What does ICE’s Deportation Data Project dataset show for Florida arrests in 2025 when filtered by conviction status and offense code?
Executive summary
The Deportation Data Project’s ICE release — which reproduces individual-level ICE arrest records through October 15, 2025 — shows that Florida experienced both large-scale, publicized operations targeting people with criminal convictions and a simultaneous growth in arrests of people with no criminal convictions or only minor offenses; local reporting found non‑criminal immigration arrests in Florida rising to exceed arrests of people with criminal convictions by early June 2025 [1] [2] [3]. Analysts caution the dataset has coding gaps and missing location identifiers that complicate precise counts by conviction status and offense code, so conclusions require careful caveating [4] [2].
1. What the raw Deportation Data Project files contain and their timeframe
The Deportation Data Project republishes individual‑level ICE tables for arrests, detainer requests, and detentions covering October 2011 through mid‑October 2025, with the most recent ICE update extending to October 15, 2025 — the dataset is the product of FOIA litigation and multiple releases in 2025 [5] [1] [6] [2]. The project also provides documentation describing the variables (including conviction status and offense codes) and a webinar to guide users, but analysts note that ICE’s internal coding practices shifted during 2025, affecting comparability [2] [6].
2. What the dataset shows for Florida when filtered by conviction status
Across national and Florida reporting that used the Deportation Data Project files, a clear pattern emerges: arrests of people without known criminal convictions rose substantially in 2025 and, in Florida specifically, non‑criminal immigration arrests had climbed to outnumber arrests of people with criminal charges or convictions by early June 2025, according to WUSF’s analysis of the Deportation Data Project extract [3]. Nationally, complementary trackers reported a large share of people in ICE custody had no criminal conviction — for example, a November 2025 snapshot showed roughly 73.6% of detainees lacked a criminal conviction — a statistic used in Florida coverage as context [7] [8]. These figures reflect conviction status fields in the Deportation Data Project but are sensitive to how ICE classifies “conviction,” “pending,” and “immigration‑only” cases [9] [2].
3. What the dataset shows when filtered by offense code (types of offenses)
When records are filtered by offense code, the dataset aligns with ICE’s own historical profile that many arrested had convictions for DUI, drug possession, assault and non‑criminal traffic offenses such as hit‑and‑run; at the same time, many arrests in 2025 were tied only to immigration violations or low‑level traffic infractions rather than serious violent crime [9] [3]. Local Florida reporting and official statements reveal two simultaneous narratives in the data: federal and state operations publicly highlighted arrests of people with histories of robbery, assault, gang affiliation and repeated illegal re‑entry, while the underlying individual records in many counties show numerous arrests coded as immigration offenses, traffic offenses, or “no conviction” [10] [11] [12] [3].
4. Limits, caveats and competing interpretations in the dataset
Researchers and reporters warn that ICE’s 2025 release contains missing identifiers and inconsistent location coding that can obscure where arrests actually occurred and whom they represent, making it difficult to match some official operation tallies to individual records in the dataset [4] [6]. Advocates and local outlets emphasize the high share of non‑convicted detainees to argue over‑broad enforcement, while DHS and state officials point to targeted sweeps that netted many people with serious criminal convictions — both claims can be tied to different slices of the same Deportation Data Project files and to separately issued press tallies [10] [11] [3]. Analysts therefore recommend combining the Deportation Data Project extracts with local jail records and court data before drawing definitive conclusions about the criminality profile of Florida arrests [4] [5].
5. Bottom line for researchers using the dataset on Florida arrests in 2025
The Deportation Data Project dataset for 2025 demonstrates a dual reality in Florida: public operations and press releases emphasize arrests of “criminal aliens,” while the underlying ICE records show a sizable and growing number of arrests of people without criminal convictions or with minor offense codes; the dataset supports both observations but requires caution because of coding shifts and missing data that limit one‑to‑one validation of operation claims [3] [10] [11] [4]. Researchers seeking precise tallies by conviction status and offense code should expect to do additional cleaning and local cross‑validation rather than treat the published ICE fields as final truth [2] [5].