Censorship on social media: The gatekeeping functions of shadowbans in the American Twitterverse
60 Pages Posted: 29 Apr 2022
Date Written: April 20, 2022
Algorithms play a critical role in steering online attention on social media, and many have alleged that algorithms can perpetuate bias. This study audited the shadowbans on over 25,000 Twitter users who posted 3.84 million tweets in 2020-2021 to identify the type of user and tweet characteristics that predict a `shadow ban' -- where a user or their content is temporarily hidden on Twitter. We find that users with bot-like behavior are more likely to be missing in searches (search bans), search suggestions (search suggestion bans), and reply threads (ghost bans). Users with verified accounts were less likely to be shadowbanned. The replies by Twitter users who posted offensive tweets, and tweets about politics (from both the left and the right) were more likely to be downtiered. The findings have implications for algorithmic transparency and accountability and the design of future audit studies of social media platforms.
Keywords: platforms, Twitter, censorship, audit, shadowbans, text analysis
Suggested Citation: Suggested Citation