id., at 702, two centuries of experience confirm that a properly tailored criminal subpoena will not normally hamper the performance of the President’s constitutional duties. If anything, we expect that in the mine run of cases, where a President is subpoenaed during a proceeding targeting someone else, as Jefferson was, the burden on a President will ordinarily be lighter than the burden of defending against a civil suit. The President, however, believes the district attorney is investigating him and his businesses. In such a situation, he contends, the “toll that criminal process . . . exacts from the President is even heavier” than the distraction at issue in Fitzgerald and Clinton, because “criminal litigation” poses unique burdens on the President’s time and will generate a “considerable if not overwhelming degree of mental preoccupation.” Brief for Petitioner 16–18, 30 (internal quotation marks omitted). But the President is not seeking immunity from the diversion occasioned by the prospect of future criminal liability. Instead he concedes—consistent with the position of the Department of Justice—that state grand juries are free to investigate a sitting President with an eye toward charging him after the completion of his term. See Reply Brief 19 (citing Memorandum from Randolph D. Moss, Assistant Atty. Gen., Office of Legal Counsel, to the Atty. Gen.: A Sitting President’s Amenability to Indictment and Criminal Prosecution, 24 Op. OLC 222, 257, n. 36 (Oct. 16, 2000)). The President’s objection therefore must be limited to the additional distraction caused by the subpoena itself. But that argument runs up against the 200 years of precedent establishing that Presidents, and their official communications, are subject to judicial process, see Burr, 25 F. Cas., at 34, even when the President is under investigation, see Nixon, 418 U. S., at 706. 2 The President next claims that the stigma of being subpoenaed will undermine his leadership at home and abroad. Notably, the Solicitor General does not endorse this argument, perhaps because we have twice denied absolute immunity claims by Presidents in cases involving allegations of serious misconduct. See Clinton, 520 U. S., at 685; Nixon, 418 U. S., at 687. But even if a tarnished reputation were a cognizable impairment, there is nothing inherently stigmatizing about a President performing “the citizen’s normal duty of . . . furnishing information relevant” to a criminal investigation. Branzburg v. Hayes, 408 U.S. 665, 691 (1972). Nor can we accept that the risk of association with persons or activities under criminal investigation can absolve a President of such an important public duty. Prior Presidents have weathered these associations in federal cases, supra, at 6–10, and there is no reason to think any attendant notoriety is necessarily greater in state court proceedings. To be sure, the consequences for a President’s public standing will likely increase if he is the one under investigation. But, again, the President concedes that such investigations are permitted under Article II and the Supremacy Clause, and receipt of a subpoena would not seem to categorically magnify the harm to the President’s reputation. Additionally, while the current suit has cast the Mazars subpoena into the spotlight, longstanding rules of grand jury secrecy aim to prevent the very stigma the President anticipates. See S. Beale et al., Grand Jury Law and Practice §5:1, p. 5–3 (2d ed. 2018) (“[T]he federal system and most states have adopted statutes or court rules” that “impose sharp restrictions on the extent to which matters occurring before a grand jury may be divulged” to outside persons.). Of course, disclosure restrictions are not perfect. See Nixon, 418 U. S., at 687, n. 4 (observing that news media reporting made the protective order shielding the fact that the President had been named as an unindicted co-conspirator “no longer meaningful”). But those who make unauthorized disclosures regarding a grand jury subpoena do so at their peril. See, e.g., N. Y. Penal Law Ann. §215.70 (West 2010) (designating unlawful grand jury disclosure as a felony). 3 Finally, the President and the Solicitor General warn that subjecting Presidents to state criminal subpoenas will make them “easily identifiable target[s]” for harassment. Fitzgerald, 457 U. S., at 753. But we rejected a nearly identical argument in Clinton, where then-President Clinton argued that permitting civil liability for unofficial acts would “generate a large volume of politically motivated harassing and frivolous litigation.” Clinton, 520 U. S., at 708. The President and the Solicitor General nevertheless argue that state criminal subpoenas pose a heightened risk and could undermine the President’s ability to “deal fearlessly and impartially” with the States. Fitzgerald, 457 U. S., at 752 (internal quotation marks omitted). They caution that, while federal prosecutors are accountable to and removable by the President, the 2,300 district attorneys in this country are responsive to local constituencies, local interests, and local prejudices, and might “use criminal process to register their dissatisfaction with” the President. Brief for Petitioner 16. What is more, we are told, the state courts supervising local grand juries may not exhibit the same respect that federal courts show to the President as a coordinate branch of Government. We recognize, as does the district attorney, that harassing subpoenas could, under certain circumstances, threaten the independence or effectiveness of the Executive. See Tr. of Oral Arg. 73. Even so, in Clinton we found that the risk of harassment was not “serious” because federal courts have the tools to deter and, where necessary, dismiss vexatious civil suits. 520 U. S., at 708. And, while we cannot ignore the possibility that state prosecutors may have political motivations, see post, at 15 (Alito, J., dissenting), here again the law already seeks to protect against the predicted abuse. First, grand juries are prohibited from engaging in “arbitrary fishing expeditions” and initiating investigations “out of malice or an intent to harass.” United States v. R. Enterprises, Inc., 498 U.S. 292, 299 (1991). See also, e.g., Virag v. Hynes, 54 N.Y.2d 437, 442–443, 430 N.E.2d 1249, 1252 (1981) (recognizing that grand jury subpoenas can be “challenged by an affirmative showing of impropriety,” including “bad faith” (internal quotation marks omitted)). These protections, as the district attorney himself puts it, “apply with special force to a President, in light of the office’s unique position as the head of the Executive Branch.” Brief for Respondent Vance 43. And, in the event of such harassment, a President would be entitled to the protection of federal courts. The policy against federal interference in state criminal proceedings, while strong, allows “intervention in those cases where the District Court properly finds that the state proceeding is motivated by a desire to harass or is conducted in bad faith.” Huffman v. Pursue, Ltd., 420 U.S. 592, 611 (1975). Second, contrary to Justice Alito’s characterization, our holding does not allow States to “run roughshod over the functioning of [the Executive B]ranch.” Post, at 22. The Supremacy Clause prohibits state judges and prosecutors from interfering with a President’s official duties. See, e.g., Tennessee v. Davis, 100 U.S. 257, 263 (1880) (“No State government can . . . obstruct [the] authorized officers” of the Federal Government.). Any effort to manipulate a President’s policy decisions or to “retaliat[e]” against a President for official acts through issuance of a subpoena, Brief for Respondent Vance 15, 43, would thus be an unconstitutional attempt to “influence” a superior sovereign “exempt” from such obstacles, see McCulloch, 4 Wheat., at 427. We generally “assume[ ] that state courts and prosecutors will observe constitutional limitations.” Dombrowski v. Pfister, 380 U.S. 479, 484 (1965). Failing that, federal law allows a President to challenge any allegedly unconstitutional influence in a federal forum, as the President has done here. See 42 U. S. C. §1983; Ex parte Young, 209 U.S. 123, 155–156 (1908) (holding that federal courts may enjoin state officials to conform their conduct to federal law). Given these safeguards and the Court’s precedents, we cannot conclude that absolute immunity is necessary or appropriate under Article II or the Supremacy Clause. Our dissenting colleagues agree. Justice Thomas reaches the same conclusion based on the original understanding of the Constitution reflected in Marshall’s decision in Burr. Post, at 2, 5–6. And Justice Alito, also persuaded by Burr, “agree[s]” that “not all” state criminal subpoenas for a President’s records “should be barred.” Post, at 16. On that point the Court is unanimous. B We next consider whether a state grand jury subpoena seeking a President’s private papers must satisfy a heightened need standard. The Solicitor General would require a threshold showing that the evidence sought is “critical” for “specific charging decisions” and that the subpoena is a “last resort,” meaning the evidence is “not available from any other source” and is needed “now, rather than at the end of the President’s term.” Brief for United States as Amicus Curiae 29, 32 (internal quotation marks and alteration omitted). Justice Alito, largely embracing those criteria, agrees that a state criminal subpoena to a President “should not be allowed unless a heightened standard is met.” Post, at 16–18 (asking whether the information is “critical” and “necessary . . . now”). We disagree, for three reasons. First, such a heightened standard would extend protection designed for official documents to the President’s private papers. As the Solicitor General and Justice Alito acknowledge, their proposed test is derived from executive privilege cases that trace back to Burr. Brief for United States as Amicus Curiae 26–28; post, at 17. There, Marshall explained that if Jefferson invoked presidential privilege over executive communications, the court would not “proceed against the president as against an ordinary individual” but would instead require an affidavit from the defense that “would clearly show the paper to be essential to the justice of the case.” Burr, 25 F. Cas., at 192. The Solicitor General and Justice Alito would have us apply a similar standard to a President’s personal papers. But this argument does not account for the relevant passage from Burr: “If there be a paper in the possession of the executive, which is not of an official nature, he must stand, as respects that paper, in nearly the same situation with any other individual.” Id., at 191 (emphasis added). And it is only “nearly”—and not “entirely”—because the President retains the right to assert privilege over documents that, while ostensibly private, “partake of the character of an official paper.” Id., at 191–192. Second, neither the Solicitor General nor Justice Alito has established that heightened protection against state subpoenas is necessary for the Executive to fulfill his Article II functions. Beyond the risk of harassment, which we addressed above, the only justification they offer for the heightened standard is protecting Presidents from “unwarranted burdens.” Brief for United States as Amicus Curiae 28; see post, at 16 (asking whether “there is an urgent and critical need for the subpoenaed information”). In effect, they argue that even if federal subpoenas to a President are warranted whenever evidence is material, state subpoenas are warranted “only when [the] evidence is essential.” Brief for United States as Amicus Curiae 28; see post, at 16. But that double standard has no basis in law. For if the state subpoena is not issued to manipulate, supra, at 16–17, the documents themselves are not protected, supra, at 18, and the Executive is not impaired, supra, at 12–15, then nothing in Article II or the Supremacy Clause supports holding state subpoenas to a higher standard than their federal counterparts. Finally, in the absence of a need to protect the Executive, the public interest in fair and effective law enforcement cuts in favor of comprehensive access to evidence. Requiring a state grand jury to meet a heightened standard of need would hobble the grand jury’s ability to acquire “all information that might possibly bear on its investigation.” R. Enterprises, Inc., 498 U. S., at 297. And, even assuming the evidence withheld under that standard were preserved until the conclusion of a President’s term, in the interim the State would be deprived of investigative leads that the evidence might yield, allowing memories to fade and documents to disappear. This could frustrate the identification, investigation, and indictment of third parties (for whom applicable statutes of limitations might lapse). More troubling, it could prejudice the innocent by depriving the grand jury of exculpatory evidence. Rejecting a heightened need standard does not leave Presidents with “no real protection.” Post, at 19 (opinion of Alito, J.). To start, a President may avail himself of the same protections available to every other citizen. These include the right to challenge the subpoena on any grounds permitted by state law, which usually include bad faith and undue burden or breadth. See, e.g., Virag, 54 N. Y. 2d, at 442–445, 430 N. E. 2d, at 1252–1253; In re Grand Jury Subpoenas, 72 N.Y.2d 307, 315–316, 528 N.E.2d 1195, 1200 (1988) (recognizing that grand jury subpoenas can be challenged as “overly broad” or “unreasonably burdensome” (internal quotation marks omitted)). And, as in federal court, “[t]he high respect that is owed to the office of the Chief Executive . . . should inform the conduct of the entire proceeding, including the timing and scope of discovery.” Clinton, 520 U. S., at 707. See id., at 724 (Breyer, J., concurring in judgment) (stressing the need for courts presiding over suits against the President to “schedule proceedings so as to avoid significant interference with the President’s ongoing discharge of his official responsibilities”); Nixon, 418 U. S., at 702 (“[W]here a subpoena is directed to a President . . . appellate review . . . should be particularly meticulous.”). Furthermore, although the Constitution does not entitle the Executive to absolute immunity or a heightened standard, he is not “relegate[d]” only to the challenges available to private citizens. Post, at 17 (opinion of Alito, J.). A President can raise subpoena-specific constitutional challenges, in either a state or federal forum. As previously noted, he can challenge the subpoena as an attempt to influence the performance of his official duties, in violation of the Supremacy Clause. See supra, at 17. This avenue protects against local political machinations “interposed as an obstacle to the effective operation of a federal constitutional power.” United States v. Belmont, 301 U.S. 324, 332 (1937). In addition, the Executive can—as the district attorney concedes—argue that compliance with a particular subpoena would impede his constitutional duties. Brief for Respondent Vance 42. Incidental to the functions confided in Article II is “the power to perform them, without obstruction or impediment.” 3 J. Story, Commentaries on the Constitution of the United States §1563, pp. 418–419 (1833). As a result, “once the President sets forth and explains a conflict between judicial proceeding and public duties,” or shows that an order or subpoena would “significantly interfere with his efforts to carry out” those duties, “the matter changes.” Clinton, 520 U. S., at 710, 714 (opinion of Breyer, J.). At that point, a court should use its inherent authority to quash or modify the subpoena, if necessary to ensure that such “interference with the President’s duties would not occur.” Id., at 708 (opinion of the Court). * * * Two hundred years ago, a great jurist of our Court established that no citizen, not even the President, is categorically above the common duty to produce evidence when called upon in a criminal proceeding. We reaffirm that principle today and hold that the President is neither absolutely immune from state criminal subpoenas seeking his private papers nor entitled to a heightened standard of need. The “guard[ ] furnished to this high officer” lies where it always has—in “the conduct of a court” applying established legal and constitutional principles to individual subpoenas in a manner that preserves both the independence of the Executive and the integrity of the criminal justice system. Burr, 25 F. Cas., at 34. The arguments presented here and in the Court of Ap-peals were limited to absolute immunity and heightened need. The Court of Appeals, however, has directed that the case be returned to the District Court, where the President may raise further arguments as appropriate. 941 F. 3d, at 646, n. 19.[6] We affirm the judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion. It is so ordered. Notes 1 This maxim traces at least as far back as Lord Chancellor Hardwicke, in a 1742 parliamentary debate. See 12 Parliamentary History of England 693 (1812). 2 The grand jury subpoena essentially copied a subpoena issued to Mazars in April 2019 by the Committee on Oversight and Reform of the U. S. House of Representatives, which is at issue in Trump v. Mazars USA, LLP, post, p. ___. The principal difference is that the instant subpoena expressly requests tax returns. 3 See generally N. Isenberg, Fallen Founder: The Life of Aaron Burr 271–365 (2007); J. Smith, John Marshall: Definer of a Nation 348–374 (1996); M. Lomask, Aaron Burr: The Conspiracy and Years of Exile, 1805–1836, pp. 222–298 (1982). 4 Wilkinson was secretly being paid by Spain for information and in-fluence. In the wake of Burr’s trial, he was investigated by Congress and later court-martialed. But he was acquitted for want of evidence, and his duplicity was not confirmed until decades after his death, when Spanish archival material came to light. 5 While the subpoena was directed to the President’s accounting firm, the parties agree that the papers at issue belong to the President and that Mazars is merely the custodian. Thus, for purposes of immunity, it is functionally a subpoena issued to the President. 6 The daylight between our opinion and Justice Thomas’s “dissent” is not as great as that label might suggest. Post, at 12. We agree that Presidents are neither absolutely immune from state criminal subpoenas nor insulated by a heightened need standard. Post, at 6, 11, n. 3. We agree that Presidents may challenge specific subpoenas as impeding their Article II functions. Post, at 6–7. And, although we affirm while Justice Thomas would vacate, we agree that this case will be remanded to the District Court. Post, at 12.

591.US.2019_19-46

A generic name—the name of a class of products or services—is ineligible for federal trademark registration. Respondent Booking.com, an enterprise that maintains a travel-reservation website by the same name, sought federal registration of marks including the term “Booking.com.” Concluding that “Booking.com” is a generic name for online hotel-reservation services, the U. S. Patent and Trademark Office (PTO) refused registration. Booking.com sought judicial review, and the District Court determined that “Booking.com”—unlike the term “booking” standing alone—is not generic. The Court of Appeals affirmed, finding no error in the District Court’s assessment of how consumers perceive the term “Booking.com.” The appellate court also rejected the PTO’s contention that, as a rule, combining a generic term like “booking” with “.com” yields a generic composite. Held: A term styled “generic.com” is a generic name for a class of goods or services only if the term has that meaning to consumers. Pp. 6–14. (a) Whether a compound term is generic turns on whether that term, taken as a whole, signifies to consumers a class of goods or services. The courts below determined, and the PTO no longer disputes, that consumers do not in fact perceive the term “Booking.com” that way. Because “Booking.com” is not a generic name to consumers, it is not generic. Pp. 6–7. (b) Opposing that determination, the PTO urges a nearly per se rule: When a generic term is combined with a generic Internet-domain-name suffix like “.com,” the resulting combination is generic. The rule the PTO proffers is not borne out by the PTO’s own past practice and lacks support in trademark law or policy. Pp. 7–14. (1) The PTO’s proposed rule does not follow from Goodyear’s India Rubber Glove Mfg. Co. v. Goodyear Rubber Co., 128 U.S. 598. Goodyear, the PTO maintains, established that adding a generic corporate designation like “Company” to a generic term does not confer trademark eligibility. According to the PTO, adding “.com” to a generic term—like adding “Company”—can convey no source-identifying meaning. That premise is faulty, for only one entity can occupy a particular Internet domain name at a time, so a “generic.com” term could convey to consumers an association with a particular website. More- over, an unyielding legal rule that entirely disregards consumer perception is incompatible with a bedrock principle of the Lanham Act: The generic (or nongeneric) character of a particular term depends on its meaning to consumers, i.e., do consumers in fact perceive the term as the name of a class or, instead, as a term capable of distinguishing among members of the class. Pp. 8–11. (2) The PTO’s policy concerns do not support a categorical rule against registration of “generic.com” terms. The PTO asserts that trademark protection for “Booking.com” would give the mark owner undue control over similar language that others should remain free to use. That concern attends any descriptive mark. Guarding against the anticompetitive effects the PTO identifies, several doctrines ensure that registration of “Booking.com” would not yield its holder a monopoly on the term “booking.” The PTO also doubts that owners of “generic.com” brands need trademark protection in addition to existing competitive advantages. Such advantages, however, do not inevitably disqualify a mark from federal registration. Finally, the PTO urges that Booking.com could seek remedies outside trademark law, but there is no basis to deny Booking.com the same benefits Congress accorded other marks qualifying as nongeneric. Pp. 11–14. 915 F.3d 171, affirmed. Ginsburg, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Alito, Sotomayor, Kagan, Gorsuch, and Kavanaugh, JJ., joined. Sotomayor, J., filed a concurring opinion. Breyer, J., filed a dissenting opinion.

This case concerns eligibility for federal trademark registration. Respondent Booking.com, an enterprise that maintains a travel-reservation website by the same name, sought to register the mark “Booking.com.” Concluding that “Booking.com” is a generic name for online hotel-reservation services, the U. S. Patent and Trademark Office (PTO) refused registration. A generic name—the name of a class of products or services—is ineligible for federal trademark registration. The word “booking,” the parties do not dispute, is generic for hotel-reservation services. “Booking.com” must also be generic, the PTO maintains, under an encompassing rule the PTO currently urges us to adopt: The combination of a generic word and “.com” is generic. In accord with the first- and second-instance judgments in this case, we reject the PTO’s sweeping rule. A term styled “generic.com” is a generic name for a class of goods or services only if the term has that meaning to consumers. Consumers, according to lower court determinations uncontested here by the PTO, do not perceive the term “Booking.com” to signify online hotel-reservation services as a class. In circumstances like those this case presents, a “generic.com” term is not generic and can be eligible for federal trademark registration. I A A trademark distinguishes one producer’s goods or services from another’s. Guarding a trademark against use by others, this Court has explained, “secure[s] to the owner of the mark the goodwill” of her business and “protect[s] the ability of consumers to distinguish among competing producers.” Park ’N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189, 198 (1985); see S. Rep. No. 1333, 79th Cong., 2d Sess., 3 (1946) (trademark statutes aim to “protect the public so it may be confident that, in purchasing a product bearing a particular trade-mark which it favorably knows, it will get the product which it asks for and wants to get”). Trademark protection has roots in common law and equity. Matal v. Tam, 582 U. S. ___, ___ (2017) (slip op., at 2). Today, the Lanham Act, enacted in 1946, provides federal statutory protection for trademarks. 60Stat. 427, as amended, 15 U. S. C. §1051 et seq. We have recognized that federal trademark protection, supplementing state law, “supports the free flow of commerce” and “foster[s] competition.” Matal, 582 U. S., at ___, ___–___ (slip op., at 3, 4–5) (internal quotation marks omitted). The Lanham Act not only arms trademark owners with federal claims for relief; importantly, it establishes a system of federal trademark registration. The owner of a mark on the principal register enjoys “valuable benefits,” including a presumption that the mark is valid. Iancu v. Brunetti, 588 U. S. ___, ___ (2019) (slip op., at 2); see §§1051, 1052. The supplemental register contains other product and service designations, some of which could one day gain eligibility for the principal register. See §1091. The supplemental register accords more modest benefits; notably, a listing on that register announces one’s use of the designation to others considering a similar mark. See 3 J. McCarthy, Trademarks and Unfair Competition §19:37 (5th ed. 2019) (hereinafter McCarthy). Even without federal registration, a mark may be eligible for protection against infringement under both the Lanham Act and other sources of law. See Matal, 582 U. S., at ___–___ (slip op., at 4–5). Prime among the conditions for registration, the mark must be one “by which the goods of the applicant may be distinguished from the goods of others.” §1052; see §1091(a) (supplemental register contains “marks capable of distinguishing . . . goods or services”). Distinctiveness is often expressed on an increasing scale: Word marks “may be (1) generic; (2) descriptive; (3) suggestive; (4) arbitrary; or (5) fanciful.” Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 768 (1992). The more distinctive the mark, the more readily it qualifies for the principal register. The most distinctive marks—those that are “ ‘arbitrary’ (‘Camel’ cigarettes), ‘fanciful’ (‘Kodak’ film), or ‘suggestive’ (‘Tide’ laundry detergent)”—may be placed on the principal register because they are “inherently distinctive.” Wal-Mart Stores, Inc. v. Samara Brothers, Inc., 529 U.S. 205, 210–211 (2000). “Descriptive” terms, in contrast, are not eligible for the principal register based on their inherent qualities alone. E.g., Park ’N Fly, Inc. v. Dollar Park & Fly, Inc., 718 F.2d 327, 331 (CA9 1983) (“Park ’N Fly” airport parking is descriptive), rev’d on other grounds, 469 U.S. 189 (1985). The Lanham Act, “liberaliz[ing] the common law,” “extended protection to descriptive marks.” Qualitex Co. v. Jacobson Products Co., 514 U.S. 159, 171 (1995). But to be placed on the principal register, descriptive terms must achieve significance “in the minds of the public” as identifying the applicant’s goods or services—a quality called “acquired distinctiveness” or “secondary meaning.” Wal-Mart Stores, 529 U. S., at 211 (internal quotation marks omitted); see §1052(e), (f ). Without secondary meaning, descriptive terms may be eligible only for the supplemental register. §1091(a). At the lowest end of the distinctiveness scale is “the generic name for the goods or services.” §§1127, 1064(3), 1065(4). The name of the good itself (e.g., “wine”) is incapable of “distinguish[ing] [one producer’s goods] from the goods of others” and is therefore ineligible for registration. §1052; see §1091(a). Indeed, generic terms are ordinarily ineligible for protection as trademarks at all. See Restatement (Third) of Unfair Competition §15, p. 142 (1993); Otokoyama Co. v. Wine of Japan Import, Inc., 175 F.3d 266, 270 (CA2 1999) (“[E]veryone may use [generic terms] to refer to the goods they designate.”). B Booking.com is a digital travel company that provides hotel reservations and other services under the brand “Booking.com,” which is also the domain name of its website.[1] Booking.com filed applications to register four marks in connection with travel-related services, each with different visual features but all containing the term “Booking.com.”[2] Both a PTO examining attorney and the PTO’s Trademark Trial and Appeal Board concluded that the term “Booking.com” is generic for the services at issue and is therefore unregistrable. “Booking,” the Board observed, means making travel reservations, and “.com” signifies a commercial website. The Board then ruled that “customers would understand the term BOOKING.COM primarily to refer to an online reservation service for travel, tours, and lodgings.” App. to Pet. for Cert. 164a, 176a. Alternatively, the Board held that even if “Booking.com” is descriptive, not generic, it is unregistrable because it lacks secondary meaning. Booking.com sought review in the U. S. District Court for the Eastern District of Virginia, invoking a mode of review that allows Booking.com to introduce evidence not presented to the agency. See §1071(b). Relying in significant part on Booking.com’s new evidence of consumer perception, the District Court concluded that “Booking.com”—unlike “booking”—is not generic. The “consuming public,” the court found, “primarily understands that BOOKING.COM does not refer to a genus, rather it is descriptive of services involving ‘booking’ available at that domain name.” Booking.com B.V. v. Matal, 278 F. Supp. 3d 891, 918 (2017). Having determined that “Booking.com” is descriptive, the District Court additionally found that the term has acquired secondary meaning as to hotel-reservation services. For those services, the District Court therefore concluded, Booking.com’s marks meet the distinctiveness requirement for registration. The PTO appealed only the District Court’s determination that “Booking.com” is not generic. Finding no error in the District Court’s assessment of how consumers perceive the term “Booking.com,” the Court of Appeals for the Fourth Circuit affirmed the court of first instance’s judgment. In so ruling, the appeals court rejected the PTO’s contention that the combination of “.com” with a generic term like “booking” “is necessarily generic.” 915 F.3d 171, 184 (2019). Dissenting in relevant part, Judge Wynn concluded that the District Court mistakenly presumed that “generic.com” terms are usually descriptive, not generic. We granted certiorari, 589 U. S. ___ (2019), and now affirm the Fourth Circuit’s decision. II Although the parties here disagree about the circumstances in which terms like “Booking.com” rank as generic, several guiding principles are common ground. First, a “generic” term names a “class” of goods or services, rather than any particular feature or exemplification of the class. Brief for Petitioners 4; Brief for Respondent 6; see §§1127, 1064(3), 1065(4) (referring to “the generic name for the goods or services”); Park ’N Fly, 469 U. S., at 194 (“A generic term is one that refers to the genus of which the particular product is a species.”). Second, for a compound term, the distinctiveness inquiry trains on the term’s meaning as a whole, not its parts in isolation. Reply Brief 9; Brief for Respondent 2; see Estate of P. D. Beckwith, Inc. v. Commissioner of Patents, 252 U.S. 538, 545–546 (1920). Third, the relevant meaning of a term is its meaning to consumers. Brief for Petitioners 43–44; Brief for Respondent 2; see Bayer Co. v. United Drug Co., 272 F. 505, 509 (SDNY 1921) (Hand, J.) (“What do the buyers understand by the word for whose use the parties are contending?”). Eligibility for registration, all agree, turns on the mark’s capacity to “distinguis[h]” goods “in commerce.” §1052. Evidencing the Lanham Act’s focus on consumer perception, the section governing cancellation of registration provides that “[t]he primary significance of the registered mark to the relevant public . . . shall be the test for determining whether the registered mark has become the generic name of goods or services.” §1064(3).[3] Under these principles, whether “Booking.com” is generic turns on whether that term, taken as a whole, signifies to consumers the class of online hotel-reservation services. Thus, if “Booking.com” were generic, we might expect consumers to understand Travelocity—another such service—to be a “Booking.com.” We might similarly expect that a consumer, searching for a trusted source of online hotel-reservation services, could ask a frequent traveler to name her favorite “Booking.com” provider. Consumers do not in fact perceive the term “Booking.com” that way, the courts below determined. The PTO no longer disputes that determination. See Pet. for Cert. I; Brief for Petitioners 17–18 (contending only that a consumer-perception inquiry was unnecessary, not that the lower courts’ consumer-perception determination was wrong). That should resolve this case: Because “Booking.com” is not a generic name to consumers, it is not generic. III Opposing that conclusion, the PTO urges a nearly per se rule that would render “Booking.com” ineligible for registration regardless of specific evidence of consumer perception. In the PTO’s view, which the dissent embraces, when a generic term is combined with a generic top-level domain like “.com,” the resulting combination is generic. In other words, every “generic.com” term is generic according to the PTO, absent exceptional circumstances.[4] The PTO’s own past practice appears to reflect no such comprehensive rule. See, e.g., Trademark Registration No. 3,601,346 (“ART.COM” on principal register for, inter alia, “[o]nline retail store services” offering “art prints, original art, [and] art reproductions”); Trademark Registration No. 2,580,467 (“DATING.COM” on supplemental register for “dating services”). Existing registrations inconsistent with the rule the PTO now advances would be at risk of cancellation if the PTO’s current view were to prevail. See §1064(3). We decline to adopt a rule essentially excluding registration of “generic.com” marks. As explained below, we discern no support for the PTO’s current view in trademark law or policy. A The PTO urges that the exclusionary rule it advocates follows from a common-law principle, applied in Goodyear’s India Rubber Glove Mfg. Co. v. Goodyear Rubber Co., 128 U.S. 598 (1888), that a generic corporate designation added to a generic term does not confer trademark eligibility. In Goodyear, a decision predating the Lanham Act, this Court held that “Goodyear Rubber Company” was not “capable of exclusive appropriation.” Id., at 602. Standing alone, the term “Goodyear Rubber” could not serve as a trademark because it referred, in those days, to “well-known classes of goods produced by the process known as Goodyear’s invention.” Ibid. “[A]ddition of the word ‘Company’ ” supplied no protectable meaning, the Court concluded, because adding “Company” “only indicates that parties have formed an association or partnership to deal in such goods.” Ibid. Permitting exclusive rights in “Goodyear Rubber Company” (or “Wine Company, Cotton Company, or Grain Company”), the Court explained, would tread on the right of all persons “to deal in such articles, and to publish the fact to the world.” Id., at 602–603. “Generic.com,” the PTO maintains, is like “Generic Company” and is therefore ineligible for trademark protection, let alone federal registration. According to the PTO, adding “.com” to a generic term—like adding “Company”—“conveys no additional meaning that would distinguish [one provider’s] services from those of other providers.” Brief for Petitioners 44. The dissent endorses that proposition: “Generic.com” conveys that the generic good or service is offered online “and nothing more.” Post, at 1. That premise is faulty. A “generic.com” term might also convey to consumers a source-identifying characteristic: an association with a particular website. As the PTO and the dissent elsewhere acknowledge, only one entity can occupy a particular Internet domain name at a time, so “[a] consumer who is familiar with that aspect of the domain-name system can infer that BOOKING.COM refers to some specific entity.” Brief for Petitioners 40. See also Tr. of Oral Arg. 5 (“Because domain names are one of a kind, a significant portion of the public will always understand a generic ‘.com’ term to refer to a specific business . . . .”); post, at 7 (the “exclusivity” of “generic.com” terms sets them apart from terms like “Wine, Inc.” and “The Wine Company”). Thus, consumers could understand a given “generic.com” term to describe the corresponding website or to identify the website’s proprietor. We therefore resist the PTO’s position that “generic.com” terms are capable of signifying only an entire class of online goods or services and, hence, are categorically incapable of identifying a source.[5] The PTO’s reliance on Goodyear is flawed in another respect. The PTO understands Goodyear to hold that “Generic Company” terms “are ineligible for trademark protection as a matter of law”—regardless of how “consumers would understand” the term. Brief for Petitioners 38. But, as noted, whether a term is generic depends on its meaning to consumers. Supra, at 6. That bedrock principle of the Lanham Act is incompatible with an unyielding legal rule that entirely disregards consumer perception. Instead, Goodyear reflects a more modest principle harmonious with Congress’ subsequent enactment: A compound of generic elements is generic if the combination yields no additional meaning to consumers capable of distinguishing the goods or services. The PTO also invokes the oft-repeated principle that “no matter how much money and effort the user of a generic term has poured into promoting the sale of its merchandise . . . , it cannot deprive competing manufacturers of the product of the right to call an article by its name.” Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4, 9 (CA2 1976). That principle presupposes that a generic term is at issue. But the PTO’s only legal basis for deeming “generic.com” terms generic is its mistaken reliance on Goodyear. While we reject the rule proffered by the PTO that “generic.com” terms are generic names, we do not embrace a rule automatically classifying such terms as nongeneric. Whether any given “generic.com” term is generic, we hold, depends on whether consumers in fact perceive that term as the name of a class or, instead, as a term capable of distinguishing among members of the class.[6] B The PTO, echoed by the dissent, post, at 10–12, objects that protecting “generic.com” terms as trademarks would disserve trademark law’s animating policies. We disagree. The PTO’s principal concern is that trademark protection for a term like “Booking.com” would hinder competitors. But the PTO does not assert that others seeking to offer online hotel-reservation services need to call their services “Booking.com.” Rather, the PTO fears that trademark protection for “Booking.com” could exclude or inhibit competitors from using the term “booking” or adopting domain names like “ebooking.com” or “hotel-booking.com.” Brief for Petitioners 27–28. The PTO’s objection, therefore, is not to exclusive use of “Booking.com” as a mark, but to undue control over similar language, i.e., “booking,” that others should remain free to use. That concern attends any descriptive mark. Responsive to it, trademark law hems in the scope of such marks short of denying trademark protection altogether. Notably, a competitor’s use does not infringe a mark unless it is likely to confuse consumers. See §§1114(1), 1125(a)(1)(A); 4 McCarthy §23:1.50 (collecting state law). In assessing the likelihood of confusion, courts consider the mark’s distinctiveness: “The weaker a mark, the fewer are the junior uses that will trigger a likelihood of consumer confusion.” 2 id., §11:76. When a mark incorporates generic or highly descriptive components, consumers are less likely to think that other uses of the common element emanate from the mark’s owner. Ibid. Similarly, “[i]n a ‘crowded’ field of look-alike marks” (e.g., hotel names including the word “grand”), consumers “may have learned to carefully pick out” one mark from another. Id., §11:85. And even where some consumer confusion exists, the doctrine known as classic fair use, see id., §11:45, protects from liability anyone who uses a descriptive term, “fairly and in good faith” and “otherwise than as a mark,” merely to describe her own goods. 15 U. S. C. §1115(b)(4); see KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 122–123 (2004). These doctrines guard against the anticompetitive effects the PTO identifies, ensuring that registration of “Booking.com” would not yield its holder a monopoly on the term “booking.” Booking.com concedes that “Booking.com” would be a “weak” mark. Tr. of Oral Arg. 66. See also id., at 42–43, 55. The mark is descriptive, Booking.com recognizes, making it “harder . . . to show a likelihood of confusion.” Id., at 43. Furthermore, because its mark is one of many “similarly worded marks,” Booking.com accepts that close variations are unlikely to infringe. Id., at 66. And Booking.com acknowledges that federal registration of “Booking.com” would not prevent competitors from using the word “booking” to describe their own services. Id., at 55. The PTO also doubts that owners of “generic.com” brands need trademark protection in addition to existing competitive advantages. Booking.com, the PTO argues, has already seized a domain name that no other website can use and is easy for consumers to find. Consumers might enter “the word ‘booking’ in a search engine,” the PTO observes, or “proceed directly to ‘booking.com’ in the expectation that [online hotel-booking] services will be offered at that address.” Brief for Petitioners 32. Those competitive advantages, however, do not inevitably disqualify a mark from federal registration. All descriptive marks are intuitively linked to the product or service and thus might be easy for consumers to find using a search engine or telephone directory. The Lanham Act permits registration nonetheless. See §1052(e), (f ). And the PTO fails to explain how the exclusive connection between a domain name and its owner makes the domain name a generic term all should be free to use. That connection makes trademark protection more appropriate, not less. See supra, at 9. Finally, even if “Booking.com” is generic, the PTO urges, unfair-competition law could prevent others from passing off their services as Booking.com’s. Cf. Genesee Brewing Co. v. Stroh Brewing Co., 124 F.3d 137, 149 (CA2 1997); Blinded Veterans Assn. v. Blinded Am. Veterans Foundation, 872 F.2d 1035, 1042–1048 (CADC 1989). But federal trademark registration would offer Booking.com greater protection. See, e.g., Genesee Brewing, 124 F. 3d, at 151 (unfair-competition law would oblige competitor at most to “make more of an effort” to reduce confusion, not to cease marketing its product using the disputed term); Matal, 582 U. S., at ___ (slip op., at 5) (federal registration confers valuable benefits); Brief for Respondent 26 (expressing intention to seek protections available to trademark owners under the Anticybersquatting Consumer Protection Act, 15 U. S. C. §1125(d)); Brief for Coalition of .Com Brand Owners as Amici Curiae 14–19 (trademark rights allow mark owners to stop domain-name abuse through private dispute resolution without resorting to litigation). We have no cause to deny Booking.com the same benefits Congress accorded other marks qualifying as nongeneric. * * * The PTO challenges the judgment below on a sole ground: It urges that, as a rule, combining a generic term with “.com” yields a generic composite. For the above-stated reasons, we decline a rule of that order, one that would largely disallow registration of “generic.com” terms and open the door to cancellation of scores of currently registered marks. Accordingly, the judgment of the Court of Appeals for the Fourth Circuit regarding eligibility for trademark registration is Affirmed. Notes 1 A domain name identifies an address on the Internet. The rightmost component of a domain name—“.com” in “Booking.com”—is known as the top-level domain. Domain names are unique; that is, a given domain name is assigned to only one entity at a time. 2 For simplicity, this opinion uses the term “trademark” to encompass the marks whose registration Booking.com seeks. Although Booking.com uses the marks in connection with services, not goods, rendering the marks “service marks” rather than “trademarks” under 15 U. S. C. §1127, that distinction is immaterial to the issue before us. 3 The U. S. Patent and Trademark Office (PTO) suggests that the primary-significance test might not govern outside the context of §1064(3), which subjects to cancellation marks previously registered that have “become” generic. See Reply Brief 11; Tr. of Oral Arg. 19. To so confine the primary-significance test, however, would upset the understanding, shared by Courts of Appeals and the PTO’s own manual for trademark examiners, that the same test governs whether a mark is registrable in the first place. See, e.g., In re Cordua Restaurants, Inc., 823 F.3d 594, 599 (CA Fed. 2016); Nartron Corp. v. STMicroelectronics, Inc., 305 F.3d 397, 404 (CA6 2002); Genesee Brewing Co. v. Stroh Brewing Co., 124 F.3d 137, 144 (CA2 1997); Trademark Manual of Examining Procedure §1209.01(c)(i), p. 1200–267 (Oct. 2018), http://tmep.uspto.gov. We need not address today the scope of the primary-significance test’s application, for our analysis does not depend on whether one meaning among several is “primary.” Sufficient to resolve this case is the undisputed principle that consumer perception demarcates a term’s meaning. 4 The PTO notes only one possible exception: Sometimes adding a generic term to a generic top-level domain results in wordplay (for example, “tennis.net”). That special case, the PTO acknowledges, is not presented here and does not affect our analysis. See Brief for Petitioners 25, n. 6; Tr. of Oral Arg. 25–26. 5 In passing, the PTO urges us to disregard that a domain name is assigned to only one entity at a time. That fact, the PTO suggests, stems from “a functional characteristic of the Internet and the domain-name system,” and functional features cannot receive trademark protection. Brief for Petitioners 32. “[A] product feature is functional, and cannot serve as a trademark,” we have held, “if it is essential to the use or purpose of the article or if it affects the cost or quality of the article.” TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23, 32 (2001) (internal quotation marks omitted); see §1052(e) (barring from the principal registrar “any matter that, as a whole, is functional”). This case, however, does not concern trademark protection for a feature of the Internet or the domain-name system; Booking.com lays no claim to the use of unique domain names generally. Nor does the PTO contend that the particular domain name “Booking.com” is essential to the use or purpose of online hotel-reservation services, affects these services’ cost or quality, or is otherwise necessary for competitors to use. In any event, we have no occasion to decide the applicability of §1052(e)’s functionality bar, for the sole ground on which the PTO refused registration, and the sole claim before us, is that “Booking.com” is generic. 6 Evidence informing that inquiry can include not only consumer surveys, but also dictionaries, usage by consumers and competitors, and any other source of evidence bearing on how consumers perceive a term’s meaning. Surveys can be helpful evidence of consumer perception but require care in their design and interpretation. See Brief for Trademark Scholars as Amici Curiae 18–20 (urging that survey respondents may conflate the fact that domain names are exclusive with a conclusion that a given “generic.com” term has achieved secondary meaning). Moreover, difficult questions may be presented when a term has multiple concurrent meanings to consumers or a meaning that has changed over time. See, e.g., 2 J. McCarthy, Trademarks and Unfair Competition §12:51 (5th ed. 2019) (discussing terms that are “a generic name to some, a trademark to others”); id., §12:49 (“Determining the distinction between generic and trademark usage of a word . . . when there are no other sellers of [the good or service] is one of the most difficult areas of trademark law.”). Such issues are not here entailed, for the PTO does not contest the lower courts’ assessment of consumer perception in this case. See Pet. for Cert. I; Brief for Petitioners 17–18. For the same reason, while the dissent questions the evidence on which the lower courts relied, post, at 7–8, 9, we have no occasion to reweigh that evidence. Cf. post, at 1–2 (Sotomayor, J., concurring).

591.US.2019_19-177

In the United States Leadership Against HIV/AIDS, Tuberculosis, and Malaria Act of 2003, as relevant here, Congress limited the funding of American and foreign nongovernmental organizations to those with “a policy explicitly opposing prostitution and sex trafficking.” 22 U. S. C. §7631(f). In 2013, that Policy Requirement, as it is known, was held to be an unconstitutional restraint on free speech when applied to American organizations. Agency for Int’l Development v. Alliance for Open Society Int’l, Inc., 570 U.S. 205. Those American organizations now challenge the requirement’s constitutionality when applied to their legally distinct foreign affiliates. The District Court held that the Government was prohibited from enforcing the requirement against the foreign affiliates, and the Second Circuit affirmed. Held: Because plaintiffs’ foreign affiliates possess no First Amendment rights, applying the Policy Requirement to them is not unconstitutional. Two bedrock legal principles lead to this conclusion. As a matter of American constitutional law, foreign citizens outside U. S. territory do not possess rights under the U. S. Constitution. See, e.g., Boumediene v. Bush, 553 U.S. 723, 770–771. And as a matter of American corporate law, separately incorporated organizations are separate legal units with distinct legal rights and obligations. See, e.g., Dole Food Co. v. Patrickson, 538 U.S. 468, 474–475. That conclusion corresponds to Congress’s historical practice of conditioning funding to foreign organizations, which helps ensure that U. S. foreign aid serves U. S. interests. Plaintiffs’ counterarguments are unpersuasive. First, they claim that because a foreign affiliate’s policy statement may be attributed to them, American organizations themselves possess a First Amendment right against the Policy Requirement’s imposition on their foreign affiliates. First Amendment cases involving speech misattribution between formally distinct speakers, see, e.g., Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, Inc., 515 U.S. 557, 574–575, however, are premised on something missing here: Government compulsion to associate with another entity. Even protecting the free speech rights of only those foreign organizations that are closely identified with American organizations would deviate from the fundamental principle that foreign organizations operating abroad do not possess rights under the U. S. Constitution and enmesh the courts in difficult line-drawing exercises. Second, plaintiffs assert that the Court’s 2013 decision encompassed both American organizations and their foreign affiliates. That decision did not facially invalidate the Act’s funding condition, suggest that the First Amendment requires the Government to exempt plaintiffs’ foreign affiliates or other foreign organizations from the Policy Requirement, or purport to override longstanding constitutional law and corporate law principles. Pp. 3–9. 911 F.3d 104, reversed. Kavanaugh, J., delivered the opinion of the Court, in which Roberts, C. J., and Thomas, Alito, and Gorsuch, JJ., joined. Thomas, J., filed a concurring opinion. Breyer, J., filed a dissenting opinion, in which Ginsburg and Sotomayor, JJ., joined. Kagan, J., took no part in the consideration or decision of the case.

In 2003, Congress passed and President George W. Bush signed the United States Leadership Against HIV/AIDS, Tuberculosis, and Malaria Act, known as the Leadership Act. 117Stat. 711, as amended, 22 U. S. C. §7601 et seq. Aiming to enhance America’s response to the ravages of the global HIV/AIDS crisis, the Leadership Act launched “the largest international public health program of its kind ever created.” §7601(29). The Act has helped save an estimated 17 million lives, primarily in Africa, and is widely viewed as the most successful American foreign aid program since the Marshall Plan. To advance the global relief effort, Congress has allocated billions of dollars to American and foreign nongovernmental organizations that combat HIV/AIDS abroad. As relevant here, Congress sought to fund only those organizations that have, or agree to have, a “policy explicitly opposing prostitution and sex trafficking.” §7631(f ); see also §7631(e); 45 CFR §89.1 (2019). Congress imposed that condition on funding, known as the Policy Requirement, because Congress found that prostitution and sex trafficking “are additional causes of and factors in the spread of the HIV/AIDS epidemic” and that prostitution and sex trafficking “are degrading to women and children.” §7601(23). Plaintiffs are American nongovernmental organizations that receive Leadership Act funds to fight HIV/AIDS abroad. Plaintiffs have long maintained that they do not want to express their agreement with the American commitment to eradicating prostitution. Plaintiffs consider a public stance of neutrality toward prostitution more helpful to their sensitive work in some parts of the world and also to their full participation in the global efforts to prevent HIV/AIDS. After enactment of the Leadership Act, plaintiffs challenged the Policy Requirement, alleging that it violated the First Amendment. In 2013, this Court agreed, concluding that the Policy Requirement ran afoul of the free speech principle that the Government “may not deny a benefit to a person on a basis that infringes his constitutionally protected . . . freedom of speech.” Agency for Int’l Development v. Alliance for Open Society Int’l, Inc., 570 U.S. 205, 214 (2013) (internal quotation marks omitted). Therefore, the Policy Requirement no longer applies to American organizations that receive Leadership Act funds, meaning that American organizations can obtain Leadership Act funds even if they do not have a policy explicitly opposing prostitution and sex trafficking. But as has been the case since 2003, foreign organizations that receive Leadership Act funds remain subject to the Policy Requirement and still must have a policy explicitly opposing prostitution and sex trafficking. Following this Court’s 2013 decision barring the Government from enforcing the Policy Requirement against American organizations, plaintiffs returned to court, invoking the First Amendment and seeking to bar the Government from enforcing the Policy Requirement against plaintiffs’ legally distinct foreign affiliates. The U. S. District Court for the Southern District of New York agreed with plaintiffs and prohibited the Government from enforcing the Policy Requirement against plaintiffs’ foreign affiliates. The U. S. Court of Appeals for the Second Circuit affirmed. Judge Straub dissented. He described as “startling” the proposition that the First Amendment could extend to foreign organizations operating abroad. 911 F.3d 104, 112 (2018). The Second Circuit’s decision was stayed pending this Court’s review, meaning that foreign organizations currently remain subject to the Policy Requirement. We granted certiorari, 589 U. S. ___ (2019), and now reverse the judgment of the Second Circuit. Plaintiffs’ position runs headlong into two bedrock principles of American law. First, it is long settled as a matter of American constitutional law that foreign citizens outside U. S. territory do not possess rights under the U. S. Constitution. Plaintiffs do not dispute that fundamental principle. Tr. of Oral Arg. 58–59; see, e.g., Boumediene v. Bush, 553 U.S. 723, 770–771 (2008); Hamdi v. Rumsfeld, 542 U.S. 507, 558–559 (2004) (Scalia, J., dissenting); United States v. Verdugo-Urquidez, 494 U.S. 259, 265–275 (1990); Johnson v. Eisentrager, 339 U.S. 763, 784 (1950); United States ex rel. Turner v. Williams, 194 U.S. 279, 292 (1904); U. S. Const., Preamble. As the Court has recognized, foreign citizens in the United States may enjoy certain constitutional rights—to take just one example, the right to due process in a criminal trial. See, e.g., Verdugo-Urquidez, 494 U. S., at 270–271; Plyler v. Doe, 457 U.S. 202, 210–213 (1982); Kwong Hai Chew v. Colding, 344 U.S. 590, 596 (1953); Bridges v. Wixon, 326 U.S. 135, 148 (1945); Yick Wo v. Hopkins, 118 U.S. 356, 369 (1886); cf. Bluman v. Federal Election Comm’n, 800 F. Supp. 2d 281, 286–289 (DC 2011), aff ’d, 565 U.S. 1104 (2012). And so too, the Court has ruled that, under some circumstances, foreign citizens in the U. S. Territories—or in “a territory” under the “indefinite” and “complete and total control” and “within the constant jurisdiction” of the United States—may possess certain constitutional rights. Boumediene, 553 U. S., at 755–771. But the Court has not allowed foreign citizens outside the United States or such U. S. territory to assert rights under the U. S. Constitution. If the rule were otherwise, actions by American military, intelligence, and law enforcement personnel against foreign organizations or foreign citizens in foreign countries would be constrained by the foreign citizens’ purported rights under the U. S. Constitution. That has never been the law. See Verdugo-Urquidez, 494 U. S., at 273–274; Eisentrager, 339 U. S., at 784.[1]* To be sure, Congress may seek to enact laws that afford foreign citizens abroad statutory rights or causes of action against misconduct by U. S. Government officials, or laws that otherwise regulate the conduct of U. S. officials abroad. See Verdugo-Urquidez, 494 U. S., at 275; cf. 10 U. S. C. §§2734(a), 2734a(a); 18 U. S. C. §2340A; 21 U. S. C. §904; 22 U. S. C. §§2669, 2669–1; 42 U. S. C. §2000dd; but see 28 U. S. C. §2680(k) (Federal Tort Claims Act’s exception for torts “arising in a foreign country”). Plaintiffs did not raise any such statutory claim in this case. Second, it is long settled as a matter of American corporate law that separately incorporated organizations are separate legal units with distinct legal rights and obligations. See Dole Food Co. v. Patrickson, 538 U.S. 468, 474–475 (2003); Cedric Kushner Promotions, Ltd. v. King, 533 U.S. 158, 163 (2001); P. Blumberg, K. Strasser, N. Georgakopoulos, & E. Gouvin, Corporate Groups §§6.01, 6.02, 6.05 (2020 Supp.). Plaintiffs’ foreign affiliates were incorporated in other countries and are legally separate from plaintiffs’ American organizations. Even though the foreign organizations have affiliated with the American organizations, the foreign organizations remain legally distinct from the American organizations. Plaintiffs do not ask this Court to pierce the corporate veil, nor do they invoke any other relevant exception to that fundamental corporate law principle. Tr. of Oral Arg. 54. Those two bedrock principles of American constitutional law and American corporate law together lead to a simple conclusion: As foreign organizations operating abroad, plaintiffs’ foreign affiliates possess no rights under the First Amendment. That conclusion corresponds to historical practice regarding American foreign aid. The United States supplies more foreign aid than any other nation in the world. Cong. Research Serv., Foreign Assistance: An Introduction to U. S. Programs and Policy (2020) (Summary). Acting with the President in the legislative process, Congress sometimes imposes conditions on foreign aid. See 22 U. S. C. §§2271, 2272, 2371, 7110(g)(2). Congress may condition funding on a foreign organization’s ideological commitments—for example, pro-democracy, pro-women’s rights, anti-terrorism, pro-religious freedom, anti-sex trafficking, or the like. Doing so helps ensure that U. S. foreign aid serves U. S. interests. By contrast, plaintiffs’ approach would throw a constitutional wrench into American foreign policy. In particular, plaintiffs’ approach would put Congress in the untenable position of either cutting off certain funding programs altogether, or instead funding foreign organizations that may not align with U. S. values. We see no constitutional justification for the Federal Judiciary to interfere in that fashion with American foreign policy and American aid to foreign organizations. In short, plaintiffs’ foreign affiliates are foreign organizations, and foreign organizations operating abroad have no First Amendment rights. To overcome that conclusion, plaintiffs advance two main arguments. But neither persuades us. First, plaintiffs theorize that the foreign affiliates’ required statement of policy against prostitution and sex trafficking may be incorrectly attributed to the American organizations. Therefore, the theory goes, the American organizations themselves possess a First Amendment right against imposition of the Policy Requirement on their foreign affiliates. As support, plaintiffs point to First Amendment cases involving speech misattribution between formally distinct speakers. See, e.g., Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, Inc., 515 U.S. 557, 574–575 (1995); Pacific Gas & Elec. Co. v. Public Util. Comm’n of Cal., 475 U.S. 1, 15 (1986) (plurality opinion); cf. PruneYard Shopping Center v. Robins, 447 U.S. 74, 87 (1980). But the constitutional issue in those cases arose because the State forced one speaker to host another speaker’s speech. See Hurley, 515 U. S., at 572–573; Pacific Gas, 475 U. S., at 15; cf. PruneYard, 447 U. S., at 85, 87. Here, by contrast, the United States is not forcing plaintiffs to affiliate with foreign organizations. Plaintiffs are free to choose whether to affiliate with foreign organizations and are free to disclaim agreement with the foreign affiliates’ required statement of policy. Any alleged misattribution in this case and any effect on the American organizations’ message of neutrality toward prostitution stems from their choice to affiliate with foreign organizations, not from U. S. Government compulsion. Because the First Amendment misattribution cases are premised on government compulsion to associate with another entity, those cases do not apply here. In support of their misattribution argument, plaintiffs also cite Regan v. Taxation With Representation of Wash., 461 U.S. 540, 544–545, and n. 6 (1983). But as relevant here, that case simply explained that a speech restriction on a corporate entity did not prevent a separate affiliate from speaking, a point that is not disputed in this case. We appreciate that plaintiffs would prefer to affiliate with foreign organizations that do not oppose prostitution. But Congress required foreign organizations to oppose prostitution in return for American funding. And plaintiffs cannot export their own First Amendment rights to shield foreign organizations from Congress’s funding conditions. Stressing that their position is limited, plaintiffs emphasize that the Court could narrowly decide to protect the free speech rights of only those foreign organizations that are closely identified with American organizations—for example, those foreign affiliates that share similar names, logos, and brands with American organizations. According to plaintiffs, those “closely identified” scenarios greatly increase the risk of misattribution. But again, the First Amendment cases involving speech misattribution arose when the State forced one speaker to host another speaker’s speech. No compulsion is present here. Moreover, plaintiffs’ proposed line-drawing among foreign organizations would blur a clear rule of American law: Foreign organizations operating abroad do not possess rights under the U. S. Constitution. Plaintiffs’ carve-out not only would deviate from that fundamental principle, but also would enmesh the courts in difficult line-drawing exercises—how closely identified is close enough?—and leave courts without any principled basis for making those judgments. We discern no good reason to invent a new and legally unmoored exception to longstanding principles of American constitutional and corporate law. Second, plaintiffs argue that the Court’s 2013 decision in this case encompassed both plaintiffs’ American organizations and their foreign affiliates, meaning that, in plaintiffs’ view, the Court has already resolved the issue before us. That is not correct. The plaintiffs in the 2013 case were these same American organizations. It is true that the Court considered the possibility that an American organization could work through affiliates to potentially avoid the burdens of the otherwise-unconstitutional application of the Policy Requirement. But the Court rejected that alternative, which in essence would have compelled the American organizations to affiliate with other organizations. The Court instead ruled that the Policy Requirement may not be applied to plaintiffs’ American organizations. Therefore, plaintiffs’ current affiliations with foreign organizations are their own choice, not the result of any U. S. Government compulsion. Stated simply, in the prior decision, the Court did not facially invalidate the Act’s condition on funding. The Court did not hold or suggest that the First Amendment requires the Government to exempt plaintiffs’ foreign affiliates or other foreign organizations from the Policy Requirement. And the Court did not purport to override the longstanding constitutional law principle that foreign organizations operating abroad do not possess constitutional rights, or the elementary corporate law principle that each corporation is a separate legal unit. The dissent emphasizes that this case concerns “the First Amendment rights of American organizations.” Post, at 1 (opinion of Breyer, J.). We respectfully disagree with that characterization of the question presented. The Court’s prior decision recognized the First Amendment rights of American organizations and held that American organizations do not have to comply with the Policy Requirement. This case instead concerns foreign organizations that are voluntarily affiliated with American organizations. Those foreign organizations are legally separate from the American organizations. And because foreign organizations operating abroad do not possess constitutional rights, those foreign organizations do not have a First Amendment right to disregard the Policy Requirement. In sum, plaintiffs’ foreign affiliates are foreign organizations, and foreign organizations operating abroad possess no rights under the U. S. Constitution. We reverse the judgment of the U. S. Court of Appeals for the Second Circuit. It is so ordered. Justice Kagan took no part in the consideration or decision of this case. Notes 1 * As Justice Jackson stated for the Court in Eisentrager: “If the Fifth Amendment confers its rights on all the world . . . , the same must be true of the companion civil-rights Amendments, for none of them is limited by its express terms, territorially or as to persons. Such a construction would mean that during military occupation irreconcilable enemy elements, guerrilla fighters, and ‘werewolves’ could require the American Judiciary to assure them freedoms of speech, press, and assembly as in the First Amendment, right to bear arms as in the Second, security against ‘unreasonable’ searches and seizures as in the Fourth, as well as rights to jury trial as in the Fifth and Sixth Amendments. “Such extraterritorial application of organic law would have been so significant an innovation in the practice of governments that, if intended or apprehended, it could scarcely have failed to excite contemporary comment. Not one word can be cited. No decision of this Court supports such a view. Cf. Downes v. Bidwell, 182 U.S. 244. None of the learned commentators on our Constitution has even hinted at it.” 339 U. S., at 784–785.

590.US.2019_18-1584

Petitioner Atlantic Coast Pipeline, LLC (Atlantic), sought to construct an approximately 604-mile natural gas pipeline from West Virginia to North Carolina along a route that traversed 16 miles of land within the George Washington National Forest. As relevant here, Atlantic secured a special use permit from the United States Forest Service, obtaining a right-of-way for a 0.1-mile segment of pipe some 600 feet below a portion of the Appalachian National Scenic Trail (Appalachian Trail or Trail), which also crosses the National Forest. Respondents filed a petition for review in the Fourth Circuit, contending, inter alia, that the issuance of the special use permit for the right-of-way under the Trail violated the Mineral Leasing Act (Leasing Act). Atlantic intervened. The Fourth Circuit vacated the permit, holding that the Leasing Act did not empower the Forest Service to grant the right-of-way because the Trail became part of the National Park System when the Secretary of the Interior delegated its authority over the Trail’s administration to the National Park Service, and that the Leasing Act prohibits pipeline rights-of-way through lands in the National Park System. Held: Because the Department of the Interior’s decision to assign responsibility over the Appalachian Trail to the National Park Service did not transform the land over which the Trail passes into land within the National Park System, the Forest Service had the authority to issue the special use permit. Pp. 3–18. (a) These cases involve the interaction of multiple federal laws. The Weeks Act provided for the acquisition of lands for inclusion in the National Forest System, stating that such lands “shall be permanently reserved, held, and administered as national forest lands.” 16 U. S. C. §521. The Forest Service, with authority granted by the Secretary of Agriculture, has jurisdiction over the National Forest System, including the George Washington National Forest. The National Trails System Act (Trails Act) establishes national scenic and national historic trails, 16 U. S. C. §1244(a), including the Appalachian Trail, §1244(a)(1). It also empowers the Secretary of the Interior to establish the Trail’s location and width by entering into “rights-of-way” agreements with other federal agencies, States, local governments, and private landowners. §§1246(a)(2), (d), (e). The Leasing Act enables any “appropriate agency head” to grant “[r]ights-of-way through any Federal lands . . . for pipeline purposes,” 30 U. S. C. §185(a), defining “Federal lands” as “all lands owned by the United States,” except (as relevant) lands in the National Park System, §185(b). The National Park System is, in turn, defined as “any area of land and water now and hereafter administered by the Secretary of the Interior, through the National Park Service for park, monument, historic, parkway, recreational, or other purposes.” 54 U. S. C. §100501. Pp. 3–5. (b) An examination of the interests and authority granted under the Trails Act shows that the Forest Service “right-of-way” agreements with the National Park Service for the Appalachian Trail did not convert “Federal lands” under the Leasing Act into “lands” within the “National Park System.” Pp. 5–13. (1) A right-of-way is a type of easement. And easements grant only nonpossessory rights of use limited to the purposes specified in the easement agreement: They are not land; they merely burden land that continues to be owned by another. The same principles that apply to right-of-way agreements between private parties apply here, even though the Federal Government owns all lands involved. A right-of-way between two agencies grants only an easement across the land, not jurisdiction over the land itself. Read in light of basic property law principles, then, the plain language of the Trails Act and the agreement between the two agencies did not divest the Forest Service of jurisdiction over the lands crossed by the Trail. Pp. 7–10. (2) The various duties described in the Trails Act—that the Secretary of the Interior (through the National Park Service) administers the Trail “primarily as a footpath,” 16 U. S. C. §1244(a)(1); can designate Trail uses, provide Trail markers, and establish interpretative and informational sites, §1246(c); and can regulate the Trail’s “protection, management, development, and administration,” §1246(i)—reinforce the conclusion that the agency responsible for the Trail has the limited role of administering a trail easement, but that the underlying land remains within the Forest Service’s jurisdiction. Pp. 10–11. (3) This conclusion is also reinforced by the fact that Congress spoke in terms of rights-of-way in the Trails Act rather than in terms of land transfers, as it has unequivocally and directly done in multiple other statutes when it has intended to transfer land from one agency to another. See, e.g., Wild and Scenic Rivers Act, 16 U. S. C. §1281(c). Pp. 12–13. (c) Respondents’ theory—that the National Park Service administers the Trail, and therefore the lands that the Trail crosses—depends on presuming, with no clear congressional command, a vast expansion of the Park Service’s jurisdiction and a significant curtailment of the Forest Service’s express authority to grant pipeline rights-of-way on “lands owned by the United States.” 30 U. S. C. §185(b). It also has striking implications for federalism and private property rights, especially given that Congress has used express language in other statutes when it has intended to transfer lands between agencies. Pp. 13–17. 911 F.3d 150, reversed and remanded. Thomas, J., delivered the opinion of the Court, in which Roberts, C. J., and Breyer, Alito, Gorsuch, and Kavanaugh, JJ., joined, and in which Ginsburg, J., joined except as to Part III–B–2. Sotomayor, J., filed a dissenting opinion, in which Kagan, J., joined. Notes 1 Together with No. 18–1587, Atlantic Coast Pipeline, LLC v. Cowpasture River Preservation Association et al., also on certiorari to the same court.

state” lands, 67 Fed. Reg. 8479 (2002), the National Park Service has concluded that the regulations governing the Trail pointed to by the dissent “do not apply on non-federally owned lands,” 36 CFR 1.2(b) (2019); see also 48 Fed. Reg. 30253 (1983); Dept. of Interior, W. Janssen, Appalachian National Scenic Trail, Superintendent’s Compendium of Designations, Closures, Permit Requirements and Other Restrictions Imposed Under Discretionary Authority §5, p. 3 (2019) (“The rules contained in this Compendium apply to all persons entering, using, visiting or otherwise present on federally owned lands”). Thus, the dissent points to nothing indicating that the National Park Service has ever adopted its novel theory, with its attendant federalism concerns. 8 Objections that a pipeline segment interferes with rights of use enjoyed by the National Park Service would present a different issue. See Bruce, Law of Easements and Licenses in Land §1:1. These cases do not present anything resembling such a scenario. Under the current proposal, the workstations for laying the challenged segment of the pipeline will be located on private land, approximately 1,400 feet and 3,400 feet respectively from the Trail. Atlantic plans to use a method of drilling that will not require the company to clear any land or dig on the Trail’s surface. The entry and exit sites will not be visible from the Trail, nor will any detour be required. And, the final pipeline will lie approximately 600 feet below the Trail.

590.US.2019_19-67

Respondent Evelyn Sineneng-Smith operated an immigration consulting firm in San Jose, California. She assisted clients working without authorization in the United States to file applications for a labor certification program that once provided a path for aliens to adjust to lawful permanent resident status. Sineneng-Smith knew that her clients could not meet the long-passed statutory application-filing deadline, but she nonetheless charged each client over $6,000, netting more than $3.3 million. Sineneng-Smith was indicted for multiple violations of 8 U. S. C. §1324(a)(1)(A)(iv) and (B)(i). Those provisions make it a federal felony to “encourag[e] or induc[e] an alien to come to, enter, or reside in the United States, knowing or in reckless disregard of the fact that such coming to, entry, or residence is or will be in violation of law,” §1324(a)(1)(A)(iv), and impose an enhanced penalty if the crime is “done for the purpose of commercial advantage or private financial gain,” §1324(a)(1)(B)(i). In the District Court, she urged that the provisions did not cover her conduct, and if they did, they violated the Petition and Free Speech Clauses of the First Amendment as applied. The District Court rejected her arguments and she was convicted, as relevant here, on two counts under §1324(a)(1)(A)(iv) and (B)(i). Sineneng-Smith essentially repeated the same arguments on appeal to the Ninth Circuit. Again she asserted a right under the First Amendment to file administrative applications on her clients’ behalf, and she argued that the statute could not constitutionally be applied to her conduct. Instead of adjudicating the case presented by the parties, however, the court named three amici and invited them to brief and argue issues framed by the panel, including a question never raised by Sineneng-Smith: Whether the statute is overbroad under the First Amendment. In accord with the amici’s arguments, the Ninth Circuit held that §1324(a)(1)(A)(iv) is unconstitutionally overbroad. Held: The Ninth Circuit panel’s drastic departure from the principle of party presentation constituted an abuse of discretion. The Nation’s adversarial adjudication system follows the principle of party presentation. Greenlaw v. United States, 554 U.S. 237, 243. “In both civil and criminal cases, . . . we rely on the parties to frame the issues for decision and assign to courts the role of neutral arbiter of matters the parties present.” Id., at 243. That principle forecloses the controlling role the Ninth Circuit took on in this case. No extraordinary circumstances justified the panel’s takeover of the appeal. Sineneng-Smith, represented by competent counsel, had raised a vagueness argument and First Amendment arguments homing in on her own conduct, not that of others. Electing not to address the party-presented controversy, the panel projected that §1324(a)(1)(A)(iv) might cover a wide swath of protected speech, including abstract advocacy and legal advice. It did so even though Sineneng-Smith’s counsel had presented a contrary theory of the case in her briefs and before the District Court. A court is not hidebound by counsel’s precise arguments, but the Ninth Circuit’s radical transformation of this case goes well beyond the pale. On remand, the case is to be reconsidered shorn of the overbreadth inquiry interjected by the appellate panel and bearing a fair resemblance to the case shaped by the parties. Pp. 3–9. 910 F.3d 461, vacated and remanded. Ginsburg, J., delivered the opinion for a unanimous Court. Thomas, J., filed a concurring opinion.

This case concerns 8 U. S. C. §1324, which makes it a federal felony to “encourag[e] or induc[e] an alien to come to, enter, or reside in the United States, knowing or in reckless disregard of the fact that such coming to, entry, or residence is or will be in violation of law.” §1324(a)(1)(A)(iv). The crime carries an enhanced penalty if “done for the purpose of commercial advantage or private financial gain.” §1324(a)(1)(B)(i).[1] Respondent Evelyn Sineneng-Smith operated an immigration consulting firm in San Jose, California. She was indicted for multiple violations of §1324(a)(1)(A)(iv) and (B)(i). Her clients, most of them from the Philippines, worked without authorization in the home health care industry in the United States. Between 2001 and 2008, Sineneng-Smith assisted her clients in applying for a “labor certification” that once allowed certain aliens to adjust their status to that of lawful permanent resident permitted to live and work in the United States. §1255(i)(1)(B)(ii). There was a hindrance to the efficacy of Sineneng-Smith’s advice and assistance. To qualify for the labor- certification dispensation she promoted to her clients, an alien had to be in the United States on December 21, 2000, and apply for certification before April 30, 2001. §1255(i)(1)(C). Sineneng-Smith knew her clients did not meet the application-filing deadline; hence, their applications could not put them on a path to lawful residence.[2] Nevertheless, she charged each client $5,900 to file an application with the Department of Labor and another $900 to file with the U. S. Citizenship and Immigration Services. For her services in this regard, she collected more than $3.3 million from her unwitting clients. In the District Court, Sineneng-Smith urged unsuccessfully, inter alia, that the above-cited provisions, properly construed, did not cover her conduct, and if they did, they violated the Petition and Free Speech Clauses of the First Amendment as applied. See Motion to Dismiss in No. 10–cr–414 (ND Cal.), pp. 7–13, 20–25; Motion for Judgt. of Acquittal in No. 10–cr–414 (ND Cal.), pp. 14–19, 20–25. She was convicted on two counts under §1324(a)(1)(A)(iv) and (B)(i), and on other counts (filing false tax returns and mail fraud) she does not now contest. Throughout the District Court proceedings and on appeal, she was represented by competent counsel. On appeal from the §1324 convictions to the Ninth Circuit, both on brief and at oral argument, Sineneng-Smith essentially repeated the arguments she earlier presented to the District Court. See Brief for Appellant in No. 15–10614 (CA9), pp. 11–28. The case was then moved by the appeals panel onto a different track. Instead of adjudicating the case presented by the parties, the appeals court named three amici and invited them to brief and argue issues framed by the panel, including a question Sineneng-Smith herself never raised earlier: “[W]hether the statute of conviction is overbroad . . . under the First Amendment.” App. 122–124. In the ensuing do over of the appeal, counsel for the parties were assigned a secondary role. The Ninth Circuit ultimately concluded, in accord with the invited amici’s arguments, that §1324(a)(1)(A)(iv) is unconstitutionally overbroad. 910 F.3d 461, 485 (2018). The Government petitioned for our review because the judgment of the Court of Appeals invalidated a federal statute. Pet. for Cert. 24. We granted the petition. 588 U. S. ___ (2019). As developed more completely hereinafter, we now hold that the appeals panel departed so drastically from the principle of party presentation as to constitute an abuse of discretion. We therefore vacate the Ninth Circuit’s judgment and remand the case for an adjudication of the appeal attuned to the case shaped by the parties rather than the case designed by the appeals panel. I In our adversarial system of adjudication, we follow the principle of party presentation. As this Court stated in Greenlaw v. United States, 554 U.S. 237 (2008), “in both civil and criminal cases, in the first instance and on appeal . . . , we rely on the parties to frame the issues for decision and assign to courts the role of neutral arbiter of matters the parties present.” Id., at 243. In criminal cases, departures from the party presentation principle have usually occurred “to protect a pro se litigant’s rights.” Id., at 244; see, e.g., Castro v. United States, 540 U.S. 375, 381–383 (2003) (affirming courts’ authority to recast pro se litigants’ motions to “avoid an unnecessary dismissal” or “inappropriately stringent application of formal labeling requirements, or to create a better correspondence between the substance of a pro se motion’s claim and its underlying legal basis” (citation omitted)). But as a general rule, our system “is designed around the premise that [parties represented by competent counsel] know what is best for them, and are responsible for advancing the facts and argument entitling them to relief.” Id., at 386 (Scalia, J., concurring in part and concurring in judgment).[3] In short: “[C]ourts are essentially passive instruments of government.” United States v. Samuels, 808 F.2d 1298, 1301 (CA8 1987) (Arnold, J., concurring in denial of reh’g en banc)). They “do not, or should not, sally forth each day looking for wrongs to right. [They] wait for cases to come to [them], and when [cases arise, courts] normally decide only questions presented by the parties.” Ibid. The party presentation principle is supple, not ironclad. There are no doubt circumstances in which a modest initiating role for a court is appropriate. See, e.g., Day v. McDonough, 547 U.S. 198, 202 (2006) (federal court had “authority, on its own initiative,” to correct a party’s “evident miscalculation of the elapsed time under a statute [of limitations]” absent “intelligent waiver”).[4] But this case scarcely fits that bill. To explain why that is so, we turn first to the proceedings in the District Court. In July 2010, a grand jury returned a multicount indictment against Sineneng-Smith, including three counts of violating §1324, three counts of mail fraud in violation of 18 U. S. C. §1341, and two counts of willfully subscribing to a false tax return in violation of 26 U. S. C. §7206(1). Sineneng-Smith pleaded guilty to the tax-fraud counts, App. to Pet. for Cert. 78a–79a, and did not pursue on appeal the two mail-fraud counts on which she was ultimately convicted. We therefore concentrate this description on her defenses against the §1324 charges. Before trial, Sineneng-Smith moved to dismiss the §1324 counts. Motion to Dismiss in No. 10–cr–414 (ND Cal.). She asserted first that the conduct with which she was charged—advising and assisting aliens about labor certifications—is not proscribed by §1324(a)(1)(A)(iv) and (B)(i). Being hired to file lawful applications on behalf of aliens already residing in the United States, she maintained, did not “encourage” or “induce” them to remain in this country. Id., at 7–13. Next, she urged, alternatively, that clause (iv) is unconstitutionally vague and therefore did not provide fair notice that her conduct was prohibited, id., at 13–18, or should rank as a content-based restraint on her speech, id., at 22–24. She further asserted that she has a right safeguarded to her by the Petition and Free Speech Clauses of the First Amendment to file applications on her clients’ behalf. Id., at 20–25. Nowhere did she so much as hint that the statute is infirm, not because her own conduct is protected, but because it trenches on the First Amendment sheltered expression of others. The District Court denied the motion to dismiss, holding that Sineneng-Smith could “encourag[e]” noncitizens to remain in the country, within the meaning of §1324(a)(1)(A)(iv), “[b]y suggesting to [them] that the applications she would make on their behalf, in exchange for their payments, would allow them to eventually obtain legal permanent residency in the United States.” App. to Pet. for Cert. 73a. The court also rejected Sineneng-Smith’s constitutional arguments, reasoning that she was prosecuted, not for filing clients’ applications, but for falsely representing to noncitizens that her efforts, for which she collected sizable fees, would enable them to gain lawful status. Id., at 75a. After a 12-day trial, the jury found Sineneng-Smith guilty on the three §1324 counts charged in the indictment, along with the three mail-fraud counts. App. 118–121. Sineneng-Smith then moved for a judgment of acquittal. She renewed, “almost verbatim,” the arguments made in her motion to dismiss, App. to Pet. for Cert. 65a, and the District Court rejected those arguments “[f ]or the same reasons as the court expressed in its order denying Sineneng-Smith’s motion to dismiss,” ibid. She simultaneously urged that the evidence did not support the verdicts. Motion for Judgt. of Acquittal in No. 10–cr–414 (ND Cal.), at 1–14. The District Court found the evidence sufficient as to two of the three §1324 counts and two of the three mail-fraud counts. App. to Pet. for Cert. 67a.[5] Sineneng-Smith’s appeal to the Ninth Circuit from the District Court’s §1324 convictions commenced unremarkably. On brief and at oral argument, she reasserted the self-regarding arguments twice rehearsed, initially in her motion to dismiss, and later in her motion for acquittal. Brief for Appellant in No. 15–10614 (CA9), at 9–27, 35–41; Recording of Oral Arg. (Apr. 18, 2017), at 37:00–39:40; see supra, at 5. With the appeal poised for decision based upon the parties’ presentations, the appeals panel intervened. It ordered further briefing, App. 122–124, but not from the parties. Instead, it named three organizations—“the Federal Defender Organizations of the Ninth Circuit (as a group)[,] the Immigrant Defense Project[,] and the National Immigration Project of the National Lawyers Guild”—and invited them to file amicus briefs on three issues: “1. Whether the statute of conviction is overbroad or likely overbroad under the First Amendment, and if so, whether any permissible limiting construction would cure the First Amendment problem? “2. Whether the statute of conviction is void for vagueness or likely void for vagueness, either under the First Amendment or the Fifth Amendment, and if so, whether any permissible limiting construction would cure the constitutional vagueness problem? “3. Whether the statute of conviction contains an implicit mens rea element which the Court should enunciate. If so: (a) what should that mens rea element be; and (b) would such a mens rea element cure any serious constitutional problems the Court might determine existed?” Ibid. Counsel for the parties were permitted, but “not required,” to file supplemental briefs “limited to responding to any and all amicus/amici briefs.” Id., at 123 (emphasis added). Invited amici and amici not specifically invited to file were free to “brief such further issues as they, respectively, believe the law, and the record calls for.” Ibid. The panel gave invited amici 20 minutes for argument, and allocated only 10 minutes to Sineneng-Smith’s counsel. Reargument Order in No. 15–10614 (CA9), Doc. No. 92. Of the three specified areas of inquiry, the panel reached only the first, holding that §1324(a)(1)(A)(iv) was facially overbroad under the First Amendment, 910 F. 3d, at 483–485, and was not susceptible to a permissible limiting construction, id., at 472, 479. True, in the redone appeal, Sineneng-Smith’s counsel adopted without elaboration counsel for amici’s overbreadth arguments. See Supplemental Brief for Appellant in No. 15–10614 (CA9), p. 1. How could she do otherwise? Understandably, she rode with an argument suggested by the panel. In the panel’s adjudication, her own arguments, differently directed, fell by the wayside, for they did not mesh with the panel’s overbreadth theory of the case. II No extraordinary circumstances justified the panel’s takeover of the appeal. Sineneng-Smith herself had raised a vagueness argument and First Amendment arguments homing in on her own conduct, not that of others. Electing not to address the party-presented controversy, the panel projected that §1324(a)(1)(A)(iv) might cover a wide swath of protected speech, including political advocacy, legal advice, even a grandmother’s plea to her alien grandchild to remain in the United States. 910 F. 3d, at 483–484.[6] Nevermind that Sineneng-Smith’s counsel had presented a contrary theory of the case in the District Court, and that this Court has repeatedly warned that “invalidation for [ First Amendment] overbreadth is ‘strong medicine’ that is not to be ‘casually employed.’ ” United States v. Williams, 553 U.S. 285, 293 (2008) (quoting Los Angeles Police Dept. v. United Reporting Publishing Corp., 528 U.S. 32, 39 (1999)). As earlier observed, see supra, at 4, a court is not hidebound by the precise arguments of counsel, but the Ninth Circuit’s radical transformation of this case goes well beyond the pale. * * * For the reasons stated, we vacate the Ninth Circuit’s judgment and remand the case for reconsideration shorn of the overbreadth inquiry interjected by the appellate panel and bearing a fair resemblance to the case shaped by the parties. It is so ordered. Addendum of cases, 2015–2020, in which this Court called for supplemental briefing or appointed amicus curiae This Court has sought supplemental briefing: to determine whether a case presented a controversy suitable for the Court’s review, Trump v. Mazars USA, LLP, post, p. ___ (ordering briefing on application of political question doctrine and related justiciability principles); Frank v. Gaos, 586 U. S. ___ (2018) (ordering briefing on Article III standing); Wittman v. Personhuballah, 576 U.S. 1093 (2015) (same); Docket Entry in Gloucester County School Bd. v. G. G., O. T. 2016, No. 16–273 (Feb. 23, 2017) (ordering briefing on intervening Department of Education and Department of Justice guidance document); Kingdomware Technologies, Inc. v. United States, 577 U.S. 970 (2015) (ordering briefing on mootness); to determine whether the case could be resolved on a basis narrower than the question presented, Zubik v. Burwell, 578 U. S. ___ (2016) (ordering briefing on whether the plaintiffs could obtain relief without entirely invalidating challenged federal regulations); and to clarify an issue or argument the parties raised, Google LLC v. Oracle America, Inc., post, p.___ (ordering further briefing on the parties’ dispute over the standard of review applicable to the question presented); Babb v. Wilkie, 589 U. S. ___ (2020) (ordering briefing on an assertion counsel made for the first time at oral argument about alternative remedies available to the plaintiff ); Sharp v. Murphy, reported sub nom. Carpenter v. Murphy, 586 U. S. ___ (2018) (ordering briefing on the implications of the parties’ statutory interpretations). In rare instances, we have ordered briefing on a constitutional issue implicated, but not directly presented, by the question on which we granted certiorari. See Jennings v. Rodriguez, 580 U. S. ___ (2016) (in a case about availability of a bond hearing under a statute mandating detention of certain noncitizens, briefing ordered on whether the Constitution requires such a hearing); Johnson v. United States, 574 U.S. 1069 (2015) (in a case involving interpretation of the Armed Career Criminal Act’s residual clause, briefing ordered on whether that clause is unconstitutionally vague). But in both cases, the parties had raised the relevant constitutional challenge in lower courts; the question was not interjected into the case for the first time by an appellate forum. In Jennings, moreover, the parties’ statutory arguments turned expressly on the constitutional issue. Jennings v. Rodriguez, 583 U. S. ___ (2018). And in Johnson, although this Court had interpreted the Act’s residual clause four times in the preceding nine years, there still remained “pervasive disagreement” in the lower courts about its application. Johnson v. United States, 576 U.S. 591, 601 (2015). We have appointed amicus curiae: to present argument in support of the judgment below when a prevailing party has declined to defend the lower court’s decision or an aspect of it, Seila Law LLC v. Consumer Financial Protection Bureau, 589 U. S. ___ (2019); Holguin-Hernandez v. United States, 588 U. S. ___ (2019); Culbertson v. Berryhill, 584 U. S. ___ (2018); Lucia v. SEC, 583 U. S. ___ (2018); Beckles v. United States, 579 U. S. ___ (2016); Welch v. United States, 577 U.S. 1098 (2016); McLane Co. v. EEOC, 580 U. S. ___ (2016); Green v. Brennan, 576 U.S. 1087 (2015); Reyes Mata v. Lynch, reported sub nom. Reyes Mata v. Holder, 574 U.S. 1118 (2015); and to address the Court’s jurisdiction to decide the question presented, Montgomery v. Louisiana, 575 U.S. 933 (2015). Notes 1 For violations of 8 U. S. C. §1324(a)(1)(A)(iv), the prison term is “not more than 5 years,” §1324(a)(1)(B)(ii); if “the offense was done for . . . private financial gain,” the prison term is “not more than 10 years,” §1324(a)(1)(B)(i). 2 Sineneng-Smith argued that labor-certification applications wereoften approved despite expiration of the statutory dispensation, and that an approved application, when submitted as part of a petition for adjustment of status, would place her clients in line should Congress reactivate the dispensation. See Motion for Judgt. of Acquittal in No. 10–cr–414 (ND Cal.), p. 16. 3 See Kaplan, Civil Procedure—Reflections on the Comparison of Systems, 9 Buffalo L. Rev. 409, 431–432 (1960) (U. S. system “exploits the free-wheeling energies of counsel and places them in adversary confrontation before a detached judge”; “German system puts its trust in a judge of paternalistic bent acting in cooperation with counsel of somewhat muted adversary zeal”). 4 In an addendum to this opinion, we list cases in which this Court has called for supplemental briefing or appointed amicus curiae in recent years. None of them bear any resemblance to the redirection ordered by the Ninth Circuit panel in this case. 5 The court sentenced Sineneng-Smith to 18 months on each of the remaining counts; three years of supervised release on the §1324 and mail-fraud counts; and one year of supervised release on the filing of false tax returns count, all to run concurrently. She was also ordered to pay $43,550 in restitution, a $15,000 fine, and a $600 special assessment. 6 The Solicitor General maintained that the statute does not reach protected speech. Brief for United States 32. In the Government’s view, §1324(a)(1)(A)(iv) should be construed to prohibit only speech facilitating or soliciting illegal activity, thus falling within the exception to the First Amendment for speech integral to criminal conduct. Id., at 22–26, 31 (citing United States v. Williams, 553 U.S. 285, 298 (2008)).