Where Does the Hate Flow? The Impact of Multihoming on User Responses to Content Moderation

Abstract

Despite the significant resources that online platforms allocate to content moderation, both regulatory authorities and users consider these efforts to be insufficient in protecting against online hate and harassment (H&H). A challenge arises from spillover effects, where users respond to platform sanctions by increasing harmful behaviors elsewhere. This complexity is heightened by users’ tendency to multihome - simultaneously engaging with multiple online platforms - and by social contagion within online networks. Sanctioned users may also migrate to external alternative platforms, further complicating matters. This study investigates how external platform adoption, multihoming, and contagion affect content moderation’s externalities. Using a robust empirical methodology, we examine the aftermath of banning a group high in H&H. By matching users across platforms, we observe that the ban significantly increased the adoption of an external alternative platform and multihoming. Moreover, the ban triggered H&H spillovers on both the focal (in ideologically aligned groups) and external platforms. Notably, multihoming reshaped these spillovers, causing users to also target ideologically opposed groups on the focal platform. Our findings reveal that these H&H spillovers were contagious, spreading to users uninvolved with the banned group. Spillovers initiated by multihomers in ideologically opposed groups exhibited the highest contagion. These results contribute to the Information Systems literature on content moderation, multihoming, and online social contagion. We underscore the societal significance of addressing externalities in content moderation policies and offer actionable design recommendations for platforms to mitigate these effects while accounting for external platform adoption, multihoming, and contagion dynamics.

Keywords: Content moderation policies, hate and harassment, multihoming, contagion