Social Learning and Strategic Pricing with Rating Systems

Rating systems, widely used in online transactions, often reduce buyers' diverse opinions to summary statistics. To explore the consequences of this coarse aggregation, we analyze a dynamic adverse selection model where buyers share anonymous evaluations via a rating system. With heterogeneous buyers, the seller is tempted to secretly lower prices to attract favorable ratings from price-sensitive buyers. That leads to sporadic flash sales. The seller's incentive to manipulate ratings is, however, self-defeating. Our analysis illustrates how the rating system shapes the allocation of surplus and offers insights for platform and product design. (JEL D11, D82, L11, L81)