Polarization and echo chambers in Reddit's political discourse

2025-10-31

AI Summary

This study challenges the prevailing assumption that Reddit is more resistant to political polarization than Twitter or Facebook, focusing on political discourse during the 2016 U.S. presidential election. Method: Leveraging multi-source data—including user interactions, news consumption patterns, and linguistic features—we construct a statistically validated two-dimensional (interaction-topic) community model to dynamically track ideological community evolution. Our approach advances beyond conventional single-network analysis by jointly quantifying inter-community interaction strength and intra-community linguistic diversity. Contribution/Results: Contrary to expectations, the election period induced only transient cross-ideological engagement while significantly reducing intra-community linguistic diversity—empirically confirming strong echo chamber effects. We robustly identify three distinct clusters: Democratic, conservative, and banned subreddits—demonstrating that Reddit functions as a highly polarized discursive arena. The study underscores the critical role of network-validation techniques in comparative platform research.

Technology Category

• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Natural Language Processing: Discourse, Pragmatics & Argument Mining
• Application Domains: Humanities & Computational Social Science

Application Category

• Social Networks and Social Media: Media and governance, opinion dynamics, filter bubbles, polarization
• Web Mining and Content Analysis: Models for Web evolution
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Political debate nowadays takes place mainly on online social media, with election periods amplifying ideological engagement. Reddit is generally considered more resistant to polarization and echo chamber effects than platforms like Twitter or Facebook. Here, we challenge this assumption through a case study across the 2016 US presidential election. We use statistical validation techniques to extract ideologically distinct communities of subreddits, in terms of their contributing user base and news consumption, which we use to analyze the dynamics of political debate. We thus reveal clear polarization in both interaction-based and topic-based communities, with clusters of Democratic, Conservative, and Banned subreddits. Election periods intensify cross-group engagement, align Banned and Conservative content, and reduce linguistic diversity within groups. Overall we characterize Reddit as a polarized environment marked by the presence of echo chambers, highlighting network validation as a key method for identifying behavioral and interaction patterns on online social media.

Problem

• Analyzing political polarization and echo chambers on Reddit
• Examining ideological communities during the 2016 US election
• Identifying reduced linguistic diversity in polarized groups

Innovation

• Used statistical validation to extract ideological communities
• Analyzed polarization through interaction and topic clusters
• Applied network validation to identify behavioral patterns