Is voter fraud a problem in the US?

Voter fraud claims get a lot of attention — I'll check the evidence on how common it is in the US.

In the United States, voter fraud exists but extensive research, audits and government investigations have consistently found it to be rare and not at a scale that routinely affects election outcomes.

What the evidence shows

> I do not believe the available evidence supports the claim that widespread voter fraud is a serious problem in the United States. Individual cases of fraud do occur and should be investigated and prosecuted, but extensive audits, investigations and academic research have consistently found such cases to be rare

> There is no evidence for widespread voter fraud in U.S. elections, including in recent national elections

> Empirical evidence across decades and multiple methodologies leads to one clear, repeatable conclusion: proven voter fraud is extremely rare relative to the number of votes cast — measured in orders of magnitude of 0.0000X percent — and no rigorous body of research has shown fraud at the scale necessary to flip major U.S. elections

Studies distinguish between:

How rare is it - estimated rates

One widely cited Brennan Center for Justice analysis found that voter impersonation occurs at a rate between 0.0003% and 0.0025%

The Brennan Center's report reviewed elections that had been meticulously studied for voter fraud, and found incident rates between 0.0003 percent and 0.0025 percent

Other studies found even lower rates:

Research into the 2016 and 2020 presidential elections similarly found no evidence of widespread fraud. One study investigating claims about non-citizen voters, out-of-state voting, voting machines changing votes, and absentee ballots found no evidence of fraud after investigation

Studies also find no evidence that voting by mail increases the risk of voter fraud

What case databases show

The most-cited public tally of proven cases comes from the Heritage Foundation's Election Fraud Database:

> the Heritage Foundation's Election Fraud Database catalogs roughly 1,561 instances that produced about 1,325 convictions across many decades

Other summaries of the same database list approximately 1,465 proven cases over roughly 44 years, or 1,546 election fraud cases of all kinds accumulated over decades, but only 68 involving noncitizen voting going back to the 1980s

When put in context, that averages to about 30-40 proven cases per year nationwide among more than 150 million votes cast in a presidential election. The Brennan Center analysis of an early version noted there were only 10 cases involving in-person voter-impersonation fraud and only 41 involving noncitizen voting in that sample

Government investigations and court findings

Election security assessments for recent elections

The Cybersecurity and Infrastructure Security Agency (CISA), the federal agency responsible for election security, said the 2020 election was "the most secure in American history," with no evidence that voting systems deleted, lost or compromised votes

Georgia and Pennsylvania conducted audits of the 2024 general election and officials in both states confirmed the accuracy of election results

Why perceptions differ from evidence

Research notes that rates are extremely low, but perception is influenced by other factors:

> The psychological effects of winning an election and messaging by political leaders may alter voter confidence and public perceptions of voter fraud beyond its demonstrated occurrence

The "winner effect" describes voters who voted for the winner having more confidence that there was not significant fraud than those who voted for the loser

Most researchers also note that many reported incidents turn out to be clerical errors or bad data matching practices rather than intentional fraud, and that when residents move between states they can end up with registration in both states, creating the appearance of greater fraud than actually exists

Bottom line