5% of People Detained By ICE Have Violent Convictions, 73% No Convictions
Executive summary
The headline that “5% of people detained by ICE have violent convictions, 73% have no convictions” is grounded in multiple data analyses of ICE’s recent detention snapshots: analyses by Cato and reporting summarized by MinnPost and TRAC find about 5% of detainees convicted of violent crimes and roughly 73% of the detained population lacking any U.S. criminal conviction as of late 2025 [1] [2] [3]. Those headline percentages are accurate to the published datasets cited, but they require important contextual qualifiers about definitions, timing, geographic variation and the distinction between arrests, charges, and convictions [4] [5].
1. How the numbers are being calculated and why they match
The 5% violent-conviction figure comes from Cato’s analysis of ICE’s public detention data that breaks detainees into categories by criminal-history status; Cato reported that about 5% (or similarly low single-digit shares) of people in ICE custody were convicted of violent offenses in the mid-to-late 2025 period [1] [6]. The roughly 73% no-conviction figure is reported in TRAC’s aggregation of ICE detention counts—TRAC found about 48,377 of 65,735 detainees (73.6%) had no criminal conviction as of Nov. 30, 2025—and Cato’s summaries show consistent large shares without U.S. convictions [3] [1].
2. Important definitional and timing caveats
ICE’s public categories distinguish people with convictions, people with pending charges, and those held solely for immigration (civil) violations, and the agency’s arrest dataset is not identical to the snapshot of people in custody—so “arrests” and “people in detention” can yield different percentages [4] [1]. Multiple outlets note rapid shifts in the composition of custody during 2025: the share of detainees with violent convictions fell from earlier months to under 5% by October in some datasets, while the number held with no conviction rose sharply in short windows, so the precise percentage is sensitive to the date range and which ICE dataset is queried [5] [1].
3. What “no conviction” actually means in practice
People classified as having “no conviction” may include those with pending charges that have not produced convictions, those charged only with immigration violations (e.g., illegal entry or reentry), and transfers from CBP at the border—counts do not necessarily capture foreign convictions or sealed records, and some pending cases later result in convictions while many are dismissed or resolved without a conviction [4] [2]. Reporting and analyses emphasize that many detainees who do have convictions are for lower-level offenses such as traffic violations, DUI, or misdemeanor drug possession rather than serious violent felonies [4] [3].
4. Regional and operational variation matters
ICE field offices differ: local enforcement spikes, worksite raids, or policy directives can rapidly change the mix of detainees in a particular AOR, as Dallas reporting and TRAC regional data document, meaning national averages can obscure pockets where criminal-conviction shares are higher or lower [7] [3]. Analysts have also pointed to enforcement prioritization changes that drove a disproportionate increase in people without convictions entering detention during 2025, so the national snapshot reflects both policy choices and operational tempo [8] [9].
5. Disputes, alternative tallies, and limits of available reporting
Some compilations produce different shares—public summaries like “about 47% had criminal charges or convictions” (a dataset cited by TheWorldData) conflict with TRAC/Cato snapshots because of differing cutoffs, data pulls, and inclusion criteria; these divergences underscore how methodological choices change headline percentages and why careful source-reading is necessary [10] [3]. Reporting also highlights consequences beyond arithmetic—advocates and watchdogs point to humanitarian and legal concerns tied to mass detention, while enforcement proponents argue the agency must detain immigration violators to uphold the law; both perspectives are supported by parts of the record but hinge on which populations are being emphasized in the data [11] [9].
Conclusion: the headline is supportable but incomplete
The concise claim that 5% of ICE detainees have violent convictions and 73% have no convictions is supported by the cited public-data analyses from late 2025 [1] [2] [3], but it omits key qualifiers about timing, the difference between arrests and custody, pending charges, regional variation, and how convictions are defined—factors that materially affect interpretation and policy implications [4] [5]. The available reporting accurately records the snapshot numbers but cannot, on its own, resolve debates over enforcement priorities or the justice of specific detention decisions without further case-level and temporal detail [1] [7].