. Project Summary: The purpose of this work is to demonstrate that a coordination process can improve group decision making about shared goods. The long term goal of this research is to develop decision making tools for use in a cooperative group setting, based on economic theory and experimental economics testing. Groups in social, government, and business situations frequently have to decide on the nature of a shared good and how its costs are allocated among group members. A garbage facility shared among several counties is one local government example, and a shared computer system among divisions in a firm is another. Our underlying thesis is that successful group outcomes may be impeded by complexity of the group situation, particularly when there is heterogeneity in group member types and the cost function is nonlinear. An exacerbating factor is group size: in a large group, direct cornmunication among members to find a group optimum is simply not possible. Equal cost sharing is frequently observed as a group outcome even when group members differ in preferences, perhaps as a response to complexity. Voluntary contribution has been shown experimentally to under provide shared goods. At worst, a group simply may not be able to reach a decision on a voluntary basis. Our coordination process uses a specific algorithm tO find the group solution for the shared good and its finance through cost sharing. Each group member sends the quantity demanded as a signal to the coordinator in response to a price signal; the process repeats until a stationary signal is obtained. Thus, the coordination process provides structured interaction among group members and, more importantly, provides information signals for individual decisions relative to the group that help guide the group to an equilibrium outcome. The coordination process that we will test is based on the Cost Share Equilibrium which gives an efficient group solution as a Nash equilibrium. The Quantity Adjustment Process has been given preliminary testing in an experimental economics setting at the Economic Science Laboratory at the University of Arizona. In striking contrast to other voluntary contribution experiments, group performance in terms of efficiency improves with repetitions as group members learn about how the process works. The current project tests, using experimental economics techniques, coordination (Quantity Adjustment Process) compared to a base case (Voluntary Bid) for a worst case scenario: no communication and heterogeneity with a nonlinear cost function. We construct a voluntary mechanism to parallel our shared good situation with a cost function. Results are be used to compare statistically the improvement in efficiency provided by the coordination process. Individual strategic behavior as related to type (endowment and reward) is also investigated.
