Chapter 4 — Clustering and Polarization in Social Networks

Abstract

This chapter investigates how homophily shapes network formation and the evolution of political positions among agents with heterogeneous views. Agents form pairwise-stable networks by linking preferentially with similar others, and political positions evolve through a modified DeGroot updating process.

Numerical analysis demonstrates that such networks naturally give rise to clusters of like-minded agents, producing persistent opinion polarization and impeding the emergence of global consensus. The results illustrate how initial opinions and network structure jointly influence the long-run distribution of political positions, offering insight into the mechanisms that sustain polarized communities.