News Release

How do social media influence ethnic polarization?

New study in Bosnia and Herzegovina finds deactivation of Facebook accounts linked to lower regard for out-groups

Peer-Reviewed Publication

New York University

Those who deactivated their Facebook profiles report a lower regard for other ethnic groups, and this effect was more prevalent among people living in more ethnically homogenous areas, shows a new study of users in Bosnia and Herzegovina (BiH). The findings run counter to a commonly held view that social media usage exacerbates societal polarization.

The work, conducted by researchers at New York University's Center for Social Media and Politics (CSMaP), appears in the Proceedings of the National Academy of Sciences (PNAS).

"For all our attention to the online drivers of polarization, we should not forget about the importance of offline factors as well," observes Nejla Asimovic, a doctoral candidate in NYU's Department of Politics and the lead author of the paper.

While a majority of Americans see social media as having a negative effect on the way things are going in the U.S. today, according to a recent survey by the Pew Research Center, and see it as fomenting polarization, the impact of social media on inter-ethnic attitudes has yet to be rigorously evaluated.

In the PNAS research, the paper's authors conducted an experiment in early July of 2019, coinciding with the 24th anniversary of the Srebrenica genocide. This period commemorating the 1995 atrocities--resulting in the deaths of over eight thousand Bosniak Muslims at the hands of Bosnian Serb forces--was chosen because of the heightened discourse around the past conflict during the studied days (July 7-July 14).

Participants were recruited through Facebook advertisements in BiH using both the Cyrillic and Latin alphabets. The more than 350 participants included those who identified themselves as Bosniaks (58.9 percent), as Serbs (15.7 percent), and as Croats (6.5 percent). Approximately 13 percent of respondents chose to identify as Bosnians and nearly 5 percent opted to not report their ethnic identification.

The subjects were randomly assigned to two groups: one whose Facebook accounts remained active during the studied period (the control group) and one whose accounts were deactivated during this time. Deactivation was confirmed through the monitoring of users' Facebook URLs; the control group was instructed to continue to use the platform as it normally would.

After the studied period, users filled out a questionnaire in which they were asked not only about attitudes toward those of other ethnic groups in the region (out-groups), but also about their knowledge of current events ("news knowledge") and about their well-being (e.g., feelings of loneliness, isolation, and joy).

Surprisingly, those in the group who deactivated their Facebook accounts reported more negative attitudes about ethnic out-groups than did those in the group who continued to use the platform. In response to these unexpected findings, the researchers turned to a question that had asked participants what they did in the time they were off Facebook.

"The most popular response was that people spent more time with friends and family," said NYU Professor Joshua A. Tucker, a co-author of the study. "This led us to suspect that perhaps our findings were being driven by people who were spending more time offline with people of their own ethnic group."

To test this intuition, the researchers decided to examine whether the effect of Facebook deactivation in driving worse out-group attitudes was more prevalent among people living in ethnically homogenous areas of the country. Notably, this is exactly what they found: these effects were largely concentrated among those who live in more ethnically homogeneous environments--and whose offline environments were therefore likely to be more ethnically homogeneous than their online environments. Moreover, these effects were not found among users living in parts of the country that were more ethnically mixed.

"Our research suggests that social media experience can be particularly influential in shaping out-group attitudes where the experiences of offline contact is low, especially in contexts of limited media fragmentation and no language barrier between groups," notes Asimovic. "We should keep in mind that offline environments or states' rhetoric may be as divisive, if not more, than online environments that may still allow people to engage--directly or indirectly--with the out-group."

In addition, they found that Facebook deactivation led to a significant decrease in the levels of news knowledge, but an improvement in users' subjective well-being--consistent with an earlier study of U.S. Facebook users.

"Our findings suggest that simply deactivating from social media is not a panacea to ethnic polarization, especially if the offline environment provides little to no opportunities for positive intergroup contact," says Asimovic. "Given these results, future work should be mindful in making assumptions about social media's impact and consider, with it, contextual factors and opportunities for intergroup contacts."

###

The paper's other authors were Jonathan Nagler, a professor in NYU's Department of Politics, and Richard Bonneau, a professor in NYU's Department of Biology and Courant Institute of Mathematical Sciences. Tucker, Nagler, and Bonneau are co-directors of the NYU Center for Social Media and Politics, where Asimovic is a graduate research associate.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.