Which U.S. states saw the largest increases or decreases in ICE arrest rates during 2025 and why?
Executive summary
A wave of interior ICE enforcement in 2025 produced the largest percentage increases in arrest rates in the District of Columbia, New Mexico, Idaho, Oregon and Virginia — while states with the biggest absolute counts of arrests were Texas, Florida and California — changes driven by White House pressure to scale up community raids, expanded use of local jails and new detention capacity rather than a shift toward arresting only people with criminal records [1] [2] [3] [4]. Regional exceptions — notably much lower arrest rates in parts of Northern California despite the national surge — and limits in publicly released time windows complicate across-the-board comparisons [5] [6] [3].
1. Biggest percentage increases: sudden hot spots and why they matter
The sharpest year‑over‑year jumps by rate were concentrated in places that previously had few interior arrests: the District of Columbia saw arrests climb from seven under the prior year to 1,190 in 2025, and New Mexico, Idaho, Oregon and Virginia each recorded more-than-fivefold increases in arrests compared with the Biden period, signaling new operational focus on nontraditional locales rather than a simple transfer of enforcement from border areas [1]. Those percentage increases reflect both targeted campaigns — such as the high‑profile October horse‑track operation in Idaho — and a baseline effect where small prior counts make percentage growth dramatic [1].
2. Largest absolute increases: Texas, Florida and California as enforcement epicenters
By raw numbers, Texas, Florida and California accounted for a disproportionate share of arrests: Texas, Florida and California together made up more than 41 percent of ICE arrests in the Jan.–June 2025 window, and Texas alone was the origin of roughly one-quarter of ICE’s arrests through July, reflecting the state’s size, large undocumented population and the role of local jails as a principal funnel into ICE custody [2] [4]. Texas’ position was amplified by detention capacity concentrated there and by the agency’s practice of picking people up from local jails after criminal processing — a dynamic documented in reporting that found local criminal justice systems channeling many cases into ICE custody [4] [7].
3. Policy levers that explain the rises — White House quotas, local partnerships, and detention growth
Multiple reporting pieces tie the surge to policy choices: analysts found that arrests spiked around the second Trump inauguration and again after late‑May pressure from White House staff for ICE to ramp community raids toward a 3,000‑a‑day target, and researchers standardized two distinct periods to compare pre‑ and post‑May activity [3]. The administration’s expansion of detention capacity, use of new and reopened facilities, and a sharp increase in 287(g) agreements that deputize local police for immigration enforcement created both the capability and the local pipelines for far more arrests [8] [9]. Advocates and watchdogs contend that these operational changes — plus new Congressional funding for detention — shifted tactics toward “at‑large” community arrests and worksite raids, increasing arrests of people without criminal records [9].
4. Notable decreases and exceptions: where enforcement did not mirror the national surge
Not all places saw proportional increases; Northern California’s Bay Area retained one of the nation’s lowest ICE arrest rates through mid‑October despite the national rise, demonstrating that local office practices, regional prosecutorial policies and resource constraints can blunt federal enforcement even in a high‑pressure year — while parts of San Diego/Imperial counties bucked that regional pattern with a sharp local rise [5]. State‑level statistical analyses emphasize that outcomes depend heavily on which local law enforcement partnerships and detention arrangements exist in each jurisdiction, so per‑capita arrest rates vary widely [6].
5. Caveats, competing narratives and data limits
Comparisons hinge on how time periods and denominators are chosen — researchers standardized different windows (Jan. 20–May 20 vs. May 21–Oct. 15) and used per‑100,000 or per‑1,000 noncitizen denominators to make state comparisons meaningful, so “largest” increases can shift depending on whether reporters mean percentage growth, absolute counts or arrests per noncitizen population [3] [6]. Public reporting also varies in coverage windows (some pieces stop in June or October) and in whether they emphasize administrative ICE statistics or compiled datasets, and available sources do not provide a single, fully audited national state‑by‑state yearbook for all of 2025 — a limitation readers should weigh when interpreting rank order [2] [3] [6].