Essays in microeconomic theory

This thesis uses game theory and microeconomics to investigate empirical puzzles. In Chapter 1 we provide a formal model to explain why new legislation that has widespread quickly in the last decades failed to achieve its objectives in accordance to recent evidence. In Chapter 2 we uncover a new class of equilibria in the canonical social learning setting with endogenous timing of decisions. We argue how our results can offer a social learning explanation for two applications: delays in the adoption of policy measures during the Covid-19 pandemic, and the timing of investments in the venture capital industry. In Chapter 3 we provide a formal model to explain recent evidence that correlates high levels of residential segregation based on income with low intergenerational mobility. \n \nApologies are considered as a cheap and strong mechanism to restore broken relationships. Between 1999 and 2011, the number of US states with apology laws, legislation that excludes the admissibility of apologies in court, increased from 2 to 38, along with all the Australian jurisdictions, the United Kingdom, most of the Canadian provinces, and Hong Kong. Legislators’ hope is that by passing these laws apologies will be encouraged, with the consequence that civil disputes will settle more often and lawsuits will be prevented. However, recent evidence from US shows that these laws have had the opposite effect: apology laws have increased the number of lawsuits. In Chapter 1 we provide an explanation for why apology laws fail that is consistent with the best available evidence. We show that apology laws may reduce settlements by encouraging insincere apologies which in turn induce plaintiffs not to accept apologies. We contribute to show on which type of relationships apology laws fail: apology laws preclude the settlement of cases that are socially valuable to be settled. Moreover, for the cases where these laws increase litigation we show that apology laws induce more miscarriages of justice and deter inter-party communication. \n \nIn Chapter 2 we ask: Does waiting to observe others’ action delay profitable choices? If so, for how long? We characterize long delays in a social learning environment. In contrast with previous work, we show the existence of equilibria in which agents end up adopting a profitable and risky policy with substantial delay. These results point to social learning as a plausible explanation for delays evidenced in the adoption of policy measures during the Covid-19 pandemic. Next, we allow agents to choose the quality of their information before deciding. We show how in this setting long delays may also exists, and how our equilibrium sheds light on the investment timing patterns evidenced in the venture capital industry. \n \nRecent evidence shows a negative association between social mobility and residential segregation based on income. In Chapter 3 we provide a theory that explains this link based on beliefs in a just world. Our argument is that segregated communities exhibit more polarized and pessimistic views that hard work pays off than integrated ones because families in those communities learn differently about the value of effort. This polarization and pessimism in segregated communities make in turn mobility lower, as those families with low beliefs in effort have higher income inertia. We model agents as trying to learn the relative importance of effort and predetermined factors in the generation of income. They learn from two sources, by socialization in neighbourhoods and from their dynastic income mobility experience. In a dynamic model, we characterize conditions on initial beliefs under which the society exhibits in the long run income segregation with low rates of social mobility, or income integration with high social mobility rates. We provide evidence for US that support our theoretical results. Using survey-data with beliefs in a just world we show that more segregated communities are correlated with more polarized and pessimistic views about the value for effort.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.