Can a public information campaign increase trust in American elections?

Abstract

Confidence in the conduct and results of American elections has declined and polarized along party lines in recent years. Can a nonpartisan public information campaign featuring election officials enhance trust in elections when delivered to skeptical voters during a contentious political contest? We analyze a preregistered, randomized field experiment conducted during the 2024 US presidential election in which conservative voters were randomly assigned to a control condition or to receive a multimodal messaging campaign designed by political professionals explaining protections on election integrity. In contrast to the large positive effects of informational messages found in recent survey experiments, we observed primarily null effects with respect to turnout (n = 147,488), general confidence in elections (n = 4740 to 4868), and perceptions of the frequency of fraud (n = 4925). We observed one hypothesized effect: an increase in trust in the vote-by-mail process (n = 6438). This study provides rigorous evidence from a real public information campaign on election integrity, yielding effect sizes in the field that are much smaller than those observed in prior survey settings.