How do CBP and ICE define and count 'criminal aliens' and what are the limitations of those counts?
Executive summary
CBP and ICE label and tally “criminal aliens” by matching apprehended or detained noncitizens against law‑enforcement conviction or charge records and administrative detention events in DHS databases, producing public tallies used for oversight and enforcement planning [1] [2]. Those counts are shaped by agency definitions, technical data flows between CBP, ICE and OHSS, and operational choices — and therefore systematically understate, overstate or obscure different slices of reality depending on what question one asks [3] [4].
1. How CBP defines and reports “criminal aliens” at the border
CBP’s Criminal Alien Statistics are drawn from records checks conducted after Border Patrol apprehensions and record conviction information into a CBP database; the public summaries cover Border Patrol enforcement actions and fiscal‑year aggregates based on those database matches [1]. CBP’s framing is incident‑based — an apprehension followed by a records check can produce a “criminal alien” count when prior convictions are found, but that label depends on the availability and content of external law‑enforcement databases accessed after arrest [1].
2. How ICE defines and reports “criminal aliens” in the interior
ICE’s enforcement statistics and public FAQs describe “criminal illegal aliens” as removable noncitizens who have criminal convictions or charges in the United States and who are arrested, detained or otherwise processed by ICE’s Enforcement and Removal Operations; ICE publishes detention, arrest and ERO arrest breakdowns by country and criminal‑history categories drawn from its Enforcement Integrated Database (EID) and reporting systems [2] [3]. ICE also emphasizes mission language — “worst of the worst” — in DHS press releases about high‑profile arrests, reflecting program priorities rather than a neutral statistical taxonomy [5] [6].
3. What the agencies actually count and where the numbers come from
Counts originate in operational events (CBP apprehensions, ICE book‑ins/book‑outs) and database matches: ICE extracts book‑in/book‑out events from EID using its reporting mart and uploads data to CBP’s Unified Immigration Portal monthly; OHSS produces flow reports using CBP and ICE records to summarize apprehensions, arrests, detentions and removals [3] [4]. ICE has reported large identification efforts — for example telling Congress it had identified roughly 435,000 unauthorized immigrants with criminal convictions who were not in custody — a figure derived from internal database matching and subject to the same data‑source limits [7].
4. Main limitations: double counts, database gaps, and definitional slippage
Agency counts can double‑count people with multiple book‑ins or re‑arrests because reporting tables count initial book‑ins and final book‑outs and count some events multiple times; ICE notes people with more than one initial book‑in or final release are counted multiple times in tables [3]. Databases rely on accessible conviction records and vary in whether they record convictions abroad, pending charges vs convictions, or noncriminal administrative infractions; the agencies’ public materials make clear counts reflect matches to “criminal conviction” where those records are available, not a comprehensive census of culpability [1] [2]. Transfers between ICE facilities are not new detention events and thus can be invisible in some tallies, while referrals, releases to Alternatives to Detention, and people identified but not in custody create further gaps between identified, arrested, and removed populations [3] [2].
5. How politics and rhetoric shape perception of the data
DHS and ICE press releases spotlight violent or sexual‑offense convictions and use phrases like “worst of the worst” and categorical percentages (e.g., claims that 70% of ICE arrests involve people with U.S. charges or convictions) to frame enforcement victories — messaging that emphasizes select convictions rather than methodological nuance [8] [5]. Independent trackers and reporters use ICE’s internal estimates (such as the 435,000 figure shared with Congress) to assess scope, but those media summaries can mask the underlying data assumptions and the difference between “identified with convictions” and “in custody or removed” [7].
6. Bottom line: useful operational snapshots, not a definitive crime‑burden measure
CBP and ICE counts are valuable for measuring enforcement activity and operational caseloads because they track events and database matches, but they are not a direct measure of all noncitizen criminality or community risk: methodological choices (what constitutes a countable event), database completeness, double‑counting, foreign vs. domestic convictions, and political messaging all limit what the headline numbers can reliably say about prevalence or threat [1] [3] [2] [7].