What proportion of ICE interior removals are for violent crimes versus minor offenses over the last five years?
Executive summary
Available public reporting shows that a large majority of ICE’s interior removals in recent years involved people with some kind of criminal history, but the datasets and public summaries do not allow a precise, defensible split between “violent crimes” and “minor offenses” across the last five years; the best-cited figures indicate roughly 79 percent of interior deportations in FY2021–24 involved people with a criminal record (a broad category) while independent analysts report that only a small single-digit share of detainees have violent convictions [1] [2]. GAO and other researchers warn that ICE’s public reporting lacks the detail and consistency needed to answer the question definitively [3].
1. What the agencies report — a broad-brush criminality figure, not a crime-type breakdown
ICE’s public materials and summaries classify many interior removals by broad categories — including people with convictions or pending charges, those without convictions but removable for immigration violations, and fugitives — but do not publish a consistent, public table that cleanly breaks out removals by specific offense types across every recent fiscal year, which means headline percentages about “criminal removals” conflate a wide range of conduct [4] [5]. Migration Policy Institute’s analysis of ICE interior deportations for FY2021–24 finds that 79 percent of interior deportations were of people with a criminal history, but that 79 percent is an aggregated “has a criminal record” measure and is not a direct violent-versus-minor split [1].
2. What watchdogs and analysts add — violent convictions are a small slice, but definitions vary
Independent analysts and think tanks find that people with violent felony convictions represent a relatively small fraction of those in ICE custody or removed, though the share varies by dataset and by whether the measure counts convictions, pending charges, or any contact with the criminal justice system; one data-focused analysis cited in public reporting has summarized that roughly 5 percent of people detained by ICE had violent convictions while a large share had no convictions [2]. That single-digit violent-conviction figure should be treated cautiously: it reflects detention populations at points in time and different reporting windows, not a clean, annually comparable removals-by-offense decomposition for five consecutive fiscal years [2].
3. Trends matter — removals fell and then partially rebounded, changing the composition
GAO’s review of ICE statistics documented a sharp fall in overall arrests and removals from 2019 through 2021 with an uptick in 2022, and it highlighted that public ICE reports understate some detention counts because of how temporary facility bookings are excluded from summaries [3]. Those policy and practice shifts — including enforcement priorities and use of alternatives to detention — changed who ICE targeted and thereby altered the mix of criminal versus noncriminal removals, complicating any straightforward five-year “violent vs minor” percentage calculation [3] [6].
4. Why precise violent-vs-minor percentages can’t be produced from public sources
Several credible sources flag gaps: ICE’s dashboards and releases omit consistent, accessible offense-level removals tables across the entire period and sometimes exclude people transferred through temporary facilities [4] [5] [3]. Advocacy and research projects have reconstructed and reprocessed ICE files to approximate offense mixes, but those reconstructions often exclude border expulsions, duplicate records, or are limited to specific programs like Secure Communities, so they cannot deliver a single authoritative five-year violent-versus-minor share [5] [7].
5. Bottom line and honest caveats
The best defensible statement based on available public reporting is: most interior removals in the last several years involved people with some kind of criminal history (MPI’s 79% for FY2021–24 is the clearest aggregate figure), but independent analyses and ICE’s own disclosures show that only a small fraction of ICE detainees are convicted of violent offenses — and no public ICE dataset currently provides a clean, year-by-year percentage split of interior removals into “violent crimes” versus “minor offenses,” a shortcoming GAO explicitly calls out [1] [2] [3]. Any attempt to offer a single precise proportion would therefore overclaim the evidence; instead, transparency gaps in ICE reporting are the limiting factor for answering the question definitively [3] [5].