What is the methodology behind the Homeland Security report on undocumented immigrant costs?
Executive summary
The Homeland Security family of estimates rests on a multi-step “residual” approach that first infers the size and demographic profile of the undocumented population from Census surveys and administrative records, then allocates government spending and tax flows to that population across buckets such as education, health, law enforcement, detention, and local services using agency data and program rules; those allocations depend on many strong assumptions about eligibility, usage, and which individuals are counted as “undocumented” (or part of undocumented households) [1] [2] [3]. Critics and alternative analysts differ sharply over data choices, eligibility rules, and whether to include downstream economic effects, so headline dollar totals vary widely and are shaped as much by methodology and assumptions as by raw data [4] [5] [6].
1. How DHS derives the population: the residual method and survey adjustments
The core methodological step is estimating how many people are undocumented by comparing census and survey counts of foreign‑born residents to administrative immigration records — the so‑called residual method — a technique DHS has used in its population estimates and which other analysts replicate or adapt when constructing fiscal burdens [1]. That procedure typically uses the American Community Survey microdata to construct an economic and demographic profile of the foreign‑born population and then subtracts the legally present population to infer the unauthorized share; DHS and downstream users standardize and publish these inputs through the Office of Homeland Security Statistics (OHSS) [1] [3].
2. Turning population into dollars: categories, data sources, and allocations
Once a population estimate exists, DHS‑style fiscal tallies map that cohort into expenditure and revenue categories — K‑12 education, emergency and Medicaid spending, law enforcement and detention, property damage and local services, and tax contributions — by applying per‑person cost or use rates derived from federal and state agency data, program eligibility rules, claims data, and ad hoc literature estimates [4] [7] [8]. For instance, emergency Medicaid is treated as a distinct line because federal and state rules require limited emergency coverage regardless of immigration status, and CMS figures on emergency services have been cited in congressional summaries [4] [8]. Homeland Security statistics and committee reports supply many of the counts used to scale federal enforcement and detention costs [7] [3].
3. Critical assumptions: who counts as beneficiary, household attribution, and tax treatment
These tallies hinge on disputed assumptions: whether to credit benefits received by U.S.‑born children in mixed‑status households to the undocumented adults in those households, how to treat use of stolen Social Security numbers, whether people with temporary protections or work authorization are counted as “undocumented,” and how to apportion payroll and income taxes paid by unauthorized workers — all of which materially change net fiscal estimates [2] [9] [1]. Analysts differ on whether to exclude lawful residents in mixed households from benefit tallies (a choice Heritage’s approach does explicitly) and on whether undocumented workers’ payroll contributions to Social Security and Medicare should be counted as offsets [9] [1].
4. Alternative methods and contested inputs: think tanks and advocates
Different organizations produce diverging dollar totals by changing inputs: FAIR and some conservative reports scale per‑pupil, per‑patient, and per‑detention costs upward to produce large net burdens, while groups like the American Immigration Council or ITEP widen the lens to include economic contributions, tax payments, or the costs of mass deportation scenarios, producing very different fiscal pictures and emphasizing different policy trade‑offs [5] [10] [6]. Independent critiques — for example from Cato and others — argue that some estimates double‑count costs, rely on out‑of‑date ACS slices, or embed partisan choices about eligibility and scope [11] [12].
5. Bottom line: methodological transparency, assumptions drive totals, and limits in real‑time policy use
The methodological architecture is transparent in outline — residual population estimation, allocation of per‑unit costs from agency data, and aggregation — but the resulting headline sums are extremely sensitive to definitional choices (who is “undocumented,” which services count, time horizons) and to data lags in ACS and administrative sources; congressional users sometimes plug additional assumptions (e.g., costs of released migrants or hypothetical mass deportation scenarios) that extend beyond DHS’s baseline reporting and increase variability [1] [13] [3]. Given those sensitivities, fiscal totals are useful for framing tradeoffs but must be read alongside methodological appendices and alternative studies to understand which assumptions produce which numbers [7] [6].