What data does ICE publish on detainees’ citizenship status and how can researchers get person‑level records?
Executive summary
ICE publishes aggregate and case-level fields that record an individual’s “citizenship” or “citizenship country” across its public enforcement tables, and these fields appear in datasets covering arrests, detainers, detentions, encounters and removals [1] [2]. Researchers seeking person‑level records rely on ICE’s public statistics, historical FOIA releases reposted by intermediaries such as the Deportation Data Project and TRAC, and direct FOIA requests, but must navigate known data limitations, retroactive updates, and gaps documented by watchdogs [2] [3] [4].
1. What ICE publishes on detainees’ citizenship status
ICE’s public statistics and datasets include a recorded citizenship field (often labeled “Citizenship Country” or “Citizenship”) in its enforcement tables for arrests/apprehensions, detentions/book‑ins, encounters, detainers and removals, and these data are incorporated into OHSS monthly tables and ICE’s public statistics pages [1] [5] [6]. The Deportation Data Project hosts original individual‑level tables produced by ICE that include citizenship fields for arrests, detainers, detentions, encounters and removals and notes that these fields are the agency’s recorded citizenship entries [1] [2].
2. Granularity, semantics and documented limitations
The “Citizenship Country” field is not a flawless identifier: ICE’s documentation and intermediaries note that it may omit secondary nationalities in cases of dual citizenship, it can contain errors, and ICE sometimes retroactively updates records—so the same individual’s citizenship entry may change across releases [1]. OHSS and ICE state that their publishings can be updated quarterly and that later releases may supersede earlier ones, which affects researchers trying to build stable longitudinal person files [6] [5]. TRAC’s reconstruction from FOIA releases also warns of coverage gaps—its FOIA suits revealed missing records (for example, an estimated 45,000 detainees missing in one ICE response), underscoring that published person‑level datasets are incomplete in places [3].
3. Where to obtain person‑level records
There are three practical sources: first, ICE’s public statistics and OHSS monthly tables offer aggregate and some line‑level exports maintained as the OHSS Persist dataset [6] [5]. Second, FOIA‑released, anonymized, person‑level files previously obtained by researchers and reposted by intermediaries—most prominently the Deportation Data Project and TRAC—provide downloadable, linked datasets for arrests, detainers, detentions, encounters and removals, often with identifiers that permit merging across tables [2] [1] [3]. Third, researchers can file their own FOIA requests for Enforcement Integrated Database (EID)/ENFORCE module extracts or specific OHSS Persist slices, recognizing litigation or appeals may be necessary to obtain complete person‑level records [7] [2].
4. Accuracy risks, oversight findings and hidden collection practices
Oversight reports and advocacy research raise two structural cautions: GAO found ICE and CBP do not systematically track or consistently investigate potential U.S. citizenship claims and inconsistencies in guidance, meaning citizenship flags in enforcement data may understate U.S. citizens encountered or misclassify individuals [4]. Separately, investigative documentation on ICE’s Enforcement Integrated Database shows internal systems compile extensive biographic and biometric profiles and case notes beyond what public datasets publish, meaning public citizenship fields are a simplified slice of a much larger, potentially sensitive internal record system [7].
5. Practical research workflow and caveats
Researchers should start with the Deportation Data Project and TRAC exports to access already‑cleaned, person‑level tables and use OHSS Persist tables for official counts, then triangulate with ICE’s statistics pages for as‑of dates; always document the data release date because ICE publishes with a lag and later corrections can change citizenship tallies [2] [6] [5]. Expect to merge across arrests, detainers, detentions and removals using provided linked identifiers when available, but plan for missing identifiers and retroactive edits—treat citizenship fields as recorded agency assertions, not independently verified nationality determinations [1] [3]. For more sensitive or complete person‑level extractions, FOIA requests targeting EID/ENFORCE extracts may be necessary, but these requests encounter exemptions, redactions and legal pushback given the operational nature of those systems [7].