Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
How do ICE's 2025 removal numbers compare to previous years and what trends are visible?
Executive summary
ICE’s public removals data for FY2025 show a sharp increase compared with several recent years, but the figures and trends are complicated by dataset revisions, carryover from prior fiscal years, and methodological changes; for example, the Deportation Data Project and TRAC note major additions and corrections in mid‑2025 releases that alter FY2025 and FY2024 totals [1] [2]. Independent trackers warn the first FY2025 releases include carryover data and that late‑June and late‑July 2025 datasets differ by tens of thousands of removals, so caution is required when comparing year‑to‑year counts [3] [1] [4].
1. What the raw numbers claim — apparent uptick in 2025
ICE’s official Enforcement and Removal Operations (ERO) dashboard is the primary source for arrests and removals data and is updated regularly, with the agency noting data are published one quarter in arrears and may be revised until a fiscal year is “locked” [5]. Independent analysts and projects observing ICE’s semi‑monthly and FOIA‑released datasets documented that FY2025 removals reported in mid‑2025 are substantially higher than some earlier releases — for example, the Deportation Data Project reported that the late‑July release showed over 40,000 more FY2025 removals than the late‑June release [1]. TRAC and TRAC‑linked analyses also tracked cumulative semi‑monthly removals showing rapid accrual of removals in the early part of FY2025 [2].
2. Why comparisons to prior years are tricky — revisions, carryover and methodology
Comparing FY2025 removals to previous years is not a simple apples‑to‑apples exercise because ICE’s published series can include carryover entries from the prior fiscal year in the first release of each year, and ICE and third‑party datasets have undergone retroactive methodological changes since 2023 [3] [1]. The Deportation Data Project cautioned users that several removals tables were incomplete or contained CBP removals mistakenly included in an earlier file, and thus recommended relying on the late‑July dataset as the more accurate version [4] [1]. TRAC likewise notes ICE’s semi‑monthly cumulative reporting begins in October and can mix data from administrations when a fiscal year overlaps presidencies [2].
3. Short‑term trends visible in 2025 data — acceleration and concentrated activity
Multiple trackers observed an acceleration of removals in early 2025 once reporting stabilized: TRAC’s semi‑monthly series documented new removals accumulating rapidly after January 2025, and reporting projects emphasize that much of FY2025’s early figures reflect activity after February 2025 [2]. The Guardian’s analysis underscores that the first data release of each fiscal year “includes carryover data” and more than two weeks of arrests and deportations, which can amplify apparent early‑year spikes [3].
4. Broader historical context — rises and falls since 2019
Government audits and analyses provide a longer baseline showing that ICE arrests and removals declined from 2019 through 2021 and then increased in 2022; GAO reported arrests rose from 133,541 in 2019 to 154,204 in 2022, illustrating that enforcement levels have fluctuated markedly in recent years [6]. Those historical swings mean a single‑year rise in 2025 may reflect policy shifts, operational scaling, or corrections to reporting rather than a straight trend line without additional context [6].
5. Data quality warnings and what independent projects say
Deportation Data Project and the UCLA‑linked FOIA releases repeatedly warned users about incomplete removals tables and recommended caution; they explicitly documented the presence of missing records in March and June 2025 releases and later clarified which files are more reliable [4] [1]. The Deportation Data Project’s update concludes that some earlier releases undercounted removals in FY2024 and that late‑July 2025 removals data are the preferred set for FY2025 analysis [1].
6. What this means for interpretation and reporting
Analysts should treat FY2025 removals totals as provisional and cross‑check the late‑July ICE‑based datasets recommended by independent projects before drawing firm conclusions [1] [4]. When reporting trends, journalists and researchers must note carryover effects at fiscal‑year start, the retroactive application of ICE’s 2023 counting methodology, and dataset corrections that can add or subtract tens of thousands of removals [3] [1].
7. Bottom line and recommended next steps for readers
Available public sources show a pronounced increase in reported removals in 2025 compared with some earlier releases, but independent data projects and TRAC stress that revisions and methodological issues materially affect totals; therefore, rely on the corrected late‑July datasets and ICE’s locked fiscal‑year reports for final comparisons, and cite those provenance notes when making year‑to‑year claims [1] [2] [4].