According to game theoretic results, the duration of exclusion from participation in the group should not affect contributions in a public goods game in which sanctions depend on majority vote and voting for sanctions is costly. Because of the costs associated to the voters individually, and the group collectively given the reduction in total investment, the threat of sanctioning is incredible and will thus not deter shirkers from free-riding. Assuming the existence of normative concerns among some group members, however, exclusion becomes a risk to the lowest contributor, which is more costly in the case of unlimited exclusion than in the case of one-term suspension. Avoidance of being the lowest contributor sets in motion a virtuous circle toward full contribution. Testing this argument in a laboratory experiment, we find evidence supporting the argument that irreversible exclusion generates a higher level of contributions and higher efficiency than short-term exclusion.
