Reputation mechanisms such as those employed by Amazon and eBay offer an effective way to prevent market failure in online economies. However, most of these mechanisms assume that the privately monitored transaction outcomes are honestly reported. This clearly is a simplification since buyers may have incentives to misreport. While it has been shown that the truthful elicitation of these outcomes is feasible in settings with pure adverse selection, i.e. with a purely stochastic seller, we study whether honest feedback can be elicited in settings with moral hazard, i.e. with a strategic seller. For a pure moral hazard setting motivated by the one at eBay, we find that there is no feedback mechanism that makes honest reporting a best response to truthful play by all other players. For a combined setting with both adverse selection and moral hazard, however, we retrieve a positive result and construct a payment scheme that can be used as a "feedback plug-in" for reputation mechanisms.

@MastersThesis{witkowski:2009b,
  author = 	 {Witkowski, Jens},
  title = 	 {{Truthful Feedback for Reputation Mechanisms}},
  school = 	 {Albert-Ludwigs-Universit\"{a}t Freiburg},
  year = 	 {2009},
  type = 	 {Diplomarbeit},
  month = 	 {May},
}