Abstract:
In the recent literature, several competing hypotheses have been advanced to explain the stylized fact of declining contributions in repeated public goods experiments. We present results of an experiment that has been designed to evaluate these hypotheses. The experiment elicits individual beliefs about the contributions of the partners in the repeated game and involves between-subjects variation on information feedback. The data favor the hypothesis of selfish-biased conditional cooperation as the source for the declining contributions over the competing hypotheses.