Is voter fraud a problem in the US?

No, voter fraud is not a significant or widespread problem in the US. It occurs occasionally (as individual acts like double voting, ineligible voting by felons or noncitizens, or rare absentee ballot misuse), but multiple studies, audits, prosecutions, and analyses—including from sources across the political spectrum—consistently show it is extremely rare, typically measured in tiny fractions of a percent of ballots cast, and has not been shown to alter the outcomes of modern elections at any meaningful scale.[1]

Key Evidence on Scale

•Overall rates: Reviews of closely studied elections find fraud incident rates of roughly 0.0003% to 0.0025% (or far lower in some analyses). A 2014 investigation identified about 31 credible cases of in-person impersonation fraud out of more than 1 billion ballots cast from 2000–2014. Americans are more likely to be struck by lightning than to successfully impersonate another voter at the polls. A Department of Justice review of 2002–2004 federal elections found a fraudulent ballot rate of about 0.00000013%, with no evidence of coordinated efforts.[2]

•Heritage Foundation database (a conservative source tracking proven cases leading to convictions, civil penalties, or official findings): It catalogs roughly 1,400–1,620 instances spanning decades (often back to the 1980s or earlier) across all types of election fraud. This is a tiny fraction of the hundreds of millions to billions of ballots cast in that period (recent presidential elections alone involve 130–160 million votes each; midterms add more). Brookings Institution analyses of this same data for competitive states found, for example:

•Arizona: 36 cases over 25 years and ~42.6 million ballots (rate ~0.0000845%).

•Pennsylvania: 39 cases over 30 years and over 100 million votes.

•No election outcomes altered by ballot fraud in these or similar reviews.[1]

•Recent elections and specific claims:

•2020: An Associated Press review of potential fraud cases in six swing states found fewer than 475 out of more than 25 million ballots—far too few to affect results (Biden’s combined margin there was over 300,000). Most involved isolated individuals, not coordination.[3]

•Mail-in/absentee: Brookings analysis of Heritage data for 2016–2022 found an average of about 4 fraud cases per 10 million mail ballots. States with longstanding all-mail systems (e.g., Oregon) show similarly negligible rates.[4]

•Noncitizen voting: Extremely rare. Heritage data shows dozens of convictions over decades (e.g., ~24–71 depending on the timeframe). Surveys of election officials in high-noncitizen areas (covering millions of votes) find referrals for investigation in the low dozens at most (e.g., ~30 out of 23.5 million votes in one 2016 Brennan Center survey, or ~0.0001%). Recent federal charges remain in the low hundreds amid heightened scrutiny and do not indicate systemic issues. Initial large claims of noncitizens on rolls often shrink dramatically upon review due to errors, naturalizations, or false positives.[5]

•Other investigations (e.g., by academics, state officials of both parties, the AP, and even a Trump-era effort) reach the same conclusion: fraud is “infinitesimally rare” and almost never occurs at a scale that would change outcomes. Many alleged cases turn out to be clerical errors, data mismatches, or misunderstandings rather than intentional fraud.[3]

Context and Safeguards

US elections have multiple layers of protection: voter registration checks and list maintenance, ID requirements in many states, signature verification for mail ballots, poll books, bipartisan observers, post-election audits/risk-limiting audits, chain-of-custody procedures, and criminal penalties (fines and jail time). Prosecutions do occur when fraud is detected. Isolated cases (e.g., a felon voting unknowingly or someone casting a relative’s ballot) happen and should be addressed, but they do not equate to systemic vulnerability or “rampant” fraud.[6]

Heritage itself frames its database as evidence of vulnerabilities needing reforms (e.g., better list maintenance or ID), not as proof of outcome-changing chaos. Critiques note it includes older cases, non-voter-fraud election issues, and civil matters, and that even taken at face value the numbers remain minuscule relative to total votes.[7]

Why Claims Persist

High-profile allegations (especially post-2016 and 2020) often amplify rare or unproven incidents, rely on flawed data matching (producing false positives), or lack follow-through prosecutions. Extensive research, including by the Brennan Center, Brookings, academic studies, and journalists, has repeatedly debunked claims of millions of illegal votes. False or exaggerated claims can erode trust and lead to policies that restrict access without addressing actual (rare) problems.[8]

In short, voter fraud is real but rare—like other low-level crimes in a large system—and US elections are secure against it at scale due to existing processes. Focus on integrity is valid (e.g., accurate rolls, audits), but evidence does not support it being a major problem that undermines results. For the latest case data, see sources like the Heritage database, state election offices, or nonpartisan trackers.

Citations (8)

1brookings.eduNonprofits and think tanks
https://www.brookings.edu/articles/how-widespread-is-election-fraud-in-the-united-states-not-very/
Backs: “It occurs occasionally (as individual acts like double voting, ineligible voting by felons or noncitizens, or rare absentee ballot misuse), but multiple studies, audits, prosecutions, and analyses—including from sources across the political spectrum—consistently show it is extremely rare, typically measured in tiny…”
evidence, article, points to no. Claim: Studies and audits show voter fraud is extremely rare and has not altered modern election outcomes; Brookings analysis of Heritage data found tiny case counts in Arizona and Pennsylvania.
2legalclarity.orgCommercial content
https://legalclarity.org/how-common-is-voter-fraud-what-the-evidence-shows/
Backs: “A Department of Justice review of 2002–2004 federal elections found a fraudulent ballot rate of about 0.00000013%, with no evidence of coordinated efforts.”
statistic, article, points to no. Claim: Fraud incident rates are roughly 0.0003% to 0.0025%; 31 credible impersonation cases among over 1 billion ballots; DOJ review found tiny fraudulent ballot rate.
3npr.orgNews and mediaquality 0.93
https://www.npr.org/2024/10/11/nx-s1-5147732/voter-fraud-explainer
Backs: “An Associated Press review of potential fraud cases in six swing states found fewer than 475 out of more than 25 million ballots—far too few to affect results (Biden’s combined margin there was over 300,000). Most involved isolated individuals, not coordination.”
statistic, article, points to no. Claim: AP review of 2020 found fewer than 475 potential fraud cases among over 25 million ballots in six swing states, too few to affect results.
4bbc.co.ukNews and mediaquality 0.72
https://www.bbc.co.uk/news/articles/ckeq3vezxlx1o
Backs: “States with longstanding all-mail systems (e.g., Oregon) show similarly negligible rates.”
statistic, article, points to no. Claim: Brookings analysis of Heritage data for 2016–2022 found about 4 fraud cases per 10 million mail ballots; all-mail states show negligible rates.
5nytimes.comNews and mediaquality 0.86
https://www.nytimes.com/interactive/2026/10/15/us/politics/noncitizen-voting-data.html
Backs: “Initial large claims of noncitizens on rolls often shrink dramatically upon review due to errors, naturalizations, or false positives.”
evidence, article, points to no. Claim: Noncitizen voting is extremely rare; large claims of noncitizens on rolls shrink on review due to errors, naturalizations, or false positives.
6apnews.comNews and mediaquality 1.00
https://apnews.com/article/election-2024-voter-fraud-trump-harris-a3b4c2db17217311770259193c115b80
Backs: “Isolated cases (e.g., a felon voting unknowingly or someone casting a relative’s ballot) happen and should be addressed, but they do not equate to systemic vulnerability or “rampant” fraud.”
background, article, points to no. Claim: US elections have layered safeguards and prosecutions occur; isolated cases do not equate to systemic vulnerability or rampant fraud.
7brennancenter.orgNonprofits and think tanks
https://www.brennancenter.org/our-work/research-reports/heritage-fraud-database-assessment
Backs: “Critiques note it includes older cases, non-voter-fraud election issues, and civil matters, and that even taken at face value the numbers remain minuscule relative to total votes.”
evidence, official document, points to no. Claim: Critiques of the Heritage database note it includes older cases, non-fraud issues and civil matters, and numbers remain minuscule relative to total votes.
8brennancenter.orgNonprofits and think tanks
https://www.brennancenter.org/topics/voting-elections/vote-suppression/myth-voter-fraud
Backs: “False or exaggerated claims can erode trust and lead to policies that restrict access without addressing actual (rare) problems.”
evidence, article, points to no. Claim: Research has repeatedly debunked claims of millions of illegal votes; exaggerated claims erode trust and lead to restrictive policies.