Segregation in Social Networks on Facebook
On the morning of November 9th 2016, many democratic-oriented Americans updated their Facebook timelines, unhappy and frustrated about a rather unexpected outcome in the U.S. presidential election. People were particularly shocked that they had not seen this outcome coming. Naturally, they sought explanations for their surprise. Fingers were soon pointed at online “filter bubbles” or “echo chambers,” in which individuals are increasingly surrounded by like-minded people with similar political views or similar race or ethnic background.This phenomenon, however, is far from new. Since the 1950s, sociologists have found that “birds of a feather flock together.” Over the last two decades, scientists have shown repeatedly that people forge friendships with those who are similar to them. This “homophily” has serious implications for the diversity—or lack of diversity—of opinions, attitudes, and prejudices that one may be exposed to.
Our study in the American Sociological Review examined the ethnic and gender diversity of social relationships on Facebook among 2,800 Dutch adolescents, who together had more than a million Facebook friends. What did we find out? Those who have ample opportunities to befriend similar others—for instance, those who belong to a racial or ethnic majority group—will find themselves in highly segregated Facebook networks. Additionally, making friends on Facebook is highly affected by the schools, neighborhoods, or associations one belongs to: if these are segregated, Facebook networks will segregate. Of course, this is rather intuitive: adding a Facebook friend is likely to result from having met someone in an offline context.
In this light, it is not such a surprise that people experience rather homogeneous Facebook networks. It turns out that segregation on Facebook and, hence, echo chambers or filter bubbles reproduce our highly segregated offline settings. If people surround themselves only with people of similar backgrounds, this will be mirrored in a set of Facebook friends exclusively of these backgrounds.
Bas Hofstra is a postdoctoral research fellow at Stanford University. His co-authors on the original article are Rense Corten and Frank van Tubergen in the department of sociology at the Utrecht University and Nicole Ellison, on the faculty at the University of Michigan School of Information.