Reputations and Fairness in Bargaining - Experimental Evidence from a Repeated Ultimatum Game With Fixed Opponents

The results of Ultimatum Game experiments are often quoted as evidence for the role of fairness in bargaining or in economic behaviour more generally. This paper argues that the observed fairness levels are contingent on the traditional experimental design where players are newly matched each round, and reputations are therefore excluded. Evidence from a new experiment shows that average behaviour is more competitive and conflict rates are higher when subjects play against the same opponent repeatedly. This finding is not expected by the traditional fairness hypothesis. A detailed analysis of the dynamics of pairs of players shows that different types of players coexist in the subject pool. Whereas previous experiments found evidence for the existence of "fair" players, the present study reports also a significant number of "tough" players. Hence, there is evidence that allowing for reputations in repeated ultimatum bargaining induces different patterns of behaviour that have not been observed before in this game.