Competitive autocracy goes online: analysis of the Turkish case

Research On digital repression has primarily focused on closed authoritarian regimes characterized by centralized censorship and comprehensive surveillance. Less is known about how repression operates in competitive authoritarian regimes where global social media platforms remain accessible, but enforcement is selective. This study argues that, rather than suppressing digital speech uniformly, such regimes exploit legal ambiguity and discretionary enforcement to generate asymmetrical political risk for users. This process skews political expression across user groups and reshapes the visible digital public sphere according to users’ proximity to political power. We examine Turkey’s 2022 act against disinformation, which sharply increased perceived risks surrounding online political expression. Using an original user-day panel of more than 16,000 Twitter users who follow verified local branches of political parties, we estimate user fixed-effects Poisson models for two outcomes: daily retweet volume and political tweeting. The results demonstrate a pattern in which opposition-aligned, left-wing, and high-visibility users significantly reduced online expression, while regime-aligned, right-wing, and low-visibility users do not exhibit comparable restraint and, in some cases, increase online activity. These findings suggest that, much as competitive authoritarian regimes preserve formal democratic institutions while bending them towards incumbent advantage, they preserve formally open digital spaces while tilting visibility towards regime-aligned users