Measuring Trust with Market Data

Abstract

A body of research argues that cultural biases in trust may be responsible for economic backwardness. This paper studies the role of cultural biases in a real-world market, eBay, with well-designed features intended to facilitate cooperation. We analyze the buyer’s decision to leave negative feedback as an act of mistrust towards the seller. To identify the effects of buyer characteristics on feedback choices, we exploit an identification strategy that leverages high-volume data from millions of transactions. We find that negative feedback decreases as buyers gain experience on the market. Also, we show that measures of pro-social beliefs and behavior do not explain feedback choices. Our favorite interpretation is that cultural biases in trust may have limited importance in markets with effective reputation-building mechanisms.