| 520 | |
▼a This dissertation addresses a general problem: how political actors persuade others to support their goals through manipulating information. It builds on the premise that political actors in a variety of contexts need support of others to achieve their goals: rulers depend on compliance of the people to main political order, candidates rely on sufficient votes to get into office, bureaucrats must acquire approval of their superiors to implement policies, and rebels have to mobilize swing citizens to challenge the status quo. In general, these actors have two approaches to gather support: offering others incentives or providing them with information. While the former approach is quite fully characterized in the principal-agent framework and the related literature that applies this framework, the latter approach lacks as much attention and is the topic of my dissertation. Specifically, the dissertation presents a general formal framework to analyze information manipulation in several political contexts. In particular, I investigate strategies political actors apply to induce others to undertake certain actions through information manipulation under different institutional constraints.In the first chapter, I develop a generic sender-receiver framework, in which a political actor, referred to as the sender, tries to induce another actor, referred to as the receiver, to take an action through information manipulation. Different from the previous works, I allow the receiver to have private information. For instance, the president may not know exactly how citizens think about him. When the receiver has private information, the ability of the sender to manipulate information is limited: he can only manipulate what the receiver does not yet know. I find out that in this scenario, the sender may benefit from ``discriminatory persuasion,'' communicating with the receiver andmanipulating contingently what they communicate. For example, the president can tweet his policies, inferring how citizens think about these policies from their comments, and conditioning propaganda about these policies on the comments.In the second chapter (coauthored with Arturas Rozenas), we distinguish two tactics of information manipulation the sender can possibly employ in the context of election rigging. First, the sender can manipulate how information would be produced. For instance, a president can shut down media, preventing them from reporting undesirable news. Second, the sender can manipulate how information would be transmitted to the receiver. For example, in the case media have already exposed undesirable news, the president can discredit them as ``fake news.'' By solving the model, we derive an endogenous trade-off the sender faces between these two tactics and identify the institutional conditions under which the sender prefers one tactic over the other.The framework developed in the first two chapters has the potential to be applied in a wide range of substantive contexts. To illustrate this point, I apply this framework in the third chapter to study policy experimentation and politicization in economic reforms. I ask why economic reforms follow different patterns: why some reforms are abandoned, some recklessly assume the form of a "big bang," while some proceed through careful experimentation. Departing from previous literature of economic reforms, my explanationemphasizes political incentives of competing politicians, who are decision-makers, and the effect of these incentives on the information they seek. I find that economic reforms can impose on politicians smaller or larger political stakes and that different degrees of politicization can lead to different patterns of reforms. In particular, overly high political stakes would make reform too risky even to initiate, generating stalemates |