What share of ICE removals in 2017–2024 originated from interior arrests versus border apprehensions?
Executive summary
ICE removals shifted strongly from interior-origin cases toward border-origin cases between 2017 and 2024: removals that began with CBP border apprehensions grew to constitute roughly four‑fifths of ICE ERO removals by FY2024, while removals originating from ICE interior arrests declined to the low tens of thousands annually [1] [2]. This reflects agency resource shifts, procedural changes at the border (including Title 42 expulsions), and definitional limits in public data [3] [4] [5].
1. The headline numbers: FY2024 as the pivot year
ICE’s own and independent analyses show FY2024 as illustrative: Migration Policy reports that deportations of noncitizens first arrested at the border reached roughly 224,000 in FY2024 while ICE removals from the interior have been much lower — aggregate interior removals fell to the range of roughly 38,000 per year in FY2021–24 compared with higher levels earlier in the decade [1] [6]. A closely cited breakdown of ICE’s FY2024 ERO removals records about 271,484 removals processed by ERO with roughly 47,732 traced to ICE interior arrests and about 223,752 traced to initial CBP encounters — a split of roughly 18% interior and 82% border [2].
2. The trend since 2017: interior down, border up
In the years after 2017 the composition of removals changed markedly: interior removals that once made up a larger share of ICE activity declined — Migration Policy documents a drop from earlier highs (averages over 100,000–200,000 in prior administrations) to averages of about 81,000 in FY2017–20 and plunging further to roughly 38,000 in FY2021–24 — while removals tied to border apprehensions rose as CBP encounters surged and ICE shifted agents to support border processing and removals [1] [7] [8]. GAO’s review corroborates large year‑to‑year variation and a marked fall in annual removals from 276,122 in 2019 to 81,547 in 2022, underscoring the volatility tied to policy and operational priorities [9].
3. Why the split widened: policy, capacity, and classification
Multiple explanations appear across sources: ICE and DHS recount that ERO was redeployed to assist CBP and process large border flows (including Title 42 expulsions), increasing the share of removals that began at the border [3] [1]. Analysts note that interior enforcement has been deprioritized under certain DHS guidance, and resource constraints plus immigration‑court backlogs mean fewer interior arrests lead quickly to removals [7] [9]. Data construction and classification matter: public datasets track whether the “latest apprehension” was by CBP (border) or ICE (interior), but CBP returns and expulsions processed outside ICE’s formal removal machinery can complicate comparisons, and some ICE dashboards were corrected for historical coding errors [5] [4].
4. Conflicting interpretations and data caveats
Advocacy and policy outlets frame the same numbers differently: the Center for Immigration Studies and FAIR emphasize the decline in interior removals as evidence of lax interior enforcement and highlight the FY2024 numerical split [2] [10], while Migration Policy and Econofact stress that rising border‑origin removals reflect migration flows and operational necessity rather than a simple abdication of interior enforcement [1] [8]. Independent auditors warn ICE’s public reporting understates detained counts and that reporting definitions changed over time, which limits simple year‑to‑year comparisons and complicates a clean “share” calculus across 2017–2024 [9] [4]. Where sources do not provide exhaustive annual tabulations in a consistent format, precise multi‑year percentage averages require direct extraction from ICE/OHSS dashboards or TRA microdata [3] [5].