If the coronavirus pandemic rages on, most Americans will probably vote by mail in November. But like most political issues in the U.S., voting by mail is an increasingly partisan affair, with Democrats more likely to support it than Republicans.
The fight over voting by mail isn’t new. But if the political cues around voting by mail weren’t firmly set before, they are now, with President Trump calling voting by mail “corrupt” and pushing Republicans to fight efforts to expand it. And this has many election experts concerned as the partisan fighting, last-minute scrambles and inevitable litigation could make 2020 an even more uncertain election.
But before the political fighting gets too ugly, lawmakers really ought to look at the evidence. Numerous studies have arrived at the same conclusion: Voting by mail doesn’t provide any clear partisan advantage. In fact, as states have expanded their use of mailed ballots over the last decade — including five states that conduct all-mail elections by default — both parties have enjoyed a small but equal increase in turnout.
In short: voting by mail is more convenient for some voters but more difficult for others, and these conflicting factors appear to cancel each other out, dampening any partisan advantage. Moreover, the vast majority of nonvoters don’t participate not because it’s too inconvenient to vote, but because voting isn’t a habit for them. Maybe they don’t care about politics, maybe they don’t think their vote matters, maybe they don’t like any of the candidates, or maybe it’s some combination of all of the above. But the bottom line is that these voters’ decision to vote depends more on whether somebody around them can motivate them to vote, not whether they are able to vote by mail or in person.
We’re going to walk through the evidence that demonstrates there isn’t a partisan advantage of voting by mail, then look at 2016 results from two key swing states — Florida and Wisconsin — to help illustrate this. But first, let’s quickly unpack what we mean when we say “voting by mail.”
At its most basic level, voting by mail means that instead of going to the polls, you get a ballot mailed to your home, you fill it out, and then you mail it in. (Though in some states, it’s much more complex than that.)
There are three main ways states vote by mail. First, there are “universal” vote-by-mail states, which is when all registered voters automatically receive a ballot, but this is fairly rare. Just five states currently conduct all-mail elections by default: Colorado, Hawaii, Oregon, Utah and Washington (although a handful of states allow certain elections to be conducted by mail or some of their counties to vote by mail). Next, there are “no-excuse” vote-by-mail states, which is the most common method of voting by mail in the U.S. — 29 states currently use it in federal elections. Under this system, any registered voter can vote by mail, but must first request a ballot. And finally, there are 16 states that are “excuse-only” vote by mail, which means voters must provide a reason as to why they can’t show up at the polls in order to get their ballot. Although some states might expand the types of excuses accepted in 2020 — the coronavirus is already a valid excuse for general election voters in New Hampshire, and in Texas there’s a lawsuit over whether the coronavirus should be considered a valid excuse for the state’s July primary runoff elections.
So keeping those differences in mind, let’s take a look at the voting-by-mail landscape in 2016. In that presidential election, states saw anywhere from about 2 percent of ballots cast by mail (West Virginia, Kentucky, Tennessee) to 97 percent (Washington, Oregon).
On balance, states that voted for Hillary Clinton were slightly more oriented toward voting by mail than states that voted for Trump. For instance, of the states that voted for Clinton, the median state reported 8.2 percent of ballots cast by mail (Nevada), compared to 7.7 percent of ballots cast by mail in the median state that voted for Trump (technically, the average of Texas, 7.5 percent, and South Carolina, 7.9 percent).
This might sound like evidence that voting by mail does benefit Democrats, but it’s important to keep in mind that these are small differences, and in most states, it’s still a really small share of the voting-eligible population that has cast a ballot by mail in recent elections. We’ll later address why voting by mail has had such a minimal partisan effect, but for now, let’s get a better understanding of what voting by mail looks like in key swing states. These states’ elections are so competitive that it’s unlikely their voting rules change before 2020, as any rule change would be highly contested. But because all the states we’re classifying as swing states1 allow any voter to vote absentee, we can still dive into the mail-in vote in two battleground states in 2016 and 2018: Florida and Wisconsin.
In Florida, where just under 30 percent of voters cast their ballots by mail in 2016, the first thing to know is that more Democratic counties tend to cast a higher share of ballots by mail. As you can see in the chart below, the share of voters in a county casting their ballot by mail increases as Trump’s voting share declines.
Because more Democratic counties take greater advantage of voting by mail, the first impression might be that voting by mail benefits Democrats. However, examining which ballots were returned in 2016 show that counties that Trump won actually saw a higher share of requested mailed ballots get counted.2
Florida Republicans voted by mail more successfully
Total absentee ballots requested, ballots cast and the share of ballots successfully cast by party in Florida in the 2016 presidential election and the 2018 midterm elections
|2016 presidential election||Ballots…|
|2018 midterm elections||Ballots…|
Between 2016 and 2018, the share of Florida voters casting ballots by mail increased from about 28.5 percent to 31.6 percent, with most counties seeing an uptick in the share of votes cast by mail. But as you can see in the table above, it was still Republicans, not Democrats, who had a higher rate of ballots returned successfully.
In short, Florida Republicans do a better job of completing their ballots than Florida Democrats. However, there may be reasons that this is true other than Democrats not being as diligent in mailing in their ballots. Research by political scientist Daniel Smith, conducted for the American Civil Liberties Union of Florida, found that younger voters and voters from racial and ethnic minorities are much more likely to have their absentee ballots rejected by state officials. It’s hard to know the exact impact of this, but in a close election, even a small number of rejections could be decisive.
That said, we didn’t find a relationship between the change in the share of ballots cast by mail in a county and its change in vote share from 2016 to 2018, suggesting neither party gained from an uptick in ballots cast by mail.
So in 2020, we shouldn’t expect either party to automatically benefit from a higher rate of voting by mail. Instead, we should expect the party that mobilizes its voters to better take advantage of voting by mail to benefit more. And despite Trump’s public aversion to voting by mail, Florida Republican operatives have learned how to work with, not against, voting by mail. As one Florida Republican campaign veteran recently told NPR: “Absentee ballots are typically Republicans’ friends in Florida.”
So now let’s look at Wisconsin, which has a much lower rate of voting by mail than Florida — just 9 percent voted by mail in 2016. Its recent primary fiasco brought conflicts over voting by mail into sharp relief, and Democrats ended up taking advantage of it more than Republicans, presumably because the state Democratic party clearly communicated the urgency of voting absentee. But up until the 2020 primary, there hasn’t been evidence of a county-level relationship between partisanship and ballot success rate in Wisconsin. In fact, as was the case in Florida, the change in county-level absentee voting from 2016 to 2018 had no relationship to the change in a county’s partisan vote share.
Studies that have analyzed the effects of voting by mail on partisan fortunes on a larger scale have found similar results. In a recent study, a team of Stanford University political scientists looked at county-level implementation of vote by mail in California, Utah and Washington. They found no statistically significant partisan difference between counties that had transitioned to voting fully by mail and those that had not, along with a slight increase in voter turnout. Another research team recently analyzed the effects of voting by mail in Colorado, and although they found a much higher effect on overall voter turnout than previous studies, they also had a similar null partisan finding. More voters turned out to support Democratic candidates, but more voters also turned out to support Republican candidates.
So what do we make of the lack of evidence that any one party benefits from voting by mail? The best way to think about this is to imagine the marginal voter who didn’t vote in person but would vote by mail. Because most voting is habitual, most voting rules don’t seem to make a huge change in who votes. So to the extent that voting by mail makes voting more convenient, it helps those who are already motivated to vote but who may not be able to find time on Election Day for whatever reason. This is why studies generally find that even universal voting by mail (where everyone is sent a ballot) only increases voting marginally, and probably among those who already are most likely to vote anyway.
Bottom line: By making it a little easier to vote, voting by mail probably increases the likelihood of the marginal Democratic voter engaging in the process. (Though younger and lower-income voters, who tend to vote at lower rates, also tend to not take advantage of voting by mail.) But it also makes it easier for more habitual older voters, who tend to vote more Republican than younger voters, to cast a ballot. Thus, on balance, any associated partisan effects from voting by mail have tended to cancel out.
We should be careful to apply past patterns to 2020, though. The states that moved to universal voting did so gradually, over several election cycles. So we’ve never seen anything on the scale of what we might see in 2020. The obvious implication is that efforts to expand absentee voting in a pandemic might work differently. And maybe there will be partisan differences in who chooses to vote by mail, as we saw in Wisconsin’s primary.
But we may also learn something more about how states implement voting-by-mail systems and what those impacts are. (For example, is postage prepaid? How easy is it to request a ballot? How easy is it to correct a rejected ballot?) We may also see that different campaign tactics are more effective in getting people to vote by mail than getting people to vote in person. At the very least, we’ll almost certainly see tremendous variation on both counts — variation that will give us a new cottage industry of studies that refine our understanding of how vote by mail impacts turnout, or at least how it impacted turnout in 2020.
Looking at Florida’s 2020 primary, held on March 17 (before a statewide shelter-in-place order was in place, although other states were issuing orders), may hold a few clues. Not surprisingly, the share of primary voters casting a ballot by mail increased from 31 percent in 2016 to 48 percent, according to an analysis by MIT political scientist Charles Stewart. And it was Republicans — not Democrats — who were more likely to mail their ballot (57 percent of Republicans compared with 41 percent of Democrats) even though it was Democrats who had a contested primary, and therefore, higher turnout.3 In 2016, there was no partisan difference.
But the more important clue may be Stewart’s finding that it was people who had voted by mail in the 2016 primary were much more likely to vote in the 2020 primary. In total, he found, 62 percent of voters who voted by mail in the primary in 2016 voted in 2020, compared with just 35 percent of those who voted in person on Election Day in 2016. This is in line with previous findings that voting by mail benefits habitual voters most. But this may also suggest that those who are in the habit of requesting a mail ballot may have an advantage if states shift to all-mail elections — the electoral ramifications of which aren’t entirely clear.
Likewise, the Wisconsin 2020 primary may also offer some clues as for what to expect in the general election, especially if voting by mail becomes an increasingly polarized issue. Turnout was higher than expected in the Wisconsin primary, and Democrats do appear to have taken more advantage of voting by mail than Republicans (though as with Florida, Democratic turnout was presumably higher because Democrats had a contested presidential primary, whereas Republicans didn’t). Granted, there were other down-ballot elections, including a judicial election for the state Supreme Court which was one reason the primary couldn’t be postponed, but Democratic party leaders also led a strong effort to get Democratic voters to request mail ballots.
That final point is a crucial one, too, in trying to anticipate what the effects may be on the general. Voters rely on parties for cues on how and when to vote. So if mail ballots are going to be a part of 2020 (and beyond), campaigns will adopt strategies to educate and encourage voters to request and fill out ballots — even if they don’t like the rules. And in a fluid voting environment like the one we expect for 2020, whether Democrats or Republicans benefit more will almost certainly have less to do with the actual rules, and more to do with which campaigns most effectively organize their voters to take advantage of their ability to vote from the comfort of their homes.
CORRECTION (May 15, 2020, 4:46 p.m.) A previous version of this article cited research by political scientist Daniel Smith, conducted for the American Civil Liberties Union of Florida, that suggested that the small portion of ballots received in Florida could have been attributed to ballots that were not mailed back; however, the research this drew on did not include ballots that the state did not receive.