The Washington PostDemocracy Dies in Darkness

Letting people vote at home increases voter turnout. Here’s proof.

A ballot drop-off box in Denver during Colorado’s 2014 election, the first one there where people could vote at home. (Ivan Couronne/AFP/Getty Images)

It's no secret that American democracy is in a bad way. In the last midterm elections, only about 37 percent of eligible adults voted — the worst showing since World War II. Trump fever may push that number up this November, or it may not. The Alabama Senate election in December, in which Democrat Doug Jones won a surprise victory, was heralded for its unusually high turnout for a special election — yet even that was only 40 percent of registered voters and less than 30 percent of eligible adults, showing how low our expectations have fallen. The problem is especially acute for Democrats, because participation is typically lower among minorities, the poor and the young. But it ultimately hurts both parties, because hyper-low turnout in primaries empowers the most extreme candidates at the expense of the party establishments — as the GOP learned in Alabama with Roy Moore. 

Over the past few years, Washington Monthly, where we work, has published a series of articles arguing that the simplest and most effective way to address these problems is to adopt universal vote by mail, also known as "vote at home." The idea is simple: Instead of going to a designated polling place, every registered voter automatically receives a ballot in the mail. They have a window of time, typically two weeks, to fill it out and either mail it back or drop it off at a secure site. 

By making voting dramatically more convenient and changing it from opt-in to opt-out behavior, vote at home promises to boost turnout, especially among the people who today vote the least, and to broadly expand participation in lower-profile, non-presidential elections. Yet only a handful of states have begun to use it. Republican political leaders fear that it would benefit Democrats; meanwhile, some Democrats aren't convinced that vote at home would help them. Some even worry that it could help the other side. They point to the 2014 election in Colorado, the state's first using a vote-at-home system, in which Democrat Mark Udall lost his Senate seat. 

But there's never been concrete data to test either side's interpretation. So last summer, we commissioned a study by the political research firm Pantheon Analytics, using data from Colorado's 2014 election for more than 2.8 million registered voters. The study found that vote at home increased overall turnout by 3.3 percent, and by even more among young and low-propensity voters. The implication is clear: Anyone who cares about improving turnout should make expanding vote at home a top priority.

The challenge in trying to evaluate the impact of vote at home on turnout is that there's never a control group. Turnout rates in the states where everyone can vote at home — Oregon, Washington and Colorado — have increased since the system was adopted, and they're now among the highest in the country. But because elections are complicated, and each state's demographics are unique, it's hard to prove that vote at home is the cause. 

 So Pantheon Analytics did the next best thing. The firm looked at voter files — the records of individual voters kept by state elections officials — in Colorado in 2014 and compared them with the predictions of a respected voter turnout model created by the Democratic-leaning consulting firm Clarity Campaign Labs. Turnout models use publicly available voter file information — age, sex, voting history — combined with consumer data to generate a probability score for each registered voter. 

With the benefit of hindsight, we know that the Clarity model was fairly accurate nationwide in 2014. But because it was based on past elections, it couldn't account for changes to a state's electoral system. So if the 2014 results in Colorado, which was implementing vote at home for the first time, deviated from the predictions, it would be reasonable to infer that vote at home was the reason — especially if voting was up among people with low predicted turnout. 

That's exactly what happened. Nationally, the Clarity model slightly overpredicted turnout. But in Colorado, it was the opposite: Turnout was 3.3 percentage points higher than the model forecasted. Even more significant, that increase was mostly driven by a huge boost among people with the lowest turnout probability scores. Registered voters assigned a 10 percent chance of voting, for instance, turned out at a 31 percent rate. 

  Who were these overperforming voters? Overwhelmingly, they were young. The 18-to-24-year-old bracket, forecast to turn out at 26.6 percent, actually turned out at 38.7 percent. And 25-to-34-year-olds outperformed their prediction by 7.4 percentage points. 

 Those results should have Democrats salivating. While there are many reasons they struggle to win elections — candidates and campaigns matter too, of course — Democrats' biggest hurdle is turnout. If young people voted as regularly as seniors, Democrats would win a whole lot more elections. (In Alabama last month, Jones won voters under age 30 by 22 percentage points.) A boost like Colorado's could have major implications if replicated nationwide. 

How Trump’s delusions of voter fraud will turn election results into fake news

 At the same time, the results broken down by partisanship are more ambiguous. When analyzed by party affiliation, vote at home seems to have benefited Republicans slightly more: Registered Republicans outperformed the model by 3.7 points , compared with 2.8 points for registered Democrats. But voters with no party affiliation outnumber both registered Democrats and Republicans in Colorado, and they outperformed the model by 3.1 points. 

 An analysis according to a Clarity predictive partisanship model yielded similarly inconclusive results. The model scores people from 0 to 100, where 0 is "definitely voting Republican" and 100 is "definitely voting Democrat." People pegged as near-certain Republican voters beat turnout predictions by 2.3 points, while near-certain Democrats outperformed by 1.8 points. Again, that's an apparent (but tiny) advantage for Republicans.  

But the biggest turnout boosts were toward the middle of the spectrum. Voters with a partisanship score between 60 and 70 — likely Democratic leaners — outperformed the model by 7.9 points. Those in the 40 to 50 range did so by 6.1 points. This group would be classified as weak Republican leaners, potentially still persuadable for a savvy Democratic campaign. Younger people are probably overrepresented in that middle zone, because they have less survey data and voting history to use  to assign them a partisanship score.

 Vote at home clearly seems to have pushed young people and infrequent voters into the electorate — exactly what our ailing system needs. Democrats should be fighting tooth and nail to get it implemented wherever possible. But the partisanship results are hazy enough that Republicans shouldn't be too scared: It looks like these low-frequency voters tend to be less partisan and therefore more up for grabs. 

 Besides, boosting turnout in lower-profile elections would be good for the long-term health of both parties. Low turnout may help Republican candidates prevail in general elections, but it also empowers the right-wing fringe, leading to a party that, once in power, is incapable of governing — as the GOP's inability to prevent a shutdown, despite unified control of the federal government, recently illustrated. 

 Meanwhile, vote at home offers tangible benefits for voters of any party affiliation. Because it relies on old-fashioned pen and paper, it can't be hacked. While ballots are counted by machine, those machines don't need to be connected to the Internet, and a paper trail is there for a recount. Since 2000, tens of millions of mailed-out ballots have been cast in Oregon, Washington and Colorado, without a hint of serious fraud or other mischief. (As a check against possible fraud, election officials match signatures on the ballots against those on voters' registration cards.) And counties stand to save millions of dollars per election rather than paying for poll workers and voting machines or renting voting locations, money that can be used to provide other essential services. Perhaps we would lose something intangible by doing away with brick-and-mortar polling locations and "I voted" stickers — an element of performative civic participation. But that would be a small price to pay for increasing actual civic participation. 

 So, unless you adore waiting in lines on Election Day, what's not to like? Adopting a vote at home system nationwide could go a long way toward reviving our democracy. But to make that happen, more states have to follow the example of Colorado, Oregon and Washington. 

Twitter: @giladedelman

Read more from Outlook:

Want to thank black voters for defeating Roy Moore? Tackle voter suppression.

I'm on Trump's voter commission. I sued it to find out what it was doing.

Follow our updates on Facebook and Twitter