Strategic aspects of decision making ; Aspectos estratégicos de la toma de decisiones

In: http://hdl.handle.net/10016/25423

Abstract

This dissertation consists of five independent essays: Reliability and Responsibility: A Theory of Endogenous Commitment. A common assumption in Political Science is that candidates are committed to their policy announcement. In the first chapter I present a model of political competition where candidates announce policies that might not correspond to the policies that are actually implemented. I analyze the possibility of transmitting information to voters through costless electoral campaign. I stem on Down's intuition that meaningful electoral campaigns must be based on candidates' reliability and responsibility. Through the analysis of an incomplete information model of finitely repeated elections I show that electoral promises are able to transmit reliable information to voters. Electoral campaign provides an instrument to discipline politicians linking re-election perspectives to the fulfillment of electoral promises and contributes to solve the informational asymmetries between candidate and voters. An unavoidable proportion of ambiguous politicians emerge. These results are robust to the consideration of equilibrium refinements. Implementation with State Dependent Feasible Sets and Preferences: A Renegotiation Approach. The second chapter (fruit of a joint work with Luis Corchón) provides a framework of analysis for situations in which the feasible set is unknown to the planner. The concept of reversion function is used, which describes the process that society uses to deal with unfeasibilities (punishment, renegotiation, legal system). This reduces the problem of implementation with unknown feasible to the case in which preferences are unknown. We characterize the maximal set of Social Choice Correspondences that can be implemented in a class of renegotiation functions that do not reward agents for infeasibilities. Applications to Exchange Economies, Bargaining and Bankruptcy are presented. The impossibility to take into account in the renegotiation process of individual messages imposes strong restrictions to implementation. The results are compared to the findings of literature on "feasible implementation". Ramón y Canal: Mediation and Meritocracy. The third chapter (fruit of a joint work with Antonio Romero-Medina) analyzes the centralized matching mechanism associated to the Ramon y Canal Program. This Program is used in Spain to promote the hiring of top researchers in Spanish R&D centers and academic institutions. We model the process as a bilateral matching market to study if the mechanism provides the incentives to hire good researchers. We analyze the model both under complete and incomplete information. The comparison of the theoretical findings with the available data of the program indicates that the model provides poor incentives to the agents involved and does not prevent from collusion between research centers and candidates. While the objective is to reward the best researchers the mechanism only guarantees "almost stable solutions. The Condorcet Winner: Incomplete Information, Implementation and Multiplicity. The fourth chapter studies the implementation of the Condorcet Winner under incomplete information. It is shown that a trade-off between the robustness of the Condorcet Winner and equilibrium multiplicity appears. For a non negligible set of probability distributions exact implementation of Condorcet consistent social choice functions cannot be obtained. The result is proven in single peaked environments and extended to more general setups. A positive result is provided for uniform priors. Coordination in Sequential Admission Mechanisms. The fifth chapter considers a family of admission (or hiring) mechanisms in which participant are allowed to send multiple applications, as it happens in many real life situations. This is done by extending the model presented by Alcalde and Romero- Medina (2000). The possibility of applying to different colleges imposes serious coordination problems to colleges themselves. Unstable equilibrium allocations may arise. Stability can be recovered only imposing application fees.