Using computerized exchange systems to solve an allocation problem in project management

Abstract

In this article we study the allocation problem facing the management of a large research and development project. The project management has to allocate resources among competing users to achieve the project goal. Besides the constraint of scarcity, the allocation problem is difficult because users have private parameters that project management requires to know in order to make an optimal allocation. Furthermore, users have incentives to misrepresent the information about these parameters to advance their individual agendas, which can differ from the project goal. A method to solve the allocation problem using computerized exchange institutions is introduced and analyzed. We emphasize that the rules of the exchange should be carefully selected, because different rules produce different results. We use the methodology of experimental economics to demonstrate this conclusion. This research was motivated by JPL's Cassini Mission to Saturn. A computerized exchange described in this article has been implemented by the Cassini Project to assist in the management of the resources used in the design and operation of science instruments.