Abstract:
Scholars have long noted ideological segregation: Individuals tend to gravitate toward different social media and networks that align with their preferred political stance, as seen in case studies of political polarization in, e.g., the U.S. and China. Does the presence of a single, politically diverse community help reduce segregation? We quantify ideological segregation in this setting using data from Zhihu.com, a Quora-like online community in China that features opinions either aligned with or different from the Chinese government’s “official stance” (i.e., “pro-official” and “anti-official”). We employ classical measures of segregation (e.g., the dissimilarity index) to examine users’ self-selected social interactions under articles with clear pro-official or anti-official stances. We find significant evidence of ideological segregation, even though the community itself is not polarized: Articles—whether pro-official or anti-official—primarily attract comments with the same political stances. Moreover, comments that receive more upvotes tend to align with the article’s political stance. We also discuss the characteristics of both articles and comments in the uncommon cases where users break from ideological segregation.
Dafeng Xu is an Assistant Professor in the Evans School of Public Policy and Governance at the University of Washington.
