Tolga Dizdarer, PhD Candidate, The Wharton School, Gerry Tsoukalas, Operations, Information & Decisions, The Wharton School, and Gérard Cachon, Operations, Information & Decisions, The Wharton School
Abstract: Service platforms like Uber and AirBnb create markets that connect customers with service providers. Platforms provide the medium of interaction and receive commissions from each transaction taking place between the two parties. The servers working for the platforms are commonly flexible in choosing how much they work. While some platforms like AirBnb further provide the servers the flexibility of choosing their own fares, others like Uber and Lyft assume the role of market-maker and set the prices for the whole market. What should be the extent of platform’s role in setting the prices in the market is a fundamental question for the platform and we study it as a trade-off between price efficiency and information utilization. In a market with cost information asymmetry, server pricing allows platform to utilize servers’ private cost information and increase server participation. However, this policy is risky. As the prices are determined by self-interested actors, the market is susceptible to two well-known pricing inefficiencies observed in decentralized markets: inflation of prices due to double-marginalization and deflation of prices due to competition. If the market competition is at an extreme, server pricing may significantly shrink the market size and damage platform’s profits. Platform pricing is a robust policy that alleviates the prices efficiencies; however, it cannot capitalize on servers’ cost information and results in non-optimal profits for the platform. We find that offering quantity bonuses/surcharges under server pricing eliminates price inefficiencies and achieves optimal profits for the platform.