Amount of fraud in 2020 election that has been verified
Executive summary
Verified instances of voter fraud tied to the 2020 U.S. presidential election were vanishingly small compared with the tens of millions of ballots cast: major post‑election reviews and reporting projects found fewer than 475 potential cases in the six battleground states most disputed by former President Trump, and courts repeatedly rejected claims of widespread fraud for lack of evidence [1] [2] [3] [4].
1. What “verified” means and why precision matters
"Verified" in reporting and litigation generally refers to cases that election officials, prosecutors, independent auditors, or courts have investigated and corroborated as improper or illegal voting conduct, not merely allegations shared online or in partisan filings, and the most comprehensive journalistic and academic efforts after 2020 aimed to map those verified instances rather than unproven claims [2] [5] [6].
2. The headline number: AP’s review of six battleground states
An Associated Press review that remains the most granular public accounting examined every potential case flagged in six battleground states and identified fewer than 475 potential instances of voter fraud—an amount the AP and multiple outlets concluded would not remotely have changed the outcome of the presidential race given the more than 25 million votes cast in those states and nationwide totals far larger [2] [1] [7].
3. Courts, audits and election officials: repeated failure to substantiate mass fraud claims
Dozens of lawsuits brought after the election were dismissed or lost because plaintiffs failed to meet evidentiary standards or showed only clerical or human‑error irregularities rather than deliberate, systemic fraud; judges across the political spectrum and bipartisan state officials certified results and rejected claims of a stolen election [4] [6] [7].
4. Academic and institutional assessments: fraud is rare, not system‑wide
Scholarly analyses and research institutions concluded that statistical evidence and broad empirical claims of millions of fraudulent ballots, manipulated machines, or thousands of votes cast by dead people were unsupported; peer‑reviewed analysis and policy centers described election fraud in the U.S. as uncommon and insufficient to alter national outcomes [5] [8] [9].
5. Proven cases exist but are isolated and historically sparse
Databases like the Heritage Foundation’s Election Fraud Map and other government and media compilations document proven instances of wrongdoing—but those cases are isolated, often individual actors rather than coordinated conspiracies, and analysts have stressed that even aggregating historical proven cases finds an extremely small fraction of total ballots and no evidence that such cases flipped the 2020 result [10] [8].
6. Limits of the public record and sober conclusion
Public reporting and litigation produce strong, consistent findings: verified wrongdoing associated with the 2020 election was extremely limited in number (the AP’s under‑475 figure in six key states is the most cited working total), courts and audits found no support for widespread fraud, and scholarly work characterizes systemic fraud claims as unsupported by data—however, this account is bounded by what has been investigated and publicly disclosed, and there is no single nationwide tally that captures every local prosecution or administrative finding beyond aggregated journalistic and institutional reviews [2] [4] [5] [1].