How does ICE define “criminal” in its arrest statistics, and how do pending charges factor in?
Executive summary
ICE’s public reporting and agency messaging treat “criminal” in multiple, overlapping ways: sometimes meaning people with criminal convictions, sometimes including people with pending criminal charges, and sometimes grouping civil immigration violators (like re‑entry after removal) with criminal categories for enforcement purposes [1] [2] [3]. Outside analysts who parsed ICE data separately from DHS messaging have found sharply different shares of detainees with convictions versus no convictions or pending charges, highlighting that the term “criminal” in ICE statements often conflates distinct legal statuses [4] [5] [2].
1. How ICE’s public categories are structured: convictions, pending charges, and immigration violations
ICE’s statistics explicitly distinguish people with criminal convictions, people with pending criminal charges, and people with no convictions or pending charges but who are held for immigration violations such as unlawful reentry or final orders of removal, meaning the agency tracks multiple legal states even as public statements often collapse them into “criminal” [1] [4].
2. Where the confusion starts: agency messaging versus raw data
DHS and ICE public statements and press releases commonly describe arrests as targeting “criminal illegal aliens” or the “worst of the worst,” language that the agency has used while asserting high percentages of arrests are of criminals — claims that have been contradicted by independent analyses showing large shares without convictions [6] [4] [5]. Independent reviews of ICE datasets, including work cited by FactCheck.org and Roll Call, found the proportion of people in ICE custody with no convictions or pending charges rose substantially, underlining a gap between rhetoric and the underlying categories [4] [5].
3. Pending charges: counted as “criminal” by some, not equivalent to conviction
Advocates and researchers note that DHS and ICE often include individuals with pending criminal charges in their “criminal” tallies even though pending charges are not convictions and are frequently dropped or dismissed; critics argue this practice inflates the appearance that ICE is detaining principally convicted violent offenders [2] [7]. Cato and other analyses documented that many people counted as “criminal” in enforcement messaging had only pending charges or minor offenses, while only a small share had violent convictions [2] [7].
4. Administrative arrests and legal nuance: civil vs. criminal enforcement
Much interior ICE enforcement occurs through administrative (civil) arrests for immigration violations rather than criminal arrests; the Congressional Research Service and reporting explain that administrative arrests do not convert immigration violations into federal crimes, though ICE can also make criminal arrests under specific statutory circumstances (e.g., warrantless arrests where a felony is believed to have been committed) [3]. Thus an ICE “arrest” in a dataset can be an administrative detention based on immigration status even if the person has no U.S. criminal conviction.
5. Independent data shows a different picture of who is detained
Multiple independent trackers and advocacy groups reported that a majority of people in ICE detention in recent periods had no criminal convictions — figures ranging from roughly two‑thirds to over 70 percent depending on the dataset and date — and that the detained population’s growth came largely from those without convictions or charges [8] [9] [10] [11]. These findings are consistent with analyses that separate convictions from pending charges, rather than combining them under a single “criminal” label [4] [5].
6. Bottom line and limits of the public record
“Criminal” in ICE arrest statistics is not a single, legally precise label: ICE’s own categories distinguish convictions, pending charges, and immigration‑only violations, but DHS/ICE public messaging and some press releases often treat pending charges and immigration offenses as part of a broad “criminal” category — a practice that inflates the perceived share of convicted violent offenders in custody [1] [2] [6]. Available reporting and dataset analyses make clear the disagreement between agency claims and disaggregated data, but public transparency remains limited: where datasets separate convictions from charges, independent analysts find far fewer convicted violent offenders than agency rhetoric suggests [4] [5]. Where source material does not resolve specific classification questions about particular arrests, that ambiguity should be acknowledged rather than assumed away [3].