For years, courts and legislatures have treated criminal restitution as if it were something softer than punishment—a kind of civil payback to victims that somehow floats outside the usual constitutional protections. In Ellenberg v. United States, the U.S. Supreme Court just shattered that illusion.

In a unanimous decision, the Court held that restitution imposed under the Mandatory Victims Restitution Act (MVRA) is plainly criminal punishment for purposes of the Ex Post Facto Clause. That single sentence has big consequences for anyone sentenced under the MVRA for conduct that occurred before the law existed, and it forces the legal system to confront the way massive, long‑term financial obligations function as a “second sentence” after prison.

But the ruling is not the revolution some advocates are hoping for. It doesn’t overturn Smith v. Doe, the 2003 case that upheld sex‑offense registration as “civil.” It doesn’t magically convert every collateral consequence into punishment. And as one of the commentators in the transcript warns, Congress could rewrite the restitution scheme to make it look more “civil” again.

This article unpacks what Ellenberg actually decided, how the Court got there, what Justice Thomas is signaling in his separate opinion, and what it all means for the future of ex post facto challenges and criminal justice reform.

What Is Restitution, Really?

At its core, restitution is money a person convicted of a crime must pay because of that offense. In federal cases under the MVRA:

  • The defendant is ordered to pay a fixed amount.
  • The state collects it.
  • The victim ultimately receives the money.

On paper, that can sound like a purely civil remedy—one person compensating another for a loss. But functionally, restitution is wrapped into the criminal process:

  • It is imposed at sentencing, alongside imprisonment and fines.
  • Only someone convicted of a crime can be ordered to pay it.
  • The government, not the victim, appears as the party opposing the defendant.

For decades, courts often treated restitution like a hybrid: civil in label, punitive in feel. Ellenberg forces the system to say out loud what’s been obvious in practice: this is punishment.

The Human Story Behind Ellenberg

The case doesn’t arise in a vacuum. It comes out of what happens after a prison sentence ends.

The petitioner, Ellenberg, was ordered to pay $7,567.25 in restitution. Over decades, with interest, that amount nearly doubled. He had committed his crime before Congress enacted the MVRA in 1996, but he was sentenced after it took effect—and sentenced under the MVRA’s mandatory scheme.

Advocates like Andrew Burrell describe this as a “second invisible sentence.” You serve your time, but you walk out of prison into:

  • Decades of accruing interest
  • Collection efforts by the government
  • Ongoing financial obligations that you may never realistically satisfy

As the National Association of Criminal Defense Lawyers (NACDL) put it, thousands of people have been “shackled by debt long after their prison cells were opened.” The Court’s decision meets that reality head‑on: if something walks, talks, and bites like punishment, you can’t simply label it “civil” to escape the Constitution.

The Legal Issue: Ex Post Facto and the MVRA

The Ex Post Facto Clauses of the Constitution prohibit laws that retroactively increase the punishment for a crime. Two key points matter:

  1. The law must be retroactive (applied to conduct that occurred before the law was enacted).
  2. The law must impose punishment.

Ellenberg’s argument was simple:

  • He committed his crime before the MVRA.
  • He was later sentenced under the MVRA.
  • The MVRA made restitution mandatory and more severe.
  • Therefore, applying it to him violated the Ex Post Facto Clause—if restitution counts as punishment.

The Eighth Circuit held that restitution under the MVRA was not criminal punishment, treating it as a civil, compensatory device. That meant the Ex Post Facto Clause didn’t apply. The Supreme Court disagreed and reversed.

An Unusual Twist: When Both Sides Say the Lower Court Was Wrong

One striking feature of Ellenberg is procedural rather than substantive.

On the very first page of the opinion, the Court notes that both Ellenberg and the United States agreed the Eighth Circuit had erred and that MVRA restitution is criminal punishment.

When both true parties to a case abandon the judgment they’re supposed to be defending, the Court can’t just declare victory and move on. Someone has to stand up for the lower court’s ruling so the justices can hear both sides.

So the Court appointed John F. Bash as amicus curiae to defend the Eighth Circuit’s judgment. Bash argued that Ellenberg’s restitution wasn’t actually imposed under the MVRA at all, despite the timing. The Court acknowledged he did a capable job but ultimately rejected that framing because the lower courts had clearly treated the case as an MVRA case.

That cleared the way for the Court to tackle the central question: Is MVRA restitution punishment?

How the Court Decided Restitution Is Punishment

To decide whether a law violates the Ex Post Facto Clause, the Court first asks whether the law imposes a criminal or penal sanction, or instead a civil remedy. This is a question of statutory construction, guided by text, structure, and context.

The Court identified several features of the MVRA that point unmistakably toward punishment:

  1. Text labels it a penalty. The statute explicitly refers to restitution as a “penalty for a criminal offense.”
  2. Tied to a criminal conviction. Only someone convicted of a qualifying federal crime can be ordered to pay MVRA restitution.
  3. Imposed at sentencing. Restitution is ordered at sentencing, alongside prison terms and fines.
  4. Government as opposing party. In restitution proceedings, the government, not the victim, is the party adverse to the defendant—just as in ordinary criminal sentencing.
  5. Placement in Title 18. The entire restitution regime is codified in Title 18: Crimes and Criminal Procedure, specifically in a chapter labeled “Miscellaneous Sentencing Provisions.”
  6. Mandatory sentencing language. A related statute, 18 U.S.C. § 3556, states: “The court shall order restitution in imposing sentence for a covered offense.”

Put together, the Court concluded, the text and structure of the MVRA “make abundantly clear” that restitution under that act is criminal punishment.

Where Smith v. Doe Fits In

Many reform advocates focused on one citation in the opinion: Smith v. Doe, the 2003 Supreme Court case that held Alaska’s sex‑offense registration statute was a civil, non‑punitive regulatory scheme. Because Ellenberg cites Smith, some in the community leapt to the hope that the Court might be ready to revisit or even overturn Smith.

That’s almost certainly an overread of this particular decision.

Here’s what the Court actually does with Smith:

  • It uses Smith as a template for how to analyze whether a law is civil or punitive.
  • Smith upheld a registration law as civil because the legislature adopted “distinctly civil procedures” and expressly declared a non‑punitive, public‑safety purpose.
  • In Ellenberg, the Court contrasts that with the MVRA, where Congress explicitly labeled restitution a penalty, located it in the criminal code, tied it to conviction, and made it part of sentencing.

Then the Court effectively says: Smith does not control here. This is not a soft attack on Smith; it’s a way of saying, “Even if you accept Smith’s framework at face value, MVRA restitution clearly comes out punitive.”

So, Ellenberg is important, but it is not a stealth reversal of Smith v. Doe.

Why Sex‑Offense Registration Is Different (For Now)

The hosts in the transcript repeatedly compare MVRA restitution with PFR (sex‑offense) registration. The contrast is instructive:

  • Labeled a penalty
  • Codified in Title 18 and a sentencing chapter
  • Imposed at sentencing
  • Government is the adverse party
  • Nonpayment can lead to resentencing or further punishment

Sex‑offense registration schemes (under Smith)
– Legislatures declare non‑punitive, regulatory intent
– Processes and procedures are designed to look civil
– Registration duties generally considered collateral consequences, not part of the sentence
– Often codified in public safety or administrative titles, not sentencing chapters

Under the current doctrine, courts first ask: what did the legislature intend? If lawmakers say “this is civil and regulatory,” courts then examine the effects to see if those effects are so punitive that they override the label. That two‑step intent–effects analysis comes from cases like Kennedy v. Mendoza‑Martinez and is sometimes called the Kennedy–Mendoza–Martinez test.

In MVRA, Congress made the Court’s job easy: it intended to punish, and it said so. That ends the inquiry. With registration, legislatures have gone out of their way to say the opposite. That’s why Ellenberg doesn’t, by itself, crack open Smith.

The Limits—and the Power—of the Decision

The Supreme Court is explicit about the scope of its holding:

“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.”

That qualification matters. It means:

  • This ruling applies directly to MVRA restitution.
  • Other restitution schemes could still be drafted and defended as civil.
  • Congress is effectively invited to rewrite the law if it wants to avoid Ex Post Facto issues.

As one commentator in the transcript puts it, that’s an “invitation for you to fix it.” All Congress has to do is change the text and structure to emphasize a civil, compensatory aim and place it outside the core criminal sentencing framework.

Still, for people like Ellenberg and others whose crimes predated the MVRA but who were sentenced (or re‑sentenced) under it, the ruling is very real. It means retroactive application of the MVRA can violate the Ex Post Facto Clause—and those claims now have the Supreme Court’s blessing.

Justice Thomas’s Concurrence: A Bigger Vision of Punishment

If the majority opinion is significant, Justice Clarence Thomas’s separate concurrence is potentially transformative—down the road.

Thomas agrees fully with the Court’s reasoning and result. But he writes separately to talk about something deeper: the foundation of Ex Post Facto doctrine.

He points back to a 1798 case, Calder v. Bull, where the Court said the Ex Post Facto Clauses forbid only retroactive laws that impose punishment for a crime. So far, that sounds uncontroversial. What Thomas emphasizes is how punishment would have been understood in 1798:

“Punishment for a crime would have been understood to refer to any coercive penalty for a public wrong.”

That is a much broader conception of punishment than the modern Court typically applies. Over the past two centuries, Ex Post Facto doctrine has grown more complicated, layering on multifactor tests and heavy deference to legislative labels. Thomas is skeptical of that trend.

Reading his concurrence fairly, he appears to be saying:

  • If a law coerces you or burdens you because you committed a public wrong (a crime), it should count as punishment.
  • Many laws that are nominally civil today would have been understood as punishment under Calder.
  • We should restore Calder’s broader approach to the Ex Post Facto Clauses.

If that view ever becomes the majority, it could have sweeping consequences for collateral consequences, including registration, lifetime bans, and other post‑sentence disabilities.

For now, though, he stands alone. None of his colleagues joined his concurrence in Ellenberg. That makes his opinion a roadmap, not a rule—but for reformers, it’s a roadmap worth studying.

What This Means Going Forward

Ellenberg v. United States does three main things:

  1. Reclassifies MVRA restitution as punishment. This squarely brings MVRA restitution under the Ex Post Facto umbrella. Retroactive application is now constitutionally suspect.

  2. Exposes the reality of post‑sentence financial bondage. The Court’s analysis implicitly acknowledges that decades of accruing restitution debt operate as a serious criminal sanction, not mere housekeeping.

  3. Signals a live debate over what counts as punishment. The majority works within the existing Smith / Kennedy–Mendoza–Martinez framework. Justice Thomas sketches a more historical, expansive view that could reshape that framework in the future.

Practical Implications

For defendants, lawyers, and advocates:

  • Ex Post Facto claims. If the underlying conduct predates the MVRA, Ellenberg strengthens arguments that applying the MVRA’s mandatory restitution scheme violates the Ex Post Facto Clause.
  • Litigation strategies. Lawyers can now cite Supreme Court authority that restitution is punishment when challenging other statutory schemes modeled on the MVRA.
  • Legislative reform. Advocates can use the Court’s language to push legislatures toward more realistic, restorative restitution systems that don’t trap people in unpayable debt.

For policymakers:

  • There is a clear warning: you cannot simply label a sanction “civil” to escape constitutional scrutiny when everything about its structure says otherwise.
  • There is also a clear escape hatch: if you truly want a civil‑like compensation scheme, write and place it like one—and accept that it may be less coercive.

Actionable Takeaways

  1. For defense attorneys: Review older cases where clients were sentenced under the MVRA for pre‑1996 conduct. Ellenberg may support renewed Ex Post Facto challenges or resentencing efforts.

  2. For advocates and organizers: Use this decision in messaging and policy work to highlight how financial obligations function as punishment, and push for caps, interest reform, and realistic payment structures that support reentry.

  3. For legal scholars and reformers: Track and build on Justice Thomas’s concurrence. Develop arguments and scholarship that reconnect Ex Post Facto doctrine with the broader, historical understanding of “any coercive penalty for a public wrong.”

Ellenberg doesn’t fix restitution or collateral consequences overnight. But it does something rare: it calls a punishment a punishment, and in doing so, reshapes the constitutional conversation about how long the state’s hand can lawfully stay on someone’s shoulder after they have served their time.