User Suspension and its Consequences for Online Behavior
Speaker: Gloria Gennaro (University College London) with Laurenz Derksen (ETH), Andy Gues (Princeton University), and Dominik Hangartner(ETH)
Wednesday, January 22, 14:00–14:45 (Irish time)
Please register (opens in a new window)here to receive the link and password to the online meeting and information on the room at UCD.
Abstract: Banning toxic content creators from online platforms is a widespread yet contested policy. On the one hand, banning reduces the chances that new harmful speech is produced, on the other it entails a restriction on free speech. In this study, we aim at better understanding the first component of this trade-off: we look at whether and how suspending toxic users extends beyond its first-order effect, to affect their followers’ behavior on Twitter. We track 1000 ordinary (i.e. non-celebrity) toxic users for over 3 months, and observe how their followers’ posting patterns change after their suspension, compared to other unaffected users. Focusing on the suspensions of ordinary accounts allows us to infer suspension effects that are not confounded by media events surrounding the suspensions. For the average follower we do not find significant effects of account suspensions. For followers that are more highly exposed to suspended seed accounts, we find that experiencing a suspension reduces users’ activity and, to a lesser degree, their toxicity. Those effects apply to all type of posts, including replies and original tweets. We conclude that, although user suspension effectively prevents the creation of further toxic content, it also significantly silences users.
About the speaker: Gloria Gennaro is an assistant professor in Public Policy and Data Science at University College London’s Department of Political Science. Prior to that, she was a postdoctoral fellow at the Public Policy Group and Immigration Policy Lab at ETH Zurich. She holds a PhD in social and political sciences from Bocconi University. Her research in comparative political economy explores electoral behavior in democratic societies, using causal inference and computational social science.