The Effects of Group Sanctions on Participation and Toxicity: Quasi-experimental Evidence from the Fediverse

TL;DR

This study investigates defederation, a decentralized group sanction in the Fediverse, and its effects on participation and toxicity. Using a quasi-experimental design with coarsened exact matching and difference-in-differences across 214 defederation events, the authors analyze weekly posting activity and toxicity (via the Perspective API) for 661 accounts on 275 servers. They find asymmetric effects: accounts on blocked servers reduce activity after defederation, while accounts on blocking servers show little change, and there is no detectable impact on toxicity for either group. The findings suggest decentralized sanctions can curb cross-server engagement without broadly suppressing on-server participation or increasing toxic behavior, offering practical insights for governance in federated online ecosystems.

Abstract

Online communities often overlap and coexist, despite incongruent norms and approaches to content moderation. When communities diverge, decentralized and federated communities may pursue group-level sanctions, including defederation (disconnection) to block communication between members of specific communities. We investigate the effects of defederation in the context of the Fediverse, a set of decentralized, interconnected social networks with independent governance. Mastodon and Pleroma, the most popular software powering the Fediverse, allow administrators on one server to defederate from another. We use a difference-in-differences approach and matched controls to estimate the effects of defederation events on participation and message toxicity among affected members of the blocked and blocking servers. We find that defederation causes a drop in activity for accounts on the blocked servers, but not on the blocking servers. Also, we find no evidence of an effect of defederation on message toxicity.

Figures (8)

• Figure 1: Illustration of how defederation disconnects two servers and thereby disconnects the subnetworks of people using each server. The top row shows the network of servers before (left) and after (right) defederation. The bottom row shows the corresponding networks of users. On the left, an edge connects a user on one server with a user on a different server. Defederation (right) disconnects them so they can no longer exchange messages.
• Figure 2: The y-axis shows the cumulative number of blocked and blocking accounts included in our analysis over our study period.
• Figure 3: A screenshot shows a cross-server interaction on Mastodon. Note that the account name of the replying user indicates their home server (@user@server).
• Figure 4: A covariate balance plot shows the standardized mean difference between treatment and control groups for each measure used in our matching procedure before (unadjusted) and after (adjusted) matching. Our procedure effectively found a group of matched controls similar to the treated accounts along these measures.
• Figure 5: Box and whisker plots visualize the distributions of our dependent variables within the blocked and blocking groups of user accounts and their matched controls before and after defederation. The lines correspond to the median, the boxes to the inter-quartile range (IQR), the whiskers to the range of the data within 1.5 * IQR, and the dots to data points outside the range of the whiskers.