Is voter fraud a problem in the US?

Based on extensive research from academic institutions, think tanks across the political spectrum, and government reviews, here's what the evidence shows:

The Short Answer: Fraud exists, but is extremely rare

Voter fraud does occur in the U.S., but multiple independent studies—including from sources on both the left and right—consistently find it happens at exceedingly low rates and has never been shown to alter a national election outcome.

What the Data Shows

Scale of documented cases: Even the Heritage Foundation's own database, often cited by those concerned about fraud, is telling. Looking at the seven closest states, in Arizona there were 36 elections held and 42,626,379 ballots cast, in order to come up with 36 cases of fraud. The percentage of fraudulent votes was a minuscule .0000845%, and no election outcome was altered by ballot fraud throughout that time period. Similarly, in the highly contested state of Pennsylvania, Heritage data goes back 30 years and covers 32 elections with over 100 million votes cast and found only 39 cases of voter fraud.

Academic consensus: Research conducted during the past two decades has concluded that there is no evidence of significant election fraud in contemporary American national elections.

In-person impersonation fraud (the type voter ID laws target) is almost nonexistent: A study of 2,068 alleged election-fraud cases in 50 states between 2000 and 2012 found the level of fraud was infinitesimal compared with the 146 million voters registered over the 12-year period. The analysis 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 analysis reached a similar conclusion: it looked for fraud specifically in states where politicians have argued that fraud is a pernicious problem, and found zero successful prosecutions for impersonation fraud in five states from 2012–2016.

Mail voting fraud is also minimal: A recent Brookings analysis found that of the minuscule 0.000043 percent of mail voting found to be fraudulent out of all mail ballots cast in the 2016–2022 general elections, after adjusting for total mail ballots cast by type, 48.5 percent involved no-excuse absentee systems, 37.9 percent involved excuse-based absentee systems, and 13.5 percent involved universal vote-by-mail systems.

2020 election specifically: A count by the Associated Press of voter fraud cases in the six battleground states in 2020 that were disputed by Donald Trump found fewer than 475 votes in all six states combined—less than two-tenths of one percent of Biden's margin of victory in those states. Most of those were not counted, and not all were for one candidate. Importantly, the AP study also found no systematic collusion or "rigging" in those cases. Nearly every one was an individual acting alone.

Expert Characterization

Brookings' Darrell West summarized the general findings: "There are a few cases of election fraud in every campaign," but "it's not systemic. It's not widespread. It's generally not organized, and I don't think ever has affected the actual outcome, but people often pick on isolated cases and turn it into a systemic and widespread pattern."

A Notable Gap: Perception vs. Reality

Interestingly, there's a significant disconnect between public belief and the empirical evidence. Research shows each of these types of election fraud is considered to be common or occasional for a third or more of American registered voters, despite two decades of academic research concluding there is no evidence of significant election fraud in contemporary American national elections.

Bottom Line

The empirical research—spanning studies from Brennan Center, Brookings, News21, academic political scientists, and even Heritage Foundation's own database—converges on a consistent finding: individual instances of fraud do occur, but at rates so low (often measured in thousandths or hundred-thousandths of a percent) that they pose no realistic threat to election outcomes at scale. This is a well-documented empirical conclusion, though it remains a politically contested topic where public perception often diverges from the research consensus.