Essays on coordination problems in economics

• Ano de defesa: 2016
• Autor(a) principal: Pereira, Ana Elisa Gonçalves
• Orientador(a): Guimarães, Bernardo de Vasconcellos
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Palavras-chave em Português: Coordination; Bank runs; Global games; Information disclosure; Dynamic games; Heterogeneous agents
• Palavras-chave em Inglês: Timing frictions
• Link de acesso: http://hdl.handle.net/10438/16668

There are several economic situations in which an agent’s willingness to take a given action is increasing in the amount of other agents who are expected to do the same. These kind of strategic complementarities often lead to multiple equilibria. Moreover, the outcome achieved by agents’ decentralized decisions may be inefficient, leaving room for policy interventions. This dissertation analyzes different environments in which coordination among individuals is a concern. The first chapter analyzes how information manipulation and disclosure affect coordination and welfare in a bank-run model. There is a financial regulator who cannot credibly commit to reveal the situation of the banking sector truthfully. The regulator observes banks’ idiosyncratic information (through a stress test, for example) and chooses whether to disclose it to the public or only to release a report on the health of the entire financial system. The aggregate report may be distorted at a cost – higher cost means higher credibility. Investors are aware of the regulator’s incentives to conceal bad news from the market, but manipulation may still be effective. If the regulator’s credibility is not too low, the disclosure policy is state-contingent and there is always a range of states in which there is information manipulation in equilibrium. If credibility is low enough, the regulator opts for full transparency, since opacity would trigger a systemic run no matter the state. In this case only the most solid banks survive. The level of credibility that maximizes welfare from an ex ante perspective is interior. The second and the third chapters study coordination problems in dynamic environments. The second chapter analyzes welfare in a setting where agents receive random opportunities to switch between two competing networks. It shows that whenever the intrinsically worst one prevails, this is efficient. In fact, a central planner would be even more inclined towards the worst option. Inefficient shifts to the intrinsically best network might occur in equilibrium. When there are two competing standards or networks of different qualities, if everyone were to opt for one of them at the same time, the efficient solution would be to choose the best one. However, when there are timing frictions and agents do not switch from one option to another all at once, the efficient solution differs from conventional wisdom. The third chapter analyzes a dynamic coordination problem with staggered decisions where agents are ex ante heterogeneous. We show there is a unique equilibrium, which is characterized by thresholds that determine the choices of each type of agent. Although payoffs are heterogeneous, the equilibrium features a lot of conformity in behavior. Equilibrium vii thresholds for different types of agents partially coincide as long as there exists a set of beliefs that would make this coincidence possible. However, the equilibrium strategies never fully coincide. Moreover, we show conformity is not inefficient. In the efficient solution, agents follow others even more often than in the decentralized equilibrium.