Resources

While the evidence is clear that 2020 voters shifted away from Election Day voting in favor of vote-by-mail and early voting, very is known about how health risk versus party polarization around risk assessment influenced how and when to vote. The research finds that age and party were large factors in vote mode decisions in 2020, but not in 2016 or 2018.

The research finds that counties that moved to send registered voters mail-in ballots ahead of Election Day experienced 2.6 percent higher turnout compared to counties that made no change, although lesser reforms may have hindered turnout. Additionally no evidence is found that making voting by mail easier conferred a partisan advantage.

These results suggest that making it easier to vote by mail—especially mailing every voter a ballot—generally does increase turnout, both before and during the 2020 election. By contrast, the same policies do not have robust partisan effects, and in many models, they tilt the results in a more Republican direction.

The research finds that information about possible coronavirus exposures decreases comfort with voting in-person yet does not increase comfort with voting by mail.

Drawing on a large survey of validated Florida voters, including those who regularly vote by mail, the research finds that retrospective and prospective misreporting of vote method prior to the 2020 General Election was driven primarily by support for Trump. The president’s supporters who were most politically aware were most likely to disavow their own voting by mail and misreport their anticipated vote method in the November election.

Using surveys of registered voters conducted in April and May 2020 designed to assess the support for various electoral reforms, the research shows that the overall support for expanding VBM in April 2020 falls sharply after just six weeks because Republicans became less worried about catching COVID-19, and unconcerned Republicans became far more opposed to VBM. These differences not only persisted, but actually increased even further between May and Election Day according to a survey done using a different methodology in November 2020.

This paper challenges the HAVA minimum of one ADA-compliant device per polling location as insufficient and recommends that local election officials determine the number of accessible devices based on the proportion of voters with disabilities in their jurisdiction.

In this paper, authors provide recommendations from disability voting rights advocates on how to improve the physical accessibility of polling locations and the usability of accessible voting equipment.

This paper employs discrete-event simulation to model Milwaukee's in-person voting system during COVID-19. It reveals that poll worker shortages, social distancing measures, and PPE requirements can lead to very long voter wait times. The evaluation considers various design strategies to reduce pandemic-related effects, such as adding check-in locations, expanding early voting, and preventing the consolidation of polling sites.

Hostetter examines whether the use of electronic poll books affects voter wait times, finding mixed results that depend on context, including photo ID requirements and precinct demographics.

In this PhD dissertation, Bernardo develops a simulation-based framework to examine how in-person election system conditions affect voter wait times and throughput. Three core contributions: (1) observational time studies of voting operations across multiple Rhode Island elections to calibrate arrival-rate and service-time parameters; (2) discrete-event simulation models of polling-place operations under varying equipment types, layouts, and resource levels; and (3) application of the models to COVID-19 social-distancing scenarios, precinct-consolidation decisions, and minimum requirements for accessible equipment.