What portion of deportations under recent administrations involved interior arrests by ICE versus border apprehensions by CBP?
Executive summary
Across recent administrations the bulk of U.S. removals have come from people taken into custody at or very near the border or by CBP and later processed for removal, while ICE’s “interior” administrative arrests that begin inside communities account for a substantially smaller share — on the order of a single-digit to low‑20 percent slice of removals in the most recent multi‑year data (roughly 14–15% by one measure) [1] [2].
1. What the datasets say: border-linked removals dominate
Aggregated enforcement reporting and independent analysis show that removals in recent years are overwhelmingly tied to border apprehensions or to people ICE removed after a border arrest, not to ICE-initiated community arrests; Migration Policy notes ICE removals after a border arrest reached 224,000 in FY2024 while ICE removals that originated from interior administrative arrests fell to about 38,000 in FY2021–24, which implies roughly 14–15% of those ICE removals began as interior arrests and the rest were border‑linked [1].
2. The long arc: interior arrests have declined while border removals rose
The decline of ICE-origin interior removals is stark: historically ICE averaged about 155,000 interior removals a year in FY2009–16, dropped to about 81,000 in FY2017–20, and plunged again to roughly 38,000 in FY2021–24 — even as ICE’s removals tied to border apprehensions grew substantially in the most recent four‑year window [1].
3. How agency roles and transfers shape the counts
Part of the reason border‑linked removals dominate is structural: CBP performs border encounters and apprehensions and often transfers people to ICE for detention and formal removal processing, and ICE’s detention population includes “transfers from CBP” as a major component — ICE’s own statistics emphasize that most detained individuals are transfers from CBP following arrests at the border [2] [3] [4].
4. Why simple percentages can be misleading: definitions and gaps
Official totals depend on fiscal‑year definitions, Title 8 vs. Title 42 processing, whether CBP or ICE is counted as the actor carrying out the final removal, and on imperfect residence/“apprehension method” coding; DHS’s monthly tables and methodological notes stress these limits, and researchers warn it is hard to perfectly isolate community arrests from jail/prison transfers in ICE data [5] [4].
5. Competing narratives and political claims
Political messaging and press releases sometimes present very different pictures: DHS and administration statements tout record deportation activity and large numbers removed in recent years, while advocacy groups and researchers focus on the measurable shift away from interior arrests and on state/local limits on ICE cooperation; both types of sources are visible in reporting, and independent analyses flag that public data from recent surges can be incomplete or delayed [6] [7] [8].
6. Bottom line answer to the central question
Using the best available public breakdowns for recent years, most deportations under recent administrations stemmed from border apprehensions by CBP or from ICE removals following border arrests, while ICE‑initiated interior administrative arrests accounted for a minority share — roughly in the mid‑teens percentage range (about 14–15% in one recent multi‑year comparison) of ICE removals in the FY2021–24 period [1] [2]. This conclusion is subject to caveats about reporting definitions, transfers between agencies, and evolving policies that alter where and how arrests occur [5] [4].