How do DHS and ICE report deportations and what methodological changes have occurred since 1993?
Executive summary
Since the 1990s the Department of Homeland Security and its component immigration-and-customs-enforcement">Immigration and Customs Enforcement have moved from patchwork counting toward a database-driven, component-aggregated reporting system, but key methodological shifts—most notably the inclusion of Border Patrol/CBP “border” cases and the centralization of records in DHS systems—have changed what counts as a “deportation,” complicating comparisons over time [1] [2] [3]. Independent reviewers and audits say DHS agencies now publish more standardized tables drawn from enterprise systems but still make opaque exclusions and late-year adjustments that affect headline removal totals [2] [4] [5].
1. How DHS and ICE currently compile deportation counts
DHS reporting is built on integrated administrative systems: ICE’s operational removals and returns come from the Enforcement Integrated Database (EID) and are extracted via the ICE Integrated Decision Support (IIDS) System/Data Mart (DSSDM), then consolidated in DHS reporting products such as the OHSS Persist dataset and other KHSM tables [2] [4] [6]. ICE’s public statistics and DHS monthly tables reflect data that are mature to varying degrees: ICE publishes statistics with a one-quarter lag and DHS “locks” fiscal‑year data at year‑end after operational data‑quality checks, meaning earlier releases can change as records “mature” [7] [4]. Independent projects, including the Deportation Data Project, complement official releases by publishing FOIA-obtained ICE datasets and documenting event‑level encounters, arrests, transports, and deportations for external analysis [8] [9].
2. The single biggest methodological change: counting border apprehensions
A crucial shift that inflates certain past‑and‑present removal tallies is that DHS began incorporating Customs and Border Protection apprehensions and other “border” dispositions into aggregated deportation reporting, a change traceable to the post‑1990s era and explicitly noted as expanding counts during the Bush and Obama years—making cross‑era comparisons misleading unless one accounts for which components’ actions are included [1]. Migration Policy and others document that CBP removals today represent a large share of DHS deportations and that ICE’s proportionate role has varied with policy emphasis—so shifts in enforcement focus (interior vs. border) change reported totals without necessarily reflecting vastly different ICE operational performance [3].
3. Data systems, exclusions, and adjustments that shape the numbers
The technical architecture—EID, DSSDM/IIDS, UIP and BorderStat snapshots—creates a standardized pipeline but also introduces definitional choices: OHSS states it only removes administrative records “ICE directs us to” to preserve consistency, and GAO found ICE excludes individuals first booked into certain temporary facilities from some detention counts, exclusions that amount to tens of thousands of events and are not fully explained in annual reporting [2] [4] [5]. DHS also allows quarterly revisions and a fiscal-year “lockdown,” meaning monthly public tallies can be revised later, and ICE notes that reported statistics are subject to future corrections [7] [4].
4. Oversight, critique, and competing narratives
Government auditors, researchers, and advocacy groups tell competing stories: GAO recommended ICE publicly report fuller detention and methodology explanations because current presentation omits key exclusions and rationales, while DHS has pushed back saying existing reports are sufficient [5]. Independent trackers such as TRAC, the Deportation Data Project, and academic explainers fill gaps—publishing event‑level data from FOIA releases, documenting fingerprint‑matching programs like Secure Communities, and analyzing shifts in interior vs. border removals—yet all warn that interpreting trends requires careful attention to changing counting rules and policy priorities [10] [9] [3].
5. What changed since 1993—and what remains uncertain
Available reporting points to two durable changes since the 1990s: migration from fragmented paper and agency‑level tallies to centralized electronic records and standardized DHS statistical products (EID/IIDS/DSSDM and OHSS tables), and methodological expansion to include border agency apprehensions and other dispositions that were not counted the same way before, producing higher, but less comparable, removal totals over time [2] [4] [1] [3]. Precise, documented step‑by‑step changes beginning exactly in 1993 are not fully traceable in the provided sources, so definitive claims about every procedural change in that specific year cannot be made from this record; auditors and FOIA datasets remain the best path to reconstruct detailed year‑by‑year methodology shifts [5] [8] [9].