Why more than 80 million Americans won’t vote on Election Day

If history is anything to go by, we can expect 80 to 100 million voting-age Americans to sit out today’s presidential election.

Consider what happened in 2012. There were 241 million people of voting age, but only 130.2 million actually cast ballots in the general election — a turnout rate of just 58.6 percent. America’s turnout rate has been remarkably stable for the past 40 years, and the betting markets don’t seem to think this year will be much different.

America’s abysmal turnout rate is one of the lowest in the developed world, as this graphic from the Pew Research Center shows. For comparison, Belgium (where voting is mandatory), had a turnout rate of 89.4 percent in its last election. The United Kingdom was at 61 percent. The only wealthy countries we beat out are Japan and Switzerland:

So what gives? Why don’t more Americans vote? I put this question to two political scientists who study voter turnout: Donald Green of Columbia University and David Nickerson of Temple University. They pointed to a couple of broad factors at work:

If you look at the Pew chart above, you’ll notice a quirk: The vast majority of Americans who are registered to vote do show up to the polls, at least in presidential years. In 2012, around 84 percent of registered voters cast a ballot.

The catch is that tens of millions of Americans simply aren’t registered to begin with. Sometimes that’s because they’re ineligible, due to residency requirements or state laws that bar ex-felons from voting. In 2008, the Census Bureau found around 4 million Americans fell into these various categories.

Even more commonly, many people just miss the deadline for registering: Unlike in Sweden or Germany, the US government doesn’t automatically register anyone who’s eligible to vote. Instead, people have to remember to register — and the rules vary from place to place. Minnesota allows you to register on Election Day. In North Dakota, you don’t have to register at all. But in many other states, you have to register well in advance. And the Census found that plenty of people seem to forget.

“A lot of people don’t really get interested in the election until the last few weeks,” says Temple’s David Nickerson. “But by then, it’s often too late.”

There are ways to fix this: One 2013 study led by Barry Burden of the University of Wisconsin Madison found that allowing people to register on Election Day can boost turnout. And, given that voter fraud is basically a nonissue, we could get rid of the bewildering array of voter ID laws that often thwart people from voting. (See this post for the many ways states try to restrict voting rights.)

Still, Nickerson cautions that these efforts, while worthwhile, may not have a huge impact on the overall turnout rate. He has conducted experiments in which everyone in randomly assigned street blocks gets registered automatically. What he’s found, across variations of this experiment, is that for every 10 people who get registered, only about three people end up voting who would’ve otherwise abstained. That suggests registration hassles are only part of the reason people don’t vote.

Voting is undeniably an inconvenience, and there are all sorts of obstacles to casting a ballot, even if you’re registered. The wait at some polling places is unconscionably long (witness these half-mile lines in Cincinnati). Election Day falls on a Tuesday, when people have to work. Many polling places don’t accommodate people with disabilities. We can, and should, make voting much easier.

But if you look at survey data, most people who sit out elections don’t say they were deterred by inconveniences. They say they simply weren’t interested in voting at all, or disliked the candidates, or didn’t care. While we should take self-reported surveys with a grain of salt, there’s something important going on here.

To state the obvious: People are far more likely to vote if they think their vote matters. And that belief is far from universal in the United States. Distrust in institutions is at an all-time high. What’s more, willingness to vote varies a lot by socioeconomic status. More affluent people vote at much higher levels.

Political scientists argue this has to do, in part, with how cultural norms around voting are developed. “In affluent households,” says Green, “the conversations that parents have with their kids are more empowering. They’re told this is something that’s expected of you, this is something that can make a difference. Whereas other people get the message that politics is a kind of abstract, dirty business. So a lot of people come to adulthood with a different understanding of their place in the political system.”

“Say you’re college educated, reasonably affluent, a homeowner with kids,” adds Nickerson. “Then chances are most of your neighbors and the people you speak to also vote — there’s a cultural norm there. Whereas if you grow up in a poorer immigrant population, where many of your neighbors might not even be citizens, there may be less social expectation and social pressure to vote.”

These unequal norms can also become self-reinforcing over time. Wealthy communities vote more, their voices are far more likely to be heard, and it’s much easier for them to think that participating in politics is worthwhile. “Whereas it’s different if you’re in a poorer community and don’t see your problems as something that government can address,” says Nickerson. “You think, our neighborhood has bad schools and a high crime rate no matter who’s in office. So does it even matter who gets elected?”

But that returns us to the question of why US voter turnout is lower than most other countries. Here, Green points to a few structural issues to consider.

First, the United States has many more elections other countries: over a four-year period, Americans are faced with a barrage of federal, state, and local elections. Not only is that fatiguing, but it makes it easier to slip into the habit of not voting at all. (Turnout tends to be significantly lower in non-presidential-year elections.)

Second, most of America’s elections are winner-take-all events in which only two major parties stand a chance. That, too, can depress turnout. Many voters may decide they don’t agree with either of the two parties and stay home. By contrast, in a multiparty parliamentary system, a voter has more options to choose from.

There’s a flip side here. In any given election, campaigns have incentive to turn out voters. But in America’s winner-take-all elections, the two parties don’t necessarily have incentive to turn out all voters. Instead, they focus most on turning out base supporters they know will be reliable votes for their side (think environmentalists or gun owners), and shy away from turning out voter blocs that might vote unpredictably. By contrast, in parliamentary systems with proportional representation, you have a plethora of parties scrambling for votes wherever they can find them.

Short of, uh, massively overhauling America’s political institutions so that they’re more responsive to voters and offer people more options, it’s hard for states and campaigns to massively increase voter turnout. But there are a few tricks here and there.

In 2008, Green and Alan Gerber of Yale University published a book, Get out the Vote, evaluating the evidence around campaign get-out-the-vote operations. One thing they found was that a personal touch matters. Going door-to-door is more effective than phone calls, which are in turn more effective than text messages, which are in turn more effective than mass emails or robocalls (which are basically worthless). Similarly, having enthusiastic volunteers conducting a phone bank is more effective than having paid professionals.

But while this knowledge has been helpful for campaigns searching for votes at the margins, it hasn’t caused America’s turnout rate to soar — at least not yet.

As for policies, some states have tried various tactics to boost turnout, from expanding early voting to Oregon’s experiment with voting by mail to allowing registration later and later in the year. But these, too, seem to be modestly effective at best. “The evidence suggests those have maybe boosted turnout by a couple of percentage points,” Green says, “but I think they’ve been a disappointment to those who originally proposed them.”

Plus, not all states are actually trying to make it easier to vote. In recent years, states like North Carolina have been going in the other direction and trying to suppress turnout. The ACLU has an interactive graph here showing all the new voter-suppression laws on the books since 2012, from Arizona to Wisconsin to Georgia.

There may be a few other offbeat ideas to try. In one 2006 study, Green and his colleagues found suggestive evidence that turning elections into giant festivals of a sort can boost turnout. Note that in 19th-century America, when turnout was much higher, elections had much more of a festive, whiskey-soaked atmosphere. The idea, presumably, is that throwing a giant party can help cement cultural norms around voting. It’s unclear if this would work on a large scale, but it might be worth exploring.

Finally, if we were really serious about getting people to vote, the United States could always try mandatory voting, as seen in Belgium or Australia. “You do see much higher turnout in countries with mandatory voting,” says Nickerson. “The odd thing is that these laws are never really enforced, or the penalties are comically small. They seem to work because they signal to people that this is expected.”

Barring that, personal appeals are about the best we’ve got.