Casting and Curing Vote-by-Mail Ballots in the 2020 General Election

• In Event: Election Administration: Impact of Voting Rules in the States

Abstract

Who casts rejected vote-by-mail (VBM) ballots and what types of voters successfully cure deficiencies with their mail ballots? Answers to these questions were a major focus of the 2020 General Election, when a record 65.6 million voters cast mail ballots, up from roughly 33.4 million in 2016. The dramatic increase in VBM ballots across the country was due in large part to health concerns stemming from the COVID-19 pandemic. Rules governing VBM ballots—who is eligible to request and cast a mail ballot, how and when must these ballots are returned, what information and signatures are required on a ballot's return envelope, and how may voters cure deficient return envelopes—vary considerably across states and in the lead-up to the November 2020 election several states, either proactively or due to lawsuits, eased their restrictions on the use of mail ballots. Not all VBM ballots cast, however, were accepted as valid by local election officials. In the 2016 election, 318,728 VBM ballots, roughly one percent, were rejected for some deficiency. We compiled data from state election officials showing (so far) at least 246,745 mail ballots were rejected in 21 states. The overall VBM rejection rate in these states is at least 0.6 percent, ranging from 0.1 percent in Colorado to 2.5 percent in Maine. In several states, we have collected individual level absentee ballot files, which allows us to probe not only why such variation exists, but more importantly whether certain voters (younger, first-time voters, racial and ethnic minorities) are more likely to have their ballots flagged for rejection. In some states these individual-level data include if VBM ballots were initially rejected and subsequently cured, allowing us to examine VBM cure rates for specific groups of voters. Our preliminary analysis finds younger, less experienced, and racial and ethnic minorities to have higher VBM rejection rates; but likely due to the national media attention and efforts on the ground by voting rights organizations, these subsets of voters were able to cure their ballots at rates comparable to other groups of mail voters. In a few states, we can further determine how long it took certain groups of voters to cure their ballots that were flagged for rejection upon initial receipt. Our findings suggest that not all voters have equal opportunities when casting, and curing, their VBM ballots.

Authors

• Daniel A. Smith, University of Florida

• Michael C. Herron, Dartmouth College

• Michael P. McDonald, University of Florida