Validating Vote-by-Mail Ballots: The Subjective Nature of Signature Verification

Abstract

Vote-by-mail (VBM) is an established electoral practice across the United States. Millions of voters cast VBM ballots every election, but roughly one percent of them are rejected by local election officials, according to data compiled by the U.S. Elections Assistance Commission. Although scholars have examined a variety of issues associated with mail voting, including postal delays and signature deficiencies, few studies have probed the subjective nature of validating VBM ballots, specifically matching a registered voter’s signature on record with a signature on a VBM return envelope. In an original survey of over 3,000 registered voters in Florida, we tasked respondents with evaluating a set of eight similar signatures in a modular grid, asking them if they thought there was a pair of matching signatures. We offer several explanations—demographics, life experience, partisan politics, and ideology—for why some individuals tasked with evaluating pairs of signatures might be less likely than others to report a pair among the set. Our findings suggest that signature validation, which serves as a primary safeguard for mail voting integrity, may be systematically influenced by underlying biases. Since the task of signature matching is largely in the eye of the beholder, the fervent support for and against mail voting is likely to endure, creating concerns for election integrity in elections to come.