Fact check: What percentage of 2025 US deportations were based on felony convictions?

1. Conflicting Portraits: Authorities Claim Criminals vs. Data Showing No Convictions

Government-aligned reporting and some advocacy outlets advanced the narrative that ICE enforcement in 2025 prioritized individuals with criminal histories, suggesting a focus on those posing public-safety risks; one source summarized that 70% of ICE arrests involved criminal aliens with convictions or pending charges ^[4]. At the same time, multiple analyses and data-driven think tanks report the opposite: a substantial proportion of those detained or arrested had no criminal record, with figures like 65%–71.7% cited for detainees lacking convictions, and only a small share categorized as violent offenders ^{[2] [1]}. These divergent figures underscore that enforcement rhetoric and aggregate detention statistics can tell very different stories.

2. What the supplied sources actually measure—and what they don’t

The materials conflate related but distinct metrics: arrests, detentions, people “held in ICE facilities,” and decisions to deport. Several pieces give counts of arrests or detainee composition without linking those counts to final removal outcomes or separating felony versus misdemeanor convictions. For example, one source lists arrests of individuals with prior records and enumerates offense categories—gang membership, sex offenses, murder, assault—without calculating a percentage of total deportations attributable to felony convictions ^[3]. Another gives the share of detainees with no convictions as of a certain date but stops short of translating that into deportation percentages ^[1]. This methodological mismatch makes a clear, single-percentage estimate impossible from the supplied items.

3. Timing matters: mid‑2025 snapshots versus later summaries

The supplied items come from mid- to late‑2025 and present time-bound snapshots: June and July data on detainee composition and enforcement shifts, and September summaries that highlight trends through late summer and early fall ^{[1] [5] [4]}. Figures like the 71.7% without convictions are explicitly dated (as of June 29, 2025) and cannot be extrapolated to the entire year without more comprehensive, time-series deportation data. Conversely, sources citing enforcement surges or policy changes emphasize procedural shifts that could alter the composition of deportations later in 2025, but they don’t retroactively quantify felony-based removals for the full year ^{[6] [7]}.

4. Disparate definitions: “criminal record,” “conviction,” “violent offense,” and “felony”

The supplied analyses use varying definitions: some track any prior criminal record including arrests and pending charges, while others count only adjudicated convictions or narrow categories like violent offenses. One report notes that 6.9% of detainees with convictions had committed violent crimes, but that metric does not equate to the percent of deportations arising from felony convictions ^[1]. Another labels people as “criminal illegal aliens” based on convictions or pending charges, a broader operationalization that inflates the appearance of felony-based enforcement if readers assume all such entries are felonies ^[4]. These definitional inconsistencies mean that headline percentages across sources are not directly comparable.

5. Policy shifts and prosecutorial discretion that complicate attribution

Several pieces document 2025 policy changes—expanded priorities, workplace raids, and termination of sensitive-locations protections—that increased encounters with individuals regardless of felony status ^[7]. Media accounts of targeted enforcement of DUI arrests and other misdemeanors among Guatemalans illustrate how administrative priorities can capture people with minor offenses and funnel them into removal processes, further blurring the line between misdemeanor- and felony-based deportations ^[5]. Thus, administrative emphasis on “criminals” does not reliably translate to a deportation tally dominated by felony convictions.

6. What would be required to answer the question precisely

A definitive percentage would require consolidated ICE/CBP data that lists, for all removals in 2025, the legal classification of each person’s criminal conviction^[8] (felony vs. misdemeanor), the conviction status at removal, and a clear accounting of removals initiated for criminal convictions versus other grounds. The supplied sources do not provide that combined dataset; they offer partial slices—arrests, detention composition, and selected offense tallies—without linking convictions to final removal actions ^{[3] [1] [2]}.

7. Bottom line: current evidence does not support a single percentage claim

Based on the supplied reporting and analyses, it is not possible to state a reliable percentage of 2025 U.S. deportations that were based on felony convictions; sources are internally inconsistent, measure different populations (arrests vs. detentions vs. removals), and use varying definitions of criminality. The most defensible conclusion from these items is that a large portion of people in ICE custody in mid‑2025 had no conviction, while other accounts portray many arrests involving prior records—but none quantify felony-based removals for the full year ^{[1] [2] [4]}.