Neither the Republicans or Democrats would gain an advantage if every voter in US elections was given the opportunity to vote by mail, according to a new statistical analysis.
The debate surrounding postal voting in the US has come to a boil this year, as the country determines how best to hold a presidential election in the midst of a pandemic.
Voting by mail has been put forward as a solution to staging an election while minimising the risk of spreading the coronavirus, but President Donald Trump has claimed it would result in voter fraud and give his Democrat opponents an unfair advantage.
Michael Barber at Brigham Young University in Utah and John Holbein at the University of Virginia devised a model to look at the effect of voting by mail, by combining county-level voting data from 1992 to 2018 with US census records.
In 2018, only 175 of the 3100 or so countries in the US had switched to mostly or entirely staging their elections by postal vote, including all those in Oregon and Colorado.
The pair first modelled how election turnout and party vote shares morphed as these counties switched to postal voting. They also compared turnout and vote share in postal-voting counties with counties that still held in-person elections. Using all this data, they were able to estimate what would happen to the average county if its election was conducted by mail.
Barber says the model works because the counties that switched to postal voting are representative of the country. “Some are more conservative, while others are more liberal,” he says. “You also have both rural and urban counties.”
The model showed that giving everyone a postal vote would increase voter turnout by 2 per cent. It also showed that the Democrats would have a non-statistically significant increase in their vote share by 0.7 per cent. As there was no effect at a county level on election outcome, the pair argue that mail-in voting would have no effect at a national level either.
“The conventional wisdom that an increased turnout gives an advantage to Democrats is incorrect,” says Barber.
He also says that the low rise in voter turnout suggests that the convenience of voting by mail will have little effect on those who wouldn’t normally vote anyway.
As this data only goes up to 2018, it is unclear how much it can tell us about the 2020 presidential election. Due to the coronavirus pandemic, at least three-quarters of voters will be eligible to receive a ballot in the mail, according to an analysis by The New York Times.
“These findings suggest that when policymakers are evaluating whether to adopt a vote by mail system, they need not focus on its consequences for their own re-election,” says Marc Meredith at the University of Pennsylvania.
Journal reference: Science Advances, DOI: 10.1126/sciadv.abc7685
More on these topics: