What estimates exist for non-citizen voter registrations in 2020 and 2024?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
Estimates of non‑citizen voter registrations for 2020 and 2024 diverge dramatically across studies and official checks: academic survey-based extrapolations put the number in the low‑to‑mid hundreds of thousands, advocacy pieces have claimed millions, and federal checking has produced only a few thousand referrals from tens of millions of records examined — all of which reflects widely different methods and significant uncertainty [1] [2] [3].
1. The academic/survey estimates: small but nonzero — roughly 100k–200k
Some researchers using the Cooperative Election Study (CES) and other survey instruments estimated that about 1% of non‑citizens report being registered to vote, which, when scaled to national non‑citizen population estimates, produces figures on the order of roughly 100,000–200,000 registrations (the BBC cited an author’s 1% figure that would equal about 117,000) — an estimate that was applied to discussions of the 2020 period and used cautiously for 2024 extrapolations [1].
2. The high estimates from advocacy/interest groups: millions claimed
By contrast, the conservative group Just Facts published a markedly different range for 2024 — claiming roughly 10% to 27% of non‑citizen adults are illegally registered, which the group translates into millions (they reported an extrapolated 2–5 million registered non‑citizens and warned 1.0–2.7 million might illegally vote absent reforms) — a figure driven by different assumptions and an amplified stress‑test of survey-derived inputs [2].
3. Administrative verification and referrals: tens of thousands checked, ~10,000 referrals
Federal and state administrative checks paint a far smaller picture: the Department of Homeland Security tool that states used to cross‑check registrations processed about 49.5 million voter records and produced roughly 10,000 cases referred to Homeland Security Investigations for potential non‑citizenship concerns — a referral rate of about 0.02% of names processed — and DOJ also cited about 10,000 registered non‑citizen referrals, though local officials warned the tool can produce false positives [3].
4. State and county reviews: isolated pockets, not mass phenomena
Several state audits and record reviews turned up only small counts in specific jurisdictions — for example, an Arizona county analysis reportedly found between roughly 1,900 and 6,500 non‑citizens registered in that state’s rolls and Ohio officials found around 600 questionable registrations in their sample — and Alabama moved 3,251 names to inactive status after federal flagging; these are concrete instances but far from national counts in the millions [1] [4].
5. Why estimates diverge: methodology, population base, and false positives
The discord owes largely to methodology: survey‑based extrapolations depend on identifying non‑citizens in polls (a small subgroup prone to sampling noise and misreporting), advocacy groups may apply stress tests or different population denominators to reach larger totals, and administrative cross‑matching can flag people erroneously (the CES authors caution against simple extrapolation of CES results to the whole population, and officials note the DHS tool misflags some citizens as noncitizens) — each approach answers a different question and produces different headline numbers [1] [3] [2].
6. Bottom line and limits of current reporting
There is no convergent, authoritative national figure that both academic researchers and officials accept: cautious academic extrapolations point to roughly a hundred‑thousand‑level phenomenon for registrations, federal administrative checks have produced roughly 10,000 referrals from tens of millions checked, and advocacy estimates claim numbers in the millions — the variation reflects divergent data sources, contested methods, and known problems with false positives and extrapolation; the Census’s main registration tables report citizen registration totals but do not provide a definitive national count of non‑citizen registrations to resolve the dispute [1] [3] [2] [5].