Measuring centralization of online platforms through size and interconnection of communities

Abstract

Decentralization of online social platforms offers a variety of potential benefits, including divesting of moderator and administrator authority among a wider population, allowing a variety of communities with differing social standards to coexist, and making the platform more resilient to technical or social attack. However, a platform offering a decentralized architecture does not guarantee that users will use it in a decentralized way, and measuring the centralization of socio-technical networks is not an easy task. In this paper we introduce a method of characterizing inter-community influence, to measure the impact that removing a community would have on the remainder of a platform. Our approach provides a careful definition of “centralization” appropriate in bipartite user-community socio-technical networks, and demonstrates the inadequacy of more trivial methods for interrogating centralization such as examining the distribution of community sizes. We use this method to compare the structure of five socio-technical platforms, and find that even decentralized platforms like Mastodon are far more centralized than any synthetic networks used for comparison. We discuss how this method can be used to identify when a platform is more centralized than it initially appears, either through inherent social pressure like assortative preferential attachment, or through astroturfing by platform administrators, and how this knowledge can inform platform governance and user trust.