What do ICE and independent datasets say about the criminal conviction status by nationality of detainees?
Executive summary
ICE’s own public data classify detainees by criminal-history categories and show a substantial and growing share of people in custody without criminal convictions, a trend corroborated by independent trackers such as TRAC, Syracuse University analyses, and several academic and policy centers [1] [2] [3]. Independent researchers also find that most of the recent increase in ICE detention is accounted for by people with no convictions or only immigration violations, though datasets differ in definitions and completeness and offer limited, inconsistent country-by-country breakdowns [4] [5] [6].
1. What ICE’s datasets publicly show about convictions
ICE presents detention tables that explicitly separate “individuals with criminal convictions” from those without convictions or with only pending charges and immigration violations, and the agency’s statistical system (SSOR / EID) is described as the authoritative administrative source for those categories [1] [7]. Public releases and the Enforcement and Removal Operations statistics therefore allow tabulation of detention populations by “convicted” versus “non‑convicted/pending” status, and agency notes clarify that the non‑convicted group can include immigration fugitives and repeat re‑entrants even absent a U.S. criminal conviction [1].
2. What independent datasets report about conviction shares
Independent analyses—most notably TRAC at Syracuse and related fact‑checks—report that a large majority of ICE detainees at several points in 2025 had no criminal conviction: TRAC’s compilation found roughly 72% of people in ICE detention had no criminal conviction as of a September snapshot, a figure echoed by fact‑checkers and other researchers [3] [2]. Multiple analyses and commentators also document that only a small fraction of detainees are convicted of violent offenses and that many convictions cited by ICE are for relatively minor offenses, including traffic violations [8] [9].
3. Recent growth driven by non‑convicted detainees
Researchers tracking month‑to‑month changes report that almost all of the surge in detention levels during FY2026 has been driven by people without criminal convictions: one analysis estimates roughly 92% of detention growth came from non‑convicted individuals, with only about 8% attributable to those with criminal convictions, and similar work finds that non‑criminal detainees accounted for the lion’s share of numerical increases [4] [5]. National reporting likewise noted that the share of arrested and detained immigrants with convictions fell sharply in the fall of 2025, with some time windows showing single‑digit percentages of detainees having convictions [10].
4. Disagreements, definitional pitfalls, and datasource limits
A central point of contention across sources is terminology: ICE sometimes flags people with pending charges as “criminal arrests,” which independent analysts warn can conflate charges with convictions, inflating the appearance of criminality in the detained population [8]. TRAC and other independent projects also stress gaps and identifier problems in publicly released data that make it difficult to cleanly map conviction status by state or country of citizenship in all cases, and some datasets explicitly note missing identifiers that hinder per‑jurisdiction analyses [6] [2].
5. What the sources do and do not say about nationality breakdowns
ICE’s public tables include a “detention by country of citizenship and criminal history” construct, indicating the agency captures nationality alongside conviction categories, but independent reporting warns that limitations in the available datasets impede clear, consistent country‑by‑country comparisons of conviction status in the public record [1] [6]. Independent studies (UCLA, TRAC, others) emphasize demographic patterns—such as a disproportionate increase in detained Latino immigrants without convictions—but the public sources provided do not contain a comprehensive, validated crosswalk in these materials that would produce a reliable, detailed matrix of conviction status by every nationality [11] [2].
6. Competing narratives and hidden incentives
Advocacy and policy organizations use the same underlying administrative data to advance contrasting narratives: civil‑liberty groups and some academic centers highlight the human‑rights and due‑process implications of large numbers of non‑convicted detainees [11] [9], while government statements framing enforcement as focused on criminals can rely on broader categories (including pending charges) to justify operations—an approach critics say obscures the underlying conviction rates [8]. Independent watchdogs repeatedly flag that presentation choices and definitional framing materially shape public perception of how “criminal” the detained population is [8] [3].