What methodologies do Pew, DHS and independent researchers use to estimate unauthorized immigrant populations?
Executive summary
Pew, DHS and most independent demographers primarily rely on variants of the "residual" estimation approach—start with survey or census counts of the foreign-born, subtract the estimated legally resident foreign-born, and treat the difference as the unauthorized population—while layering in adjustments for survey omissions, deaths, emigration and legal arrivals [1] [2]. Alternative independent approaches that emphasize administrative outflow measures and adjusted assumptions about undercounts produce markedly higher estimates and attract partisan debate; both families of methods hinge on limited data about departures and visa overstays, which is why transparency about assumptions is central to interpretation [3].
1. The residual method: Pew’s core approach and its logic
Pew’s published estimates use a residual estimation methodology that begins with national survey data (mainly the American Community Survey or CPS) to estimate the total foreign‑born population, then constructs a demographic tally of the legally resident foreign‑born from DHS admission records, mortality, emigration and status adjustments and subtracts that lawful total from the survey total; the resulting "residual" is treated as the unauthorized population [1] [2]. Pew explicitly updates legal resident counts year‑to‑year using DHS data on lawful permanent resident admissions and applies post‑survey adjustments for omissions because censuses and surveys tend to miss people—especially unauthorized immigrants—so final estimates are increased to reflect undercoverage research [4] [2].
2. Survey modifications, weighting and probabilistic assignment
Because survey microdata do not list immigration legality, researchers at Pew and others assign legal status probabilistically to noncitizen respondents based on available characteristics and external administrative counts, reweight surveys to align with intercensal population benchmarks, and compute variances using replicate weights or generalized variance formulas to avoid false precision [5] [6]. Pew’s Methodology B explains how ACS and CPS data are modified and reweighted to be consistent over long time series and to account for known survey inconsistencies, an important step for estimating trends [7].
3. Adjusting for omissions, mortality, emigration and visa overstays
All residual‑method implementations require assumptions about outflows—deaths, voluntary emigration and deportations—and increasingly about visa overstays; the Social Security Administration review and methodological literature note that estimates of outflows heavily influence final totals and that rising visa overstays have grown more salient for measurement [3]. Pew and similar groups explicitly model deaths and departures when updating lawful immigrant counts and then adjust the residual upward to account for undercounts in surveys documented by Census Bureau research [1] [2].
4. Variants and independent alternatives: MPI, CMS and outflow‑heavy estimates
Other independent researchers and institutes use related residual frameworks but differ in data linkages and status assignment—Migration Policy Institute, for example, assigns legal status within ACS using SIPP-linked information to estimate characteristics and local totals [8]. A distinct class of studies departs from the classic residual logic and produces substantially larger estimates by using alternative assumptions about outflows and administrative undercounting; a review in the SSA policy bulletin discusses such a methodology that yielded estimates roughly double residual‑method counts and notes it has been both promoted and criticized by ideologically oriented groups [3].
5. Sources of disagreement and the politics of methodological choices
Differences across estimates trace to choices about (a) which survey to use and how to reweight it, (b) how fully to adjust for omissions, (c) assumptions about emigration and mortality, and (d) treatment of visa overstays and temporary protections—choices that can be numerically consequential and that critics tie to advocacy agendas when researchers emphasize one data source or assumption over another [5] [3]. Pew and DHS variants are widely used in policy and actuarial settings because they are transparent about inputs and adjustments, while outflow‑heavy alternatives generate controversy precisely because small changes in departure assumptions yield large population swings [3] [6].
6. What the methods can—and cannot—deliver
Residual and demographic methods provide defensible, replicable benchmarks grounded in survey and administrative records, but they cannot directly observe legal status for individuals in public surveys and rely on modeled departures and undercount corrections—limitations that researchers acknowledge and document in methodological appendices [1] [9]. Where reporting goes beyond these sources, transparency about counterfactual assumptions is necessary; readers should expect estimates to come with uncertainty bounds and clear statements of the key assumptions driving differences between Pew, DHS and alternative independent estimates [6] [3].