Who Censors What, When and Why? Findings from a New Event Dataset of Censorship

• In Event: Censorship, Content Moderation, and Government Accountability

Abstract

Alongside democratic backsliding, security threats and the rise of the ‘fake news’ phenomenon, censorship has been increasingly used by various actors to control the media. Who are the censors and the censored? When and how does censorship occur? What gets censored? What is its relationship to other tools of media capture such as the economic control of outlets or bribery? Our hypotheses revolve around individual outlet characteristics, political institutions, economic development, and a substitution effect between coercion-based (censorship) and inducement-based (bribery) strategies employed by censors. The findings come from a new micro-dataset of 9,000 hand coded individual media censorship events around the world between 2001 and 2015. Outlets that are less state dependent (foreign, located abroad, or aligned with opposition parties) are at a disproportionate risk, if censored, to be targeted by the government compared to other potential censors. When informational taboos are less clear (economic news or election-time reporting), pro-government media is at increased risk of censorship. Government censorship in general seems to work through the top of the news production hierarchy (editors, owners) while individual journalists are at higher comparative risk of targeting by third parties and when reporting on certain topics such as protests or elections. On a macro-level, democracies and richer countries censor less than their autocratic counterparts, but the result is primarily driven by political content. We also find that coercion and inducements are substitutes, with censorship being employed primarily when bribes and ownership fail to control information.

Author

• Cristina Corduneanu-Huci, Central European University