Fact check: How many removals did the Trump administration record each year 2017 2018 2019 2020?

1. What the sources actually claim — conflicting headline numbers that demand attention

The collection of source analyses contains several distinct headline claims about removals by year. One ICE document is cited as reporting 226,119 removals for 2017 while later ICE ERO materials report 256,085 removals for 2018 ^{[1] [4]}. Another ICE-derived summary lists 267,258 removals in 2019 and 185,884 in 2020 ^[2]. By contrast, DHS “Table 39” and an Annual Flow-style compilation present larger totals: one entry lists 288,093 removals in 2017 and 337,287 in 2018, and another lists 348,468 removals in 2019 ^{[3] [5]}. These claims are mutually incompatible when read as a single, continuous time series, which indicates underlying methodological or definitional differences across the reports rather than straightforward reporting errors ^[4].

2. ICE ERO figures: an internal series focused on administrative removals and ERO activity

The ICE Enforcement and Removal Operations reports cited in the analyses yield a coherent internal series for ERO activity: 226,119 in 2017 (an ICE FY2017 report), 256,085 in 2018, 267,258 in 2019, and 185,884 in 2020 ^{[1] [4] [2]}. Those ICE documents emphasize enforcement actions, administrative arrests, detainers, and criminality breakdowns; one 2018 summary highlights a 13% rise in removals from ICE’s perspective and an increase in removals of convicted criminals ^[4]. When the provided analyses attribute numbers directly to ICE ERO reports, they consistently reflect that agency’s operational removals rather than broader DHS aggregates, making this series the most consistent representation of ICE’s own removal tallies within the supplied material ^{[1] [2]}.

3. DHS Table 39 and the Yearbook-style totals: larger aggregates that mix returns and removals

Separate DHS compilations cited in the materials produce larger totals that do not match ICE’s ERO-only counts. One source analyzing Table 39 reports 288,093 removals in 2017 and 337,287 in 2018, and a different DHS-style summary lists 348,468 removals in 2019 while also reporting “returns” separately ^{[3] [5]}. These Yearbook-style tables historically combine removals executed by multiple components of DHS (ICE, CBP) and sometimes count “returns” (voluntary or assisted departures) separately from formal removals; therefore the DHS totals can be substantially higher than ICE-only removal counts and are not direct one-to-one equivalents of ICE ERO numbers ^{[3] [5]}.

4. Why the figures diverge — definitions, jurisdictions, and counting rules matter

The analyses suggest, and the differing series imply, that methodological differences drive the discrepancies: ICE ERO reports reflect ICE administrative removals and operational arrests, while DHS Yearbook/Table 39 aggregates removals across DHS components and may include different categories such as voluntary returns or removals issued by CBP or legacy INS ^{[4] [3]}. One analysis explicitly notes the distinction between “removals” and “returns,” and another highlights that a large share of removals concerned non-criminal or non-violent convictions under expanded enforcement priorities, further complicating apples-to-apples comparisons ^{[6] [5]}. The presence of multiple reporting units and varied definitions means any single-year figure must be tied to the specific dataset and counting rules used to compile it ^{[1] [3]}.

5. Bottom line: the best answer from the provided materials and a recommended interpretation

Based solely on the analyses supplied, the most defensible ICE-focused annual series is: 2017 — 226,119; 2018 — 256,085; 2019 — 267,258; 2020 — 185,884 ^{[1] [4] [2]}. If the question seeks the broader DHS or Yearbook totals that combine removals and returns across agencies, the provided materials instead point to higher figures such as 2017 ≈ 288,093 and 2018 ≈ 337,287, with 2019 totals reported up to 348,468, but those numbers are drawn from different tables and are not ICE ERO-specific ^{[3] [5]}. Readers must therefore specify which metric they want — ICE ERO removals or DHS-wide “removed or returned” totals — because both sets of numbers appear in the supplied sources and they are not interchangeable ^{[1] [3]}.