But these notions are patently incorrect, according to a study that reaffirms and expands upon years of prior research: No party benefits when a state switches to universal vote-by-mail.
The question has taken on new relevance in the context of the novel coronavirus pandemic, when in-person voting raises health risks to both voters and poll workers.
Universal vote-by-mail typically works like this: Ballots are automatically sent to all registered voters, who in turn are supposed to fill them out and mail them back to election officials. In some states, such as Colorado, voters have the additional option of either dropping off ballots at specified locations on Election Day or voting in person at those same locations.
According to the National Conference of State Legislatures, Colorado, Hawaii, Oregon, Washington and Utah have universal vote-by-mail. A sixth state, California, allows its counties to implement vote-by-mail at their discretion.
In a number of these states, vote-by-mail was rolled out at the county level on a staggered basis. This setup allowed a team of political scientists at Stanford University to conduct a natural experiment: “By comparing counties that adopt a vote-by-mail program to counties within the same state that do not adopt the program, we are able to compare the election outcomes and turnout behavior of voters who have different vote-by-mail accessibility but who have the same set of candidates on the ballot for statewide races,” they write.
To test partisan effects, they compiled a data set containing county-level election results as well as public voter file data containing the party registration of voters in California and Utah. The data covered elections from 1996 to 2018.
After controlling for county-level differences, the data showed “a truly negligible effect” on partisan turnout rates. The effect on partisan vote share was similarly indistinguishable from zero.
One thing they did find was a modest boost in across-the-board turnout. “Vote-by-mail causes around a 2-percentage-point increase (estimates range from 1.9 to 2.4 percentage points) in the share of the voting-age population that turns out to vote.”
The Stanford study, however, “should increase our confidence in these views,” the authors write, “both because our data permits a stronger research design than was previously possible and because our data set runs through the 2018 midterm elections, allowing for the most up-to-date analysis available.”
The study does come with one big caveat: It can’t predict what (if any) effect universal vote-by-mail might have in the midst of a pandemic, simply because that situation throws so many additional variables into the mix. Would older people fearing for their health be more likely to participate with vote-by-mail, thus creating a partisan advantage for Republicans? Would otherwise disaffected younger voters be more likely to vote, creating an advantage for Democrats? It’s impossible to say.
Nonetheless, the findings do underscore that partisan opposition to vote-by-mail over electoral concerns is a standpoint with no basis in empirical reality.