How frequently do misidentifications occur in ICE arrests and what data tracks these incidents?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
ICE’s own and FOIA-derived datasets show thousands of interior enforcement arrests in 2025 — frequently exceeding 1,000 a day at times — and large shares of those arrested lack a criminal conviction, but the public record does not offer a simple, official tally of “misidentifications” as a distinct category (for example, U.S. citizens wrongly arrested) [1] [2] [3]. Multiple independent projects (the Deportation Data Project, TRAC, newsrooms) have published and analyzed ICE arrest, detainer and detention records made available through ICE releases and FOIA lawsuits; those are the primary sources researchers use to track errors, but available sources do not mention a single, authoritative count of misidentifications [3] [4] [5].
1. What “misidentification” means in the records: messy definitions and missing fields
ICE’s public tables and the datasets rebuilt from FOIA responses record every encounter, arrest, detainer request and book-in; they include fields for arresting agency, criminal history categories (convicted, pending charge, no conviction) and identifiers that can be linked across files, but they do not publish a clear, labeled “misidentified” flag or a routine audit of mistaken citizen arrests — so researchers must infer errors from anomalies and follow-up reporting rather than read a single column that says “wrong person” [3] [6].
2. How frequently researchers see arrests without criminal records — a proxy but not the same as mistakes
Analysts using the Deportation Data Project and ICE’s own dashboard report that a substantial share of recent ICE administrative arrests involve people with no criminal convictions: reporting finds roughly one-third nationwide and much higher shares in some localities (more than half in some targeted operations and 58% in San Diego/Imperial County reporting) [2] [7] [8]. These figures document that many arrests are civil immigration actions rather than arrests predicated on criminal convictions — they are not direct counts of misidentifications such as citizen arrests [2] [7].
3. Where journalists and researchers get the data: ICE releases, FOIA and third‑party scrapes
Primary tracking comes from three sources: ICE’s biweekly dashboard and statistics (the agency’s Arrests, Detention and Removals tables), FOIA-obtained datasets assembled by the Deportation Data Project and subsequent public analyses by newsrooms and research shops (Prison Policy, The New York Times, TRAC, local outlets) that merge and analyze those records [6] [4] [3] [1] [2]. The Deportation Data Project’s FOIA releases have been central because they contain row‑level arrest and detainer records through Oct. 15, 2025 and enable tracing individual cases across the enforcement pipeline [3] [4].
4. Limits of the data for counting genuine “mistakes” like citizen arrests
ICE’s datasets and the FOIA dumps permit detailed counts of arrests, locations and criminal-history status, but they do not systematically document follow-up adjudications that would prove an arrest was erroneous (for example, a confirmed U.S. citizen wrongly detained and later released). Local reporting has produced high‑profile accounts of mistaken arrests, but the national datasets do not provide a validated, centralized count of such errors — available sources do not mention a definitive national tally of citizen misidentifications [3] [6].
5. What independent analyses have revealed about scale and trends
Researchers using these datasets have shown ICE arrest totals rising in 2025 (peaking at over 1,000 arrests per day in aggregate analyses), with sizable shares of arrests originating at local jails and many arrestees lacking criminal convictions; those patterns help explain how misidentifications can occur (data linking, jail referrals, and hasty processing), but they stop short of quantifying confirmed mistakes [1] [9] [2].
6. Conflicting narratives and institutional incentives
ICE and DHS emphasize criminality in their public messaging and cite increases in arrests of people with convictions, yet independent data analysts and local outlets show large shares of administrative arrests of people without convictions — a competing interpretation about who ICE is arresting. The datasets themselves can be updated and “locked” only at fiscal-year close, and ICE warns that numbers fluctuate until then — an institutional limitation that benefits neither full transparency nor quick error‑correction [6] [9] [3].
7. Practical steps for someone researching misidentifications
Track the row‑level FOIA datasets maintained by the Deportation Data Project and cross‑check with ICE’s biweekly dashboard; use linked identifiers (when available) to trace detainer requests, book‑ins and removals and combine that with local court and newsroom follow‑ups to spot likely errors. Remember that the datasets can show anomalies (citizenship fields, duplicate identifiers, arrests of people with no criminal history) but do not substitute for case‑level confirmation that an arrest was a misidentification [3] [6] [4].
Limitations and bottom line: The data infrastructure to count arrests is robust and public (ICE dashboards, FOIA datasets), and independent analysts have shown high volumes and substantial shares without criminal convictions, but available sources do not provide a single, authoritative national count of misidentifications such as U.S. citizens wrongly arrested; answering “how frequently” requires painstaking case-level verification that the current public datasets were not designed to deliver automatically [3] [2] [1].