Experimental evaluation of an eBay-style self-reporting reputation mechanism

We experimentally studied the effects of a eBay-style self-reporting reputation mechanism in an double-sided exchange economy in which participants have the option of not fulfilling their contracts. We found that submitted reports quite accurately reflected their transactions and this mechanism maintaining a high contract fulfillment rate. The inaccurate reports, which were about 5% of the total, were heavily biased towards bad ratings when the transaction is successful. This is strong evidence that the inaccurate reports were not results of random errors, but derived from an underlying behavior effect. Our experimental design allowed identifying the effect of reputation mechanism on endogenous market behavior.