Can Social Media Help Build Communities?
28 Pages Posted: 19 Mar 2018 Last revised: 22 Aug 2018
Date Written: August 21, 2018
Abstract
Since the mid-1990s, political polarization has been on the rise in the U.S. as Democratic and Republican parties have become increasingly dichotomized ideological structures, where loyalists are either for or against a particular issue. While claims of differences may be more media hyperbole than realistic, affective polarization, where there is increased animosity between parties, is on the rise (Abramowitz & Webster, 2016; Iyengar, Sood, & Lelkes, 2012; Iyengar & Westwood 2015; Mason, 2015). Drawing from theories in the fields of social psychology, political science, and network science, we investigate the extent to which community-building is possible on social media platforms, where partisanship has forced Americans to choose sides on politically charged issues. Analyzing 81,316 tweets circulated on the day of and following the Day of Action for Net Neutrality on July 12, 2017, we test whether social media platforms can assuage the often heated offline tensions on politically charged topics and present brokers that can help online users work through their differences, or toward compromise. In support of research that suggests that social media platforms are mere “echo chambers” for existing groups, we also conclude that this will not change unless the construction of human social networks are considered, alongside the existing technical architecture of social media platforms that maintains the dominance of like-minded associations. Tempering the online influence of political leaders, particularly legislators and organizations may also foster more productive conversations. Thus, we conclude with suggestions for how social media companies can work towards the goal of community building by connecting users with brokers, individuals who span the divide between politically disconnected communities.
Keywords: net neutrality, social media, social network analysis, prejudice, Twitter
Suggested Citation: Suggested Citation