When Innocence Is No Longer a Reason

The Supreme Court narrows compassionate release under the First Step Act and decides that neither a sentence the country has long considered too harsh nor doubts about guilt are extraordinary and compelling enough!

What qualifies as an extraordinary and compelling reason to release someone from prison early? A terminal illness, very old age, or a family emergency? For all of these, the answer in the federal system is “maybe,” more precisely, “probably not.” But what about a sentence that Congress itself now considers too harsh, or a case in which a judge later doubts whether a person was even convicted lawfully? Since the end of 2018, the answer here had also been “maybe,” at least in some federal courts. That changed last week. The majority of the Supreme Court answered both questions with no and in doing so narrowed compassionate release in the federal system. In two separate decisions, the justices held that this form of release may serve neither to revisit doubts about a conviction nor to reduce a sentence that would now be considered excessive by today’s standards.

Compassionate release has been part of federal law since 1984, when Congress gave courts the authority to reduce prison sentences for “extraordinary and compelling” reasons. At the time, however, only the Bureau of Prisons could file such a request, and it almost never did. A 2013 report by the Department of Justice found an average of only twenty four releases per year, about one hundredth of one percent of federal prisoners. Partly to change that, Congress passed the bipartisan First Step Act in 2018, which Donald Trump signed during his first term. It allowed incarcerated people to submit their request for release directly to a judge. That did not lead to mass releases, but it did increase successful petitions roughly twentyfold, to 481 out of more than 2,500 requests in fiscal year 2024. There is a particular irony in the fact that the law Trump signed is now being narrowed by a court he helped shape.

In about one fifth of approved cases, judges cited “an unusually long sentence and a change in the law” as part of their reasoning. That path is precisely what the decision in Rutherford v. United States closes off. Daniel Rutherford committed two armed robberies in 2003, and a jury convicted him of multiple offenses, including two counts of using a firearm during a crime of violence. At the time, federal law required certain sentences for such charges to be stacked, meaning served consecutively. That resulted in a sentence of forty two and a half years, driven primarily by thirty two mandatory years for the firearm counts. The First Step Act ended this stacking practice for future cases. If Rutherford were sentenced today for the same conduct, he would face at least fourteen years. He sought release in part for that reason. Justice Amy Coney Barrett wrote for the majority that Congress’s decision not to apply new sentencing laws retroactively is common and therefore not extraordinary and does not qualify for compassionate release. Here lies the reversal that carries the entire ruling. Because the injustice is common, it is said not to be extraordinary. The fact that a man serves twenty eight years longer than the country would impose today does not become acceptable simply because many others suffer the same fate. The Court confuses the frequency of a wrong with its legitimacy.

What is remarkable is that the majority reached this conclusion even though Congress explicitly designated who should define the phrase “extraordinary and compelling,” and it was not the justices. The 1984 law created the Sentencing Commission and instructed it to determine what qualifies as a valid petition. In 2023, that commission decided that an unusually long sentence could under certain circumstances qualify as an extraordinary and compelling reason, although the decision itself was controversial within the commission. For the Court’s three liberal justices, that should have ended the matter. In her dissent, Sonia Sotomayor wrote that Congress had tasked the commission, “not this Court.” The rule did not automatically release anyone, she emphasized, because eighty percent of petitions continued to be denied. It merely allowed judges to consider sentencing disparities case by case. The Court therefore took for itself a power Congress had assigned elsewhere, and it did so in the name of restraint.

The second case, Fernandez v. United States, raised the same question in a harsher form. If an outdated sentence is not compelling, what about possible innocence? Joe Fernandez was convicted in 2014 of murder for hire conspiracy and sentenced to two life terms. He has always maintained his innocence. No court has declared him innocent to this day, but in 2022 the district judge who had presided over his original trial granted release to time served, largely because of doubts about the credibility of the government’s main witness. Fernandez had also received a far longer sentence than the alleged co conspirators who accepted plea deals.

The majority of the Supreme Court held that doubts about the validity of a conviction are also not a permissible reason. The very concept of compassionate release, Barrett wrote, shows that the law “aims to grant mercy rather than correct legal errors.” In her lone dissent, Ketanji Brown Jackson answered plainly: “An innocent man serving life in prison. Can there be a more extraordinary and compelling reason to reduce a sentence than that?” One has to place that sentence beside Barrett’s to understand the full coldness of the decision. The Court tells a possibly innocent person that innocence is the wrong kind of reason and calls that fidelity to the text of mercy.

Part of the reasoning is that the proper avenue for someone who believes a conviction is invalid is a petition for habeas corpus or federal post conviction review, not compassionate release. Habeas corpus is the old legal mechanism through which a detained person asks a court to examine whether detention is lawful, and in recent months it has received renewed attention in American law especially in relation to immigration detention. That makes the next First Step Act case particularly notable. On Monday, the justices agreed to hear Maxwell v. Thomas in the upcoming term, which will determine whether a federal prisoner may use habeas corpus to challenge how the Bureau of Prisons handles the expanded good conduct credits under the First Step Act. Those credits could move someone earlier into home confinement or transitional housing. It is as if, after one door was closed, only this one remains open.

For the sake of completeness, it should be said that legal analyst Doug Berman does not believe the Court overreached. The rulings concern the types of petitions federal prisoners may bring in the future, but he argued that one should not fail to assert that many other reasons in different contexts could still be extraordinary and compelling. The door is therefore not completely shut, only the two most serious reasons have been excluded.

In the end, a difficult question remains about the nature of mercy. The Court has drawn a line beyond which mercy may no longer touch injustice. But what is mercy worth if it must look away when a person may be innocent and if it is forbidden to act against a sentence the country itself has rejected? Mercy that may be felt but not translated into action is no mercy at all. And a court that calls the suffering of those imprisoned too harshly or unjustly ordinary has thereby shown how much that suffering means to it. Jackson named the standard, and the Court failed it.