(Slip Opinion) OCTOBER TERM, 2025 1
Syllabus
NOTE: Where it is feasible, a syllabus (headnote) will be released, as is being done in connection with this case, at the time the opinion is issued. The syllabus constitutes no part of the opinion of the Court but has been prepared by the Reporter of Decisions for the convenience of the reader. See United States v. Detroit Timber & Lumber Co., 200 U. S. 321, 337.
SUPREME COURT OF THE UNITED STATES
ELLINGBURG v. UNITED STATES
CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE EIGHTH CIRCUIT
No. 24–482. Argued October 14, 2025—Decided January 20, 2026
The Mandatory Victims Restitution Act of 1996 requires defendants convicted of certain federal crimes to pay monetary restitution to victims. Although petitioner Ellingburg committed his crime before the MVRA’s enactment, he was sentenced under the MRVA and ordered to pay $7,567.25 in restitution. Ellingburg raised an Ex Post Facto Clause challenge to his continued restitution obligation. The Eighth Circuit concluded that restitution under the MVRA is not criminal punishment subject to the Ex Post Facto Clause. Held: Restitution under the MVRA is plainly criminal punishment for purposes of the Ex Post Facto Clause. Whether a law violates the Ex Post Facto Clause requires evaluating whether the law imposes a criminal or penal sanction as opposed to a civil remedy. That question is one “of statutory construction” that requires the Court to “consider the statute’s text and its structure.” Smith v. Doe, 538 U. S. 84, 92 (quotation marks omitted). When viewed as a whole, the MVRA makes abundantly clear that restitution is criminal punishment. The MVRA labels restitution as a “penalty” for a criminal “offense.” 18 U. S. C. §3663A(a)(1). Only a criminal defendant convicted of a qualifying crime may be ordered to pay restitution. Restitution is imposed at sentencing for that offense together with other criminal punishments such as imprisonment and fines. And at the sentencing proceeding where restitution is imposed, the Government, not the victim, is the party adverse to the defendant. Further, the federal MVRA restitution regime is codified in Title 18, “Crimes and Criminal Procedure,” and the statutory provisions authorizing restitution orders are contained in Chapter 232 of that Title, entitled “Miscellaneous Sentencing Provisions.” A district court imposing restitution must follow the 2 ELLINGBURG v. UNITED STATES
procedures applicable to other criminal penalties. The Court’s precedents have understood restitution under the MVRA to be criminal punishment. See Manrique v. United States, 581 U. S. 116, 118. And the Court’s precedents on related issues further buttress the conclusion that MVRA restitution is criminal punishment. See United States v. Bajakajian, 524 U. S. 321, 328; United States v. One Assortment of 89 Firearms, 465 U. S. 354, 363– 366. Finally, while Congress intended restitution under the MVRA to both punish offenders and compensate victims, victims cannot initiate or settle the restitution process as they would if it were a civil proceeding. The text and structure of the Act demonstrate that Congress intended restitution under the Act to impose criminal punishment. Pp. 2–5. 113 F. 4th 839, reversed and remanded.
KAVANAUGH, J., delivered the opinion for a unanimous Court. THOMAS, J., filed a concurring opinion, in which GORSUCH, J., joined. Cite as: 607 U. S. ____ (2026) 1
Opinion of the Court
NOTICE: This opinion is subject to formal revision before publication in the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, pio@supremecourt.gov, of any typographical or other formal errors.
SUPREME COURT OF THE UNITED STATES _________________
No. 24–482 _________________
HOLSEY ELLINGBURG, JR., PETITIONER v. UNITED STATES ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE EIGHTH CIRCUIT [January 20, 2026]
JUSTICE KAVANAUGH delivered the opinion of the Court. Under the Mandatory Victims Restitution Act of 1996, defendants convicted of certain federal crimes must pay monetary restitution to the victims. That Act, known as the MVRA, became law on April 24, 1996. Petitioner Ellingburg was sentenced later in 1996 and ordered to pay restitution in the amount of $7,567.25. He has not yet satisfied that obligation. Ellingburg has now raised an Ex Post Facto Clause challenge to his continued restitution obligation because he committed his crime before the enactment of the MVRA. The Ex Post Facto Clause issue turns in part on the threshold question of whether restitution under the MVRA is criminal punishment. Applying Circuit precedent, the U. S. Court of Appeals for the Eighth Circuit concluded that restitution under the MVRA is not criminal punishment. 113 F. 4th 839, 841–842 (2024). We now reverse. In this Court, Ellingburg and the United States agree that the Eighth Circuit erred and that MVRA restitution is criminal punishment. The Court therefore appointed John F. Bash as amicus curiae to defend the judgment of the 2 ELLINGBURG v. UNITED STATES
Eighth Circuit. 605 U. S. 908 (2025). He has ably discharged his responsibilities. When determining whether a law violates the Ex Post Facto Clause, the Court must evaluate whether the law imposes a criminal or penal sanction as opposed to a civil remedy. Assessing whether “a statutory scheme is civil or criminal is first of all a question of statutory construction” that requires the Court to “consider the statute’s text and its structure.” Smith v. Doe, 538 U. S. 84, 92 (2003) (quotation marks omitted).1 Here, the statutory analysis is straightforward: Restitution under the MVRA is plainly criminal punishment for purposes of the Ex Post Facto Clause.2 Numerous features of the MVRA lead to that conclusion. The MVRA labels restitution as a “penalty” for a criminal “offense.” 18 U. S. C. §3663A(a)(1). A court may order restitution only with respect to a criminal “defendant” and only after that defendant’s conviction of a qualifying crime. Ibid. Restitution is imposed during “sentencing” for the offense. Ibid. At the sentencing proceeding where restitution is ordered, the Government, not the victim, is the party adverse to the defendant. At sentencing, restitution is imposed together with other criminal punishments such as imprisonment and fines. Indeed, for misdemeanors, restitution may be “in lieu of ” those punishments, making restitution the sole —————— 1 If the text and structure of a statute do not demonstrate that
Congress intended criminal punishment, the statute may still be deemed criminal or penal if the “party challenging the statute provides ‘the clearest proof’ that ‘the statutory scheme [is] so punitive either in purpose or effect as to negate [the Government’s] intention’ to deem it ‘civil.’ ” Kansas v. Hendricks, 521 U. S. 346, 361 (1997) (quoting United States v. Ward, 448 U. S. 242, 248–249 (1980)). 2 Amicus contends that Ellingburg’s restitution was not imposed under
the MVRA even though Ellingburg was sentenced after the MVRA took effect in 1996. But the Eighth Circuit decided this case on the understanding that the MVRA was applied to Ellingburg. Cite as: 607 U. S. ____ (2026) 3
punishment for a federal offense in certain circumstances. Ibid. In addition, when a defendant does not make the required restitution payments, the court may modify the terms of his supervised release or probation and impose imprisonment if the court determines that “alternatives to imprisonment are not adequate to serve the purposes of punishment and deterrence.” §§3614(b)(2), 3613A(a)(1). As a further sign that MVRA restitution is criminal punishment, the federal MVRA restitution regime is codified in Title 18, “Crimes and Criminal Procedure.” The main statutory provisions authorizing restitution orders, §3663 and §3663A, are contained in Chapter 232 of Title 18, entitled “Miscellaneous Sentencing Provisions.” Another statutory provision, §3556, states that a court “shall order restitution” in “imposing a sentence” for covered offenses. That provision is housed in Chapter 227 of Title 18, entitled “Sentences.” The statute also refers to a “sentence that imposes an order of restitution.” §3664(o). When imposing restitution, a district court must follow the procedures applicable to other criminal penalties, including the Federal Rules of Criminal Procedure. See §3664(c). And §3663(c) directs the U. S. Sentencing Commission to promulgate criminal sentencing guidelines for restitution. §208, 110 Stat. 1240; §3663(c)(7). When viewed as a whole, then, the MVRA makes abundantly clear that restitution is criminal punishment. We are not saying that all of the statutory features present here are necessary to constitute criminal punishment, but they are sufficient. Given the statutory text and structure, it comes as no surprise that this Court’s precedents have understood restitution under the MVRA to be criminal punishment. In Manrique v. United States, the Court stated that the MVRA requires courts “to impose restitution as part of the sentence.” 581 U. S. 116, 118 (2017). In Pasquantino v. United States, the Court explained that restitution under 4 ELLINGBURG v. UNITED STATES
the MVRA is designed “to mete out appropriate criminal punishment.” 544 U. S. 349, 365 (2005). And in Paroline v. United States, the Court repeated that description. 572 U. S. 434, 456 (2014). This Court’s precedents on related issues further buttress the conclusion that MVRA restitution is criminal punishment. For example, forfeiture under 18 U. S. C. §982(a)(1)—like restitution under the MVRA—occurs at the culmination of a criminal proceeding and requires conviction of an underlying crime. That forfeiture, the Court concluded, is criminal punishment. See United States v. Bajakajian, 524 U. S. 321, 328 (1998). By contrast, the Court determined that the forfeiture of firearms under 18 U. S. C. §924(d) ordered during an in rem proceeding is a civil sanction. United States v. One Assortment of 89 Firearms, 465 U. S. 354, 363–366 (1984). As the Government cautions, not everything that occurs at criminal sentencing or even that appears in a criminal judgment may necessarily be part of the punishment. See Tr. of Oral Arg. 22–23. But we need not explore that point further here because, for the reasons we have explained, MVRA restitution is criminal punishment. Amicus relies heavily on Smith v. Doe, 538 U. S. 84. There, this Court considered a law requiring a defendant convicted of certain crimes to register as a sex offender. The Court held that the registration mandate was civil. The legislature adopted “distinctly civil procedures” for the imposition of registration requirements. Id., at 96 (quotation marks omitted). By contrast, to reiterate what we said above, MVRA restitution is labeled as a penalty, is codified in the criminal code, is predicated on a criminal conviction, is imposed against a criminal defendant, is sometimes imposed in lieu of other penalties, is ordered at sentencing where the United States is the adverse party, and can result in resentencing when the defendant refuses to pay. So Smith v. Doe does not control. Cite as: 607 U. S. ____ (2026) 5
It is true, as amicus thoroughly explains, that the MVRA seeks to compensate crime victims, a nonpunitive goal. Several provisions of the MVRA reflect that objective, such as a provision requiring that restitution amounts be based on a victim’s loss. §3664(f )(1)(A); see also, e.g., §3664(j). But those provisions show only that Congress intended restitution under the MVRA to both punish and compensate. And so long as the text and structure of the Act demonstrate that Congress intended at least “to impose punishment,” that “ends the inquiry.” Smith, 538 U. S., at 92; see Paroline, 572 U. S., at 456; Pasquantino, 544 U. S., at 365. To be sure, as amicus points out, victims receive notice of sentencing proceedings and are consulted by prosecutors. But outreach to victims and some participation by victims in criminal proceedings are not unusual. See, e.g., §3771(a). With respect to restitution, the key point is that victims do not have the power to initiate or settle the restitution process as they would if it were a civil proceeding. Our ruling today does not mean that a restitution statute can never be civil. But the statutory text and structure of the MVRA demonstrate that restitution under that Act is criminal punishment. For those reasons, we reverse the judgment of the U. S. Court of Appeals for the Eighth Circuit and remand the case for further proceedings consistent with this opinion. On remand, the Court of Appeals may consider the Government’s separate arguments for affirmance of the District Court’s judgment. It is so ordered. Cite as: 607 U. S. ____ (2026) 1
THOMAS, J., concurring
HOLSEY ELLINGBURG, JR., PETITIONER v. UNITED STATES ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE EIGHTH CIRCUIT [January 20, 2026]
JUSTICE THOMAS, with whom JUSTICE GORSUCH joins, concurring. I join the Court’s opinion in full because it correctly ap- plies our precedent. I write separately to clarify the foun- dation of that precedent. This Court’s 1798 decision in Cal- der v. Bull, 3 Dall. 386, established that the Ex Post Facto Clauses forbid only those retroactive laws that impose “punishment” for a “crime.” Id., at 386, 389–391 (opinion of Chase, J.). Over the 228 years since Calder, the Court has struggled to articulate what it means for a law to impose punishment for a crime, and thus to be subject to the Ex Post Facto Clauses. The Court’s more recent precedents have implemented Calder through two multifactor tests that turn largely on whether the legislature labels the law as criminal or civil. But in 1798, “punishment” for a “crime” would have been understood to refer to any coercive penalty for a public wrong. Many laws that are nominally civil to- day would therefore have been subject to the Ex Post Facto Clauses under Calder. I would restore Calder’s approach to the Ex Post Facto Clauses. I The Constitution twice prohibits ex post facto laws. As to the Federal Government, it provides that “No Bill of Attain- der or ex post facto Law shall be passed.” Art. I, §9, cl. 3. 2 ELLINGBURG v. UNITED STATES
As to the States, it provides that “No State shall . . . pass any Bill of Attainder, ex post facto Law, or Law impairing the Obligation of Contracts.” Art. I, §10, cl. 1. In its general sense, a law is ex post facto—meaning “after the fact”— when it applies retroactively to conduct that occurred be- fore the law was enacted. The two Ex Post Facto Clauses reflected the importance of the protection to the Founding Fathers. James Iredell believed that the protection against ex post facto laws was “the most essential right of all,” which was worth “ten thou- sand declarations of rights” without it. Answers to Mr. Ma- son’s Objections to the New Constitution, in Pamphlets on the Constitution of the United States 334, 368 (P. Ford ed. 1888) (reprint 1968). “A man may feel some pride in his security,” Iredell wrote, “when he knows that what he does innocently and safely to-day in accordance with the laws of his country, cannot be tortured into guilt and danger to- morrow.” Ibid. In the Federalist Papers, Alexander Ham- ilton placed “the prohibition of ex post facto laws” among the greatest “securities to liberty and republicanism.” The Fed- eralist No. 84, p. 571 (P. Ford ed. 1898). James Madison argued that ex post facto laws are “contrary to the first prin- ciples of the social compact.” Id., No. 44, at 296. Ex post facto laws lack legitimacy because laws must pre- cede the actions that they govern. Laws regulating actions after the fact “deprive citizens of notice and fair warning and are, therefore, an affront to man’s ‘reason and freewill.’ ” Peugh v. United States, 569 U. S. 530, 561 (2013) (THOMAS, J., dissenting) (quoting 1 W. Blackstone, Com- mentaries on the Laws of England 39 (1765) (Blackstone)). As Blackstone explained, “it is impossible that the party could foresee” that his actions would become unlawful if they were lawful when he took them. Id., at 46. When a man had no notice that his actions were unlawful, “all pun- ishment for not abstaining must of consequence be cruel and unjust.” Ibid. An ex post facto law is even “more Cite as: 607 U. S. ____ (2026) 3
unreasonable,” Blackstone thought, than the reviled prac- tice of the Roman emperor “Caligula, who . . . wrote his laws in a very small character, and hung them up upon high pil- lars, the more effectually to ensnare the people.” Ibid. Many believed that no constitutional prohibition was nec- essary because an ex post facto law would be invalid as con- trary to natural law. At the Constitutional Convention, Ol- iver Ellsworth argued that “[i]t cannot . . . be necessary to prohibit” ex post facto laws because “no lawyer, no civilian . . . would not say that ex post facto laws were void of them- selves.” 2 M. Farrand, Records of the Federal Convention of 1787, p. 376 (1966) (Farrand). James Wilson thought that a constitutional prohibition on ex post facto laws would suggest that the Convention denied that such laws were al- ready forbidden by natural law, to its embarrassment. “It will bring reflexions on the Constitution—and proclaim that we are ignorant of the first principles of Legislation.” Ibid. In fact, “[a]ll the delegates who spoke explicitly or im- plicitly regarded an ex post facto law as a violation of natu- ral law, and most of them therefore thought it unnecessary to include such a basic natural law principle in the written constitution.” S. Sherry, The Founders’ Unwritten Consti- tution, 54 U. Chi. L. Rev. 1127, 1157 (1987). The Conven- tion nonetheless adopted the Ex Post Facto Clauses after it was urged that “the Judges can take hold of ” them if a leg- islature ever enacts such a law. 2 Farrand 376 (William- son). II The lodestar of this Court’s ex post facto jurisprudence is Calder v. Bull, which established that the Ex Post Facto Clauses apply only to laws that impose “punishment” for “crime[s].” 3 Dall., at 389–391. Calder—and especially Jus- tice Chase’s lead seriatim opinion—has since provided the definitive gloss on the Ex Post Facto Clauses. Justice Chase’s definition of “ex post facto laws” continues to 4 ELLINGBURG v. UNITED STATES
“provid[e] an authoritative account of the scope of the Ex Post Facto Clause.” Stogner v. California, 539 U. S. 607, 611 (2003); accord, e.g., Peugh, 569 U. S., at 532–533. This Court’s modern precedents, including today’s opinion, de- cide whether a law is subject to the Ex Post Facto Clauses based on whether it satisfies Calder’s requirement that it be “criminal or penal.” Weaver v. Graham, 450 U. S. 24, 29 (1981); see also ante, at 2. But this Court’s implementation of Calder seems to have lost sight of how Calder would have been understood when it was decided. Calder concerned a state legislature’s intervention in a traditional private-law dispute. The case arose out of the probate proceedings for the estate of a Connecticut man named Normand Morison. In 1779, Morison wrote a will leaving his property to his wife Abigail and her heirs. See 8 The Documentary History of the Supreme Court of the United States, 1789–1800, p. 89 (M. Marcus ed. 2007). Soon after Morison wrote his will, he and Abigail had a son. Ibid. When the son was young, Morison died. Ibid. Abigail started the probate process so that she could inherit Mori- son’s property as his will promised. Ibid. But the probate judge held up the process because he wanted to check whether the birth of their son affected the will’s validity. Ibid. At the time, Abigail did not mind the delay because the default intestacy rules would have given the property to her and her son anyway. Id., at 89–90. But later events jeopardized Abigail’s right to inherit Mo- rison’s property. First, Connecticut changed the default in- testacy rules so that, absent a valid will, property like Mo- rison’s would pass to his nearest blood relative, not to a surviving spouse like Abigail. Id., at 90, and n. 9. Second, the probate court, without telling Abigail, disapproved Mo- rison’s original will. Id., at 90. And third, a few years later, Abigail and Morison’s son died. Ibid. After her son died, Abigail and her new husband Caleb Bull returned to pro- bate court. Ibid. The Bulls were surprised to learn that the Cite as: 607 U. S. ____ (2026) 5
probate court had disapproved Morison’s will years earlier without telling Abigail—apparently on the theory that he would have wanted to change it after his son was born. Ibid. And, since Connecticut had enacted new default in- testacy rules in the meantime, the absence of a valid will meant that Morison’s estate would pass on to his blood rel- atives rather than to Abigail. Ibid. Morison’s blood rela- tives, the Calders, were suddenly the lawful heirs of his property. Id., at 90–91. Abigail and Caleb Bull petitioned the Connecticut Legis- lature for relief. They convinced the legislature that Mori- son did not intend to revoke his will after his son died, so the legislature issued a decree in 1795 granting them a re- hearing on the validity of Morison’s original will. Id., at 91. The legislature’s 1795 decree effectively vacated the origi- nal probate-court decision deeming the will void and granted a new hearing at which the probate court could ad- judicate the will’s validity. Id., at 91, 107. The probate court held a new hearing pursuant to the decree and con- cluded that Morison’s original will was valid after all. Id., at 91. Having now lost in probate court, the Calders ap- pealed this new decision and challenged the legislature’s decree that had granted the Bulls a new hearing. Id., at 91–92. The Calders argued, eventually to this Court, that the legislature’s decree was an unconstitutional ex post facto law because it retroactively voided the probate court’s original decision, which caused them to lose an inheritance that would otherwise have been theirs. Id., at 92. In Calder v. Bull, this Court gave two reasons why the legislature’s decree granting a new trial did not violate the State Ex Post Facto Clause. See Art. I, §10, cl. 1. The first reason was that the legislature’s decree operated on a legal decision by a court, not a past action by the Calders. There was nothing that the parties had done in the past that was “in any manner affected by the law or resolution of Connect- icut.” 3 Dall., at 392 (opinion of Chase, J.). “It does not 6 ELLINGBURG v. UNITED STATES
concern, or relate to, any act done by them.” Ibid. Instead, the decree effectively vacated a probate decision and gave the court an opportunity to correct its decision, just like ap- pellate courts do every day. Ibid. The new hearing, after all, was conducted under the same substantive laws as were in force at the original hearing. In fact, the Justices seemed to believe that the Connecticut Legislature’s “resolution or law,” id., at 387, was not a legislative but a judicial act, con- sistent with state legislatures’ traditional judicial power to review state-court decisions—and thus could not be an ex post facto “law” at all. See id., at 395–396 (opinion of Pat- erson, J.) (“[W]e may, in the present instance, consider the Legislature of the State, as having acted in their customary judicial capacity”); id., at 398 (opinion of Iredell, J.) (“an ex- ercise of judicial, not of legislative, authority”); id., at 400 (opinion of Cushing, J.) (“If the act is a judicial act, it is not touched by the Federal Constitution”). The more lasting legacy of Calder, however, was its sec- ond reason. Three of the four seriatim opinions in Calder limited the Ex Post Facto Clauses to laws imposing criminal punishment. On their view, the Clauses did not forbid other kinds of retroactive laws. Justice Chase, in what has proved to be the most influential opinion, gave a compre- hensive account of the scope of the Ex Post Facto Clauses. He understood the Clauses to be limited to retroactive laws imposing criminal punishment. He explained that the Ex Post Facto Clauses covered only those laws that “create” a “crime,” “aggravate” a crime, “encrease the punishment” for a crime, or “change the rules of evidence, for the purpose of conviction.” Id., at 391. Justice Chase’s account implied that the Clauses did not apply to other kinds of retroactive laws. First, on his view, the Clauses did not touch private law. Justice Chase did “not think” that the Ex Post Facto Clause “was inserted to Secure the citizen in his private rights, of either property, or contracts.” Id., at 390. Accordingly, the Clauses would not Cite as: 607 U. S. ____ (2026) 7
prohibit a law that retroactively paused a debtor’s obliga- tions to a creditor or voided a contract. Such laws, in Jus- tice Chase’s view, were governed by separate constitutional prohibitions on impairing the obligations of contracts and making anything but gold or silver a tender in payments of debts. Ibid.; see Art. I, §10, cl. 1. Second, because the Clauses were meant as protections against the government, they did not apply to laws that retroactively made a law more lenient. They would not, for example, forbid a law that retroactively “mollifies the rigor of the criminal law.” Calder, 3 Dall., at 391 (opinion of Chase, J.). The other two seriatim opinions in Calder that addressed the question agreed. Justice Paterson explained that “the meaning, annexed to the terms ex post facto laws, unques- tionably refers to crimes, and nothing else.” Id., at 396. Justice Iredell—an adamant proponent of the Ex Post Facto Clauses, see supra, at 2—likewise thought that they did not “extend to civil cases, to cases that merely affect the private property of citizens.” Calder, 3 Dall., at 400. Instead, the Clauses applied only to laws that retroactively “inflict a punishment” for an “offence.” Ibid. Some, myself included, have questioned whether Calder was right to limit the Ex Post Facto Clauses to laws impos- ing criminal punishment. See, e.g., Satterlee v. Matthew- son, 2 Pet. 380, 416, and n. (a) (1829) (opinion of Johnson, J.) (arguing that the Ex Post Facto Clauses protect more); O. Field, Ex Post Facto in the Constitution, 20 Mich. L. Rev. 315, 331 (1922) (“[T]here have been reputable authorities, both past and present, who incline to the view that the ex post facto provisions of the Constitution prohibited civil as well as criminal legislation”); Eastern Enterprises v. Apfel, 524 U. S. 498, 538–539 (1998) (THOMAS, J., concurring). Others have defended Calder’s limitation. See, e.g., R. Natelson, Statutory Retroactivity: The Founders’ View, 39 Idaho L. Rev. 489, 493–494 (2003); C. Nelson, Originalism and Interpretive Conventions, 70 U. Chi. L. Rev. 519, 578– 8 ELLINGBURG v. UNITED STATES
582 (2003). But often lost in that debate has been the ques- tion of what Calder meant when it limited the Clauses to laws imposing criminal punishment. III This Court’s recent precedents have attempted to imple- ment Calder’s limitation of the Clauses to criminal punish- ment. But, in doing so, they have adopted a framework that turns largely on legislative labeling, has little basis in his- tory, and is unnecessarily convoluted. A The modern precedents follow Calder’s limitation of the Ex Post Facto Clauses to laws that impose criminal punish- ment. Based on “Justice Chase’s now familiar opinion in Calder,” this Court has repeatedly held that the Clauses apply exclusively to laws that “retroactively alter the defi- nition of crimes or increase the punishment for criminal acts,” but not to other kinds of retroactive laws. Collins v. Youngblood, 497 U. S. 37, 41–43 (1990); see also, e.g., Kan- sas v. Hendricks, 521 U. S. 346, 370 (1997); Smith v. Doe, 538 U. S. 84, 92 (2003). But in defining criminal punishment—and thus the scope of Calder’s limitation—the Court’s modern framework ap- plies two multifactor tests developed in the 20th century. The first multifactor test asks whether the legislature “in- tended” the law to be viewed as criminal or penal. Smith, 538 U. S., at 92–93; see also ante, at 2, n. 1. Under this test, a law is subject to the Ex Post Facto Clauses “[i]f the inten- tion of the legislature was to impose punishment.” Smith, 538 U. S., at 92. But if “the intention was to enact a regu- latory scheme that is civil and nonpunitive,” then a law is presumptively not subject to the Ex Post Facto Clauses. Ibid. This test has included up to five factors, most of which allow a legislature to avoid the Clauses through labeling or semantics. Those five factors are whether the legislature Cite as: 607 U. S. ____ (2026) 9
had a “ ‘preference’ ” for “ ‘one label or the other,’ ” id., at 93 (quoting Hudson v. United States, 522 U. S. 93, 99 (1997)); whether the “codification” was in the criminal or civil code, Smith, 538 U. S., at 94; whether legislative history or simi- lar indicators suggested nonpunitive goals, United States v. One Assortment of 89 Firearms, 465 U. S. 354, 364 (1984); whether the legislature provided traditional “safeguards associated with the criminal process,” Smith, 538 U. S., at 96; and whether the legislature gave enforcement power to an agency with power to enforce civil laws, ibid. The second multifactor test, which applies only if the law survives the first test, asks whether the law’s other features render it criminal or penal. See id., at 92. Under this sec- ond test, a law is subject to the Ex Post Facto Clauses if it is “so punitive either in purpose or effect as to negate [the legislature’s] intention to deem it civil.” Ibid. (internal quo- tation marks omitted). The Court has at times said that this second test is a function of up to seven factors: whether the law involves an affirmative disability or restraint, im- poses what has historically been regarded as a punishment, requires a finding of scienter, promotes the traditional aims of punishment, applies to behavior which is already a crime, lacks a rational alternative purpose, or is excessive in rela- tion to that purpose. Kennedy v. Mendoza-Martinez, 372 U. S. 144, 168–169 (1963); see Smith, 538 U. S., at 97 (ex- plaining that Mendoza-Martinez factors “migrated” to ex post facto jurisprudence). So long as a law survives the first test, “ ‘only the clearest proof ’ ” under the second test can make it subject to the Ex Post Facto Clauses. Smith, 538 U. S., at 92 (quoting Hudson, 522 U. S., at 100). B This modern framework is incongruous with the histori- cal purpose of Ex Post Facto Clauses. The modern frame- work, because of how much it turns on legislative labeling and semantics, allows a legislature to manipulate when the 10 ELLINGBURG v. UNITED STATES
protection will apply. Under the first test, for example, a legislature might be able to impose a retroactive $10,000 fine on some previously innocent conduct—such as drinking coffee or going to the gym—if it labeled the law “civil,” placed it in the civil code, expressed a nonpunitive purpose, omitted protections for the accused, and let non-traditional prosecutors (such as a health agency) enforce it. See supra, at 2–3. Under the second test, the legislature might be able to prevail by aggravating the fair-notice problems that mo- tivate the Clauses, such as by having its retroactive law im- pose strict liability and then applying it only to previously innocent conduct. See Mendoza-Martinez, 372 U. S., at 168–169. It is unlikely that a constitutional prohibition de- signed to curb a “cruel and unjust” legislature, 1 Blackstone 46, can be so easily manipulated by the same legislature. See United States v. Rahimi, 602 U. S. 680, 776 (2024) (THOMAS, J., dissenting) (explaining that the Constitution does not “allow . . . majoritarian interests to determine” the scope of “constitutional rights”). The modern framework is also convoluted. It is difficult for courts, let alone ordinary citizens, to predict how this Court will weigh the modern framework’s combined twelve factors spread over two tests. This Court’s own precedents admit that the twelve factors are “neither exhaustive nor dispositive,” United States v. Ward, 448 U. S. 242, 249 (1980), and need not be given equal weights, Smith, 538 U. S., at 105. Such “multifactor balancing test[s]” invite “policy-driven, ‘arbitrary discretion.’ ” Gamble v. United States, 587 U. S. 678, 724 (2019) (THOMAS, J., concurring). We should strive to avoid such unclear rules in any context, but especially when interpreting a constitutional guarantee of clear notice. IV When Calder said that the Ex Post Facto Clauses apply only to laws imposing punishment for crimes, it was Cite as: 607 U. S. ____ (2026) 11
referring to an established category of laws. A crime meant a “public wrong,” which is an injury to the sovereign in its sovereign capacity. Calder therefore encompassed offenses against the sovereign regardless of whether they were nom- inally criminal or civil. And punishment simply referred to the law’s coercive sanction—meaning a traditional depriva- tion of life, liberty, or property—redressing that public wrong. Applying this understanding would restore the Clauses to their proper role without the problems attendant to the more recent precedents. A The English common law divided all unlawful acts into “private wrongs” and “public wrongs.” See 3 Blackstone 2. Blackstone titled the third and fourth volumes of his Com- mentaries, respectively, “Of Private Wrongs” and “Of Public Wrongs.” This dichotomy was fundamental to English and American law. Public wrongs were injuries to the sovereign. A public wrong was “a breach and violation of public rights and du- ties, which affect the whole community, considered as a community.” Ibid. Because a public wrong injured the com- munity, it was deemed an injury to the sovereign in his sov- ereign capacity. 4 Blackstone 2–7. It followed that an ac- tion to redress a public wrong was brought on behalf of the sovereign, not the victim. J. Locke, Second Treatise of Civil Government 7 (J. Gough ed. 1948) (Locke). In concrete terms, actions for public wrongs were brought in England on behalf of the King, and here on behalf of the State or the United States—typically by public prosecutors, rather than privately by the victims. The sovereign “is supposed by the law to be the person injured by every infraction of the public rights belonging to that community, and is therefore in all cases the proper prosecutor for every public offense.” 4 Blackstone 2; see also Robertson v. United States ex rel. Watson, 560 U. S. 272, 279 (2010) (ROBERTS, C. J., 12 ELLINGBURG v. UNITED STATES
dissenting) (a criminal prosecution is an action “on behalf of the sovereign, seeking to vindicate a public wrong”). Pub- lic wrongs included all wrongs deemed injuries to the sov- ereign and redressable by the sovereign, from murder to mayhem to public nuisance. 4 Blackstone 5–6. Private wrongs, in contrast, were injuries to individuals in their private capacity. An action inflicted a private wrong insofar as it infringed “the private or civil rights be- longing to individuals, considered as individuals.” 3 Black- stone 2. For example, a dispute with a fellow citizen over a land claim or a failure to pay a debt concerned a private wrong. “[I]f I detain a field from another man, to which the law has given him a right, this is a civil injury, and not a crime,” Blackstone explained, because “only the right of an individual is concerned, and it is immaterial to the public, which of us is in possession of the land.” 4 Blackstone 5. Private wrongs were redressable by the individuals whose rights were violated, not the sovereign. A single act could be both a private wrong and a public wrong. Id., at 5–6. “The same acts will generally constitute a private injury, as well as a public crime. A public punish- ment is inflicted on the part of the state, to restrain such conduct, and the party injured is entitled to a compensation for the injury he has sustained.” 2 Z. Swift, System of the Laws of the State of Connecticut 292 (1796); accord ibid. (“If one man assaults and beats another, he is punished by a fine for disturbing the peace, and compelled to compen- sate in money the person he has abused and wounded.”). The private suit brought by the individual vindicated the private wrong, and the public prosecution brought by the sovereign vindicated the public wrong. This overlap carries forward to modern law: Theft can be redressed through a tort suit brought by the victim and a criminal prosecution brought by the state. When the opinions in Calder spoke of “crimes,” they meant public wrongs. The prevailing definition of a “crime” Cite as: 607 U. S. ____ (2026) 13
around the time of Calder was “a violation of law to the in- jury of the public, a public offense.” N. Webster, A Compen- dious Dictionary of the English Language 72 (1806) (Web- ster). Blackstone used the terms “crimes” and “public wrongs” interchangeably. See 4 Blackstone 1–5. His fourth volume covered “public wrongs, or crimes and misdemean- ors.” Id., at 1.1 The “common-law conception of crime,” this Court has long recognized, was an “offense against the sov- ereignty of the government.” Heath v. Alabama, 474 U. S. 82, 88 (1985). So Calder’s limitation of the Ex Post Facto Clauses to crimes would have originally been understood to include all public wrongs. Calder itself confirms this understanding. Justice Chase distinguished crimes not with nominally civil laws enforced by the sovereign, but with laws that merely “secure[d] pri- vate rights,” like contract laws. 3 Dall., at 390. Justices Chase and Iredell used “crime” interchangeably with “of- fence.” Ibid.; id., at 400 (opinion of Iredell, J.). Justices Chase and Paterson said that their definitions tracked Blackstone’s. See id., at 391 (opinion of Chase, J.) (“Sir Wil- liam Blackstone, in his commentaries, considers an ex post facto law precisely in the same light I have done.”); id., at 396 (opinion of Paterson, J.) (similar). Justice Iredell ex- plained in a contemporaneous opinion that the Ex Post Facto Clauses could be limited to crimes because “there . . . is little reason to apprehend a legislative interference for the sake of unjustly transferring property from one man to another.” Minge v. Gilmour, 17 F. Cas. 440, 443, No. 9,631 (CC NC 1798). And, of course, the occasion for Calder’s dis- tinction was a probate dispute, where only private rights are implicated.2 —————— 1 Crimes and misdemeanors were for these purposes “syn-
onymous.” 4 Blackstone 5. 2 Professor Caleb Nelson has read Calder along the same
lines: “By ‘criminal’ laws,” he has explained, “Justice Iredell 14 ELLINGBURG v. UNITED STATES
In turn, when Calder referred to criminal “punishment,” it simply meant whatever coercive sanction—or deprivation of life, liberty, or property—the law imposed as redress for committing those crimes. Blackstone understood punish- ment to include the traditional “coercive penalties” that the law imposed as redress for a public wrong. See 4 Black- stone 7–8. Dictionaries defined “punishment” broadly as “[a]ny infliction or pain imposed in vengeance of a crime,” S. Johnson, A Dictionary of the English Language (6th ed. 1785), or “any thing inflicted for a crime,” Webster 241. The power of punishing an offender was the power to harm the offender in a way that was otherwise forbidden: “[N]o one ought to harm another in his life, health, liberty, or posses- sions”—“unless it be to do justice on an offender.” Locke 5 (emphasis added); see also id., at 6 (describing when “one man may lawfully do harm to another, which is that we call punishment”). The Ex Post Facto Clauses therefore prohibit retroactive laws that impose coercive penalties for public wrongs. Con- trary to the modern framework, it does not matter what the legislature labels the law, where it places the law, which agency it vests enforcement with, what its stated goals were, whether it provides safeguards for the accused, whether it requires a showing of scienter, or whether the conduct to which it applies is already a crime. Contra, Smith, 538 U. S., at 96; Martinez-Mendoza, 372 U. S., at 168–169. What matters is what the law does. If it punishes a public wrong—or an injury to the sovereign in its sover- eign capacity—then it is subject to the Ex Post Facto
—————— and his colleagues may simply have meant laws that de- fined offenses and imposed punishments for them,” even if “the penalties that they prescribed could be collected through civil suits.” C. Nelson, Originalism and Interpre- tive Conventions, 70 U. Chi. L. Rev. 519, 582, n. 255 (2003). Cite as: 607 U. S. ____ (2026) 15
Clauses. If it changes merely private relations, such as the rules of contract or property or probate, then it is not. Whether a law is subject to the Ex Post Facto Clauses will therefore typically depend on how it is enforced. If it is en- forced on behalf of the sovereign to redress a sovereign in- jury, then it is subject to the Clauses. See 4 Blackstone 2. If instead it is enforced by a private person to vindicate his own private rights, then it is not. In this case, for example, the law was subject to the Ex Post Facto Clauses because it was enforced against Ellingburg by the United States, not by the First Union National Bank, whose private rights El- lingburg violated when he robbed it. And the United States’s action was redressing a sovereign injury to the community as a whole, not a private wrong. Cf. 3 Black- stone 257 (explaining that the King can bring “common law actions” to redress private wrongs to himself, such as an “action for trespass for taking away his goods”).3 Calder’s understanding secures Ex Post Facto Clause pro- tection in a wide range of contexts involving nominally civil laws. Those contexts include civil proceedings seeking fines for public offenses. See, e.g., Army Corps of Engineers v. Hawkes Co., 578 U. S. 590, 600 (2016) (describing “civil pen- alties of up to $37,500 for each day [the challenger] violated the Act”). They include enforcement proceedings brought by administrative agencies. See Axon Enterprise, Inc. v. FTC, 598 U. S. 175, 196 (2023) (THOMAS, J., concurring) (describing the 20th-century rise of such proceedings). And they include municipal sanctions like speeding tickets. “If —————— 3 Sovereigns sometimes delegated enforcement of public
wrongs to private persons. But, unlike in actions redress- ing private wrongs, the private persons in those circum- stances acted “on behalf of the sovereign” and were subject to the sovereign’s control. Robertson v. United States ex rel. Watson, 560 U. S. 272, 279 (2010) (ROBERTS, C. J., dissent- ing); see ibid. (collecting authorities). 16 ELLINGBURG v. UNITED STATES
one were to commit a minor traffic offense at a time the of- fense was to be punished by a $25 fine, and the government were to then amend the statute and impose a fine of $1,000,000 dollars, it would be nonsensical to treat that fine as non-punitive simply because the offense was processed civilly.” Brief for Professor Beth Colgan as Amicus Curiae 8, n. 3. Because all of these offenses impose punishments for public wrongs, Calder would treat them all as subject to the Ex Post Facto Clauses. B This understanding of Calder long prevailed. Courts took the position that “the relevant line for the Clause’s scope” lay between “punishments imposed in response to public of- fenses whether prosecuted criminally or civilly on the one hand, and purely private disputes on the other.” Brief for Professor Beth Colgan as Amicus Curiae 5. This Court con- firmed, contrary to the modern precedents, that “the ex post facto effect of a law cannot be evaded by giving a civil form to that which is essentially criminal.” Burgess v. Salmon, 97 U. S. 381, 385 (1878). The influential commentator Thomas Cooley likewise explained that a “law will be ex post facto in the constitutional sense” if it “subject[s] an in- dividual to a pecuniary penalty” or “deprives a party of any valuable right” retroactively, even if “it does not in terms declare the acts to which the penalty is attached criminal.” Constitutional Limitations 321 (6th ed. 1890). In a similar context, when a State sought nominally civil fines against a company for doing business without proper forms, this Court explained that the statute “was in the strictest sense a penal statute.” Wisconsin v. Pelican Ins. Co., 127 U. S. 265, 299 (1888). The Court held that the nominally civil statute was penal because “[t]he cause of ac- tion was not any private injury, but solely the offence com- mitted against the State by violating her law.” Ibid. Like- wise, “[t]he prosecution was in the name of the State.” Ibid. Cite as: 607 U. S. ____ (2026) 17
The Court explained that “[t]he real nature of the case is not affected by the forms provided by the law of the State”; these forms were “immaterial” to whether the law was criminal or penal in a constitutional sense. Ibid. Thus, the Court concluded, “[i]n whatever form the State pursues her right to punish the offence against her sovereignty,” an ac- tion is criminal or penal in substance whenever it seeks to “compe[l ] the offender to pay a pecuniary fine by way of punishment for the offence.” Ibid. Other decisions followed the same reasoning. For exam- ple, when the United States brought a nominally civil ac- tion for damages based on a public offense, the defendants were entitled to other criminal-procedure protections be- cause the action “was in its nature and essence, though not its form, a penal or criminal action.” United States v. Poyllon, 27 F. Cas. 608, 611, No. 16,081 (NY 1812). And when a statute forbade “any fine or forfeiture under any pe- nal statute,” the statute applied to a nominally civil action for debt on behalf of the sovereign against someone who committed a public offense. Adams v. Woods, 2 Cranch 336, 337, 340–341 (1805) (emphases deleted). Chief Justice Marshall explained that the category of penal laws referred “not to any particular mode of proceeding, but generally to any prosecution, trial, or punishment for the offence.” Ibid. This understanding of Calder harmonizes the Ex Post Facto Clauses with their historical purpose. It means that the Clauses ensure fair notice against government enforce- ment actions, regardless of whether the legislature labels them civil. As JUSTICE GORSUCH has explained, today’s “civil” laws “regularly impose penalties far more severe than those found in many criminal statutes.” Sessions v. Dimaya, 584 U. S. 148, 184 (2018) (opinion concurring in part and concurring in the judgment). “Today’s ‘civil’ pen- alties include confiscatory rather than compensatory fines, forfeiture provisions that allow homes to be taken, remedies that strip persons of their professional licenses and 18 ELLINGBURG v. UNITED STATES
livelihoods, and the power to commit persons against their will indefinitely.’ ” Ibid. This understanding also helps rec- oncile Calder with its critics, many of whom have focused on the injustice of allowing the government to retroactively impose sanctions for public offenses that it deems civil. See, e.g., Brief for Professor Wayne A. Logan as Amicus Curiae 31–32; cf. D. Troy, Retroactive Legislation 49–55 (1998). This understanding of Calder also simplifies the law. Ra- ther than making the Ex Post Facto Clauses’ application de- pend on twelve factors with indeterminate weights, this un- derstanding makes it depend on a simple legal inquiry with a long pedigree in our legal tradition—whether the law im- poses a coercive penalty to redress a public wrong. See, e.g., Pelican Ins. Co., 127 U. S., at 299; Robertson, 560 U. S., at 278–279 (ROBERTS, C. J., dissenting). In most cases, that inquiry will just come down to who enforces the law, the sovereign or the injured private party. * * * In Collins v. Youngblood, this Court explained that any “departure from Calder’s explanation of the original under- standing of the Ex Post Facto Clause[s] was . . . unjustified.” 497 U. S., at 49. The Court understood itself to be bound by what Calder meant rather than by its intervening prece- dents that misunderstood Calder. 497 U. S., at 49–50. To- day, this Court’s precedents concerning the scope of laws imposing criminal punishment have departed from Calder’s understanding of that category. In a future case, the Court should consider returning to Calder’s understanding.