The Federal Ex Post Facto Clause

No . . . ex post facto Law shall be passed.

—Evan C. Zoldan

Introduction

Article I, Section 9, Clause 3 provides that no “ex post facto Law shall be passed” by Congress. A companion clause, Article I, Section 10, Clause 1, similarly provides that no state shall “pass any . . . ex post facto Law.”¹ The U.S. Supreme Court has interpreted these two clauses to prohibit, roughly speaking, retroactive criminal laws at both the federal and state levels. Both clauses arose out of the same historical experiences, support the same intuitive policies, and share the same modern judicial interpretations. This essay describes the conceptual background and judicial interpretations of these two clauses. Essay No. 81 describes the historical background surrounding the inclusion of these clauses in the Constitution, focusing on the events that bear most heavily on their meaning today.

Retroactivity and the Ex Post Facto Clauses

A statute is retroactive when it changes the legal consequences of conduct that occurred before the statute’s enactment. Retroactive legislation has long been viewed with suspicion and even hostility. In Ogden v. Saunders (1827), Justice Bushrod Washington denounced retroactive laws as “oppressive, unjust, and tyrannical.”² Modern scholars are no kinder. Lon Fuller’s view that retroactive laws are “truly a monstrosity” represents an almost universally held view.³ The prohibition on retroactive statutes serves some intuitively attractive policies, including protecting the value of fair notice. When individuals are assured that their conduct will be governed by laws that already are in effect, they are better able to plan their conduct with reasonable certainty of the legal consequences of their actions.⁴ Prohibiting retroactivity also reinforces the separation of powers,⁵ ensures fundamental fairness,⁶ and upholds the rule of law.⁷

Despite this consistent aversion to retroactivity, the Constitution’s Ex Post Facto Clauses have never been interpreted by the Supreme Court to apply to all retroactive legislation. One reason is textual. Reading the clauses to cover all retroactive statutes would lead them to overlap significantly with other clauses of the Constitution. For example, the Contracts Clause, which bars the states from “impairing the Obligation of Contracts,” itself prohibits some types of retroactive action.⁸ (See Essay No. 82.)

Another reason is practical. Sometimes statutes are popular precisely because of their retroactive effect. New York, for example, retroactively extended a statute of limitation for civil suits and even revived claims of child abuse that were barred by a generally applicable statutes of limitations.⁹ Perhaps for these textual and practical reasons, the Supreme Court has never read the Constitution’s Ex Post Facto clauses for all they are worth, instead interpreting them to prohibit a subset of retrospective laws—perhaps the most odious—while excluding from their scope other kinds of retrospective laws.¹⁰

Calder’s Four Categories of Prohibited Retroactive Laws

Calder v. Bull (1798) was the Supreme Court’s first and most enduring statement concerning the scope of the Ex Post Facto clauses. Justice Samuel Chase described four categories of retroactive laws that fall within the Constitution’s prohibitions. Although the Supreme Court has used other formulations throughout the centuries, it has returned time and again to these categories.

Calder’s first category includes statutes that attach criminal liability to an action that was innocent when it was done. This type of statute is perhaps the archetype of a retroactive law and calls to mind the reasons why retroactive laws have been thoroughly denounced over the centuries. This category is exemplified by Cummings v. Missouri (1866).¹¹ Cummings declared unconstitutional a state law that penalized anyone who practiced certain professions without first taking an oath that he had not engaged in conduct that was, at the time, lawful. The Court held that a law directed at acts that “were not offences at the time they were committed” is an impermissible ex post facto law. Today, statutes that fall within the first category are rare.

The second category includes statutes that “aggravate[] a crime, or make[] it greater than it was, when committed.” Calder did not clarify what it means to “aggravate” a crime, but later cases have explained that “aggravating” a crime includes making punishment available when it previously would not have been available “in the ordinary course of law.”¹² For example, Stogner v. California (2003) declared unconstitutional a statute that extended the statute of limitations for the criminal prosecution of certain offenses, even as to conduct for which the statute of limitations already had expired. The Court held that because prosecution was not available after the expiration of the statute of limitations, there was no legal consequence to the conduct “in the ordinary course of law.” Accordingly, a statute that purported to allow prosecution to resume impermissibly aggravated the crime by making prosecution available when it would not ordinarily have been available.

The third category includes laws that “change[] the punishment, and inflict[] a greater punishment, than the law annexed to the crime, when committed.” In other words, laws cannot increase the quantum of punishment compared with what was available at the time the conduct was committed. This category, unlike the first, contemplates that the conduct already was unlawful even before the change in law; unlike the second, it contemplates that at least some punishment already was available under the old law. By contrast, this third category addresses the situation in which the change in law creates a “sufficient risk” that the person subject to punishment will receive more punishment than he would have received under the law in force at the time the unlawful conduct was committed.¹³ Peugh v. United States (2013) held that amendments to the federal sentencing guidelines that raised the punishment guideline for an already committed offense violate the Ex Post Facto Clause. This was true, the Court reasoned, despite the fact that the guidelines are advisory and leave sentencing discretion to judges. Raising the punishment guideline creates a “sufficient risk” that judges would apply a harsher punishment. As a result, the retroactive application of the more severe guidelines was impermissibly ex post facto.¹⁴

The fourth category includes laws that “alter[] the legal rules of evidence” to require “less, or different, testimony, than the law required at the time of the commission of the offence” to secure a conviction. In short, the rules of evidence cannot be modified retrospectively to make it easier to convict the accused. The classic example of a statute in this category retrospectively reduces the number of witnesses required to convict from two to one. This reduction makes conviction easier by decreasing the work the state must do to convict.¹⁵ For example, in Carmell v. Texas (2000), at the time the unlawful conduct was committed, the defendant could be convicted only on a record that included both the victim’s testimony and corroborating evidence.¹⁶ By the time he was convicted, an amendment had altered the rule of evidence to require only the victim’s testimony, without corroborating evidence, to convict. The Court held that the change reduced the “quantum of evidence required to convict” and was therefore impermissibly ex post facto.¹⁷

Open Questions

• The Ex Post Facto clauses prohibit a law only if it is punitive in nature—but what precisely qualifies as punitive? A statute is punitive if either the legislature intended it to punish or the effect of the statutory scheme is to punish.¹⁸ The Court has declined to frame a bright-line rule to determine whether a statute has a punitive effect, instead considering a number of factors to resolve whether a statutory scheme is sufficiently punitive to trigger the clauses.¹⁹ These factors, because they are both malleable and indeterminate, fail to provide predictable results.²⁰
• Are Calder’s four categories exemplary or exclusive? Justice Chase suggested that they are not exclusive.²¹ However, two centuries later, Collins v. Youngblood (1990) implied that these categories are exclusive and may neither be augmented nor diminished.²² Other recent cases, like Peugh and Weaver v. Graham (1981), have taken a less categorical approach. As a result, the precise contours of the clauses’ protections are not wholly settled.
• Do the Ex Post Facto clauses apply only to legislative action? The Supreme Court has offered mixed signals on this question. Some cases suggest that the provision applies to legislative action alone,²³ but the Court has noted that “the Ex Post Facto Clause is not limited to legislative acts.”²⁴ Extending the clauses to executive and judicial conduct would not only expand the scope of the clauses’ protections but also create significant overlap between them and other constitutional clauses, like the Due Process clauses.

1. Art. I, §§ 9, 10. ↩︎
2. Ogden v. Saunders, 25 U.S. 213 (1827). ↩︎
3. Lon Fuller, The Morality of Law 53 (1969). ↩︎
4. Weaver v. Graham, 450 U.S. 24, 30 (1981). ↩︎
5. Carmell v. Texas, 529 U.S. 513, 530 (2000). ↩︎
6. Charles B. Hochman, The Supreme Court and the Constitutionality of Retroactive Legislation, 73 Harv. L. Rev. 692, 693 (1960). ↩︎
7. Carmell, 529 U.S. at 530. ↩︎
8. Landgraf v. USI Film Prods., 511 U.S. 244 (1994). ↩︎
9. Child Victims Act, N.Y.C.P.L.R. 214-g. ↩︎
10. Douglas Kmiec & John O. McGinnis, The Contract Clause: A Return to the Original Understanding, 14 Hastings Con. L. Q. 525, 528 (1987). ↩︎
11. 71 U.S. (4 Wall.) 277, 327–28 (1866). ↩︎
12. Stogner v. California, 539 U.S. 607, 613 (2003). ↩︎
13. Peugh v. United States, 569 U.S. 530, 539 (2013). ↩︎
14. Id. ↩︎
15. Carmell, 529 U.S. at 530. ↩︎
16. Id. ↩︎
17. Id. at 532. ↩︎
18. Smith v. Doe, 538 U.S. 84 (2003). ↩︎
19. Id. ↩︎
20. Wayne Logan, The Ex Post Facto Clause (2022). ↩︎
21. 3 U.S (3 Dall.) at 391. ↩︎
22. 497 U.S. at 42. ↩︎
23. United States v. Marcus, 560 U.S. 258, 264 (2010); Nicholas Quinn Rosenkranz, The Objects of the Constitution, 63 Stan. L. Rev. 1005, 1019–20 (2011). ↩︎
24. Peugh, 569 U.S. at 544–45. ↩︎