How do independent groups like MPI, Pew, and ACLU calculate deportation totals and where do their methods diverge?
Executive summary
Independent analysts calculate “deportation” totals using different inputs, definitions and counting rules: the ACLU’s Deportation Data Project analyzes ICE administrative records obtained via FOIA and highlights expedited/nonjudicial removals [1] [2], the Migration Policy Institute (MPI) combines government enforcement data with policy-context analysis and emphasizes shifts from courtroom removals to nonjudicial expulsions [3], and Pew Research Center treats deportation counts within broader survey-based estimates and methodological adjustments to its unauthorized-population estimates [4] [5].
1. What each group measures and why it matters
The ACLU’s work focuses on operational enforcement events recorded by ICE—including arrests, detentions, transports and deportations—using raw government data it obtained via FOIA and curates on the Deportation Data Project for long-term analysis [1] [2], and in investigative reports (like American Exile) it examines expedited removals that bypass immigration-court hearings [6]. MPI’s reporting and analysis spotlight the changing character of removals—documenting that an increasing share are nonjudicial and therefore don’t involve immigration-court hearings—framing deportation totals in light of policy shifts and programmatic expansion [3]. Pew, by contrast, anchors deportation discussion to its population-estimation work and public-opinion surveys, situating deportation figures in the context of estimates of unauthorized immigrants and how many have temporary protections [4] [5].
2. Data sources and transparency: government records, FOIA, and surveys
ACLU’s advantage is direct access to ICE administrative records released in response to FOIA requests and then analyzed and published, enabling event-level counts of enforcement actions over defined windows [1] [2]. MPI typically relies on DHS/ICE statistics supplemented by its own parsing and policy analysis to show trends such as the rise of expedited removals from a historical baseline [3]. Pew’s approach relies on nationally representative survey panels and demographic estimation methods—its deportation-related analysis is tied to ATP survey waves and to methodological appendices explaining sampling and weighting [5] [4]. Each source therefore trades off different strengths: granular operational counts (ACLU), policy-context trend interpretation (MPI), and population-representativeness and public-opinion framing (Pew) [1] [3] [5].
3. Key methodological divergences: definitions, inclusion rules, and units of analysis
Groups diverge on what counts as a “deportation.” ACLU tallies ICE-recorded removals and highlights categories like expedited removal that do not entail a judge [1] [6]. MPI emphasizes the distinction between judicial removals and nonjudicial expulsions and quantifies the shift toward nonjudicial processes—reporting that in recent periods about three quarters of removals may occur without a judge [3]. Pew’s work does not produce operational deportation logs but frames deportation figures within its unauthorized-population estimates (which include people under temporary protection) and the survey measures it fields on public views of deportations [4] [5]. These choices produce materially different totals: counting only formal judicial removals will understate actions recorded administratively as expedited or administrative removals, while survey-based framing ties totals to population denominators rather than enforcement events [3] [1] [4].
4. Biases, agendas and interpretive lenses
Each organization carries an implicit analytic lens: the ACLU’s FOIA-driven datasets and reports foreground civil‑liberties concerns about speed and lack of judicial review in removals [6]; MPI’s policy analyses stress systemic shifts and implications for enforcement capacity and state-local partnerships [3] [7]; Pew centers public-opinion context and representative estimates, which can de-emphasize raw operational counts in favor of population-based interpretation [5] [4]. All three disclose their methods in at least summary form (ACLU via project notes and FOIA provenance [1], MPI through reports documenting nonjudicial trends [3], Pew via detailed methodology pages [5]), but differing priorities produce divergent emphases and headline numbers.
5. What can't be settled from the reporting provided
The sources document methodological differences and data origins, but they do not provide a single reconciled algorithm for translating ICE encounter records into a universally accepted deportation total; reconciling would require explicit harmonization choices about returns vs removals, double‑counting across encounters, and treatment of state/local actions—issues noted in MPI’s and ACLU’s analyses but not reduced here to a single crosswalk in the available material [3] [1] [2].