Elections · analysis
Voter fraud is real, and vanishingly rare
It happens, and it gets prosecuted. But the largest public database of proven cases holds 1,480 since 2000, about 1 in 157k registered voters, or 1.6 per million ballots, across 939 million presidential votes. Even the highest state, Minnesota, comes to about 1 in 28,000, over a quarter-century of elections. That is not a crisis, and it is not a reason to make voting harder: the proof-of-citizenship mandates, stricter ID rules, and aggressive purges proposed in its name have, where they've been tried, blocked or flagged far more eligible citizens than fraudsters.
This section makes an argument, unlike the rest of this site: that proven voter fraud is far too rare to justify new barriers to voting. Every number is sourced, and the strongest evidence on the other side is shown, not hidden. The other side →
The electorate vs the fraud, on a logarithmic scale (each step is ten times larger), because on a normal scale the fraud bars would be invisible. Grey: the voters. Light grey: the citizens new rules would burden. Orange: all proven fraud.
On a linear scale, if the ballots bar ran the width of this page, the 1,480 proven cases would be about 1/600,000th of it, far thinner than a single pixel.
Who chose the president: every U.S. resident, 2024
All 340.1 million people living in the United States, by what each could and did do in the 2024 presidential election. Donald Trump won with the votes of 22.7% of residents. About 87 million eligible adults did not vote for president, more than voted for either candidate.
- Eligible adults who did not vote for president 87.4M (25.7%)
- Voted for Donald Trump 77.3M (22.7%)
- Voted for Kamala Harris 75.0M (22.1%)
- Voted for someone else 2.9M (0.9%)
- Adults who could not vote (mostly non-citizens) 24.3M (7.2%)
- Children under 18 73.1M (21.5%)
Population is the Census estimate for July 1, 2024; votes are the official count. “Could not vote” is the UF Election Lab’s estimate of adult non-citizens (8.2% of adults) and people barred by felony-disenfranchisement laws (2.6 million). The eligible-non-voter slice is the remainder, so it is approximate: it includes about 1.5 million people who cast a ballot but skipped the presidential race, and it runs slightly low because the vote count includes Americans living abroad, who are not in the resident population.
Every state, as a share of its voters
Sorted by cases per million ballots; click any column to re-sort. “% of registered” divides cases since 2000 by the state's 2024 registered voters (active + inactive, EAC survey). North Dakota has no voter registration, so it has no registration-based rate; its 12 cases since 2000 are in the national count, and leaving them out would put the national share at 0.00063%. Even the highest rates are tiny: West Virginia (7.12 per million), Minnesota (6.70 per million), New Hampshire (5.74 per million), North Dakota (5.18 per million), Mississippi (5.08 per million). West Virginia's is 0.0030% of its registered voters. Delaware and Vermont have no cases in the database.
| State | Cases 1982–2025 | Since 2000 | Noncitizen | Registered voters, 2024 | % of registered | One case per | Per million ballots | Most common type | Audit |
|---|---|---|---|---|---|---|---|---|---|
| West Virginia | 36 | 36 | 0 | 1,210,415 | 0.0030% | 1 in 34k | 7.12 | Vote buying | |
| Minnesota | 138 | 138 | 2 | 3,853,668 | 0.0036% | 1 in 28k | 6.70 | Felon voting | |
| New Hampshire | 29 | 29 | 0 | 1,008,952 | 0.0029% | 1 in 35k | 5.74 | Double voting | |
| North Dakota | 12 | 12 | 0 | no registration | n/a | n/a | 5.18 | Petition fraud | |
| Mississippi | 45 | 43 | 0 | 2,131,726 | 0.0020% | 1 in 50k | 5.08 | Vote buying | |
| Kentucky | 62 | 62 | 0 | 3,548,136 | 0.0017% | 1 in 57k | 4.73 | Vote buying | |
| Wisconsin | 68 | 68 | 1 | 3,933,068 | 0.0017% | 1 in 58k | 3.19 | Felon voting | |
| Indiana | 59 | 59 | 0 | 4,840,856 | 0.0012% | 1 in 82k | 3.15 | False registration | |
| Alabama | 41 | 41 | 2 | 3,868,040 | 0.0011% | 1 in 94k | 2.84 | Absentee fraud | audit |
| Idaho | 13 | 13 | 0 | 1,178,750 | 0.0011% | 1 in 91k | 2.66 | False registration | audit |
| Iowa | 29 | 29 | 3 | 2,256,774 | 0.0013% | 1 in 78k | 2.66 | Felon voting | audit |
| North Carolina | 123 | 82 | 38 | 7,854,464 | 0.0010% | 1 in 96k | 2.63 | Vote buying | audit |
| District of Columbia | 5 | 5 | 0 | 612,904 | 0.00082% | 1 in 123k | 2.54 | Double voting | |
| Arizona | 40 | 40 | 0 | 5,075,337 | 0.00079% | 1 in 127k | 2.29 | Double voting | |
| Wyoming | 4 | 4 | 0 | 296,960 | 0.0013% | 1 in 74k | 2.26 | False registration | |
| South Dakota | 6 | 6 | 0 | 690,306 | 0.00087% | 1 in 115k | 2.25 | Petition fraud | |
| Alaska | 5 | 5 | 2 | 611,078 | 0.00082% | 1 in 122k | 2.23 | Noncitizen voting | |
| Connecticut | 31 | 25 | 2 | 2,520,650 | 0.00099% | 1 in 101k | 2.18 | Absentee fraud | |
| Kansas | 17 | 17 | 1 | 2,031,119 | 0.00084% | 1 in 119k | 1.99 | Double voting | |
| New Jersey | 53 | 53 | 1 | 6,630,364 | 0.00080% | 1 in 125k | 1.96 | Absentee fraud | |
| Texas | 113 | 113 | 4 | 18,623,931 | 0.00061% | 1 in 165k | 1.84 | Absentee fraud | audit |
| Ohio | 63 | 63 | 8 | 8,074,098 | 0.00078% | 1 in 128k | 1.62 | False registration | audit |
| Missouri | 30 | 30 | 1 | 4,388,787 | 0.00068% | 1 in 146k | 1.53 | False registration | |
| Nevada | 11 | 11 | 1 | 2,256,275 | 0.00049% | 1 in 205k | 1.48 | False registration | |
| Florida | 93 | 89 | 15 | 15,740,083 | 0.00057% | 1 in 177k | 1.44 | False registration | |
| Tennessee | 26 | 26 | 0 | 4,825,601 | 0.00054% | 1 in 186k | 1.43 | False registration | |
| Illinois | 117 | 54 | 10 | 8,970,541 | 0.00060% | 1 in 166k | 1.42 | False registration | |
| Virginia | 36 | 36 | 0 | 6,380,686 | 0.00056% | 1 in 177k | 1.36 | Absentee fraud | audit |
| Oregon | 18 | 18 | 1 | 3,060,374 | 0.00059% | 1 in 170k | 1.32 | Double voting | audit |
| Colorado | 23 | 23 | 1 | 4,583,280 | 0.00050% | 1 in 199k | 1.27 | Absentee fraud | |
| New Mexico | 8 | 7 | 0 | 1,415,984 | 0.00049% | 1 in 202k | 1.25 | Absentee fraud | |
| Georgia | 36 | 28 | 0 | 8,234,335 | 0.00034% | 1 in 294k | 1.00 | Absentee fraud | audit |
| Pennsylvania | 45 | 38 | 0 | 9,175,133 | 0.00041% | 1 in 241k | 0.89 | Absentee fraud | audit |
| Arkansas | 6 | 6 | 0 | 1,835,723 | 0.00033% | 1 in 306k | 0.78 | Absentee fraud | |
| California | 71 | 70 | 2 | 25,720,597 | 0.00027% | 1 in 367k | 0.72 | False registration | |
| Washington | 15 | 15 | 0 | 5,597,156 | 0.00027% | 1 in 373k | 0.66 | Absentee fraud | |
| Rhode Island | 2 | 2 | 0 | 792,075 | 0.00025% | 1 in 396k | 0.61 | Double voting | |
| Louisiana | 8 | 8 | 0 | 3,046,376 | 0.00026% | 1 in 381k | 0.58 | Vote buying | audit |
| Montana | 2 | 2 | 0 | 800,573 | 0.00025% | 1 in 400k | 0.57 | Absentee fraud | |
| Michigan | 19 | 19 | 1 | 8,440,236 | 0.00023% | 1 in 444k | 0.55 | Absentee fraud | audit |
| Maryland | 10 | 10 | 4 | 4,555,217 | 0.00022% | 1 in 456k | 0.54 | Noncitizen voting | |
| Nebraska | 3 | 3 | 0 | 1,263,487 | 0.00024% | 1 in 421k | 0.52 | Double voting | |
| Oklahoma | 5 | 5 | 0 | 2,442,211 | 0.00020% | 1 in 488k | 0.50 | Ineligible voting | |
| New York | 28 | 26 | 0 | 13,579,416 | 0.00019% | 1 in 522k | 0.48 | Absentee fraud | |
| Utah | 3 | 3 | 0 | 2,039,862 | 0.00015% | 1 in 680k | 0.39 | False registration | audit |
| Maine | 2 | 2 | 0 | 1,223,468 | 0.00016% | 1 in 612k | 0.38 | Double voting | |
| Hawaii | 5 | 1 | 0 | 861,333 | 0.00012% | 1 in 861k | 0.31 | False registration | |
| Massachusetts | 4 | 4 | 0 | 5,142,343 | 0.00008% | 1 in 1.3M | 0.18 | Absentee fraud | |
| South Carolina | 1 | 1 | 0 | 3,851,187 | 0.00003% | 1 in 3.9M | 0.07 | Absentee fraud | |
| Delaware | 0 | 0 | 0 | 788,441 | 0% | n/a | 0.00 | n/a | |
| Vermont | 0 | 0 | 0 | 500,986 | 0% | n/a | 0.00 | n/a | |
| United States | 1,619 | 1,480 | 100 | 232,363,362 | 0.00064% | 1 in 157k | 1.58 | Absentee fraud |
The strongest case on the other side
What cuts against this argument
Undetected fraud can't be counted, and supporters of new rules argue that is the point: the Heritage list is a floor. The most rigorous national study of strict photo-ID laws (Cantoni & Pons, Quarterly Journal of Economics, 2021) found no significant drop in turnout, for any group, though a GAO study found turnout fell 2–3 points more in Kansas and Tennessee after their laws. Only about 1% of citizens (2.6 million) lack any government photo ID. And noncitizen ballots are real, if rare: Iowa counted 35 in 2024, and Texas and Kansas charged noncitizen voters in 2026.
Why the argument still holds
Grant all of it. The same QJE study found strict ID laws had no effect on fraud either: unsurprising, since the one kind ID stops, impersonation at the polls, is 34 proven cases in 43 years. Proof-of-citizenship rules are a far bigger burden than photo ID: 21.3 million citizens can't readily produce the papers. Where such rules and purges have actually been run, they have flagged or blocked eligible citizens by the hundreds or thousands for every noncitizen found, and the fraud that does occur is caught and prosecuted under existing law. A rule whose costs land on thousands of eligible voters to stop dozens of cases is a bad trade, whichever party proposes it.
The cost of the cure
Kansas blocked 31,089 would-be voters; it had found 39 noncitizens on its rolls since 1999. What proof-of-citizenship rules, ID laws, and purges have cost eligible voters, measured against the fraud they caught.
View →Election fraud, counted
All 1,619 proven cases by type, year, and state: mostly absentee schemes and ineligible voters, caught under existing law, and what the 2020 reviews found.
View →Noncitizen voting: rare, and the checks hit citizens
Every state citizenship audit: noncitizens reported at 0.0002%–0.013% of the rolls, while the checks flag eligible citizens by the thousands.
View →