No, voter fraud is not a significant problem in the US. It occurs at extremely low rates that do not affect election outcomes, based on extensive audits, academic studies, court reviews, and official investigations.
Evidence on Scale
•Documented rates are tiny: Analyses of billions of votes cast over decades find fraud (in-person impersonation, double voting, non-citizen voting, etc.) in the range of 0.0001% to 0.0025% or lower in most jurisdictions. The Heritage Foundation’s database (which compiles proven cases and is often cited by those concerned about fraud) lists roughly 1,500 instances nationwide across many years—serious when they occur, but negligible against hundreds of millions of ballots per cycle. Academic reviews (e.g., from Loyola Law School, MIT Election Lab, and others) and state audits consistently show the same order of magnitude.
•2020 and recent elections: Claims of widespread fraud were examined by Republican and Democratic election officials, attorneys general, the Department of Justice, CISA (which called the 2020 election “the most secure in American history”), bipartisan state canvassing boards, and over 60 courts (including judges appointed by both parties). Virtually all were dismissed for lack of evidence of outcome-changing fraud. Isolated irregularities and errors exist in every large election; coordinated or systemic fraud does not.
•Common alleged vectors are rare or mitigated:
•Non-citizen voting: Multiple studies (including from the Brennan Center and state-level checks) find it vanishingly rare; voter rolls and eligibility checks catch most issues.
•Dead voters / double voting: Usually administrative lag or clerical error, not intentional fraud; cross-checks and provisional ballots limit impact.
•Mail ballots: Fraud rates remain low where tracked; chain-of-custody, signature verification, and bipartisan observation are standard. Expansion of mail voting in 2020 did not produce detectable widespread fraud.
•In-person impersonation: The hardest and least common form, precisely because it is high-risk and low-reward.
Why the Perception Differs from the Data
High-profile anecdotes, partisan rhetoric, and isolated prosecutions create an impression of scale that statistics do not support. Error (wrong address, outdated rolls, mismarked ballots) is routinely conflated with intentional fraud. Safeguards—voter ID in many states, provisional ballots, post-election audits, paper trails, poll watchers, and felony penalties—already make large-scale fraud difficult and detectable. Improving those processes (better roll maintenance, risk-limiting audits, clearer chain of custody) is worthwhile; claiming the system is routinely “stolen” is not supported by evidence.
Election administration has real challenges (long lines in some places, varying state rules, occasional local incompetence or machine glitches), and maintaining public trust requires transparency. But the empirical record from officials, courts, and researchers across the political spectrum is clear: voter fraud is real, prosecuted when found, and not a material threat to the integrity of US election results.