New Site. Please click here.

Ryan Tierney

PhD Candidate
Department of Economics
University of Rochester

I will be available for interviews at the ASSA meetings in Boston

Contact: ryanetierney at gmail.com

CV

Research

Managing Multiple Commons. (Job Market Paper) There are several locations, each of which is endowed with a resource that is specific to that location. Examples include coastal fisheries, oil fields, etc. Each agent will go to a single location and harvest some of the resource there. Several agents may go to each location. Selling the commons for money is not desirable, either because agents have equal right to use the resources or because control of the commons would give unacceptable market power to its owner. Thus we will assign harvesting rights based on preferences alone, though the model can be extended to accommodate private endowments of money. We find the best allocation rule in the class of rules that are weakly pairwise strategy-proof, anonymous, and that satisfy a weak continuity property. The rule is defined via a simulated price equilibrium, wherein agents buy their desired resource with tokens distributed by the social planner. Equilibria of this form are not unique as full distribution of the resources is not required. However, equilibrium price vectors form a lower semilattice and thus there is a unique minimal price vector. The equilibria associated with the minimal price vector are called min-price Walrasian equilibria. These equilibria form an essentially single-valued correspondence, and this correspondence is the rule we characterize.

Dense Manipulability of Efficient Exchange Rules We study a classical model. There are two divisible goods. Each agent has an endowment of the goods and continuous, monotone, convex preferences over bundles. Agents may benefit from trade. An exchange rule is a mapping that, for each profile of preferences, calculates for each agent a trade that he finds acceptable, given his preferences. Material balance is preserved: the sum of these trade vectors is the zero vector. It is known that no strategy-proof exchange rule always yields Pareto efficient outcomes. Strategy-proofness, however, is quite strong. We may instead ask the opposite question: if we insist upon Pareto efficiency, how frequently will the exchange rule fail to be strategy-proof? Unfortunately, we find a dense subset of a large open set on which any efficient exchange rule fails to be strategy-proof.

Strategy-proof Exchange and Simple Communication Revelation mechanisms are impossible to execute on large preference domains. This is certainly true of the classical domain, the space of monotone and strictly convex preferences over finite-dimensional Euclidean space. Thus, any plausible game-form defined on this domain must encode preferences into a simpler language. Clearly the outcome function of such a game-form can be sensitive only to information that can be communicated in the language. Thus we may study outcome functions directly and the richness of the message space they require. We study the outcome functions for classical exchange economies (requiring budget balance) that need at most a finite-dimensional message space. Such a function we call simple. We find that all strategy-proof, individually rational, anonymous, simple, non-exclusionary and continuous outcome functions are fixed-price mechanisms: Each such outcome function has, as a parameter, a price vector. This vector cannot vary with preferences. If any reallocation of the goods takes place, it is confined to the price hyperplane.