What national datasets exist that link immigration status to arrest and conviction records, and what are their limitations?
Executive summary
Nationally available datasets that link immigration status to arrests and convictions are fragmented and mostly originate from immigration enforcement agencies—DHS/ICE and CBP—and a few state-level exceptions; widely used criminal justice datasets (UCR, NCVS, NIBRS) do not record immigration status, leaving major gaps for researchers and policymakers [1] [2] [3] [4]. The most useful linkages come from fingerprint-based crosschecks into DHS biometric systems (IDENT/ADIS) and enforcement reporting, but these sources are biased toward people who have had contact with immigration or enforcement systems and vary by jurisdiction, completeness, and transparency [2] [5] [6].
1. DHS and ICE enforcement tables: a national but enforcement‑focused lens
The Office of Homeland Security Statistics (OHSS) at DHS publishes monthly and annual tables that include encounters, detention book‑ins and book‑outs, removals, credible fear screenings, and related enforcement actions—data that ties immigration encounters to administrative outcomes but primarily reflects enforcement activity rather than a population‑wide linkage of status to criminal convictions [1]. ICE’s Enforcement and Removal Operations (ERO) statistics report arrests by country of citizenship and categorize individuals by criminal history (convictions, pending charges, or none), but the agency’s framing focuses on removability and prioritized enforcement, and DHS public tables often lack the granular breakdowns needed to assess conviction types or match to local arrest records comprehensively [5].
2. CBP “criminal alien” statistics: aggregation from record checks, not a complete criminal‑justice ledger
CBP publishes “criminal alien” statistics derived from records checks following apprehensions and records in CBP databases, which can indicate prior convictions in the U.S. or abroad, but these datasets are inherently limited to people stopped by CBP and reflect pre‑existing law‑enforcement records available at interdiction; they are not equivalent to nationwide criminal‑justice data that uniformly annotates immigration status [6]. Because CBP’s counts derive from post‑apprehension database checks, they capture enforcement contacts and existing convictions rather than offering a representative linkage between status and the universe of arrests or convictions [6].
3. Texas: the rare statewide criminal records system that retains immigration status
Texas stands out as a unique laboratory: its Department of Public Safety checks fingerprints against DHS’s IDENT and retains the immigration status returned by DHS in its criminal arrest records, enabling researchers to compare convictions and arrests by documented, undocumented, and native‑born status—an exception that has produced influential research precisely because most national crime reporting systems omit status [2] [3]. That uniqueness underscores the limitation: most states and national crime surveys do not record immigration status, so Texas is an outlier rather than a replicable national model [2] [3].
4. What the major crime data collections and third‑party portals do and do not provide
Widely used crime data sources—the FBI’s Uniform Crime Reports (UCR), the National Crime Victimization Survey (NCVS), and NIBRS—do not include immigration status, a structural omission that prevents direct national estimation of immigrant involvement in arrests or convictions from those series [2] [3]. Third‑party aggregators like the Center for Immigration Studies Data Portal consolidate government immigration statistics (I‑94, visa counts, program stats) but draw from administrative sources that are not designed to link comprehensively to arrest/conviction records across jurisdictions [7].
5. Core limitations: selection bias, incomplete identifiers, policy barriers, and analytic opacity
Fingerprint cross‑checks into DHS systems are useful for identifying individuals who have prior immigration records, but they only work when someone has had prior contact with immigration or law enforcement—there is no definitive national list of undocumented persons, and people who have never interacted with immigration systems won’t appear in DHS databases, producing undercounts and selection bias [2] [4]. Sanctuary policies and varying state practices limit data sharing and retention, and enforcement datasets frequently emphasize removals or “criminal aliens” without standardized crime‑type breakdowns needed for scholarly comparison, a gap noted by fact‑check and policy analyses showing many ICE arrestees have no U.S. criminal convictions and that public DHS counts don’t always permit fine‑grained assessment [8] [5] [4]. Migration Policy and other analysts conclude that while administrative linkages exist, national inference is constrained by these structural and technical shortcomings [9] [10].