By John Carey, Brendan Nyhan, Brian Fogarty,  Jason Reifler

This study reports several experiments testing the e ects of corrective messages debunking false claims of widespread fraud in the 2020 and 2022 U.S. elections as well as a complementary experiment after the 2022 Brazilian presidential election. We find evidence that prebunking false voter fraud claims with substantive information about election security can reduce misperceptions more effectively than corrections from credible sources. Other results indicate that corrections of specific voter fraud claims fail to generate broader changes in perceptions of election integrity and that party (but not putative candidate race) is the major factor in perceptions of voter fraud at the Congressional race level. 

Cambridge, MA: MIT Election Lab, 2024.