Estimating Structural Models of Platform Competition in Two-sided Markets

In this paper I estimate theory-based structural models of two-sided markets with competing platforms. Given platforms’attributes and (two) prices, agents of each side decide which platforms to join. Platforms set prices to maximize the sum of profits that they earn from the two sides. There are two demand equations to be estimated, and both should be used in estimating platform markup due to cross-group externalities. I consider a two-sided single-homing model as well as a competitive bottleneck model. In the single-homing model agents of both groups use only one platform (single-home). In the competitive bottleneck model one set of agents single-homes while the other uses as many platforms as it likes (multi-home). Using data on entertainment magazines in Germany I show that magazines typically set their copy price below marginal cost and earn profit from selling advertising pages.