Has crime gone down in places where ICE has been deployed in the last two years?
Executive summary
Official federal statements tie a measurable decline in certain violent crimes in the first half of 2025 to intensified ICE operations, but independent reporting and data analyses present mixed evidence and serious caveats about causation, who was actually detained, and whether local crime drops align with ICE deployment patterns [1] [2] [3].
1. What the DHS claim actually says and the evidence it cites
The Department of Homeland Security publicized a report—citing a Council on Criminal Justice analysis—claiming homicide fell 17% and that gun assaults, aggravated assault, sexual assault and carjacking dropped between about 10% and 24% from January–June 2025 versus the year prior, and DHS credited ICE removals for those declines [1].
2. Arrest volumes and where ICE was active
Independent trackers and academic teams documented a surge in ICE interior arrests after the 2024 election and into 2025, with state-by-state variation and policy shifts producing sharp increases in community raids and jail-based arrests that amplified overall book-ins [4] [5]. ICE’s own dashboards and public datasets show expanded arrest and detention activity through 2024 and into 2025, while public projects archive these releases for analysis [6] [7].
3. Who was being detained — and why that matters for public-safety claims
Multiple analyses show that a large share of those ICE arrested in 2025 did not have violent criminal convictions: contemporaneous reporting found that fewer than half of ICE arrests under the Trump administration through mid-2025 were of people with prior convictions, and some datasets indicate that a substantial portion of increased arrests involved people without convictions or only immigration-related infractions [3] [8] [2]. Other analyses put the share with violent convictions much lower — for example, an independent fact-check noted roughly 29% of detainees in 2025 had convictions and about 5% had violent convictions in one dataset — complicating the proposition that removals targeted “the worst of the worst” [9] [10].
4. Correlation is not causation — limitations in attributing crime declines to ICE actions
While DHS links short-term crime declines to ICE activity, analysts caution that crime rates fluctuate for many reasons, and the available sources show mixed evidence rather than a direct, localized causal chain that connects specific arrests to measured reductions in crime where ICE was deployed [1] [2]. UCLA and other researchers emphasize state and regional variation in ICE arrests, which undermines a one-size-fits-all causal narrative and signals that local context matters [5].
5. Independent and investigative reporting that tempers the official line
Journalistic and advocacy outlets reported that despite record detentions, oversight dropped, facilities were strained, and deaths in custody spiked in 2025 — all facts that complicate a simple law-and-order success story and raise questions about priorities and human costs even amid enforcement surges [11] [12]. Coverage summarizing mixed evidence noted DHS’s crime figures but highlighted that many detainees lacked criminal convictions, which weakens claims that crackdowns directly produced public-safety gains [2].
6. Bottom line: did crime go down in places ICE was deployed?
The data show that certain violent-crime measures fell nationally in early 2025 and DHS connected that trend to ICE removals, but independent analyses and datasets reveal mixed and regionally variable evidence, a substantial share of ICE arrests involving people without violent convictions, and methodological limits that prevent confidently declaring causation between ICE deployment and local crime declines [1] [2] [3] [5]. The reporting reviewed does not provide clear, localized studies tying deployment events to sustained crime reductions where ICE operated, and it documents significant debate about enforcement priorities and oversight [7] [12].