Which government reports and datasets provide breakdowns of ICE arrests by criminal history?

1. ICE’s public enforcement/detention statistics (Initial Book‑Ins; Currently Detained by Criminality)

ICE publishes a suite of enforcement and removal statistics that explicitly include categories for criminal history — for example, the Initial Book‑Ins by Arresting Agency and Month table and the Currently Detained by Criminality table, which report arrests and current detainees broken down by conviction status, pending charges, and various offense categories ^{[2] [1]}.

2. ICE archival datasets and older FOIA releases

Beyond the live statistics, ICE’s archived statistics pages and older FOIA‑produced tables contain historical counts and criminality fields (for example, FY2020 summaries and archived ERO enforcement totals), and those archived files are commonly used to reconstruct multi‑year comparisons of convictions, charges and administrative arrests ^{[5] [1]}.

3. Deportation Data Project’s government‑provided datasets with linked identifiers

The Deportation Data Project republishes historical ICE datasets produced in response to FOIA requests and provides codebooks and linked identifiers that permit researchers to trace individuals through arrest, detention and removal records while preserving anonymity; reporters and analysts have relied on these FOIA releases to disaggregate criminal‑history information at case level ^[3].

4. Third‑party aggregators and civic data projects that parse ICE tables

Independent organizations such as TRAC and USAFacts routinely extract ICE’s tables (and FOIA dumps) to produce digestible breakdowns — for example TRAC’s periodic “Initial Book‑Ins” analyses and USAFacts’ summary reporting on the number of administratively arrested people with prior convictions and average convictions per person — and their work points readers back to the underlying ICE tables ^{[6] [4]}.

5. How news organizations have used these datasets to report criminal‑history breakdowns

Major news outlets and data reporters have relied on the FOIA datasets and ICE tables — The New York Times and The Guardian cited Deportation Data Project material and ICE’s book‑in tables when reporting falling shares of detainees with convictions; these articles demonstrate that criminal‑history breakdowns in public reporting are derived from the exact ICE tables described above ^{[7] [2]}.

6. Key limitations and caveats in the government data

Users should note important constraints in the government sources: ICE’s arrest figures may undercount incidents that don’t lead to book‑ins (ICE reports only arrests that result in someone entering ICE detention), ICE’s categorization of “criminal” can include pending charges and foreign convictions, and ICE changed reporting practices in 2025 (for example re‑coding many Criminal Alien Program cases as “Custodial Arrests”), which complicates longitudinal comparisons ^{[2] [8] [7]}.

7. Why analysts cross‑validate with FOIA and independent databases

Because of those limitations — shifting definitions, aggregation choices and undercounts — researchers routinely cross‑check ICE’s public tables with FOIA releases (via the Deportation Data Project), archived datasets and independent scrapes to produce more complete criminal‑history breakdowns and to link arrests to later detention, court outcomes or removals ^{[3] [2]}.

8. Bottom line for researchers seeking criminal‑history breakdowns

The primary government sources to consult first are ICE’s current statistics pages and the specific tables “Initial Book‑Ins by Arresting Agency and Month” and “Currently Detained by Criminality,” supplemented by ICE archival statistics; for case‑level detail and richer linkage, use the FOIA datasets republished by the Deportation Data Project and validated by third‑party analysts such as TRAC and USAFacts — but always account for coding changes, inclusion of pending charges, and the fact that ICE only reports arrests that result in book‑ins ^{[1] [2] [3] [6] [4]}.