What are independent per‑removal cost estimates (Penn Wharton, American Immigration Council) and how do their methodologies differ?
Executive summary
Two independent estimates commonly used in recent debate arrive at very different per‑removal figures because they answer different questions with different tools: the Penn Wharton Budget Model (PWBM) reports an average cost of about $70,236 per deportee by combining prior studies and plugging that figure into a dynamic macroeconomic simulation [1], while the American Immigration Council (AIC) produces operational, scenario‑based cost totals — for example estimating roughly $7 billion per year to carry out one million removal orders plus billions more to expand court capacity — based on microdata and line‑item enforcement costs [2] [3].
1. How each estimate is reported and the headline numbers
PWBM’s brief presents a headline “total cost per deportee” of $70,236 and says that figure comes from averaging estimates from other studies (notably the American Immigration Council and the American Action Forum) and then multiplying that average per‑deportee cost by the projected number of deportees each year to compute annual budget impacts [1]. The AIC’s report frames costs around operational scenarios: it estimates roughly $7 billion per year to remove one million people (operational removal costs such as transportation and flights), and separately estimates it would require about $12.6 billion per year to expand immigration court capacity enough to process that scale of cases — figures embedded in a larger report that models fiscal and economic impacts of removing roughly 11 million people [2] [3] [4].
2. The underlying methodological approaches — dynamic macro vs. micro‑operational accounting
PWBM approaches the question from a macroeconomic and budgetary modeling perspective: it uses a dynamic model and rich microsimulation of demographics to evaluate how deportation policies would affect GDP, wages, and tax revenues over time, and incorporates a per‑deportee cost number by averaging external estimates and multiplying by projected removals to feed into the macro model [5] [1]. AIC builds its estimates bottom‑up from public microdata (American Community Survey among others) and current enforcement cost data to price the stages of deportation — arresting, detaining, processing, and removing — and to model scenario costs and necessary infrastructure expansion [6] [3] [2].
3. What’s included (and excluded) in the cost calculations
PWBM’s use of a single “average cost per deportee” is a simplifying input to a broader macro simulation that emphasizes long‑run GDP and fiscal effects [1] [5], whereas AIC explicitly separates direct operational expenditures — transport, detention, adjudication capacity — from broader economic impacts, and provides scenario detail (e.g., air‑charter flight costs and court expansion needs) to justify its per‑scenario totals [2] [3]. AIC’s report also notes that its fiscal accounting treats detention and enforcement capacity conservatively and models tax‑base effects of removing workers from the economy [3].
4. Why the numbers diverge — key methodological drivers
The divergence stems from three sources: first, PWBM’s $70,236 figure is an average created by combining past studies (including AIC and AAF) and then embedded in a dynamic model that emphasizes indirect macro effects [1] [5]; second, AIC’s scenario costing focuses on immediate, observable operational costs and the infrastructure uplift needed to sustain mass removals, producing per‑removal implications that look much lower on the operational line items but substantially higher when court and systemic expansions are included [2] [3]; third, legacy estimates such as the AAF’s 2015 work and other academic estimates vary widely (for example AAF/2015-derived numbers in later treatments translate to tens of thousands per removal in today’s dollars), and PWBM’s averaging step amplifies those upstream differences [7] [1].
5. Caveats, alternative viewpoints and implicit agendas
Both analyses are transparent about their scope but serve different audiences: PWBM is an academic policy model centered on macro fiscal forecasting and uses averaged per‑deportee inputs to model nationwide GDP and deficit effects [5] [1], while AIC — an advocacy and research organization — foregrounds operational burdens and social‑economic consequences of mass deportation scenarios to argue against such policies [3] [2]; readers should therefore treat the numbers as complementary rather than equivalent and inspect which costs (operational vs. systemic vs. macroeconomic) are being tallied in each exercise [3] [1]. Each methodology has limits in the public record: PWBM’s brief documents the averaging step and macro linkages but does not itself provide a granular line‑item accounting that would match AIC’s operational tables, and AIC’s scenario work necessarily relies on assumptions about scale, detainment practices, and adjudication expansion that shape per‑removal averages [1] [2] [3].