Labelling, shadow bans and community resistance: did meta's strategy to suppress rather than remove COVID misinformation and conspiracy theory on Facebook slow the spread?

Abstract

In this paper, we ask how effective Meta's content moderation strategy was on its flagship platform, Facebook, during the COVID-19 pandemic. We analyse the performance of 18 Australian right-wing/anti-vaccination pages, posts and commenting sections collected between January 2019 and July 2021, and use engagement metrics and time series analysis to analyse the data, mapping key policy announcements against page performance. We combine this with content analysis of comments parsed from two public pages that overperformed in the time period. The results show that Meta's content moderation systems were partially effective, with previously high-performing pages showing steady decline. Nonetheless, some pages not only slipped through the net but overperformed, proving this strategy to be piecemeal and inconsistent. The analysis identifies trends that content labelling and ‘shadow banning’ accounts was resisted by these communities, who employed tactics to stay engaged on Facebook, while migrating some conversations to less moderated platforms.