What do FOIA-released ICE datasets show about trends in convictions among those deported from 2019–2025?
Executive summary
FOIA-released ICE datasets compiled and published by the Deportation Data Project show that a substantial and growing share of people deported or detained between 2019 and mid‑October 2025 had no criminal convictions or only minor offenses, and that 2025 brought a sharp uptick in removals and detentions as the agency shifted tactics away from jails and toward community arrests [1] [2] [3] [4]. Independent analyses of those FOIA data find that most people in ICE custody had no criminal conviction and that only a small fraction were convicted of violent crimes, though the datasets have gaps and coding changes that complicate precise trend estimates [5] [3] [6].
1. What the FOIA datasets are and their scope
The set of ICE records now published through the Deportation Data Project and related FOIA efforts covers individual‑level arrest, detention and detainer records with updates through October 15, 2025, and those raw government files are the basis for multiple third‑party analyses cited here [1] [2] [7]. Those ICE files were produced in response to litigation and FOIA requests and have been repackaged and documented by the Deportation Data Project and partners so reporters and researchers can analyze patterns over time [2] [8].
2. The headline pattern: many deported people lack criminal convictions
Across FOIA‑released ICE data, independent reviewers report that a large share of people ICE detained or deported had no criminal conviction on record — for example, TRAC’s summary of related datasets found roughly 73.6% of people in ICE detention had no criminal conviction as of late November 2025, and Marshall Project analyses of the FOIA data estimated that only about 12% of deportees were convicted of violent or potentially violent crimes in their sample periods [5] [3]. These findings align with ICE’s own public categorizations that list sizable groups removed for immigration violations, re‑entry, or administrative reasons rather than U.S. criminal convictions [9].
3. Recent trend: a 2025 surge and a shift in enforcement focus
Multiple analyses of the FOIA releases and related dashboards document a substantial rise in detentions and deportations during 2025 and a strategic shift by ICE toward so‑called at‑large arrests in communities rather than primarily arresting people in local jails, which contributed to the sharp increase in people without serious or any criminal convictions entering removal pipelines [10] [4]. Reporting based on the released ICE files and processed subsets shows that the monthly counts of deportations for very minor convictions (for example certain traffic offenses) grew notably in 2025 compared with earlier years [3].
4. What the data don’t and can’t tell us without caution
The FOIA datasets are powerful but imperfect: analysts warn of missing location identifiers, shifting ICE coding practices across 2025 releases, and siloing between ICE records and immigration court (EOIR) data that prevents straightforward linkage of every ICE arrest to final court outcomes, which limits certainty about conviction histories for particular deportations [6] [11]. The Deportation Data Project and other intermediaries have flagged omitted or corrected tables (for example removals/encounters tables) in late 2025 releases, and researchers stress that many published counts are best read as estimates rather than immutable totals [2] [6].
5. Competing interpretations and hidden agendas in the reporting
Advocates and researchers using the FOIA data argue the trend shows enforcement sweeping up people with minimal or no criminal history, challenging official rhetoric about targeting the “worst of the worst,” while ICE and some policymakers emphasize national‑security or reentry concerns in public statements and internal categorizations that can blur administrative immigration violations with felony convictions [3] [9]. Analysts also caution that some organizations’ headlines reflect advocacy priorities and that technical data issues—documented by Prison Policy and Deportation Data Project notes—can unintentionally exaggerate or obscure local patterns like where arrests occurred or the seriousness of listed charges [6] [1].
6. Bottom line for the 2019–2025 period
Drawing only on the FOIA‑released ICE datasets and independent analyses of those releases, the clearest trend from 2019 through mid‑October 2025 is that a majority of people detained or deported in recent years had no criminal conviction or only minor offenses, that the share and absolute numbers increased markedly in 2025 as ICE expanded community‑based arrests, and that data quality and linkage limitations mean precise percentages and local attributions require cautious interpretation [3] [5] [4] [6] [11].