Manipulation and implementation of resource allocation mechanisms
Abstract
Hurwicz, Maskin, Thomson, and others have investigated the dual issues of manipulation and implementation of resource allocation mechanisms. Their primary focus has been on manipulation via misrepresenting preferences and implementation of a social choice correspondence via a game form. Postlewaite has considered manipulation by withholding initial endowments. This dissertation presents a unified framework which allows us to analyze (1) both preference and endowment manipulation, (2) mechanisms employing outcome functions and outcome correspondences, and (3) manipulation games and quasi-games. The main impossibility results are revisited and extended within our "dichotomously" informationally decentralized structure. It is shown that for any mechanism satisfying individual rationality, we have truth-telling as a Nash equilibrium if the endowment allocation is Pareto efficient for the true economy. We also show that under a normality assumption, for any Pareto efficient and individually rational mechanism, truth-telling as a Nash equilibrium implies that initial endowment must be Pareto efficient. The "Mr. 1" mechanism of Thomson (1986) is examined for both types of manipulation. We correct his main result and offer a complete proof. We also extend Thomson's impossibility result on partial recovering endowment manipulation to general (i.e., not necessarily essentially single valued) outcome correspondences. A brief survey of implementation concludes.
Degree
Ph.D.
Advisors
Foster, Purdue University.
Subject Area
Economic theory
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.