Use this URL to cite or link to this record in EThOS:
Title: Essays in Bayesian implementation
Author: Dintcheva-Bis, Darina
Awarding Body: University of Warwick
Current Institution: University of Warwick
Date of Award: 2013
Availability of Full Text:
Access from EThOS:
Access from Institution:
The purpose of this thesis is to analyze mechanisms that implement a social objective for two environments in which agents have incomplete information about the others’ characteristics. Agents’ beliefs about the characteristics of any particular agent are common knowledge. We consider the case where monetary transfers or costly signals are undesirable or unavailable. Chapter 1 gives an overview of the contents of the thesis. Chapter 2 studies mechanisms for resolution of bilateral conflict over a prize of common value. This conflict may be settled by a peaceful agreement or may lead to a socially inefficient outcome of war. We model explicitly the cost of war as dependent upon opponents' types which are private information. The social choice function is the probability of peaceful resolution. We assess the chances for peace in the case of no communication and a simultaneous choice by agents whether to agree to a given split proposal. We compare these chances with the probability of peaceful settlement achieved by a mechanism which solicits partial disclosure of private information. We require the truthful revelation of this information to be a dominant strategy for agents in the game induced by the mechanism. In this framework we show that unmediated communication always improves the probability of peace upon the agreement game. In chapter 3 we study a cardinal mechanism for allocation of heterogeneous indivisible goods among agents with private valuations. We assume that agents and the mechanism designer hold the same beliefs about the ex ante distribution of the multidimensional types. We relax the dominant strategy requirement for the truthful revelation to the requirement of Bayesian incentive compatibility. We provide a necessary condition for Bayesian incentive compatibility of such mechanisms for any finite number of goods and agents. We characterize the set of Bayesian incentive compatible mechanisms for the case of three objects and three agents and we analyze efficiency and fairness properties of these mechanisms. In particular, we show that in this framework an ex post efficient and envy-free mechanism may not exist for some systems of beliefs.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: HB Economic Theory