Mandatory voting by mail modestly increases voter turnout without advantaging one party over the other, according to a new study published in the August 28 issue of Science Advances.
The results, which are based on an analysis of 30 years of U.S. county-level data and more than 40 million individual-level voter records from Washington and Utah, support the idea that this alternative to in-person voting could offer a safe pathway to fair elections during the ongoing COVID-19 pandemic.
Based on their analyses of this hefty data pool, the researchers determined that mandatory voting by mail increased Democratic shares of the vote by 0.7% — an effect that was not statistically significant. This indicates the remote style of voting did not have even a modest effect on either party's election performance.
Additionally, the researchers found that mandatory voting by mail increased voter turnout by between 1.8% and 2.9% — a finding consistent with previous research, but on a larger scale.
"We hope that our results will contribute to the discussion currently being had at all levels of government about how to vote in the upcoming elections and beyond in a way that is safe and preserves the values at the heart of our representative democracy," said John Holbein, a researcher at the University of Virginia and the co-author of the study.
As the 2020 U.S. presidential election draws nearer while the COVID-19 pandemic continues transforming everyday activities and public obligations into health risks, the idea of mandatory voting by mail has ballooned into a heated partisan issue. While many experts have suggested mandatory voting by mail could help minimize the spread of the virus during election season while still enabling voters to have their voices heard, opponents suggest it would tip the scales in one party's favor. Recent polls reveal stark divisions along party lines, with only four in 10 Republicans supporting all-mail elections compared with more than eight in 10 Democrats.
While previous studies have investigated the effects of both mandatory and voluntary vote-by-mail initiatives on voter turnout, there has been little investigation into how vote-by-mail policies affect the outcome of elections.
"It seems like a natural question to ask when considering a large public policy change in the way elections are administered," said Michael Barber, an associate professor of political science at Brigham Young University and a co-author of the study. "This, however, is a common problem in studies of the effect of election laws. Many of them only explore voter turnout overall and avoid the partisan implications of these reforms."
To dive beneath the surface of partisan-fueled debates and determine what effect, if any, mandatory voting by mail exerts on election outcomes, Barber and Holbein applied a variety of models to measure voter turnout and partisan vote margins. They pooled nationwide county-level turnout numbers and party vote shares for House, Senate, governor, and presidential elections between 1992 and 2018 from the widely used Dave Leip's Atlas of Elections and measures of total county population and citizen voting age population from the U.S. Census Bureau for the same time period and elections. Together, this formed a data set of more than 42,000 county-year observations. The researchers also analyzed voting record data from Washington (spanning the years 2002 to 2016) and Utah (spanning the years 2012 to 2018).
The researchers honed in on Washington and Utah since these are the only two states that have recently transitioned to a fully-implemented mandatory vote-by-mail system, allowing Barber and Holbein to study the effects of the policy on overall voter turnout and turnout by political party with greater precision and validity than would be possible from broader aggregates of U.S. counties. While these states may not mirror national voter demographics, pairing them allowed the researchers to draw estimates from one state, Utah, where a majority of voters identify as or lean Republican (52.6%) and one state, Washington, where a majority identify as or lean Democrat (50.3%).
Studies that rely on observational rather than experimental data must contend with omitted variable bias, which occurs when individuals in the treatment group (in this case, voters in locations that use mandatory vote-by-mail systems) are different from individuals in the control group (voters in locations that do not use mandatory vote-by-mail). This study's methods were designed to account for variables that might easily be overlooked. For example, the researchers ran a test to ensure that mandatory voting by mail within communities in Utah and Washington had little effect on whether or not someone chose to register. Since it did not, the researchers were assured that differential registration bias was not clouding their results.
"Our rich combination of county-level historical data on election results as well as individual-level historical data on voting allows us to rule out a host of potential sources of omitted variable bias," said Barber.
Although they only identified a modest increase in voter turnout under mandatory voting by mail, Barber and Holbein note that levels of in-person voter participation could otherwise stagnate or decline if citizens must choose between staying home to reduce their risk of contracting COVID-19 or making it out to the polls to vote. Mandatory vote-by-mail policies could permit elections to continue safely as planned without public health consequences.
Barber and Holbein plan to continue their research by investigating which types of voters are most mobilized by voting by mail.
"It remains to be seen whether vote-by-mail narrows or exacerbates gaps in voter turnout by age, race, and socioeconomic status," said Holbein.
[Credit for associated image: Lindsey Bieda/ Flickr]