Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
Fact check: How do Department of Homeland Security and Census Bureau estimates of the unauthorized immigrant population differ?
Executive Summary
The Department of Homeland Security (DHS) and other prominent groups produce differing totals for the U.S. unauthorized immigrant population because they use different residual and demographic modeling approaches, data inputs, and cut-off dates, which yield divergent counts—DHS reported about 10.99 million on January 1, 2022, while other organizations such as Pew and Migration Policy Institute reported higher figures near or above 13–14 million in subsequent years. These gaps reflect methodological choices about which administrative records to subtract from survey counts, how to treat undercount and recent arrivals, and the particular snapshot date used for the estimate [1] [2] [3].
1. Why DHS numbers sit lower: the government’s conservative residual method, explained
DHS publishes estimates using a residual method that starts with the Census Bureau’s American Community Survey (ACS) count of the foreign-born population and subtracts those legally resident according to administrative records, producing an estimate of the unauthorized population; DHS presents a 10.99 million figure for January 1, 2022, and reports year-to-year series showing a modest decline from some prior years [4] [1]. This approach explicitly relies on administrative data for the legally resident denominator, which can produce lower totals when administrative records undercount certain legal statuses or when the ACS undercounts unauthorized residents; DHS’s method also emphasizes demographic breakdowns by entry period, origin, and state, yielding a conservative, replicable official series used for federal planning [4] [1]. The choice to anchor to administrative legal-status counts and a fixed snapshot date is central to why DHS totals can be lower than some external estimates.
2. Why independent groups often report higher totals: modeling, timing, and different assumptions
Independent analysts such as the Migration Policy Institute and Pew Research Center apply related residual approaches or more expansive demographic models but incorporate different adjustments for undercount, recent arrivals, and administrative gaps, producing higher estimates; for example, Migration Policy’s work estimated roughly 13.7 million unauthorized immigrants by mid‑2023, while Pew published an estimate near 14 million in 2023—both notably above DHS’s January 2022 total [2] [3]. These organizations often update assumptions about trends in migration flows, border encounters, and ACS response rates, and they may blend administrative records with survey-based imputation or demographic trend models. The timing difference—DHS’s 2022 snapshot versus outside groups’ 2023 estimates—magnifies disparities because net migration and undocumented entries rose in that interval, and analytic teams treated recent flows and removals with varying weights [2] [3].
3. How methodological differences translate into millions of people: specifics that matter
Small technical differences translate into large absolute differences because the base population is large; decisions about which administrative categories to remove, how to correct for ACS undercounts, how to classify recent border crossers and parolees, and whether to include long-term visa overstays create systematic divergences. DHS’s transparency about relying on administrative lists produces replicable lower bounds, while external groups sometimes model visa overstays and add corrections for likely undercount—yielding higher totals around 13–14 million [4] [2] [1]. The contrast is not merely academic: it affects public debate, resource allocation, and perceived trends. Analysts note the DHS series showed a drop from 11.6 million in 2010 to lower counts later, while others see an expansion in recent years; these are different stories driven by different measurement choices and snapshots [1].
4. Competing narratives and institutional incentives: read the numbers in context
Different organizations have different institutional aims that can shape methodological emphasis: DHS produces an operationally useful, administratively anchored series for enforcement and planning, while research organizations aim to capture broader demographic realities, including undocumented inflows not yet in administrative systems. These roles lead to distinct methodological priorities—DHS prioritizes alignment with federal records; research groups prioritize demographic completeness and trend sensitivity [4] [2]. Observers should treat DHS numbers as an authoritative administrative baseline and independent estimates as complementary demographic bounds; both are factually defensible but reflect different policy and research priorities, which explains why media and policymakers often cite both to bracket plausible ranges [5] [6].
5. What the differences mean for policy and for interpreting future numbers
Disagreement between DHS and external estimates signals uncertainty that matters for policy design: funding formulas, enforcement resource planning, and legal or humanitarian responses depend heavily on whether the unauthorized population is closer to 11 million or 14 million. Policymakers should consider both the DHS administrative baseline and independent demographic estimates when assessing trends and allocating resources, and analysts should routinely report the date and major assumptions behind each estimate. The public debate will continue to hinge on methodological transparency—DHS’s residual approach provides a conservative, administratively grounded count (10.99 million on Jan 1, 2022), while external groups using adjusted residual or demographic models report higher 2023 estimates (about 13.7–14 million)—and those differences reflect real variation in data treatment, timing, and inclusion criteria rather than simple errors [1] [2] [3].