Political Institution, Information, and Authoritarianism

This dissertation consists of three distinct articles that address two important but understudied questions in authoritarian politics: (1) how the effects of nominally democratic institutions on regime survival are conditioned by other traditional dictatorial toolkits and (2) how institutional changes are associated with information manipulation and media reports.In the first paper, I examine how different authoritarian institutions interact by showing that subnational elections’ effect on regime survival is conditional on the free flow of information. I argue that the benefit of subnational elections for regime survival is conditional on a lack of media freedom: As the level of media freedom increases, the positive influence of holding subnational elections on regime survival decreases. This is because subnational elections provide local politicians with opportunities to build good reputations, and when good reputations formed at the local level spread to other jurisdictions via relatively free media, citizens can use them as a focal point to coordinate against the regime. I find empirical support for my theory using the quantitative analysis of Time-Series Cross-Sectional data.The second paper examines the conventional wisdom that autocrats engage in more repression after successfully circumventing term limits, a popular personalization tool in contemporary autocracies. I argue that the answer is only a partial yes. First, the evasion is followed by an increase in covert repression (information manipulation) but not overt repression because the impending threat comes from diffuse and less explicitly identified masses. Moreover, and somewhat paradoxically, this increase is more likely to occur after term limit extensions, the less severe form of term limit evasion, than after term limit removals, the more severe form of evasion. This is because the removal serves as a costly signal about regime capacity that dissuades the masses from protesting, substituting repression. Using time-series cross-sectional data on authoritarian countries with term limits and leveraging a difference-in-differences estimator with matched sets that address endogeneity issues carefully, I find empirical support for my theory.In the last paper, I discuss the relationship between autocratization and delegitimizing propaganda. Autocratization increases threats from marginalized opposition elites whose parties are more incentivized to mobilize the masses. What do autocrats do to counter this threat? I argue that autocrats increase delegitimizing propaganda, exaggerating the disunity of opposition parties strategically. It undermines opposition parties’ legitimacy as a competent alternative that conveys consistent and credible information. Moreover, this propaganda complements existing censorship. To validate my argument, I compare how regime-controlled newspapers in the South Korean dictatorship cover two very similar internal conflicts in the opposition party that occurred before and after autocratization. Using word embeddings that quantify delegitimizing propaganda, I find that newspaper reports after autocratization were more likely to associate the opposition party with negative words related to disunity