CAN PUBLIC GOODS EXPERIMENTS INFORM ENVIRONMENTAL POLICIES?

Understanding behavior in experimental public goods games is fundamental to the work of environmental, behavioral, institutional and policy-oriented economists. Although much research has been devoted to explaining the dynamics of such experiments, the conclusions drawn to date are contradictory. Through the use of a novel experimental design, a theoretical model of behavior, and appropriate econometric methods, we address weaknesses in the current literature and resolve much of the conflicting claims about motives in public goods experiments. Our analysis demonstrates that herders and strong reciprocators are the main contributors to the public good, whereas the role of interdependent utility and warm-glow altruism is weak at best. Further, the oft-observed decay in contributions over rounds is driven by the revocation of cooperation by disappointed strong reciprocators coupled with the herding behavior of confused subjects. We find no evidence that confused subjects learn the dominant strategy over time. The data instead imply that a substantial proportion of subjects do not recognize the tension between the privately optimal strategy and the socially optimal strategy. These results offer insights into improving environmental policy, but also suggest that public goods experiments cannot achiev