Social Learning and Market Structure: The Case of Airbnb

Abstract

In digital markets, consumers rely on product reviews to gather information on the quality of a product. However, they do not internalize the positive informational externality of their reviews on future consumers. Consequently, new products are purchased and reviewed at an inefficiently low speed. The lack of experimentation by consumers has repercussions on sellers, as it hampers their ability to generate initial reviews on the platform, effectively representing an entry hurdle. To analyze the welfare consequences of insufficient experimentation, we estimate a dynamic structural model of demand and supply for Airbnbs in Manhattan, NY, from 2014 to 2019. We use our estimates to compute the welfare effect of a counterfactual subsidy on the price of unreviewed Airbnb listings.