Explaining Individual Contributions in Public Goods Games Using (only) Reciprocity and Overoptimism

We explain contributions in public goods games with the help of the reciprocity model of Dufwenberg and Kirchsteiger (2004) by applying some plausible modifications: Most importantly, we assume that subjects overestimate the kindness of their group members. In combination with the finding that subjects are typically imperfect conditional cooperators, equilibrium contributions to public goods can be derived. We test our model experimentally and find robust evidence for our modifications: In the experiment, we directly link reciprocal preferences elicited in contribution schedules to beliefs and show that behavior is indeed primarily driven by reciprocity and overoptimism. Although we find distinctly heterogeneous behavior on the subject level, our model can predict such behavior if subjects' reciprocal inclinations are known. Thereby, the false consensus effect additionally fosters cooperation because it lets conditional cooperators overestimate the level of reciprocity in the subject pool