What data sources back the claim that 70% of ICE arrests are of noncitizens charged or convicted of crimes?

1. What DHS/ICE are citing when they say “70%”

The 70% claim appears in multiple DHS press releases touting recent enforcement actions; those releases state “70% of ICE arrests are of criminal illegal aliens charged or convicted of a crime in the U.S.” but do not include a detailed, linked breakdown in the releases themselves ^{[1] [2]}. Available sources do not provide the exact internal table or query DHS used to generate that headline percentage within those press releases; the claim is presented as an agency summary ^{[1] [2]}.

2. How ICE publishes arrest and detention categories

ICE's Enforcement and Removal Operations (ERO) statistics explicitly categorize arrests and detained populations by criminal history buckets that include: (a) people with convictions or pending charges; (b) immigration violators such as re‑entrants and fugitives; and (c) those without known criminal convictions or pending charges. ICE’s public statistics and dashboards are the primary raw source for counting who falls into each category ^[3].

3. Independent trackers and journalists see different mixes

The Guardian, using ICE’s biweekly detention releases, and TRAC’s aggregation work both show that at times the number of people ICE holds without criminal convictions is large—TRAC reported that 47,964 of 65,135 people in ICE detention (73.6%) had no criminal conviction as of Nov. 16, 2025, and The Guardian reported more than 21,000 people with no criminal record were arrested and detained during a recent period ^{[6] [7] [4]}. These independent counts rely on ICE’s published spreadsheets and archive efforts rather than DHS press summaries ^{[4] [5]}.

4. Why the same data can produce different headline percentages

Analysts warn ICE data can be sliced in different ways: counts can be limited to “arrests that result in ICE detention” versus all enforcement encounters; tables can be filtered by arresting agency (CBP vs ICE ERO); and the definition of “criminal” can vary (conviction vs pending charge vs removable offense). The Deportation Data Project and journalists explain that aggregating and cleaning ICE’s raw arrest files affects the resulting percentages ^{[5] [8]}. ICE itself documents that some categories include non‑conviction immigration violators such as re‑entrants and people with final orders ^[3].

5. Recent reporting highlighting conflicts with the 70% claim

News outlets and analysts have pointed to specific operations where the proportion of people with serious criminal histories was much smaller than agency statements suggested—for example, reporting on a Chicago list of 614 arrests where only 16 were flagged as high public‑safety risks (2.6%) and other coverage noting tens of thousands arrested without convictions during a shutdown period ^{[9] [7]}. These pieces use ICE's posted lists and biweekly spreadsheets to question the sweeping 70% framing.

6. What independent data resources exist to verify the claim

Primary public resources to check the 70% number are ICE’s Enforcement and Removal Operations statistics and downloadable spreadsheets, the Deportation Data Project’s processed arrest files, TRAC’s detention snapshots, and biweekly ICE releases archived and analyzed by outlets like The Guardian ^{[3] [5] [6] [4]}. However, researchers note gaps: some ICE dashboards have not been kept up to date, and different datasets (arrests vs detained population) yield different percentages ^{[10] [8]}.

7. Bottom line for readers

DHS/ICE press releases repeatedly state a 70% figure ^{[1] [2]}, but independent analyses of ICE’s published arrest/detention tables show substantial shares of people in ICE custody or arrested who do not have criminal convictions, and different reasonable methods of slicing ICE data produce very different headline percentages ^{[6] [7] [4] [5]}. For a verifiable reconciliation, review the specific ICE spreadsheet^[11] and filters the agency used to compute 70%—those exact queries are not included in the cited press releases, and independent trackers urge careful, transparent methods when translating ICE’s raw data into summary claims ^{[1] [2] [5] [8]}.