Last week, the Cato Institute published new, nonpublic data showing that Immigration and Customs Enforcement (ICE) is mostly detaining non-criminals and that most arrested immigrants with criminal convictions are not violent offenders. Thanks to a Freedom of Information Act request by DeportationData.org, a group that gathers FOIA data on ICE arrests, another dataset has become available, confirming ICE’s unprecedented crackdown on peaceful people.
The two main takeaways from these new data are
ICE is arresting enormous numbers of immigrants who have no criminal history.
ICE is taking to the streets to make arrests like never before.
This means that Americans are much more likely to see ICE violently arrest nonviolent people in their communities. Combined with ICE’s decision to hide their identities and faces, this action increases the likelihood of unnecessary confrontations between ICE and the public. It also means that ICE is more likely to make mistakes, and US citizens and legal residents are ensnared in the crackdown.
The graph below compares ICE arrests of people without criminal convictions during fiscal years 2017 (from Obama to Trump) and 2025 (from Biden to Trump). The difference in approach is staggering. By June 2025, ICE was arresting over 6,000 immigrants without criminal convictions in a week. That is a significant increase from just over 1,000 during the same time in 2017—a roughly 500 percent increase in arrests.
The graph below shows ICE’s non-custodial arrests, or “street arrests,” though some occur at worksites, courthouses, or other locations. A custodial arrest occurs when a person is already in the custody of law enforcement. By early June 2025, ICE was making over 2,500 arrests per week. That is a significant increase from just 370 per week in early June 2017—a 606 percent increase. This dramatic shift in tactics explains the chaotic images of ICE arrests on the streets.
These non-custodial arrests target people without any criminal conviction. By early June 2025, ICE was arresting more than 2,000 noncriminals weekly on the streets. That is a sharp increase from 151 in 2017—a nearly 1,300 percent increase. With agents in masks arresting people in public, ICE is detaining peaceful individuals, which is leading to chaotic scenes on the streets and at worksites.
Since January 20, ICE’s street enforcement efforts mainly target noncriminals. By early June, 81 percent of ICE’s weekly non-custodial arrests involved people with no criminal convictions—up 17 percentage points since January 20.
The data Cato published last week focused on people that ICE booked into detention. This new dataset concentrates on ICE arrests. Someone can be arrested by Customs and Border Protection and still be booked into ICE detention, and someone can be arrested by ICE but never be booked into detention. The table below compares these two measures. In early June, 69 percent of ICE book-ins and 71 percent of ICE arrests had no criminal convictions.
The share of ICE arrests without criminal convictions from January 20, 2025, to June 9, 2025, was 59 percent, indicating a significant escalation in the non-criminal share of ICE arrests in recent weeks.
ICE also arrests and detains individuals with pending criminal charges. From January 20 to June 9, 32 percent had pending charges, 41 percent had convictions, and 27 percent of ICE arrests had no criminal convictions or pending charges. However, between June 1 and June 9, ICE records show that 47 percent had no charges or convictions. ICE often calls those with pending charges “criminals.” But by arresting them, it is denying them their day in court. Moreover, ICE’s documentation does not detail what crimes these people are accused of, and we know many charges are eventually dismissed.
Moreover, if we look at arrests of people without convictions or pending criminal charges, the increase is nearly as extreme as the total arrests of people without convictions. In early June, the US government was arresting, on average, about 4,000 people per week who had no criminal charges or convictions. That compares to about 250 per week under the Biden administration and around 1,000 per week under the first Trump administration. The scale of the current arrests of people with no contact with law enforcement has no precedent.
Not surprisingly, the number of immigrants entering ICE detention based on an ICE arrest who have had no contact with law enforcement is also rising. The number of people in detention with no criminal convictions or pending criminal charges has increased by 1,271 percent, or nearly 14-fold, since January 11. At no point in US history has ICE ever detained this many people with no criminal involvement of any kind.
I have also received additional data from ICE indicating that less than 6 percent of ICE book-ins involved individuals with criminal convictions for violent offenses in early June. More than 90 percent of those booked by ICE were neither violent nor property crime offenders. The most serious conviction for those arrested was usually either an immigration offense, a traffic infraction, or a nonviolent vice crime.
In other words, the government is not targeting its arrests towards public safety threats and criminals. The broad classifications of crimes are in the appendix table.
The upshot of the new data is that ICE is not being transparent about who it is targeting. For example, during the Los Angeles protests, the department kept claiming it was arresting criminals and that protesters and California politicians wanted to protect serious criminals. For instance, here is Assistant Secretary of Homeland Security Tricia McLaughlin claiming that California Governor Gavin Newsom and Los Angeles Mayor Karen Bass were upset with ICE for arresting a convicted rapist.
Yet the new data reveal how misleading McLaughlin’s claim was. During the week ending June 9, 72 percent of those arrested in ICE’s Los Angeles Area of Responsibility had no criminal convictions of any kind, let alone for violent rapes, and 59 percent had criminal convictions or pending charges, which is higher than the 41 percent nationwide. We don’t know the breakdown of violent crimes for Los Angeles, but if it follows the nationwide data, then only about 5 percent, or one in twenty, ICE arrests in Los Angeles in early June involved individuals with violent convictions like the case the assistant secretary highlighted.
The shift in tactics and arrest numbers in late May and early June stems from White House Deputy Chief of Staff Stephen Miller’s May orders to deprioritize criminals and focus on noncriminals. ICE officials told the conservative Washington Examiner that Miller said, “What do you mean you’re going after criminals? Why aren’t you at Home Depot? Why aren’t you at 7‑Eleven?”
They told the Wall Street Journal that he effectively ordered them to start profiling. Miller said agents could “go out on the streets of Washington, DC, and arrest 30 people right away. Who here thinks they can do it?” We have now seen ICE and Border Patrol repeatedly arresting people on the streets, tackling them for simply refusing to answer questions or for running away. Several US citizens and legal residents have already been tackled, detained, and arrested in recent weeks.
ICE’s deportation campaign is not focused on public safety. ICE agents told the conservative New York Post, “All that matters is numbers, pure numbers. Quantity over quality.” ICE sources told the Post that it is “leading them to leave some dangerous criminal illegal migrants on the streets and instead look for anyone they can get their hands on at the local Home Depot or bus stop.”
Most of the people ICE has arrested are individuals who would lead peaceful, productive lives in the United States if they were allowed to. The Trump administration should stop arresting people who entered legally, stripping them of their legal status, canceling asylum hearings, and prioritizing noncriminals for arrest and detention. Congress should permit them to obtain permanent lawful status so that they can contribute to this country without fear of deportation.
