The Social Network Roots of Partisan Polarization in the United States, and Beyond

Partisan polarization is one of the principal features of contemporary US politics.  The research that will be reviewed in this presentation shows that the partisan homogeneity of social networks is related to it and that their importance has increased over time.  The study draws upon national post-election surveys of the US electorate in 1992, 2004, and 2012 and a discussion network battery of questions developed in the Comparative National Election Project (CNEP).  Network partisanship is measured from respondents’ reports of the party favored by their core network: spouse/partner (where they exist) and two additional personal discussants selected using a “most important matters” name generator.  Partisan polarization is measured by votes for president and the House of Representatives and relative evaluations of the presidential candidates.  For each, the more homogeneously partisan the core network, the more partisan are the votes and candidate evaluations of Democratic and Republican partisans.  In the United States, partisans’ network homogeneity (homophily) and its relationship to partisan behavior both have increased between 1992 and 2004-2012, contributing to the growth of partisan polarization in its politics.  Drawing upon cross-national surveys in the CNEP, I also find that the role of social networks has become stronger in the U.S. in recent years than in many democracies, even some with a recent history of deep partisan cleavages.