Sanctioning Reputation Mechanisms in Online Trading Environments with Moral Hazard

This paper offers a systematic exploration of reputation mechanism design in trading environments with opportunistic sellers, imperfect monitoring of a seller's actions and two possible seller effort levels, one of which has no value to buyers. The objective of reputation mechanisms in such settings is to induce sellers to exert high effort as often as possible. I study the impact of various mechanism parameters (such as the granularity of solicited feedback, the format of the public reputation profile, the policy regarding missing feedback, and the rules for admitting new sellers) on the resulting market efficiency. I find that the maximum efficiency that is attainable through reputation mechanisms is bounded away from the hypothetical first-best case where sellers could credibly pre-commit to full cooperation by a factor that is related to the probability that cooperating sellers may receive "unfair" bad ratings. Furthermore, the maximum efficiency is independent of the length of past history summarized in a seller's public reputation profile. I apply my framework to a model of eBay's feedback mechanism and conclude that eBay's simple mechanism is capable of inducing the maximum theoretical efficiency independently of the number of recent ratings that are being summarized in a seller's profile. I also derive optimal policies for dealing with missing feedback and easy online identity changes