Is voter fraud a problem in the US?

Short answer: No, not as a widespread or systemic problem — but it's not literally zero, either.

Decades of research, data from both nonpartisan and conservative sources, and numerous court/government investigations consistently find that voter fraud occurs, but at an extremely small scale that doesn't affect election outcomes.

What the data shows

Even conservative-leaning tallies find tiny numbers. The Heritage Foundation—a conservative think tank that maintains the most-cited database of fraud cases—has documented 1,561 instances of voter fraud that resulted in 1,325 convictions. But this is spread across decades and hundreds of millions of ballots: in Arizona, Heritage had to go back 25 years, over which there were 36 elections held and 42,626,379 ballots cast, to find 36 cases of fraud — a minuscule .0000845%, with no election outcome altered. In Pennsylvania, Heritage data covers 32 elections over 30 years with over 100 million votes cast and found only 39 cases of voter fraud.

Academic and legal reviews reach the same conclusion. A Brennan Center report reviewing meticulously studied elections found incident rates between 0.0003 and 0.0025 percent, concluding it's more likely an American will be struck by lightning than 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. Research conducted over the past two decades has concluded there is no evidence of significant election fraud in contemporary American national elections.

Even investigations launched specifically to find fraud mostly come up empty. In Ohio, Republican Secretary of State Frank LaRose announced in 2022 his office had found 75 voters who allegedly cast a ballot in the state and another state in 2020, out of almost 6 million ballots cast — and his office still said voter fraud continues to be exceedingly rare in the state. After the 2020 election, the AP found 475 potential voter fraud cases across six swing states — an extraordinarily small percentage of more than 25 million ballots cast.

Mail-in voting is slightly more vulnerable, but still rare. Experts say fraud occurs more often with mailed-in votes than with in-person voting, though it remains quite rare. Brookings researchers examining fraud cases in 2016, 2018, 2020, and 2022 estimated about four cases of fraud for every 10 million mail ballots cast, on average.

Historically, it was a bigger problem. In the last half-century, there have been only scattered examples of electoral fraud affecting outcomes, mostly at the local level — electoral fraud was significantly more prevalent in the 19th and early 20th centuries.

Where it does matter: local elections

While a handful of fraudulent votes isn't enough to overturn federal elections, it can affect local elections like school board races, where every vote matters more.

The perception gap

There's a notable disconnect between public belief and documented evidence. A 2024 nationwide NPR/PBS News/Marist poll found 58% of Americans were concerned about voter fraud in the 2024 presidential election, including 88% of Trump supporters and 29% of Harris supporters. Yet "research has been consistent over time that voter fraud is infinitesimally rare and almost never occurs on a scale that would affect an election outcome," according to Brennan Center senior counsel Alice Clapman.

Bottom line

Fraud does happen—people are prosecuted for it every election cycle—but virtually every rigorous study, across partisan lines, finds it's a tiny fraction of ballots cast and has not changed any U.S. election outcome in modern history, aside from isolated local races.