How many people died in ICE custody each fiscal year from 2003 to 2019 according to ICE and independent tallies?
Executive summary
A definitive, year-by-year tally from Fiscal Year (FY) 2003 through FY2019 is not available in a single, easily citable table in the public record supplied here; ICE maintains FOIA lists and began routine online “Detainee Death Reports” in 2018, and independent news organizations and researchers have compiled overlapping counts that sometimes differ from ICE’s raw listings [1] [2] [3]. Independent tallies cited in reporting put the cumulative total of deaths since the formation of DHS (Oct 2003) at 188 through early 2019, and contemporaneous reporting and ICE releases identify FY2018 as 6 deaths and FY2019 as 8 deaths [4] [5].
1. What the official ICE records show and their limits
ICE has published FOIA lists enumerating individual detainee deaths dating back to October 2003, including a document covering 10/01/2003–06/05/2017 and another covering October 2003–December 2013, and since FY2018 the agency has posted individual Detainee Death Reports pursuant to congressional requirements [1] [6] [2]. Those ICE documents are primary sources of names, dates and circumstances but are distributed as incident-by-incident reports rather than a clean year-by-year summary in the materials provided here, so constructing an exact FY-by-FY table requires aggregating those incident lists [1] [2].
2. What independent tallies reported through early 2019 show
An NBC News analysis cited 188 detainee deaths since 2003 as of a January 2019 report, and that same article noted that 22 of those deaths occurred in the two most recent years covered by the piece (which NBC framed as the first two years of the Trump administration) — an independent cumulative tally that reporters used to contextualize ICE’s record [4]. That 188 figure is an independent media aggregation built from ICE lists, advocacy group releases and reporting rather than a single ICE-produced FY-by-FY table [4].
3. FY2018–FY2019: where reporting and ICE data converge
Because of the 2018 congressional requirement and ICE’s subsequent policy to post Detainee Death Reports, FY2018 onward is better documented: reporting cites six deaths in FY2018 and eight deaths in FY2019, counts drawn from ICE releases and contemporaneous news coverage [5] [2]. Independent researchers and journalists cross-referenced ICE’s death reports and media/advocacy notices to confirm these numbers, and multiple outlets reported the FY2019 total as eight deaths [5].
4. Why exact FY 2003–2017 counts are not restated here as year-by-year numbers
The available ICE FOIA lists (covering 2003–2017 and overlapping ranges) are cited sources for the raw incidents, but the documents in the reporting set are not presented here as a pre-aggregated FY-by-FY table; creating a guaranteed-accurate per‑fiscal‑year breakdown from 2003–2017 would require programmatic aggregation of those FOIA lists or reliance on an independent dataset such as TRAC or peer-reviewed compilations, neither of which provides a year-by‑year breakdown in the supplied snippets [1] [6] [7]. Scholarly reviews of ICE death reporting since FY2018 have noted additional gaps — for example, some people hospitalized while in ICE custody were released and later died, and those cases may not always appear in the official “Detainee Death Reports” depending on timing and classification — which complicates clean counts [3].
5. Bottom line and how to get a precise FY-by-FY table
The best-supported specific claims in the provided reporting are the independent cumulative total of 188 deaths from 2003 through early 2019 reported by NBC and the year counts of six deaths in FY2018 and eight deaths in FY2019 reported in ICE releases and contemporaneous coverage [4] [5] [2]. For a rigorously auditable FY‑by‑FY table from 2003–2019, one must aggregate ICE’s FOIA death lists (10/2003–2017 and related PDFs) and cross‑check with independent tallies such as media compilations and TRAC; those FOIA lists are the authoritative raw material cited above [1] [6] [7]. Where independent analyses exist they sometimes yield slightly different totals because of classification differences and deaths that occur shortly after release from custody [3] [4].