• by Erik Sass , Staff Writer @eriksass1, October 21, 2014

Conventional wisdom has it that the American political scene is becoming more polarized, with liberals and conservatives increasingly entrenched in their own views and unwilling to consider those of the opposing side. This is attributed in large part to a tendency to “choose your news,” as partisans on both sides retreat into media echo chambers where they only hear reporting that agrees with their opinions.

Indeed this week the Pew Research Center released survey data showing that Americans with consistently liberal or conservative opinions (estimated at about a fifth of the population) tend to get their news from very different sources -- almost all of which are distrusted by the opposite side. No surprise, Fox News and Rush Limbaugh dominate among right-leaning folk, while left-leaning crunchy types gravitate to MSNBC, NPR, and The New York Times.

Of course this ideological self-segregation is not super news for democracy, insofar as it makes political compromise more difficult, if not impossible. But there’s a glimmer of hope from an unexpected place: social media. That’s according to a new study by an NYU PhD student, Pablo Barberá, titled, “How Social Media Reduces Mass Political Polarization. Evidence from Germany, Spain, and the U.S.”

In essence the study found that, contrary to what many experts think, social media actually exposes you to a broader range of opinion -- and this exposure occurs precisely because you don’t know many of your online “friends” very well.  

Barberá writes: “First, social media platforms like Facebook or Twitter increase incidental exposure to political messages shared by peers. Second, these sites facilitate exposure to messages from those with whom individuals have weak social ties… which are more likely to provide novel information. This induces political moderation at the individual level and, counter intuitively, helps to decrease mass political polarization.”

To prove this Barberá developed a new method to measure the ideological positions of Twitter users over time, which he then applied to users in the U.S., Germany, and Spain. This complex model assumes that Twitter users prefer to follow political figures who are close to their own political orientation. Barberá compiled a list of Twitter users who follow at least one political figure, as well as a list of all followers for every political account in each country, then matched these with existing voter files to weed out fake profiles and eliminate duplicates. Finally he compiled a list of all the accounts these Twitter users followed, in order to create a full portrait of their total network. Altogether in his analysis Barberá looked at 1.2 billion tweets from 4.4 million users in the U.S., 175,000 users in Germany, and 825,000 users in Spain.

Barberá found that, “contrary to the conventional wisdom, most Twitter users in Germany, Spain, and the United States are exposed to a high degree of political diversity in their personal networks. In fact, one-third (33%) of the average U.S. Twitter users network consisted of people who varied from them ideologically, and over 75% are embedded in networks where at least 25% of the connections vary ideologically.

Furthermore, Barberá found that “individuals who receive politically diverse messages become less extreme over time.” Here, he analyzed the individuals in his sample at two points in time -- January 2013 and July 2014 -- and found that U.S. users who were exposed to opposing views from just 15% of their networks became more moderate over time (again, as measured by which political figures they choose to follow on Twitter).