Cascades Across Networks Are Sufficient for the Formation of Echo Chambers: An Agent-Based Model

Abstract

Investigating how echo chambers emerge in social networks is increasingly crucial, given their role in facilitating the retention of misinformation, inducing intolerance towards opposing views, and misleading public and political discourse (e.g., disbelief in climate change). Previously, the emergence of echo chambers has been attributed to psychological biases and inter-individual differences, requiring repeated interactions among network-users. In the present work we show that two core components of social networks—users self-select their networks, and information is shared laterally (i.e. peer-to-peer)—are causally sufficient to produce echo chambers. Crucially, we show that this requires neither special psychological explanation (e.g., bias or individual differences), nor repeated interactions—though these may be exacerbating factors. In fact, this effect is made increasingly worse the more generations of peer-to-peer transmissions it takes for information to permeate a network. This raises important questions for social network architects, if truly opposed to the increasing prevalence of deleterious societal trends that stem from echo chamber formation.