The Hidden Benefits of Control : Evidence from a Natural Field Experiment

An important dialogue between theorists and experimentalists over the past few decades has raised the study of the interaction of psychological and economic incentives from academic curiosity to a bona fide academic field. One recent area of study within this genre that has sparked interest and debate revolves around the “hidden costs” of certain incentives, such as a principal not exercising control over an agent. While the literature highlights the importance of such effects, what has been missing is clean evidence from the field to support such claims. This study overlays randomization on a naturally-occurring environment in a series of temporally-linked field experiments to advance our understanding of the economics of charity and to test if such benefits exist in the field. This approach permits us to examine why people initially give to charities, and what factors keep them committed to the cause. Several key findings emerge. First, there are hidden benefits of conditional incentives that would have gone undetected had we maintained a static theory and an experimental design that focused on short run substitution effects rather than dynamic interactions. Second, we can reject the pure altruism model of giving. Third, we find that public good provision is maximized in both the short and long run by using conditional, rather than unconditional, incentives