An Experimental Investigation of the Impact of Exogenous Matching Methods on Contribution to Public Goods : do People Cooperate more with Same-Spirit Partners or Same Performing Partners ?

In this paper we focus on identifying the determinants of cooperation through a laboratory experiment. Randomly paired participants are asked to solve simple mathematical problems and to determine which proportion of their performance they want to contribute to a team account (the rest being contributed to an individual account). The experiment is composed of several steps corresponding to different matching methods: participants can be matched with a randomly selected participant, with a participant whose cooperative behavior is similar to theirs or with a participant whose performance is close to theirs. We find that participants are more willing to contribute to the team account (to cooperate) when matched to participants as cooperative as them. We find that being matched with participants with similar efficiency is of lower importance. Participants with reasonable levels of overconfidence should have contributed their full earned endowment to the team account when matched with a partner with the same team-spirit in order to maximize their payoffs. However, their contribution is only 33% higher than when matched with a random partner. The contribution to the team account is not affected by the impact one's performance and team-spirit could have on his/her teammate's payoffs.