Approximately, How many "illegal" (according to DHS) immigrants are there?
Executive summary
The U.S. Department of Homeland Security (DHS) — through its Office of Homeland Security Statistics (OHSS) — has produced the government’s standard series of estimates of the unauthorized (often called “illegal”) immigrant population through 2022; independent, peer organizations that use similar residual methods put the contemporary range at roughly 11 million to as high as 14 million, depending on date and method (DHS OHSS; Pew; CMS) [1] [2] [3]. The best short answer based on DHS-linked reporting: roughly 11 million unauthorized immigrants is the central DHS-aligned figure cited by analysts, with a clear caveat that other rigorous estimates and partisan tallies diverge substantially [4] [2] [3].
1. What “illegal” means in the statistics and what DHS actually counts
DHS and leading researchers typically use the neutral technical term “unauthorized immigrant” (also called undocumented or illegal) and estimate that population with a “residual” method that compares census and survey counts of the foreign-born to administrative counts of lawful entrants and residents — a method DHS’s OHSS has published annually through 2022 and that underpins many downstream estimates [1] [5]. That methodology produces a stock estimate of people living in the U.S. without legal status; it does not try to count short-term “encounters,” “gotaways,” or single-year border crossings, which are often conflated with population size in political reporting [5] [3].
2. DHS’s baseline number and how analysts cite it
Analysts who rely on DHS data or DHS-style residual methods commonly reference an estimate near 11 million unauthorized residents as the baseline for recent years; for example, an EPI summary references a DHS estimate of about 11.0 million and CMS provisional estimates put the figure in the 11–12 million neighborhood for 2023 [4] [3]. DHS’s OHSS remains the federal statistical series used as the historical backbone for other forecasts and budget projections, which is why the “about 11 million” figure is often treated as the government-aligned central estimate [1] [6].
3. Why other reputable estimates deviate — method, timing, and scope
Reputable organizations produce different numbers because of timing, data sources, and methodological choices: Pew’s residual-method update reported a record 14 million for 2023 using administrative additions through late 2024 and their application of the residual approach; the Center for Migration Studies produced provisional estimates near 11.7 million for mid‑2023; and Migration Policy and CBO rely on DHS OHSS historical series for projections but may model different flows and adjustments, producing variation [2] [3] [6]. The differences matter: some estimates emphasize recent border releases and parolees counted in DHS administrative data, others emphasize undercount adjustments in survey data, and still others model “gotaways” and overstays differently, which can inflate or lower totals [2] [5] [7].
4. Political and advocacy estimates — why they can be outliers
Non‑governmental actors frequently publish much higher or lower numbers for political purposes: FAIR’s 2025 update claims an 18.6 million “illegal” population using DHS enforcement tallies and extrapolations, a figure far above academic and DHS-linked estimates and reflecting a policy-driven agenda to emphasize larger inflows [7]. Conversely, some analyses and official messaging that emphasize enforcement activity or short-term declines in crossings may understate the resident unauthorized stock because removals and encounters are not a one‑for‑one proxy for population change [8] [3].
5. Final assessment and the honest uncertainty
The most defensible DHS‑aligned answer is: about 11 million unauthorized immigrants is the central estimate used by DHS-linked analyses for recent years, but peer researchers using similar residual techniques report a range from roughly 11 to 14 million for 2023–2024, and some partisan estimates put the number significantly higher or lower depending on which flows they count or extrapolate [4] [3] [2]. Measurement is inherently uncertain: the residual method is transparent and widely used, but small differences in assumptions, time cutoffs, and adjustments for survey undercount produce materially different headline numbers that drive political debate more than statistical consensus [5] [2].