Is voter fraud a problem in the US?

I will examine evidence from official reports, academic studies, and election integrity analyses to assess the scale of voter fraud in US elections.Next I will review a few key primary analyses and reports in more depth for precise figures and context.No, voter fraud is not a significant or widespread problem in U.S. elections. It occurs at extremely low rates—typically measured in thousandths or millionths of a percent of ballots cast—and there is no credible evidence that it has altered the outcome of modern federal (presidential or congressional) elections. Isolated cases do happen, are investigated and prosecuted when detected, and safeguards exist to prevent and detect them, but the scale is minuscule relative to the hundreds of millions of votes cast.[1]

Key Evidence and Statistics

Multiple independent analyses, including those using data from the conservative Heritage Foundation’s Election Fraud Database (which tracks proven cases via convictions, civil findings, or official determinations), academic studies, government investigations, and journalistic reviews, consistently show this:

•Heritage Foundation database: Compiles roughly 1,500–1,600 proven instances spanning decades (often back to the 1980s or earlier), involving various types of fraud (e.g., double voting, absentee ballot issues, ineligible voting). Brookings Institution analysis of this data for contested states shows tiny rates. For example:

•Arizona (25 years, 36 elections, ~42.6 million ballots): 36 cases → ~0.0000845%.

•Pennsylvania (30 years, 32 elections, >100 million ballots): 39 cases.

•No cases altered federal election outcomes in the periods examined. The database is a sampling of proven cases (not exhaustive), yet even taken at face value against hundreds of millions/billions of ballots, the rate is negligible.[1]

•2020 election specifically: Associated Press review of six battleground states found fewer than 475 potential cases out of >25 million ballots. Courts rejected dozens of lawsuits alleging widespread fraud due to lack of evidence. Statistical claims of anomalies were debunked in peer-reviewed analysis.[2]

•Mail/absentee voting: Fraud is rarer still. Brookings analysis found an average of ~0.000043% (about 4 cases per 10 million mail ballots) across 2016–2022 general elections. Universal vote-by-mail systems show no elevated risk.[3]

•In-person impersonation: Extremely rare—studies find rates like 31 credible instances from 2000–2014 out of >1 billion ballots, or fewer than 10 alleged cases in multi-year reviews. It is often noted as less likely than being struck by lightning.[4]

•Noncitizen voting: Vanishingly rare. A Brennan Center study of 42 jurisdictions (23.5 million votes in 2016) found ~30 suspected incidents referred for investigation (0.0001%). Heritage lists a small number of such cases over decades. State audits (e.g., recent ones in Georgia, Utah, Ohio) find tiny numbers of improper registrations relative to total rolls, many of which do not result in votes, and data-matching errors are common.[5]

•Prosecutions: DOJ and state efforts yield low numbers. Historical DOJ reviews of federal elections found fraud rates like 0.00000013%. Recent charges (including under different administrations) involve dozens of individuals for illegal voting/registration—still tiny against ~150+ million votes in presidential years (e.g., ~158 million in 2020; high turnout ~64% of eligible in 2024). Many cases involve mistakes (e.g., felons unaware of restoration status) rather than intentional schemes.[4]

•Other reviews: News21/Arizona State, GAO literature reviews, state investigations (e.g., Kansas, Iowa, Colorado), and academic work (e.g., University of Chicago/Stanford) reach the same conclusion: overall fraud is “infinitesimal,” not systemic or organized at a scale to change results. A national investigative compilation found 513 proven cases since 2016 out of >685 million votes, with only a handful overturning local/small elections.[6]

U.S. elections involve ~150–160 million ballots in recent presidential cycles. Fraud cases number in the dozens to low hundreds per multi-year period, not millions.

Context and Caveats

•Types of fraud: Includes double voting, voting by ineligible people (felons, noncitizens, deceased via error), absentee ballot misuse, false registration, and rare ballot stuffing or official misconduct. Most are individual acts, not coordinated conspiracies. Clerical errors, data mismatches, and good-faith mistakes are often misreported as fraud.

•Local vs. federal: Fraud has occasionally affected small local races (tight margins, fewer votes), leading to overturned results or new elections in a small number of cases. Federal races have far larger vote totals and multi-layered checks (bipartisan observers, audits, signature verification, chain-of-custody, ERIC for roll maintenance, etc.).

•Safeguards and detection: States maintain voter rolls, use ID requirements (varying by state), track ballots, conduct risk-limiting audits, and prosecute. Penalties are severe (felonies). Undetected fraud exists in theory but is constrained by these systems and the difficulty of scaling it secretly without detection.

•Heritage and other views: The database usefully documents real cases and vulnerabilities (e.g., in absentee processes), supporting arguments for tighter rules like voter ID or same-day registration limits. Critics note it includes older/tangential cases and is not comprehensive of all potential issues, but even its proponents’ data do not support “widespread” claims when contextualized against total votes.[7]

•Claims vs. reality: High-profile allegations (e.g., millions of illegal votes in 2016/2020) have not held up under scrutiny, audits, recounts, or litigation. Public concern is often higher than the data warrant, fueled by rhetoric.

In short, while any fraud undermines trust and should be prevented/prosecuted, the empirical record from conservative trackers, progressive advocacy groups, nonpartisan researchers, journalists, and officials shows it is not a meaningful threat to election integrity at the national level. Election administration improvements (better technology, clearer eligibility rules, robust audits) remain worthwhile for confidence, but the data do not indicate a systemic crisis.

Citations (7)

1brookings.eduNonprofits and think tanks
https://www.brookings.edu/articles/how-widespread-is-election-fraud-in-the-united-states-not-very/
Backs: “elections. It occurs at extremely low rates—typically measured in thousandths or millionths of a percent of ballots cast—and there is no credible evidence that it has altered the outcome of modern federal (presidential or congressional) elections. Isolated cases do happen, are investigated and prosecuted when…”
2npr.orgNews and mediaquality 0.93
https://www.npr.org/2024/10/11/nx-s1-5147732/voter-fraud-explainer
Backs: “2020 election specifically: Associated Press review of six battleground states found fewer than 475 potential cases out of >25 million ballots.”
3brookings.eduNonprofits and think tanks
https://www.brookings.edu/articles/mail-voting-in-the-us-data-points-to-very-low-fraud-and-significant-benefits-to-voters/
Backs: “Statistical claims of anomalies were debunked in peer-reviewed analysis. Mail/absentee voting: Fraud is rarer still. Brookings analysis found an average of ~0.000043% (about 4 cases per 10 million mail ballots) across 2016–2022 general elections.”
4brennancenter.orgNonprofits and think tanks
https://www.brennancenter.org/our-work/research-reports/debunking-voter-fraud-myth
Backs: “In-person impersonation: Extremely rare—studies find rates like 31 credible instances from 2000–2014 out of >1 billion ballots, or fewer than 10 alleged cases in multi-year reviews.”
5ballotpedia.orgEncyclopedias and referencequality 0.89
https://ballotpedia.org/Noncitizen_voting_election_fraud
Backs: “Heritage lists a small number of such cases over decades. State audits (e.g., recent ones in Georgia, Utah, Ohio) find tiny numbers of improper registrations relative to total rolls, many of which do not result in votes, and data-matching errors are common.”
6koat.comNot yet classified
https://www.koat.com/article/us-election-voter-fraud-investigation/73775562
Backs: “Other reviews: News21/Arizona State, GAO literature reviews, state investigations (e.g., Kansas, Iowa, Colorado), and academic work (e.g., University of Chicago/Stanford) reach the same conclusion: overall fraud is “infinitesimal,” not systemic or organized at a scale to change results.”
7heritage.orgNonprofits and think tanks
https://www.heritage.org/press/heritage-foundation-responds-medias-false-claims-election-fraud-database
Backs: “Heritage and other views: The database usefully documents real cases and vulnerabilities (e.g., in absentee processes), supporting arguments for tighter rules like voter ID or same-day registration limits. Critics note it includes older/tangential cases and is not comprehensive of all potential issues, but even its…”