A Federal Judge Just Named the DOJ's Retaliation Problem. Don't Look Away.

A federal judge in Nashville on Thursday threw out the Justice Department’s human-smuggling case against Kilmar Abrego Garcia and named the reason in plain English. The evidence before the court, U.S. District Judge Waverly D. Crenshaw Jr. wrote, “sadly reflects an abuse of prosecuting power.” Absent Abrego Garcia’s successful Supreme Court fight to be returned from a Salvadoran prison after the government illegally deported him, the judge found, “the Government would not have brought this prosecution.”

That sentence is the news. A sitting federal judge, applying the Fifth Amendment’s Due Process Clause, has formally concluded that Main Justice constructed a criminal case to punish a man for winning in court. The government’s response — an immediate vow to appeal, with a department spokesperson calling the order “wrong and dangerous” — is not an answer to the problem. It is a confirmation that the problem is durable.

The argument we want to make here is narrow and uncomfortable. It does not depend on what one thinks of immigration policy, of the merits of any underlying traffic stop, or of Abrego Garcia personally. It is that vindictive prosecution — the use of the criminal law to retaliate against a citizen, or a noncitizen, for the lawful exercise of a right — is the single most dangerous misuse of federal power, and that when a court catches it in real time, the country has to take the finding seriously rather than fold it into the news cycle’s next argument about whose side wins.

What the judge found, in order

The facts the court worked with are not in serious dispute. Abrego Garcia, a Salvadoran man with a 2019 withholding-of-removal order forbidding his return to El Salvador, was deported there in March 2025 in what the government itself eventually conceded was an “administrative error.” He was held at the Terrorism Confinement Center, the same Salvadoran prison the administration was using to detain other people removed under the wartime authority the Supreme Court has since trimmed back. A district court in Maryland ordered his return; on April 10, 2025, the Supreme Court unanimously directed the government to “facilitate” it.

He came back in June 2025 — and was promptly indicted in the Middle District of Tennessee on two counts of human smuggling, based on a November 2022 traffic stop that no one had charged him with in the three years it sat in a file. Then-Deputy Attorney General Todd Blanche publicly tied the reopened investigation to Abrego Garcia. Crenshaw found that the timing, Blanche’s “now unrebutted public statements,” and the involvement of senior Main Justice officials in driving the case were enough to trigger a legal presumption of vindictiveness — and that the government had not rebutted it.

A presumption of vindictiveness is not a casual phrase. It is a doctrine federal courts use sparingly because it second-guesses the executive branch’s charging discretion, which judges normally treat as nearly untouchable. Crenshaw applied it anyway, and the reason he could is that the record made it impossible not to. “Instead of investigating the November 2022 traffic stop to identify who was responsible for the human smuggling,” he wrote, Blanche “started the investigation to implicate Abrego.” That is the inversion that makes the doctrine apply: a prosecutor who picks the target first and looks for the law second is no longer enforcing the law.

The constitutional injury runs both ways

Vindictive prosecution is the Fifth Amendment problem with the longest shadow because it punishes the very act of going to court. A government that can criminalize the assertion of rights does not really concede those rights; it lends them out and recalls them. The framers, who had watched English Crown prosecutors do exactly that, wrote a due-process guarantee that runs to “any person” — citizen and noncitizen alike, sympathetic defendant and unsympathetic one, the man Americans cheered for in the deportation case and the man they did not.

That is why this ruling cannot be safely categorized as a “win” for one political side. The Justice Department of any administration has the tools to do what Crenshaw described. The protection is the doctrine he applied. A future DOJ that learns the lesson the wrong way — that the trick is to keep its prosecutors quieter and its memos cleaner — has not been chastened; it has been trained. The protection only works if the public, and the legal profession, hold the rule’s edges visible in cases like this one. That includes calling it what it is when the target is unsympathetic to one’s politics, and remembering it next time when the target is sympathetic.

We have made this argument before in different settings. Congress’s power of the purse exists to stop exactly the kind of executive workaround the Anti-Weaponization Fund attempts, and the principle is the same one at work here: structural constraints on executive power are load-bearing whether the executive in office is to one’s liking or not. A sealed judicial order binds the prosecutors who write it as much as anyone else, because a rule that runs only one direction is a tool, not a rule. Vindictive-prosecution doctrine is in the same family. It is a rule the department imposes on itself only when courts insist.

The DOJ’s appeal will test the bar, not the politics

Within hours of the ruling, the Justice Department announced it would appeal. It is free to do so; that is how an adversarial system stress-tests a finding. But the appeal will turn on the same record Crenshaw saw — the indictment timing, the public statements, the chain of command that ran through the deputy attorney general. The appellate question is not whether the prosecution was wise. It is whether the district court was wrong to conclude that the government had not overcome a presumption it had earned through its own conduct.

The honest case for caution here, the one we have to acknowledge, is that an indictment can survive even an ugly motive if the underlying evidence is independently strong. If Abrego Garcia in fact committed the crimes charged, the argument goes, the system should still be able to reach that conduct. There is a version of that argument that is principled. But Crenshaw’s ruling is built on a different point — that the government’s framing of the “new evidence” did not hold up as a matter of law, and that the reopening of the investigation, not just its conclusion, was the act tainted by motive. An appellate court that accepts the framing the department has now adopted in public — that any dismissal for vindictiveness is “dangerous” — will be saying something larger than this case: that the presumption itself is too costly to apply. That would be the actual constitutional change to watch for.

The standard we want

Strip away the names and the politics and the country is left with a question about what it wants its Justice Department to be. The answer is not that the department should never prosecute anyone who has won a lawsuit against it; nothing in this ruling holds that. The answer is that when the department reopens an old file shortly after a citizen — or a noncitizen with constitutional protection — prevails against it in court, the department bears the burden of showing that the timing is coincidence and that its leadership did not direct the case as payback. That burden is not punitive. It is the bare minimum the rule against vindictive prosecution requires, and it is the bare minimum any future administration of either party should be held to.

The same skepticism that we asked readers to apply to a DOJ indictment of a former FBI director over a beach photo should attach here, in the same direction: examine the record, not the press release. When a Republican judge appointed by President Obama, as Crenshaw was, writes that the evidence reflects an abuse of prosecuting power, the burden is on those who disagree to engage with the record he describes. It is not enough to call the order “wrong and dangerous.” The order is a finding. The country deserves an answer to it on the merits, in court — and in the meantime, the rule the order vindicated needs to be defended by people who are not certain they will always like its results.