How many arrests of noncitizens for sexual assault and homicide have ICE and DHS reported annually since 2020?
Executive summary
ICE and DHS reporting shows thousands of sexual‑assault and homicide‑related charges or convictions tied to noncitizens in recent years, but the agency’s publications count “associated charges/convictions” rather than a simple tally of arrests solely for sexual assault or homicide — and the available public reports provide clear year-by-year figures for FY2022–FY2024 but not for FY2020–FY1 (reporting gaps and shifting definitions complicate simple comparisons) [1] [2] [3].
1. FY2024: the largest aggregate figure released, but it reads as associated charges not single‑crime arrests
ICE’s Fiscal Year 2024 materials state that 81,312 noncitizens with criminal histories were arrested and that the combined set of charges/convictions associated with those arrests included 21,379 sexual assault and sex‑offense entries and 4,085 homicide entries — language that the report frames as counts of charges or convictions associated with the arrested population, not necessarily the number of arrests whose primary offense was sexual assault or homicide [1] [3].
2. FY2023: ICE’s press releases give lower per‑category counts tied to that year’s arrests
ICE’s FY2023 summaries and news releases report that ERO arrested 73,822 noncitizens with criminal histories in FY2023 and that the associated roster of charges/convictions for that cohort included 4,390 sex and sexual‑assault charges and 1,713 homicide charges or convictions — again presented as associated charges/convictions among the arrested population [4] [5].
3. FY2022: the agency’s published removals data show larger sex/homicide counts in associated charge tables
ICE’s FY2022 reporting and statistical appendices list removed noncitizens’ associated charges and convictions and show figures that include 7,370 sex offenses/sexual‑assault entries and roughly 1,315 homicide‑related entries as part of the total 183,251 charges/convictions tied to removed noncitizens — these are presented at the charge/conviction level, not as a headcount of single‑offense arrests [2].
4. FY2020–FY2021: public reporting gaps and figure ambiguity
The materials available in the provided reporting include FY2020–FY2022 charts and references but do not supply explicit, standalone annual tallies for sexual‑assault and homicide arrests for FY2020 and FY2021 in the snippets provided here; ICE’s public statistics pages indicate broader trends and categories but the specific per‑year counts for those two fiscal years are not present in the supplied sources, so they cannot be asserted from these materials [2] [6].
5. Why the numbers are not a straightforward “arrests for sexual assault/homicide” count
ICE and DHS commonly report the number of charges or convictions associated with arrested or removed noncitizens (often multiple charges per person), and they publish both aggregate docket counts and year‑specific arrest cohorts; that means figures such as “21,379 sexual assault charges” in FY2024 describe associated charges within a larger arrested population rather than a clean count of arrests whose sole or primary offense was sexual assault, and DHS/ICE emphasize that their data mix convictions, pending charges and historical records [1] [3] [6].
6. Bottom line — what can be stated with confidence from the supplied reporting
From the provided ICE/DHS documents and releases, the most direct, attributable annual figures available are: FY2022 — about 7,370 sex‑offense entries and ~1,315 homicide‑related entries (charges/convictions associated with removed noncitizens) [2]; FY2023 — 4,390 sex/sexual‑assault entries and 1,713 homicide entries associated with arrested noncitizens [4] [5]; FY2024 — 21,379 sexual‑assault/sex‑offense entries and 4,085 homicide entries among the combined charges/convictions tied to arrested noncitizens [1] [3]. The dataset’s construction (charges/convictions per arrested person, multiple counts per individual, and mixture of convictions versus pending charges) and the absence of explicit FY2020–FY2021 per‑category counts in the supplied sources prevent a simple headcount answer for calendar years 2020–2021 from these materials alone [1] [2] [3].