Search 200 common claims about vaccines, climate, evolution and more. Open any claim to see a short, evidence-based response.

• Claim Voting machines can be silently reprogrammed to flip results. Election science

  Response

  U.S. voting machines used in elections are extensively tested, certified, and in most jurisdictions produce a paper trail that allows independent hand-count audits comparing physical ballots to machine tallies; post-election audits across many states and elections have consistently confirmed machine counts matched paper records. A silent, undetected reprogramming would need to survive exactly this kind of independent physical verification, which audits are specifically designed to catch.

  Copy link
• Claim Statistical anomalies in vote counts are proof of fraud. Election science

  Response

  Vote count patterns that look unusual to a casual observer, like late-reported batches skewing one way, often have mundane explanations, such as when different counties or voting methods (mail-in versus in-person) are reported at different times. Statisticians and election forensics experts evaluate these patterns using established methods, and claims of anomalies from the 2020 U.S. election were reviewed and rejected in dozens of court cases, including by judges appointed by both parties.

  Copy link
• Claim Mail-in ballots are inherently impossible to secure. Election science

  Response

  Mail-in and absentee voting have been used for decades, including for military and overseas voters, with security measures like signature verification, barcodes, and chain-of-custody tracking; multiple studies of mail-in voting have found fraud rates to be extremely low. States with long-standing all-mail voting systems, like Oregon, have not documented widespread fraud.

  Copy link
• Claim Signature verification and paper ballot audits are too unreliable to catch fraud. Election science

  Response

  Signature verification combined with paper ballot audits is specifically designed as a multi-layered check, and post-election risk-limiting audits, which recount statistically significant samples of paper ballots against machine totals, have been conducted in numerous states and consistently confirmed reported results. These processes exist precisely to catch discrepancies, and they've been used to do so when real errors occur.

  Copy link
• Claim Any election result that doesn't match pre-election polling is inherently suspicious. Election science

  Response

  Polls are estimates with margins of error, based on samples and assumptions about who will actually vote, and they have a well-documented history of missing actual results by meaningful margins for reasons unrelated to fraud, like differential response rates or late shifts in voter turnout. A poll-result mismatch reflects polling's known limitations far more often than it reflects fraud.

  Copy link