What proportion of ICE detainees nationwide have violent felony convictions versus administrative immigration violations according to recent ICE datasets?

1. Data snapshot: what the recent numbers say

A Cato Institute review of ICE’s more recent interior-arrest datasets found that roughly 5 percent of those taken into ICE custody had violent criminal convictions, while about 73 percent had no criminal conviction in U.S. records over the period it examined ^{[1] [3]}, and independent reporting and analyses placed the share of detainees with violent convictions in a similar small range—The New York Times’ earlier analysis found about 7 percent for a January–October window and Stateline reported the share of violent convictions dropping from 9 percent in January to under 5 percent by October ^{[5] [4]}.

2. Administrative arrests versus criminal arrests: definitions matter

ICE’s Enforcement and Removal Operations conducts both administrative arrests for civil immigration violations and criminal arrests tied to criminal warrants or prosecutions, and agency public reports and guidance make clear that “administrative” arrests are for civil immigration violations and are the norm for interior enforcement ^{[6] [7]}; ICE’s FY2024 figures showed 113,430 administrative arrests and reported that about 49 percent of those arrested that fiscal year had no prior criminal convictions ^[2].

3. Why different sources give different percentages

The apparent contradictions across reports stem from differences in time periods, which population is measured (single-day detentions, arrests, detentions pending deportation, or deportees), and the datasets used—Cato and the Deportation Data Project analyzed leaked or compiled ICE arrest records for interior operations and found large shares without convictions, while ICE’s public historic summaries sometimes emphasize the share with convictions depending on the fiscal year and the definitional window ^{[3] [8] [9]}.

4. Competing narratives and institutional agendas

Administration officials have framed enforcement as prioritizing the “worst of the worst,” a claim echoed by DHS releases and case examples, but independent analyses show the agency’s expanded arrest activity pulled in many people without violent criminal records—reporting and advocacy groups note that the uptick in arrests coincided with policy pressure to increase interior arrests, which shifted the mix toward administrative violators and people without U.S. convictions ^{[10] [8] [11]}.

5. Key caveats: what the datasets do not resolve

DHS and ICE public data often indicate whether a detainee had a conviction or pending charge but do not always fully disaggregate offense types or capture convictions from home countries; independent reviewers therefore rely on combining ICE datasets, local jail records, and other compilations, which means precise national percentages vary by methodology and reporting window and that the datasets do not answer every question about prior foreign convictions or pending local charges ^{[5] [11]}.

6. Bottom line answer

Based on recent ICE datasets and independent analyses cited above, the share of ICE detainees with convictions for violent felonies is small—on the order of about 5 percent (with independent estimates spanning roughly 4–7 percent depending on the window and method)—while the share held for administrative immigration violations or with no U.S. criminal conviction constitutes a much larger portion of the detained population, reported between roughly 49 percent (ICE FY2024 administrative-arrest data) and 73 percent (the Cato analysis of later interior-arrest data) depending on the dataset and timeframe ^{[1] [2] [3] [4]}.