Skip to main content

Simple Ways to Increase Voter Turnout

Two political science experiments suggest that a prick of social pressure and a dash of old-fashioned Election Day partying could go a long way toward getting America voting again.
  • Author:
  • Updated:

It’s often said that we live in an era of civic decline. Where Have All the Voters Gone? asks one recent book. Another chronicles The Vanishing Voter. Academics and social commentators shake their heads at the cynicism and apathy and wonder, as another book title puts it, Does American Democracy Still Work?

How, then, to get America voting again? Maybe all it takes is a return to the good old-fashioned idea of elections as community events — just like they were in the late 19th century, when upwards of 90 percent of eligible American voters participated. At least, that’s what two recent political science experiments point toward.

One experiment shows that just holding Election Day poll parties would notably increase turnout. But more-significant results come from another experiment, in which a piece of direct mail informed voters that their participation was a matter of public record and that their neighbors would know whether or not they voted.

This prick of social pressure increased voting rates by eight percentage points over the baseline rate, a finding that surprised even the professors behind it — Yale political scientists Donald P. Green and Alan S. Gerber (University of Northern Iowa assistant professor of political science Christopher W. Larimer also worked on the experiment). After all, scholars and campaigns had already studied direct mail extensively and found that it didn’t matter how colorful the mailing was or what it said — nobody could find an effect of more than a percentage point.

But they hadn’t studied social pressure.

“We analyzed it for a week to see if there was some mistake or something was missing,” Green said. “Nothing changed.”

The experiment worked like this: During the 2006 Michigan primary elections, about 200,000 voters got no mailing. This control group voted at a rate of 29.7 percent.

Then came the four mailings, each to 38,000 voters. Group 1 was told, “Do your civic duty and vote!” This increased turnout to 31.5 percent. Group 2 also got the civic duty reminder and then was told, “You are being studied!” (Members of Group 2 were informed that researchers would be watching them, though the results would remain confidential). This increased turnout to 32.2 percent.

Group 3 got the civic duty reminder as well, plus information on whether they voted in the past election. They were also told, “Who votes is public information.” This increased turnout to 34.5 percent.

Finally, and most notably, Group 4 received the civic duty reminder, plus a list of their neighbors’ voting histories — all public information. They were asked: “What if your neighbors knew whether you voted?” This group voted at a rate of 37.8 percent — almost 8 percent more than the control group.

A good part of the explanation seems to lie in the power of social norms. “It’s possible that people simply felt that they were a little more attuned to their civic duty norm of participation once they had the sense they were compiling a track record, and I think that probably had a pretty substantial effect on their incentive,” Gerber said, suggesting that these mailings “might prick the civic conscience of a voter.”

Green called the technique “lightning in a bottle” but warned, “You gotta be very careful. If campaign exerts social pressure, there could be quite a backlash.”

For those who prefer a more carrotlike approach, another set of experiments suggests that hosting an Election Day celebration can also increase turnout — mostly because it gives people another reason to get to the polls, tipping their cost-benefit analysis in favor of participation.

In the spring of 2005, Green, along with James M. Glaser, dean of undergraduate education and a political science professor at Tufts, and Elizabeth M. Addonizio, a political science doctorate student at Yale, organized an “Election Day Poll Party” in a randomly chosen precinct in the quiet town of Hooksett, N.H., during its municipal elections. The event wasn’t anything fancy — some free sandwiches, a cotton-candy machine and a professional DJ playing “upbeat” music, all on the lawn of the local middle school that doubled as a polling place.

But it worked.

Turnout went up.

The trio also sponsored parties in the New Haven, Conn., municipal elections that spring and inspired Working Assets, a long-distance phone company that supports progressive causes, to conduct similar festivals in several cities in 2006. And sure enough, these events brought more people to the polls than otherwise expected. Controlling for past turnout rates, the researchers calculated that a simple poll party in a precinct where 50 percent of voters typically turn out would increase turnout by 6.5 percent — a highly significant result.

“I think we did the best we could with limited resources,” said Glaser. “But if you get this idea to catch on and get some significant funding possible to do this on bigger scale, with the parties being more of a draw, there is a lot of room for this to grow.”

Making politics fun is largely uncharted territory. In part, the professors say, it’s because analysts and academics have been fixated on the idea that the reason people didn’t vote was mostly the costs of voting (i.e., the time and effort it takes), as opposed to the benefits.

“When I was in graduate school, the talk of the town was increasing voter turnout by making it easier for people to register,” said Green, who earned his doctorate from the University of California, Berkeley in 1988. “There were policy innovations of the sort that extended voting during weekends and increased absentee ballots. But right now the sense is the most recent wave of extended voting and mail voting has not even had a 1 percent impact on voter turnout.”

Green is now convinced that the cost side of the voting equation is pretty minimal. “It’s really much more the motivation side, the benefits side, that is a large part of why people go off and vote,” he said.

Making politics fun also goes against the lingering Progressive-era view that elections should be serious, sober affairs, where informed and independent voters come to rational choices — as opposed to the raucous cash- and booze-infused elections that dominated the Gilded Age. But while progressives succeeded in making elections less corrupt, they also squeezed out the excitement, leaving behind what Green called a “morguelike experience.”

“There are other places that celebrate voting in ways that we don’t,” said Glaser. “There are Latin-American countries where Election Day is much more celebratory. And there is reason to celebrate if you’re living in a healthy democracy. I don’t think we’re calling for a return to beer taverns and crookedness of old systems. We are trying to say, ‘Here’s a creative idea about how to promote participation and promote community and put these things together.’”

All the scholars note that these results are preliminary. They expect campaigns, nonprofit groups and other researchers to begin experimenting further with these approaches, just as they did when Green and Gerber published research in 2000 that found that door-to-door canvassing was significantly more effective than phone calls or direct mail. “It’s gonna be darn interesting,” Green said.