Is voter fraud a problem in the US?

Short answer: No, not in any significant or widespread way — extensive research consistently finds actual voter fraud is extremely rare in the U.S., even though public perception of it as a problem is much higher than the evidence supports.

What the research shows

Multiple independent studies spanning decades converge on the same conclusion:

Impersonation fraud (the type voter ID laws target) is almost nonexistent. A landmark review of elections "meticulously studied for voter fraud" found incident rates between 0.0003 percent and 0.0025 percent, leading researchers to conclude it is more likely an American "will be struck by lightning than that he will impersonate another voter at the polls."

A comprehensive News21 review of 2,068 alleged fraud cases across all 50 states from 2000–2012 found only 10 cases of voter impersonation, the only kind of fraud that could be prevented by voter ID at the polls. A follow-up review of states where politicians specifically claimed fraud was rampant (Arizona, Ohio, Georgia, Texas, Kansas) between 2012-2016 similarly found hardly any impersonation cases: zero successful prosecutions for impersonation fraud in five states from 2012–2016.

Mail-in/absentee voting fraud is similarly tiny. Brookings Institution research examined cases of fraud reported during elections in 2016, 2018, 2020, and 2022 and estimated there were about four cases of fraud for every 10 million mail ballots cast, on average.

Even using the most fraud-friendly conservative data source (the Heritage Foundation's database, which casts a wide net), the numbers remain minuscule. An analysis of this database by a Brookings scholar found that the share of reported cases of fraud over the past 13 to 38 years across several key swing states is less than 1 percent. In Pennsylvania specifically, the Heritage Foundation analyzed 30 years of data, across which 32 elections were held. Just 39 cases of voter fraud were identified — 0.000039 percent of the more than 100 million ballots cast in those elections.

Double voting and noncitizen voting are also negligible. One working paper concluded that the upper limit on double voting in the 2012 election was 0.02%, though the incident rate was likely much lower given that "many, if not all, of these apparent double votes could be a result of measurement error." Separately, a 2014 academic paper concluded that "the likely percent of non-citizen voters in recent US elections is 0."

Where "fraud" claims actually come from

Researchers repeatedly find that most alleged fraud cases dissolve under scrutiny. Most reported incidents of voter fraud are actually traceable to other sources, such as clerical errors or bad data matching practices. Similarly, political scientist Lorraine Minnite's influential research examined thousands of voter-fraud complaints in four states—California, Minnesota, New Hampshire and Oregon—and found only scattered verifiable instances of wrongdoing.

Recent examples continue this pattern: in New Jersey, Governor Mikie Sherrill announced that about 6,600 people who identified as noncitizens were erroneously registered to vote in 2023 and 2024, through no fault of their own, due to a software error — not fraud, but an administrative glitch, and of those, fewer than 400 of those people actually voted.

But there's a notable gap between evidence and public belief

Despite this overwhelming evidence of rarity, perception differs sharply. Ahead of the 2024 election, six in ten U.S. adults were concerned or very concerned about it, with this fear particularly pronounced among Republican-leaning voters at 86 percent, versus 55 percent of Independents and just 33 percent of Democratic-leaning voters. Academic research on this gap finds that partisanship correlates strongly with endorsement of types of election fraud, with Republicans consistently more likely to state that types of election fraud are common... and conspiratorial thinking is strongly correlated with belief in the occurrence of types of election fraud, even after controlling for party.

Bottom line

Isolated, individual cases of fraud do happen — no voting system of hundreds of millions of ballots is perfectly error-free — but every rigorous study, across different methodologies, time periods, and even using data compiled by conservative organizations, finds the actual rate is a tiny fraction of a percent and has no realistic capacity to affect election outcomes. The gap between this evidence and public concern about fraud is itself a well-documented phenomenon, driven significantly by political rhetoric and media coverage rather than by the underlying data.