How do independent trackers like TRAC disaggregate ICE removals and what discrepancies exist with ICE’s official counts?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
Independent trackers such as the Transactional Records Access Clearinghouse (TRAC) disaggregate ICE removals by compiling case‑level records obtained via FOIA and litigation and then classifying each removal by factors such as actual deportation date, apprehension method (e.g., Border Patrol, CAP, 287(g)), and criminal history, producing time series and cross‑tabulations that differ from ICE’s aggregated semi‑monthly and fiscal‑year cumulative tallies [1] [2] [3]. Those methodological differences — plus lags in ICE’s public releases and differing choice of date fields — produce measurable discrepancies between TRAC’s counts and ICE’s official statistics, and fuel contrasting narratives about whether removals have risen or fallen under successive administrations [4] [5] [6].
1. How TRAC builds a disaggregated removals dataset
TRAC’s public “All ICE Removals” tool is based on individual case records it has repeatedly requested from ICE through FOIA and court action, allowing TRAC to tabulate every recorded deportation from FY 2003 through the most recent month ICE supplies and to break those records down by the actual deportation date, agency of apprehension, program affiliation (like CAP or 287(g)), and other labels TRAC derives or maps from ICE text fields [1] [2]. TRAC updates this dataset by submitting monthly FOIA requests and then reprocessing the newly released case‑level files in order to produce time series, tables and dashboards that make it possible to compare administrations and to slice enforcement by criminal history or apprehension method [1] [7].
2. Key points of contrast with ICE’s official figures
ICE’s public reporting uses semi‑monthly detention and removal reports that present removals as a single cumulative fiscal‑year number (fiscal years begin in October), creating a rolling cumulative series that can obscure short‑term changes and depends on the particular cumulative subtraction method analysts use to derive period removals [3]. By contrast, TRAC’s approach uses the actual deportation date recorded in case files — a different date field — which TRAC emphasizes produces a more immediate periodization than ICE’s semi‑monthly cumulative presentation and can therefore yield different counts for the same nominal period [1] [3].
3. Discrepancies documented by TRAC and their magnitude
In its analyses, TRAC has reported that claims of dramatic increases in removals under the Trump administration were not supported by ICE’s published semi‑monthly figures once those cumulative numbers were parsed, finding, for example, that Trump administration removals were effectively near or below Biden‑era daily averages in early 2025 — with statements such as “one percent below” or “10 percent below” appearing in separate TRAC reports that compare derived period removals [4] [3] [5]. TRAC’s accounting across FY 2025 and FY 2026 produced totals (for the period it examined) such as 290,603 removals attributed to the Trump administration, a sum TRAC notes is only modestly higher than the prior year’s total despite reported enforcement expansions [6].
4. Why those discrepancies arise — data fields, timing, and scope
Discrepancies stem from three concrete methodological choices: which date is used (actual deportation date versus ICE’s semi‑monthly cumulative reporting anchor), whether the dataset is derived from aggregated agency tallies or case‑level FOIA releases, and the handling of program labels and apprehension attribution (TRAC’s composite categories for CAP, 287(g), Secure Communities, etc.) that ICE’s summaries may not expose in the same way [1] [2] [3]. TRAC also flags that ICE has at times failed to supply updated FOIA data for months, which complicates real‑time comparisons and can make TRAC’s most recent derived series lag or require interpolation when ICE delays releases [4].
5. What the differences mean for interpreting enforcement trends
Because TRAC’s case‑level disaggregation reveals a growing share of arrests and detentions involving people without criminal convictions, it reframes whether enforcement expansions are primarily criminal‑targeted or focused on immigration violations — a policy‑relevant distinction that aggregated ICE cumulative counts do not easily make visible [8] [5]. Independent monitoring therefore serves as both a methodological cross‑check and a narrative counterweight: TRAC’s granular counts often moderate headline claims based solely on agency press lines or cumulative tallies, while acknowledging limits when ICE withholds or delays raw records [5] [4].