Social Media Echo Chambers and Political Polarization: A Network Analysis

Abstract

The echo chamber hypothesis—that social media algorithms and user homophily create ideologically segregated online environments that amplify polarization—has become one of the most influential and contested claims in contemporary sociology and political science. Originating in Sunstein’s (2001, 2017) concept of echo chambers and Pariser’s (2011) filter bubble, the hypothesis predicts that social media users primarily encounter politically congruent information, reinforcing pre-existing beliefs while limiting exposure to opposing views. However, a growing body of computational social science research suggests that the empirical evidence for echo chambers is more nuanced than the popular narrative implies: while ideological sorting is measurable on social networks, its magnitude may be smaller than feared, and cross-cutting exposure remains prevalent for many users. This paper critically reviews the network analysis literature on echo chambers, distinguishes algorithmic from behavioral drivers of ideological segregation, and proposes a framework for measuring echo chamber strength across platforms.