How have state and local non-cooperation policies affected the geographic distribution of ICE arrests and removals since January 2025?
Executive summary
State and local non‑cooperation policies have measurably reshaped where ICE arrests and removals occur since January 2025: jurisdictions that limit detainer compliance and information‑sharing show relatively lower counts of jail‑based arrests, while ICE has shifted resources toward more cooperative states and alternative tactics—tracking people in the community and using ICE facilities—to achieve removals [1] [2] [3]. The result is a geographic re‑concentration of interior enforcement and an uneven national map of who gets arrested and deported, although limits in publicly released removal tables and some agency reporting mean the picture is still being reconstructed by independent projects [4] [5].
1. Non‑cooperation changed the arrest venue: from local jails to ICE‑led operations
When jurisdictions curtailed honoring ICE detainers or transported fewer people from local jails to federal custody, ICE arrests inside jails fell in those places and rose elsewhere, because the Criminal Alien Program (CAP) and jail‑based screening historically generate many ICE apprehensions—data show a surge in community and ICE‑office arrests as jail referrals became less reliable [6] [7] [2]. Prison Policy Institute’s tracking found that the scale of arrests depends “at least partially” on state and local cooperation choices and documented growing use of local lockups by ICE in 2025 overall, but also noted the variation tied to jurisdictional policies [1].
2. Cooperative states absorbed more interior enforcement activity
Analyses of state variation show arrest spikes concentrated in places with stronger legal and practical cooperation with ICE; UCLA and other state‑level work tie higher per‑capita ICE arrests to jurisdictions that fully cooperate with the administration’s priorities, reinforcing a geographic concentration of enforcement where state policy permits handoffs and detainers [8] [1]. Migration Policy Institute’s explainer underscores that the degree of cooperation materially determines how many noncitizens are placed into removal proceedings from a given locality, confirming the mechanism by which policies drive geography [6].
3. ICE tactics adapted: more at‑large and ICE‑facility arrests where cooperation waned
Independent reporting and data projects indicate ICE responded to non‑cooperation by shifting tactics: increasing at‑large community arrests, using ICE offices and civil courthouses, and moving detainees between facilities—including distant federal detention centers—so enforcement did not simply vanish where sheriffs or states limited collaboration [2] [9] [10]. This tactical pivot is visible in state case studies (Hawaii’s arrests and bookings rose sharply despite local constraints) and in agency and research tallies showing more arrests executed outside the jail‑handoff pipeline [9] [10].
4. Removals reflect both policy geography and capacity constraints
Removals increased overall in FY2025, but the ability to convert arrests into deportations depended on detention bed capacity, fast‑track authorities, and logistical agreements—factors that interact with where arrests happen; Migration Policy notes limited bed capacity relative to the administration’s goals, and TRAC and other trackers document large but uneven increases in removals that correlate with interior arrest patterns [3] [11]. Deportation Data Project and ICE releases provide the core numbers but also acknowledge gaps and missing linked tables, which complicates precise state‑by‑state removal attribution [4] [5].
5. Who gets arrested shifted as geography shifted
As ICE leaned on CAP and expanded community arrests, the composition of arrestees changed: a growing share lacked violent criminal convictions, and CAP‑sourced arrests from jails still accounted for many cases but with lower conviction rates than administration talking points suggested—researchers found 42% of CAP jail arrests had convictions in early 2025, and the share of violent‑conviction arrestees declined over the year [7] [12]. That composition varied by place because cooperative jurisdictions produced more CAP referrals, while non‑cooperative areas saw ICE rely on different identification and locating methods [7] [2].
6. Limits of the public record and open questions
All of these conclusions rest on agency reports, independent compilations, and journalism; key datasets (removals/encounters tables and fully linked person‑level flows) remain incomplete or selectively released, meaning precise mapping of removals to local non‑cooperation policies is still best‑effort rather than definitive [4] [5]. Alternative interpretations exist—some local officials argue limited cooperation merely delays enforcement rather than redirects it—but the preponderance of data and state‑level analyses shows the policies materially reallocate where arrests and removals occur [1] [8] [2].