The Washington PostDemocracy Dies in Darkness

Voting only by mail can decrease turnout. Or increase it. Wait, what?

Voters in Denver, Colo., cast their ballots at the Denver Elections Division Building on Nov. 4, 2014. (Marc Piscotty/Getty Images)
Placeholder while article actions load

Voting by mail — and only by mail — has become an option in the United States. Will it spread?

According to the National Conference of State Legislatures, all states will mail an absentee ballot to voters who request one. While 20 states require a reason, 27 states permit “no-excuse” absentee voting.

And three states now use mail-only voting.

Oregon’s Ballot Measure 60 kicked off in 1998, making Oregon the first state  to conduct its elections exclusively by mail. In 2011, Washington’s legislature moved the state to an entirely vote-by-mail system. Colorado joined in during the 2014 general election. In 2015, California launched a limited all-mail pilot as a test run. Lawmakers will use that pilot to learn how such an election would work in California.

Supporters hope that voting by mail means more citizens will vote. Is it so?

Generally, the answer is both “no” and “yes,” but with important qualifications.

The recent research on how all-mail voting affects turnout 

Here’s how voting by mail works. For these elections, all registered voters automatically receive a ballot in the mail. The voter marks the ballot, puts it into a separate mailing envelope, signs an affidavit on the exterior of the mailing envelope and returns the package via mail. Ballots are mailed out well ahead of Election Day, typically about a month in advance. Ballots must be postmarked or returned on Election Day.

Some early research in Oregon claimed that voting by mail increased turnout by 10 percentage points. However, since then, scholars have been unable to reproduce those results. Apparently that boost to Oregon’s turnout grew from a “novelty effect” and recurred only in special elections.

In Washington, researchers found that switching to all-mail elections increased overall participation by about three percentage points in presidential and midterm elections. In the California pilot, after the Nov. 3 elections, the San Mateo County elections office received 105,325 ballots out of the approximately 353,000 that were mailed. That’s 29.5 percent voter turnout, or 4.1 percent more than a similar off-year polling place election in 2013, when 25.4 percent of registered voters cast their ballots.

That pilot is a big deal. Just two counties out of California’s 58 were allowed to test this approach. If it’s successful, it could be rolled out to the most populous state in the union, with 15 million registered voters.

The media was quick to attribute the “eye-popping” increase in voter turnout to simply switching to vote-by-mail. But it’s not that simple.

Mail-only balloting actually decreases voting

My research found that when you can only vote by mail, voter turnout actually drops by about 13 percent. I examined what happens to turnout if voting by mail is compulsory. I studied more than 90,000 voters who could vote only by mail across four elections from 2006 through 2008 in five of the most populous urban counties in California. (In that state, if a precinct has fewer than 250 voters, elections officials are allowed to forego a polling place and accept ballots only by mail.)

That decline may seem counterintuitive. Presumably voting by mail is easier and more convenient than going to the polls. So why doesn’t turnout go up?

According to a 50-state study that examined elections over a 30-year period, voter turnout is less about convenience than academics once thought. Most voting reforms, like all-mail balloting, do not attract new voters.

What’s more, alternative voting methods are most likely to be launched in states that already have high voter turnout.

Why does voting by mail decrease turnout? Because mail voters have a longer voting “window,” they receive less stimulus to vote. Scholars have found that reductions in stimulation to vote are greater than the modest positive benefits of additional convenience from mail voting.

But reminders make a difference

Reminders are critical. My research found that when the elections office communicates more often with voters, more of them vote. In particular, four official communications can wipe out the 13 percent decrease in turnout that I found. ‘‘Official communications’’ include such documents as a Sample Ballot, a Voter Guide, letters on county letterhead and postcards from the Registrar of Voters. Each additional communication improved the odds of voting by 4 percent. And a voter who received five communications was 4 percent more likely to vote than a voter who received no mailings.

That’s what happened during the pilot in San Mateo in California. County officials sent out plenty of notices about the changed rules for November’s election, letting people know they would be getting a ballot in the mail whether they liked it or not. According to Mark Church, the county’s chief elections officer, that sealed their success.

That response varied by demographics. Asians and Latinos saw the largest gains from the switch to all-mail voting in San Mateo. Again, Church commented on the county’s efforts related to boost voting in all quarters of the population:

San Mateo County made a big push to reach voters who prefer languages other than English, particularly Spanish and Chinese. There were radio and TV ads, social media campaigns and billboards. Bilingual staff members were sent to community events to encourage people to register and participate.

Switching to mail-only balloting could have reduced Asian turnout by about 30 percent and Hispanic turnout by about 27 percent, according to my findings. Obviously, what San Mateo’s election officials did mattered to Asian and Latino voters in the county.

A significant proportion of the San Mateo pilot voters decided that they preferred voting by mail. A third of those who’d been voting at a polling place switched their registration status and became mail-in voters permanently; 42 percent of voters between 18 and 24 did so. To the surprise of some, few disgruntled voters complained about getting ballots in their mailboxes.

Actually having the experience of voting by mail made an enormous difference in how a voter felt about that method. When I first asked, more than half of all polling place voters disapproved of expanding mail balloting in California. That was significantly more disapproval than I found among those who were mail voters — a gap of 25 percent between polling place and mail voters.

However, once polling place voters have actually tried voting by mail, 63.8 percent of them support mail-only elections — the same percentage I found among people who had already voted by mail. And of those who’d previously voted in a polling place exclusively, but who tried out voting by mail, 77.7 percent of voters said they planned to vote by mail in the next election.

Three states – Oregon, Washington and Colorado — have made the move to all-mail elections; 19 states are the likely next adopters as they permit all-mail elections in certain circumstances. For these states and their voters, my research suggests that compulsory vote-by-mail is not an unalloyed good. It can boost turnout, but only when officials work toward that end.

Elizabeth Bergman is associate professor of political science at California State University, East Bay.