How have 287(g) and Secure Communities enrollment trends varied by county since 2015?
Executive summary
Enrollment in Secure Communities effectively wound down around 2015 when the program was replaced at the federal level, leaving most county-level Secure Communities activity to the historical record through 2015 (data coverage is strongest through that year) [1] [2]. By contrast, 287(g) has remained a patchwork of county-by-county MOAs and local decisions—some counties expanded participation in the years after 2015 while many others resisted or curtailed programs—producing diverging county trends rather than a single national pattern [3] [4].
1. The 2015 inflection point: Secure Communities’ retreat and what the records show
Secure Communities was functionally replaced starting in mid‑2015, a federal pivot that limits meaningful county‑level enrollment growth under that brand after July 2015; academic and FOIA‑based datasets therefore concentrate county fingerprint submission and match records through 2015 [1] [2]. Scholarly reconstructions and government FOIA releases provide a near‑complete universe of fingerprint submissions, matches, and removals under Secure Communities up to 2015, enabling county‑level tallies for that period but leaving a gap for later years under that program name [2]. Multiple law‑center inventories and advocacy fact sheets compiled earlier roll‑out and county participation patterns, but they primarily document pre‑2015 expansion rather than post‑2015 enrollments [5] [6].
2. County divergence under 287(g): local politics, cost, and operational choices
287(g) never operated as a uniform national enrollment scheme; it depends on individual memoranda between ICE and local agencies, so counties have shown divergent trajectories—some, like Prince William and parts of Texas, embraced or expanded 287(g) enforcement while other jurisdictions deliberately avoided it or withdrew because of cost, community pushback, or public‑safety arguments [3] [4]. Reporting and policy reviews document that counties choosing 287(g) often did so to shift immigration enforcement to local hands, while critics argue the program was costly and diverted police resources from public‑safety priorities [3] [4] [7]. Analyses by advocacy groups and think tanks find no uniform safety benefit across counties, which helps explain why the number of active 287(g) agreements has been uneven and politically contingent rather than steadily rising nationwide [7] [8].
3. Patchwork outcomes: what county‑level trends meant on the ground
Where Secure Communities was active through 2015, county fingerprint submissions and ICE matches produced measurable increases in interior removals and detainers during the program’s peak years; the available microdata track county‑year counts of matches and removals through 2015 and show substantial variation driven by county arrest volumes and local policy choices [2]. After the federal pivot in 2015, some local agencies shifted toward 287(g) or other cooperative arrangements, producing counties with rising local deputization and others moving toward sanctuary policies—Northern Virginia offers a clear contrast between Prince William’s 287(g) embrace and Fairfax’s preference for Secure Communities participation earlier, illustrating how neighboring counties can take opposite directions [3]. Local press and advocacy pieces also document high‑profile county cases—Harris County and Orange County among them—that reflect local political climates shaping enrollment decisions [9] [10].
4. Limits of the sources, competing narratives, and what remains unanswered
The strongest empirical county‑level series in the provided reporting ends in 2015 for Secure Communities, and later characterizations rely on case studies, MOA listings, and policy summaries rather than a single national county‑by‑county time series extending to the present [2] [11]. Proponents emphasize that these programs identified hundreds of thousands of potentially removable individuals during their peaks and argue local enrollment restores federal prerogatives [12], while critics underscore costs, uneven targeting, and harms to community trust—arguments reflected in NGO fact sheets, academic critiques, and cost‑benefit summaries [5] [4] [7]. Absent comprehensive post‑2015 county‑level roll‑out data in the provided sources, the clearest, evidence‑backed conclusion is that Secure Communities’ county footprint is primarily a historical pattern through 2015, whereas 287(g) remains a decentralized, politically driven mosaic with county‑specific upticks and withdrawals rather than a unified national trend [1] [2] [4].