The Value of Feedback : An Analysis of Reputation System

Online markets, as established by many authors, are prone to asymmetric information problems. Despite this observation, many online markets have thrived and grown over time primarily due to user experience rating systems, also known as reputation systems. In this paper, we analyze how such systems work, and specifically we turn our focus on eBay. We start by analyzing the Feedback behavior of buyers and sellers over time. We use a key policy change, that sellers can no longer leave negative Feedback for buyers, as an identifier. Our data analysis points to the existence of retaliation among buyers and sellers before the policy change. Furthermore, we develop a model of Feedback behavior as a dynamic game between buyers and sellers and structurally estimate the model. The structural estimation further establishes the existence of retaliation incentives between buyers and sellers