2004): “Effects of Reputation Mechanisms on Fraud Prevention in eBay Auctions

This paper shows empirically that eBay’s feedback system doc-uments successful transactions, but often fails to inform users of unsuccessful ones. A data set containing information from 3776 auctions I collected from eBay’s platform, gives reason to assume that the feedback system exhibits a positive bias due to the na-ture of mutual feedback that is made public instantly. Sellers and buyers are able to hold feedback hostages by refusing to leave feed-back until the opposite party has provided a report. In cases of problematic or fraudulent transactions, this behavior may lead to false feedback reports, or no feedback-provision altogether for fear of retaliation. In such cases, important information to the com-munity about transactions ’ efficiency is lost. In a simple model I rationalize false feedback reports, after which I suggest a change of the current feedback mechanism that circumvents the problem of feedback-retaliation, while at the same time providing incen-tives to participate in the feedback-provision process.