Demographics of 2025 ice arrests
Executive summary
ICE arrests and detention in 2025 surged compared with the prior administration: ICE ERO was averaging roughly 1,100 arrests per day in recent weeks according to government-derived reporting [1], and ICE’s detainee population reached record highs near 66,000 with roughly 70–71% of detainees reported as having no criminal conviction in multiple accounts [2] [3]. Different outlets analyzing ICE data show that a majority or large plurality of people arrested or booked in 2025 had no criminal convictions, while violent-conviction shares remained small [4] [5].
1. Mass arrests, not meeting the quota but far above past rates
ICE arrest counts climbed dramatically in 2025; Axios reports Enforcement and Removal Operations (ERO) was arresting about 1,100 people per day in recent weeks — far short of an oft-cited 3,000-per-day administration goal but substantially higher than the Biden-era pace [1]. Other outlets and internal figures cited by CBS place ICE and partner agencies’ combined apprehensions at hundreds of thousands since January 2025, with ICE alone carrying out roughly 278,000 interior arrests in some tallies [2] [6]. These numbers show a deliberate and sustained enforcement intensification even if the most ambitious targets were not met [1] [2].
2. Most detainees arrested by ICE had no criminal conviction
Multiple independent analyses of ICE and DHS data find that a large share of people in ICE custody in 2025 had no criminal convictions. Migration Policy and other outlets report roughly 71% of ICE detainees lacked criminal convictions as of mid‑to‑late 2025 [3]. CBS News and Cato analyses similarly highlight that the fastest-growing cohort of ICE arrests were unauthorized immigrants without criminal records and that roughly two-thirds to three-quarters of detainees lacked convictions [2] [4]. These cross‑source consistencies indicate a policy shift toward broader interior enforcement beyond people with violent or serious criminal histories [3] [4].
3. Violent convictions remained a small share of arrests
Reporting across outlets finds that very few ICE arrests involved people convicted of violent crimes. The Cato Institute noted that over 93% of ICE book‑ins through mid‑June 2025 had never been convicted of a violent offense, and the New York Times found that less than 30% of people in high‑profile local operations had criminal convictions at all, with violent convictions a much smaller slice [4] [5]. Stateline’s state-level breakdowns show variation — some states’ arrest cohorts had higher shares of violent convictions (Hawaii 15%, Vermont 13%, California and Nebraska 12%) — but those were exceptions rather than the national norm [7].
4. Geographic and operational patterns: interior arrests and targeted operations
Reporting emphasizes a shift from border‑centric apprehensions to interior arrests and courthouse or workplace operations. ICE’s ERO has increasingly focused on arrests inside the U.S., and large federal operations in cities (Los Angeles, Chicago, D.C., Massachusetts) produced spikes where more than half of arrestees in some operations had no criminal record, even as nationwide figures were nearer one‑third without records in some sets [8] [5]. Cato and other analysts flag that many arrests are “non‑specific” street arrests and workplace or check‑in apprehensions, raising concerns about profiling and enforcement tactics [9] [4].
5. Discrepancies, potential biases, and data limits
Sources use different datasets and definitions. The Guardian notes ICE’s biweekly detention reports and cautions arrests may be undercounted because ICE reports only arrests that result in entering ICE detention; other databases count book‑ins or FOIA‑obtained arrest logs, and duplicates or multiple arrests of the same person can inflate totals [10] [1]. Some summaries (think tanks, advocacy sites) emphasize the share with no convictions to critique policy [4] [11], while agency releases characterize many arrestees as immigration violators with various offense histories [8]. These methodological differences mean headline percentages (65–73% with no convictions; 71% no violent convictions) are broadly consistent across reporting but not interchangeable without noting how “book‑ins,” “detainees,” and “arrests” were counted [4] [3].
6. What the numbers imply politically and legally
Journalistic and policy sources draw contrasting inferences. Critics argue the surge reflects an intentional expansion to detain and remove non‑criminal immigrants and warn of profiling, family separations, and strained court backlogs [9] [3]. Administration advocates point to operational goals and public‑safety rationales, and ICE historically asserts ERO targets those who “may present threats” while also conducting administrative arrests [8] [1]. Both interpretations rely on the same government data but emphasize different subsets: volume and non‑criminal shares for critics, arrest rates and operational reach for proponents [8] [1].
Limitations and next steps: available sources do not mention post‑Oct 15, 2025 nationwide arrest totals reconciled across ICE, CBP and HSI into a single dataset; researchers should consult ICE’s biweekly detention releases, FOIA datasets used by the Deportation Data Project, and the New York Times/UC Berkeley analyses for the most granular counts before drawing final conclusions [10] [1] [5].