| Chapter 1, Strategic Candidacy and Single-Peakedness, explores strategic candidacy on the domain of single-peaked preferences. In elections, the voting outcomes are affected by voters' preferences as well as candidates' entry decisions. We study a class of voting rules immune to strategic entries of candidates. Dutta, Jackson, and Le Breton (2001) show that such rules satisfying unanimity are necessarily dictatorial if all orderings of candidates are admissible for voters' preferences. When voters' preferences are single-peaked over one-dimensional political spectrum, there exist non-dictatorial voting rules immune to strategic candidacy. An example is the rule that selects the m-th peak from the left among the peaks of voters' preferences. We show that immunity from strategic candidacy together with basic axioms fully characterizes the family of the m-th leftmost peak rules.; Chapter 2, Inducing Participation in Lindahl Mechanisms, analyzes the problem of non-participation in Lindahl mechanisms. To solve the free-rider problem, the literature has proposed various mechanisms achieving Lindahl allocations for non-excludable public good economies. However, Saijo and Yamato (1999) show that, for any such mechanism, agents may have incentives not to participate. We propose redesigned mechanisms that overcome the non-participation problem. We incorporate a particular sequential participation process at the beginning of a given mechanism and obtain another mechanism in which provision of public goods is contingent on the set of participants. For any mechanism achieving Lindahl allocations, we can redesign the mechanism so that agents have the freedom of non-participation and the Lindahl allocations are achieved.; Chapter 3, Implementation of Nucleolus Correspondence, presents a game form that fully implements the nucleolus correspondence in subgame-perfect equilibrium. This extensive game form incorporates objections and counter-objections in the definition of the nucleolus proposed by Osborne and Rubinstein (1994). In the game form, agents do not have to report preferences, which are difficult to communicate in reality. The game form is balanced even on off-equilibrium paths. Furthermore, the game form achieves implementation for any class of allocation problems where preferences are quasi-linear and the associated coalitional form game is superadditive. |
