Humans are social creatures. They’re also, according to the people that study democracies, social voters. So if you want your friends to vote — good idea, given the high stakes of this election — shout from every mountaintop and Facebook update that you’re voting. In a post at Quartz, Olivia Goldhill points out several of the mechanisms at work, two of which we’ll get into here.
Weirdly enough, research indicates that people are less likely to go vote if they think the turnout will be low. Yale University political scientist Alan Gerber and Harvard behavioral scientist Todd Rogers argue that this is the power of “descriptive social norms”: that, essentially, as they note in a paper for the Journal of Politics, “a person’s behavior conforms to her beliefs about what people actually do in a given situation” — that they will do what they consider to be “normal,” if everybody’s doing it, they will too.
To test this, the researchers had telemarketers make get-out-the-vote phone calls in the days preceding the November 2005 gubernatorial election in New Jersey and the June 2006 primary election in California. Some would-be voters were were told that turnout would be high, others that turnout would be low. Following the descriptive-social-norms theory, people who heard the high-turnout script reported that they were more likely to vote. Having it framed as the thing people do encouraged them to do it.
An even weirder example comes from a large-scale Facebook experiment that was set up around the 2010 congressional elections and published in Nature in 2012. The design was simple enough: 611,000 (or 1 percent) of users saw an “informational message” at the top of their news feeds telling them that they should vote, giving them a link to information about their polling place, an “I voted” button they could click, and a counter showing all the Facebook users who had clicked thus far. A cool 60 million users (or 98 percent of the group) got a “social message,” which had all those elements but also half a dozen photos of randomly selected friends who themselves had clicked the “I voted” button. (A third group, the control group, was made up of the 1 percent of users who saw no message at all.)
While the people who got the informational message had the same rates of voting as the control group, the users shown the social message were, in the researchers’ analysis, .3 percent more likely to find info about their polling place and 2 percent more likely to click on “I voted” button. They were also .4 percent more likely to actually vote than the other groups.
While those percentages might seem small, the scale of Facebook yields big results: The research team, led by Robert Bond, now at Ohio State University, finds that the social message directly led to an increase of 60,000 voters. What’s more, all that clicking on “I voted” meant that their voting announcements showed up in their friends’ news feeds, indirectly leading to another 280,000 voters. This follows with what other researchers, including Yale epidemiologist Nicholas Christakis, has christened “social contagion,” a rather robust theory that indicates that behaviors radiate out through networks of people, whether it’s innovative ideas, gun violence, smoking cessation, or weight gain. (TED Talk here). In politics, as in so much of life, people mimic their friends. For better or worse, everybody’s doing it.