Online Homogeneity Can Emerge Without Filtering Algorithms or Homophily Preferences

Abstract

Ideologically homogeneous online environments - often described as "echo chambers" or "filter bubbles" - are widely seen as drivers of polarization, radicalization, and misinformation. A central debate asks whether such homophily stems primarily from algorithmic curation or users' preference for like-minded peers. This study challenges that view by showing that homogeneity can emerge in the absence of both filtering algorithms and user preferences. Using an agent-based model inspired by Schelling's model of residential segregation, we demonstrate that weak individual preferences, combined with simple group-based interaction structures, can trigger feedback loops that drive communities toward segregation. Once a small imbalance forms, cascades of user exits and regrouping amplify homogeneity across the system. Counterintuitively, algorithmic filtering - often blamed for "filter bubbles" - can in fact sustain diversity by stabilizing mixed communities. These findings highlight online polarization as an emergent system-level dynamic and underscore the importance of applying a complexity lens to the study of digital public spheres.