Regular users of Facebook may be so weary of political messages in their news feeds that they try to tune them out, or wish they could. But a recent study from the University of California at San Diego, conducted with the cooperation of Facebook, found that the social-media service could effectively influence hundreds of thousands more voters to head to the polls.
The lead author, James H. Fowler, a professor of medical genetics and political science at UC-San Diego, said that the study, published in Nature, had been devised “to see whether or not a political message might not just affect your behavior, but might spread through the network and affect your friends’ behavior, and that’s exactly what we found.”
During the 2010 Congressional elections, more than 60 million Facebook users who logged in on November 2 received a nonpartisan “get out the vote” message in their feeds, with an option to click on an “I Voted” button, a link to local polling places, and pictures of Facebook friends who had already indicated that they voted. Other Facebook users received a similar message with no pictures of friends; a control group received no message at all.
Researchers found that users who received the “social” message were 2.08 percent more likely to click the “I Voted” button than were those who received the “informational” message. By matching 6.3 million Facebook users to their public voting records, researchers also found that users who received the social message were more likely, albeit by less than 1 percent, to actually wind up in a voting booth than were those who received the informational message or no message.
Researchers estimated that encouragement via that one social message led to an additional 60,000 votes in the 2010 election.
The impact of the social message grew as the study looked at its effects on friends of recipients. Researchers found that “close friends” (as determined in part by frequency of Facebook interaction) of users who received a social message were more likely to vote than were close friends of users who received no message, even if the friends had not received the social message themselves. Once again, the percentage of increase was less than 1 percent. There was negligible effect from friends with which users interacted less frequently.
Thanks to the network effect of Facebook, with multiple social-message recipients influencing multiple close friends, researchers estimated that the carry-over effect of the social message generated an additional 282,000 votes.
While Facebook is built on interpersonal ties, many of them loose, Mr. Fowler said that the “effect on real-world behavior change was being spread through these ties that were likely to be friendships in real life. They only account for about 7 percent of the ties on Facebook, but they were responsible for 100 percent of the effect of the message spreading through the network.”
Political candidates and parties have been making use of social media to reach out to potential voters for years. Mr. Fowler said that the study indicated that a particular message may appear to fall flat among those directly contacted, but could have more of an effect via social media than is currently understood.
“I think they’re not taking advantage of information they have about people’s social networks,” Mr. Fowler said, “to estimate the true effects of their messages.”