Voluntary Contributions and Framing effect: Theory and Evidence

Overcontribution in experiments on voluntary contribution to a public good has not yet received a fully satisfactory explanation. In those experiments, subjects cooperate, in contradiction with the game-theoretic prediction. In contrast, in oligopoly-experiments and common pool resource experiments, the Nash equilibrium is usually observed. The externalities generated by the two types of experiments are of the opposite sign: in oligopoly and common pool resources experiments the subjects’ choices generate a negative externality, while in the public good experiment, the actions of the other subjects create a positive externality. In this paper we designed an experiment which allows us to compare the level of overcontribution of the two contexts. Our experiment shows that subjects in the positive context contribute (significantly) more to the public good than the subjects in the negative context. We design a model that relates the difference between the two levels of average contribution, to behavioral differences between two types of players. We provide an empirical characterization of the underlying behaviors