Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We study games where voluntary contributions can be adjusted until a steady state is reached. In consent games contributions start low and can be increased; in dissent games contributions start high and can be decreased. The equilibrium prediction is free riding in consent games but as much as social efficiency in dissent games. We test it experimentally and confirm that the dissent mechanism yields substantial welfare improvements over the consent mechanism. With experience, subjects contribute on average less than 30% of the endowment in consent games but more than 60% in dissent games. Generally, subjects match the lower of the opponents' contributions: they do not follow when single opponents increase contributions in consent games, but follow when single opponents decrease contributions in dissent games. This asymmetry in the conditional cooperation is predicted by heterogeneity of egoistic and inequity averse types, with individual types being private information.