The numbers the government is putting out don’t match the numbers the government is actually generating. That gap — stark, documented, and growing — is now at the center of a serious dispute over how the Trump administration is characterizing its sweeping immigration enforcement campaign.
The Department of Homeland Security has repeatedly claimed that nearly 70 percent of immigrants arrested by ICE have criminal records. It’s a talking point that has anchored the administration’s public defense of its enforcement surge. But multiple independent analyses of ICE’s own data tell a sharply different story — one in which the majority of people being detained and arrested have no criminal conviction at all.
What the Data Actually Shows
The Transactional Records Access Clearinghouse at Syracuse University, which tracks federal enforcement data, found that 50,259 out of 68,289 people held in ICE detention — roughly 73.6 percent — had no criminal conviction on their record. That’s not a rounding error. That’s a near-inversion of the administration’s core claim.
It gets more striking when you zoom in on recent arrests. CBS News reported that 45 percent of ICE arrests in October 2025 involved people with no criminal convictions or pending charges. In just the first two weeks of that month, that figure climbed to 71 percent. The Brennan Center, analyzing the broader enforcement picture, found that fewer than 14 percent of nearly 400,000 immigrants arrested during Trump’s return to the White House had charges or convictions for violent criminal offenses.
So where does the administration’s 70 percent figure come from? That’s the catch. The claim appears to rely on a much broader definition of “criminal record” — one that can include minor traffic violations, civil immigration infractions, or charges that were never prosecuted. Critics argue that framing is, at best, misleading.
A Trend Running in One Direction
Still, it’s worth understanding the trajectory here, not just the snapshot. Researchers at the University of Colorado Boulder documented that arrests of immigrants without criminal records have surged by 585 percent year over year. That is not a marginal uptick. That is a categorical shift in who ICE is targeting.
The Cato Institute, the libertarian think tank that’s historically been critical of both parties on immigration, noted that the percentage of ICE arrests involving people with a criminal record reached a near-historic low in 2025. The enforcement apparatus, in other words, has expanded well beyond the “criminal alien” framing the administration keeps returning to in press releases and congressional testimony.
Does that mean there are no criminals being arrested? Of course not. ICE does arrest people with serious criminal histories — murders, sexual offenses, drug trafficking. The administration highlights those cases prominently, and they’re real. But they represent a fraction of the total arrest volume, and using them to characterize the whole operation is, researchers say, statistically indefensible.
Why the Framing Matters
This isn’t just an argument about statistics. The “criminal alien” framing has real policy consequences — it shapes public support, influences court proceedings, and affects how Congress appropriates enforcement dollars. When the stated justification for a mass enforcement campaign doesn’t hold up against the agency’s own records, that’s a problem that cuts across partisan lines.
Journalists, immigration attorneys, and civil liberties groups have been raising these discrepancies for months. The administration has not publicly addressed the TRAC data or the CU Boulder analysis directly. DHS press releases continue to cite the 70 percent figure without methodological footnotes.
What the data shows, when you pull it all together, is an enforcement operation that has grown faster than its own stated rationale. The government says it’s going after criminals. The records say it’s going after a much broader population — and calling them criminals anyway.
What Comes Next
Legal challenges are mounting. Several federal courts have already scrutinized ICE’s arrest practices, and the question of who, exactly, qualifies as a public safety threat under the administration’s definitions is likely headed for prolonged litigation. Meanwhile, TRAC has flagged that the administration has at times restricted access to enforcement data — making independent verification harder, not easier.
That’s a pattern worth watching. Governments that are confident in their numbers tend to welcome scrutiny. Ones that aren’t tend to limit the data.
The administration may yet produce a fuller accounting of its methodology. Until then, the most honest summary of where things stand comes not from any advocacy group, but from ICE’s own arrest records — and those records suggest the story being told in Washington doesn’t quite match the one playing out at the door.
