Voter fraud claims are everywhere — I'll pull the latest research and court data to see what the numbers actually show.
Voter fraud in the US exists but multiple decades of studies, court reviews and government investigations have found it to be extremely rare - not at a scale that would alter national outcomes.
What the research consensus shows
Nonpartisan research and reviews consistently describe voter fraud as "vanishingly rare" and measured in tiny fractions of a percent.
•The Brennan Center's seminal report The Truth About Voter Fraud reviewed elections that had been meticulously studied and found incident rates between 0.0003 percent and 0.0025 percent. The report noted it is more likely an American "will be struck by lightning than that he will impersonate another voter at the polls".
•A comprehensive 2014 study found 31 credible instances of impersonation fraud from 2000 to 2014, out of more than 1 billion ballots cast.
•Arizona State University studies from 2012 and 2016 found negligible rates, including only 10 cases of voter impersonation fraud nationwide from 2000-2012, and zero successful prosecutions for impersonation fraud in five states from 2012-2016.
•A specialized United States Department of Justice unit formed to find instances of federal election fraud examined the 2002 and 2004 federal elections and were able to prove that 0.00000013 percent of ballots cast were fraudulent.
•Other reviews put rates even lower, in the range of millionths of a percent, such as 0.00000017% to 0.0000845% in cited analyses.
The Brennan Center and other researchers also note that most reported incidents traceable to other sources, such as clerical errors or bad data matching practices, rather than intentional fraud.
What courts and government investigations have found
•Courts: The Fifth Circuit noted "only two convictions for in-person voter impersonation fraud out of 20 million votes cast in the decade" before Texas passed its voter ID law. A federal trial court in Wisconsin found impersonation fraud is "extremely rare" and "a truly isolated phenomenon". A review of the 2016 election found four documented cases of voter fraud.
•State investigations: Kansas Secretary of State Kris Kobach's review of 84 million votes cast in 22 states yielded 14 instances of fraud referred for prosecution, which amounts to a 0.000017 percent fraud rate. In Texas, the chief law enforcement official responsible for such prosecutions knew of only one conviction and one guilty plea that involved in-person voter fraud in all Texas elections from 2002 through 2014.
•Federal election security: The Cybersecurity and Infrastructure Security Agency (CISA) called the 2020 election "the most secure in American history," stating "There is no evidence that any voting system deleted or lost votes, changed votes, or was in any way compromised". There were "many unfounded claims" and "opportunities for misinformation" about the election process.
What the Heritage Foundation database shows
The most-cited conservative compilation is the Heritage Foundation's Election Fraud Database, which is often cited as evidence that fraud is real and ongoing.
•Heritage lists roughly 1,561 "proven instances" yielding about 1,325 convictions across its time window.
•Other Heritage releases cite well over 100 new cases, documenting 1,071 instances spanning 47 states, including 938 criminal convictions, and 1,088 proven instances including 949 criminal convictions.
•Analysts note this averages 30-40 proven cases per year over roughly 44 years, out of hundreds of millions of ballots cast.
The Brennan Center analyzed the same database and found claims were "grossly exaggerated" and "devoid of context" - there were only 10 cases involving in-person voter-impersonation fraud and only 41 involving noncitizen voting in the sample of ~1,100 instances reviewed at the time.
Bottom line
•Proven voter fraud is real but rare - infinitesimal relative to total votes cast, typically a few dozen cases per several million ballots.
•No rigorous body of research has shown fraud at the scale necessary to flip major statewide or presidential outcomes, and the best available public studies place the rate at effectively zero for national elections, with no demonstrated impact on federal outcomes and transparent methodological caveats.
Supporters of stricter election integrity laws argue that even rare cases justify prevention measures like voter ID, accurate voter rolls, and prosecution, while opponents argue that claims of widespread, election-altering fraud are not supported by the data and that most allegations turn out to be mistakes.