What is the annual breakdown of ICE detention by immigration status (citizen, lawful permanent resident, undocumented)?
Executive summary
ICE’s public dashboards and independent trackers show tens of thousands in detention in 2024–2025, with multiple analyses concluding that a large plurality — and in some snapshots a majority — of people in ICE custody had no criminal convictions. For example, ICE’s own biweekly data compiled by news outlets showed 59,762 people in detention with 16,523 reported as having no criminal record in one snapshot [1], while a Cato analysis of ICE “book-ins” reported 204,297 book-ins through June 14, 2025, with 65% having no criminal convictions [2].
1. What the official numbers actually show: detention totals and criminality
ICE’s Detention Management releases and dashboards provide the agency’s baseline counts and breakouts by criminality and arresting agency; journalists and researchers have used those spreadsheets to report that overall detention reached the high tens of thousands in 2025 — for example, 59,762 in one ICE snapshot cited by The Guardian [1] and “more than 56,000” reported by KPBS [3]. Those same ICE-origin datasets and derived tables are the basis for claims about criminal versus non‑criminal detainees [4] [5].
2. Multiple independent tallies find many detainees lack convictions
Independent analyses of ICE’s data diverge slightly in methods but point to the same phenomenon: a large share of people in ICE custody have no criminal convictions. Cato’s June 2025 tally of “book-ins” counted 204,297 individuals since the fiscal year began and found 65% had no criminal convictions [2]. The Guardian’s reporting of an ICE snapshot gave specific counts — 16,523 with no record, 15,725 with convictions, 13,767 with pending charges — in a dataset showing 59,762 detained as of that release [1]. Other outlets and policy groups (Migration Policy Institute cited in the NYT opinion) reported similar high shares of non‑convicted detainees [6].
3. Why “by immigration status (citizen, LPR, undocumented)” is hard to produce from available sources
Available sources emphasize criminality categories (convicted, pending, no record) and arresting agency, not a neatly disaggregated annual breakdown of detention by immigration status such as U.S. citizen, lawful permanent resident (LPR), and undocumented. ICE and DHS tables cited by reporters focus on “aliens” subject to removal and criminality; they do publish some citizenship and removals-by-citizenship tables (OHSS monthly tables), but the public reporting we have here does not produce a clear year-by-year table that maps daily detention population to the three immigration-status groups requested [7] [8]. Therefore: available sources do not mention a straightforward annual breakdown by citizen/LPR/undocumented in the precise form you asked for [7] [8].
4. What the data do allow: proxies and partial counts
Researchers use proxies to approach your question: “book-ins” and “currently detained” spreadsheets can be filtered for citizenship or country of origin, and OHSS monthly tables include removals and some citizenship data [2] [8]. Journalistic reconstructions (The Guardian, KPBS, Prison Policy) have produced point-in-time tallies of detained people by criminality and arresting agency and tracked growth over months, but those reconstructions stop short of publishing an annualized table that breaks detention into U.S. citizen / LPR / undocumented categories for each year [1] [4] [9].
5. Competing interpretations and possible implicit agendas
Advocacy and academic groups emphasize human‑rights and due‑process harms of detaining large numbers of non‑convicted people; Cato emphasized the non‑criminal composition to argue policy inconsistencies [2]. Government pages frame detention as a tool to secure presence for proceedings and public safety (ICE Detention Management) [5]. Media outlets highlight different slices of the same datasets — some stress alarming growth in detention (Prison Policy, KPBS), others the composition shift toward non‑criminals (Guardian, NYT opinion citing MPI) — reflecting differing editorial priorities [9] [3] [1] [6].
6. Bottom line and what you can request next
The best-supported, publicly available data show large numbers detained and consistently high shares without criminal convictions (Cato: 65% of FY2025 book-ins through mid‑June; Guardian snapshot counts) [2] [1]. However, a direct annual table breaking ICE detention into U.S. citizen, lawful permanent resident, and undocumented categories is not present in the sources you supplied; to produce that, one needs either (a) the ICE/DHS “book-ins” and “currently detained” spreadsheets filtered by citizenship across a full year (OHSS/DHS Persist dataset) or (b) a FOIA or compilation from the Deportation Data Project/TRAC that assembles citizenship-by-detention over time [8] [4] [10]. If you want, I can outline precisely which ICE/DHS files to pull and how to aggregate them to construct the citizen/LPR/undocumented annual series.