Reports suggest that voter fraud is very rare in the United States and mail-in ballots will not substantially increase the risk of voter fraud.
A study conducted by Stanford University’s Democracy & Polarization Lab concluded that neither party would benefit from an entirely mail-in-voting system, and it does not affect either party’s share of turnout, rather it increases the average turnout rates. Further, a Washington Post review of data collected from the 2016 election found just four confirmed cases of voter fraud: three people who tried to vote for President Trump twice, and an election worker in Miami who was caught trying to fill in someone else’s ballot for a local mayoral candidate.
A study titled ‘An Analysis of Voter Fraud in the United States’ published by Demos in 2003 analyzed the incidence of voter fraud from 1992 to 2002 in 12 states, which collectively represent about half of the electorate. The study found that voter fraud was very rare in those 12 states and legal and news records had little evidence of significant voter fraud. In conclusion, the report indicates that fraud is a minor problem in the U.S.
Therefore, the claim that mail-in voting 'substantially' increases the risk of fraud is false.