The Role of Reputation in Open and Closed Groups: An Experimental Study of Online Trading*

Greif (1989, 1993) argues that reputations provide a solution to the agent problem deriving from information asymmetry when principal-agent relations are confined within a closed group. This raises an interesting question of whether or not reputations provide a solution to the lemon’s market problem (Akerlof, 1970) when the market has open boundaries. In order to address this question, we create an online trade market in the laboratory. In three experiments, we examine how reputation helps alleviate the lemon’s problem in open and closed markets. We manipulate openness and closeness of the market by allowing and not allowing traders to change their identities and cancel reputations. When they are allowed to change their identities, they can re-enter the same market under a different identity and a fresh reputation score. We find that a reputation system helps alleviate the lemon’s problem in both types of markets, but for different reasons. In closed market, having a reputation system greatly improves the market efficiency while individual traders who have accumulated good reputations benefit only little from their reputations. In contrast, individual traders who have acquired good reputations in open markets enjoy substantial financial advantage. We argue that the effect of reputation in the closed market is mostly based on traders’ untested beliefs, whereas the effect in the open market is based on individual traders’ experience to a much greater degree than in the closed market.