How does ICE classify detained individuals by citizenship status in its raw dashboards, and how can researchers compute 'undocumented' from that data?
Executive summary
ICE’s public dashboards and data releases categorize people it encounters and detains chiefly by country of citizenship and by criminality, not by a neat binary “documented/undocumented” flag, so researchers must infer undocumented status using combinations of fields and careful exclusions; ICE guidance and third‑party documentation warn that the citizenship field can be incomplete or updated retroactively and that other indicators (birth country, removal destination, administrative case outcomes) are needed to approximate who lacks lawful status [1] [2] [3].
1. What ICE explicitly shows in its dashboards
ICE’s Enforcement and Removal Operations dashboards publish detention, arrest and removal tables that break out records by “country of citizenship” and by criminality categories (e.g., convictions, pending charges, immigration violations) and are searchable by field office and time period; those dashboards are the agency’s primary public source for who is in custody and where they are from, and ICE states the dashboards will be updated quarterly [1] [2].
2. What “citizenship” in the raw tables actually is — and its limits
The raw ICE tables include a “Citizenship Country” column and often a “Birth Country” and “Departed Country” that researchers can cross‑reference, but ICE and independent curators note the citizenship field can omit dual nationalities, be corrected retroactively, and contain data errors — meaning a non‑U.S. citizenship value is a proxy, not a legally definitive declaration of undocumented status [3] [4] [1].
3. How ICE internally treats potential U.S. citizens and evidence of status
ICE operational guidance instructs personnel to assess indicia and probative evidence of U.S. citizenship when encountered, and the agency maintains policy documents about investigating potential U.S. citizenship — a process that can result in records being updated when new documentation is found — which underscores why raw datasets sometimes show later changes to citizenship or case status [5].
4. Why “undocumented” is not a direct field and the basic computation researchers use
Because ICE does not publish a simple “undocumented” checkbox in its public dashboards, researchers commonly compute an undocumented proxy by selecting detained/encountered individuals whose Citizenship Country is not “United States” (and whose records lack indicators of lawful status in ancillary fields), then excluding known lawful categories where available (for example, records showing removal to a person’s citizenship country, or datasets that explicitly flag lawful permanent residents or naturalized citizens); the Deportation Data Project and OHSS monthly tables recommend comparing Citizenship Country, Birth Country and Departed Country columns to identify third‑country removals and other edge cases [3] [6] [4].
5. Practical caveats, ambiguity, and recommended safeguards
Any computed “undocumented” count must be qualified: ICE’s public data can conflate nationals with lawful permanent residents and dual citizens, may be retroactively amended, and field‑level identifiers for legal status are incomplete in publicly released tables, so best practice is to document assumptions, run sensitivity checks (e.g., include/exclude dual‑citizenship flags, test for rows later amended to “U.S.”), and where possible triangulate with USCIS or FOIA records for individual cases rather than treating the dashboard as a legal determination [3] [7] [5].
6. New technologies and the risk of misclassification in field identification
Emerging ICE tools such as Mobile Fortify can return “possible citizenship status” and “possible overstay status” in real time by querying DHS databases, but those are investigative indicators rather than adjudicated findings and thus can feed into ICE records that later change — another reason researchers should treat citizenship fields in dashboards as operationally useful but imperfect for labeling someone definitively “undocumented” [8] [9].
7. Bottom line for analysts using ICE raw dashboards
Use “Citizenship Country ≠ United States” as the starting filter to estimate non‑U.S. nationals in custody, then exclude rows with explicit lawful‑status markers if present, cross‑check Birth Country and Departed Country to catch third‑country removals or dual‑citizenship edge cases, transparently report uncertainty and retrospective edits, and avoid claiming a dashboard‑derived count is a legal finding of undocumented status without corroborating adjudicative records [1] [3] [4].