This week, we are thrilled to welcome Lula Hagos to the Sentencing Matters Substack. Lula is an Associate Professor of Law at the George Washington University Law School and the founding Director of the law school’s Criminal Defense and Justice Clinic. Students in the clinic represent indigent clients charged with misdemeanors in D.C. Superior Court. For over a decade, Lula served as a public defender at the Office of the Federal Public Defender for the Eastern District of Virginia and the Public Defender Service for the District of Columbia. Lula’s scholarship explores issues around the intersection of financial punishment and race in criminal courts. She is a graduate of the University of Virginia and the Georgetown University Law Center.
-Jonathan
- - -
This term, the Supreme Court will hear the case of Ellingburg v. United States, which asks whether criminal restitution under the Mandatory Victims Restitution Act (MVRA) constitutes punishment for purposes of the Ex Post Facto Clause. The Clause prohibits retroactive increases of criminal punishments, ensuring that the government cannot change the rules after the fact to impose harsher penalties.
Criminal restitution — the money paid by a defendant to a victim — has grown into one of the most troubling, yet least examined, features of modern criminal sentencing. Unlike criminal fines and fees, it is often portrayed as fair and victim-centered. But restitution has quietly grown — both in scope and severity — into a sanction that can extend punishment for years, frequently without compensating victims. Enacted in 1996, the MVRA was instrumental in changing the landscape of criminal restitution and continues to regulate and mandate federal restitution for most offenses today. The Court’s decision in Ellingburg could finally force a reckoning with criminal restitution’s true character.
In 1995, Holsey Ellingburg, Jr. and an accomplice robbed a bank in Georgia. Mr. Ellingburg was sentenced to 27 years in prison and five years of supervised release and ordered to pay two “criminal penalties”: a $100 special assessment and $7,567 in restitution. While in prison, Mr. Ellingburg paid $2,154 toward his restitution obligation.
At the time of his offense, restitution was governed by the Victim and Witness Protection Act (VWPA), which capped enforcement at 20 years from the entry of judgment. By that rule, Mr. Ellingburg’s obligation would have expired in 2016. But then Congress enacted the MVRA, which extended enforcement to 20 years after release from imprisonment. Applying retroactively, this nearly doubled Mr. Ellingburg’s obligation and extended his restitution liability until 2042 — almost fifty years after his crime. Following his release from prison in 2022, Ellingburg started to receive text messages from his probation officer demanding continued monthly restitution payments. The Eighth Circuit upheld this retroactive application, reasoning that restitution is not punishment but a civil remedy. Yet when the case reached the Supreme Court, the Solicitor General chose not to defend that rationale, leaving the Court to decide whether restitution is indeed punishment subject to ex post facto limits.
Although criminal restitution has been around for centuries, its meaning has shifted in recent years. It is often conflated with civil restitution, which prevents unjust enrichment by requiring the accused to disgorge any ill-gotten gains. For example, if someone steals your phone, you would be entitled to the cost of replacing the phone, returning the profit that was wrongfully taken. Although criminal restitution was also originally rooted in the idea of unjust enrichment, it now bears little resemblance to its civil counterpart. By contrast, criminal restitution focuses on a victim’s purported “losses,” even when those losses far exceed a defendant’s unlawful gains from the criminal offense. Under the MVRA, the concept of loss has taken on an increasingly vague and amorphous interpretation. Courts routinely order criminal restitution for intangible and speculative harms, for losses suffered by insurers or third parties not directly harmed (likely resulting in double compensation), and even for conduct not charged, for which the defendant is acquitted, or where there is no victim at all.
By every measure — statutory design, legislative purpose, legal precedent and practical effect — restitution under the MVRA is punishment. It is imposed as part of a criminal sentence, directly tied to the underlying offense and often mandatory. The statute expressly bars courts from considering a defendant’s ability to pay when determining an appropriate amount of restitution. The sanction is entirely controlled by the state — victims have no agency over the restitution process and cannot waive or privately settle these obligations. When Congress enacted the MVRA, it acknowledged that defendants were indigent and that mandatory restitution would not meaningfully compensate victims. Yet it justified the law on penological grounds: accountability, deterrence, and ensuring defendants “pay their debt to society,” not merely to victims. Along with most circuit courts, the Supreme Court itself has repeatedly recognized restitution’s punitive aims. In Pasquantino v. United States, the Court explained that restitution under the MVRA serves “to mete out appropriate criminal punishment.” Similarly, in Paroline v. United States, the Court acknowledged the statute’s punitive purposes, explaining that it ensures that defendants are “held to account” for their conduct. And even before the MVRA, the Court recognized that federal criminal restitution is a “penal sanction” that furthers the state’s rehabilitative and deterrent goals.
Notably, the effects of unpaid restitution carry all the hallmarks of punishment. As in Mr. Ellingburg’s case, the punitive consequences that flow from a failure to pay restitution can be severe and extend one’s punishment indefinitely. Restitution debt is an enormous barrier to the reintegration process for defendants, compounding the already difficult hurdles of having a criminal conviction. When a person cannot pay their restitution, they face the threat of reincarceration, extended supervision, constant financial monitoring, and the loss of certain civil liberties. Because restitution under the MVRA is imposed without regard to a defendant’s financial circumstances, restitution disproportionally punishes the poor — the vast majority of those in the criminal legal system. The MVRA also mandates interest accrue on unpaid amounts, ensuring that debt grows faster than people can repay and that defendants remain tethered to the criminal legal system long after serving their time. The government also devotes extensive resources to go after people with restitution debt, garnishing wages, seizing assets, imposing tax liens — sometimes for decades after a person’s release.
Even apart from its punitive nature, restitution under the MVRA has shown to be inefficient and harmful to victims, the very people it is purported to help. Because the law ignores ability to pay, the vast majority of restitution obligations remain unpaid. The Government Accountability Office has reported that of the $110 billion in outstanding federal restitution debt, about $100 billion is uncollectible, overwhelmingly because defendants have a limited capacity to pay. Victims are therefore rarely made whole, while enormous governmental resources are wasted in futile enforcement.
The result is a system that neither adequately compensates victims nor imparts accountability and rehabilitation for defendants. Mr. Ellingburg’s case plainly illustrates the problem: nearly thirty years after his offense, and despite having served his sentence, he may remain indebted to the government until 2042. The Ex Post Facto Clause exists to prevent precisely this sort of retroactive punishment.
The Court will not be able to solve all criminal restitution’s woes in Ellingburg. But it should take the crucial step of acknowledging that criminal restitution is punishment subject to the Ex Post Facto Clause. Acknowledging restitution’s punitive nature would bring coherence to constitutional doctrine and prevent the government from imposing punishment without its safeguards. The Court’s decision in Ellingburg could finally recognize criminal restitution for what it is: punishment.
i have no issue with SCOTUs saying restitution attached to a sentence is punishment. But that doesn’t mean extending the collection period works an EPF violation. Thats the real issue.
Thank you for this clear discussion of the case and the punitive nature of criminal restitution.