TRUST, REPUTATION, AND EXIT IN EXCHANGE RELATIONSHIPS

Contrary to the assumptions of the supergame model, agents in exchange situations will normally not be perfectly informed about past behavio r of their partners. Moreover, they will generally be in a position to choose their partners. The formal model of this article attempts to t ake account of these facts. It is presupposed that, for any actor, the probability of finding a partner for an advantageous exchange depends on his or her past behavior. A model of reputation formation is prese nted as a formal description of this dependency, On the basis of this model, it is shown that, under certain conditions, only cooperative co nduct will pay in the long run. Thus it is confirmed that, given a sui table mechanism of trust, rational agents may act trustworthy.