Doctoral Dissertation Research: Uncertainty, reputation effects, and relational formation

Cornell University, Ithaca NY

Investigators

Abstract

For many, the Internet is still a world of anonymity, pervasive automation, and fraud. Yet, the success of large-scale online markets has inspired much interest in feedback systems as an institutional solution to the problem of market uncertainty. In eBay, buyers can rate sellers after transactions; their feedback points are tallied and posted publicly next to their screen names for other buyers to view. In Amazon, customers can likewise rate products. Interestingly, EBay and Amazon have very similar feedback systems, yet eBay is a customer-to-customer site, while Amazon is a business-to-customer site, with important differences in the types of market uncertainty. In eBay, there is a pervasive problem of trust and trustworthiness casting over transactions between anonymous strangers. In Amazon, fraud is rare. Instead, the main problem is the cost of search through an enormous inventory of products. These differences suggest that their feedback systems might work differently, but how? This proposal argues that, in markets like Amazon, reputations facilitate transactions by identifying "reputable" sellers or products. In markets like eBay, the crucial role of reputations is to discipline sellers through feedback exchange. I test this thesis using laboratory experiments. If confirmed, the research will not only extend our theoretical understanding of how reputations moderate different types of market uncertainty, but also provide practical insights for designing online markets. Whereas brick-and-mortar markets emerge slowly through bottom-up processes, online markets can be created vastly more rapidly, cheaply, and purposefully. Understanding what makes an online market effective is therefore crucial as we enter the new age of digital economy.