Asymmetric Public-Good Games - Experiments on Contribution Norms Encouraging Cooperation

Abstract

This dissertation extends the understanding of asymmetric group behavior in the provision of public goods. Previous studies, using an asymmetric framework, have so far focused on situations where all players profit form the social optimum. The discussion in the first paper breaks with this tradition and the results show how strong inequalities may decrease group contributions, as the high endowment players only contributes a small share of their endowments. This is in contrast to experiments using a weaker asymmetry, which indicate that players tend to contribute a fair-share of their endowment. This is of importance since it shows how a big rift in the income structure of a society may negatively influence the voluntary provision of public goods. The approach in the second paper joins the discussion on crowding-out of intrinsic motivation as well as the question which mandatory contribution mechanism performs the best, in an asymmetric endowment setting. It shows how progressive mandatory contributions stimulate the highest level of public good provision of all treatments, while displaying the same level of motivational crowding-out as in the treatment without mandatory contributions. Furthermore, the results suggest a norm setting character of mandatory minimum contributions. The third paper puts the emphasis on peer punishment in asymmetric groups of three players. The focus rests on the contributions as well as punishment to and punishment from the player types in majority. This experiment shows that especially high endowment players are punished more. This finding is independent of whether the high endowment type is in a majority or minority situation. The last experiment focuses on fair-share behavior in asymmetric public good games and potential effects of the way the interface in the contribution stage is presented. In each round, players have to choose the presentation style of the contribution mechanism, which can be either in absolute or relative terms. It shows that most participants prefer to contribute in absolute terms, which is consistent in all treatments. However, this has no effect on the level of contributions. Groups with weak asymmetries show overall a fair-share behavior, which vanishes if groups with mostly full or zero contributions are excluded, showing low endowment players to be contributing a larger share of their endowment than the high endowment players. Low endowment players in a strong asymmetry setting contribute a larger share compared to the high endowment players. At the same time, participants adapt their preferences of the interface depends on the actual design of the experiment. This is reflected in the observation that more participants pick the relative contribution method in the treatments with a stronger symmetry.