Authors

Grigorii Kiryukhov²; Daniil Romanov¹; Evgeny Sedashov²; Artemi Artivua²;  ¹ Oxford University, UK;  ² Personal capacity, Russian Federation

Discussion

In informational autocracies, controlling narratives and shaping public opinion are primarily achieved through propaganda and censorship. While the overt actions of the state are evident, the subtle dynamics of self-censorship in online platforms remain largely underexplored. To address this gap, our study turns to Telegram in Russia, a seminal example of informational autocracy, providing an invaluable context to delve into the intricacies of self-censorship. Historically, comprehensive studies on online censorship within Telegram have been scant, attributed to the complexities of data collection and the limited analytical tools available. Utilizing an extensive 8-terabyte dataset, which encompasses posts from a majority of Telegram channels in the Russian domain, we systematically identify instances of post deletions. These deletions serve as clear markers of self-censorship. Our analysis further extends to various channel-level factors, such as ideological orientation, audience size, competition for audience attention, and network surroundings, assessing their influence on the magnitude of self-censorship in a channel. Through this rigorous examination, our research unveils the mechanisms by which informational autocracies might indirectly exert pressure, compelling individuals or entities to self-censor. Our findings significantly contribute to a comprehensive understanding of the nuanced methods through which informational autocracies operate in digital spaces, particularly in driving self-censorship behaviors.