Economic Design of Reputation Systems

Abstract

Designing an online resource exchange system is challenging, especially when that system is decentralized and its members fully autonomous. To facilitate the system architecture and protocol design, it is important to understand the behavior of the users and the impact system parameters play on their actions. In this paper we identify key attributes, such as trust and the rate of contribution, that drive the actions of users of trading systems, whether they are cooperative, selfish, or malicious. We then present an economic model that captures the behavior of peers in a system that employs incentive and reputation schemes to mitigate the effects of freeriding and misbehavior. Using this model, we gain new insight into the operation of reputation systems.