Essays on pricing and matching on two-sided platforms

This dissertation consists of two essays on applied microeconomic theory, both focusing on an economy where a single monopolist operating a two-sided matching platform is strategically choosing prices to maximize profit. In this world, searching for a long-term partners is costly, and the monopolist lessens this inconvenience by providing a pool of potential partners to agents in exchange for a fee. Agents are searching for long-term partners while on the platform and have preferences over their potential partner's characteristics. The first essay in this dissertation studies the case where agents are horizontally differentiated. In the second essay, I study the case where agents are vertically differentiated and also care about a random horizontal component, modeled by using an idiosyncratic matching shock. As will be evident, a very important finding in the second essay is the effects of externalities on pricing.;In the first essay where types are horizontal, I assume that agent characteristics are uniformly distributed on the unit circle, a la Salop (1979). I also assume that the monopolist charges a two-part tariff. In the steady-state equilibrium, I show that per-interaction prices are optimally set to be zero. In other words, the two-part tariff pricing structure reduces to uniform pricing for all agents, regardless of type. I also show that once agents are on the platform, they obtain a long-term match, marry and leave the platform forever within finite time. In this model with horizontal types, since agents prefer others whose types are closer to their own on the unit circle, agents do not impose externalities on other agents by being on the platform. This is simply because there is no competition for ideal matches and the fact that one man is able to marry his favorite woman does not exclude another man from doing the same. In this model, faced with no externalities, agents are behaving socially optimal. The firm has no incentive to distort the matching behavior of agents by charging a non-zero per-interaction price.;In the second essay, I explore a world where agents are divided into two types, studs (high types) and duds (low types). Everybody prefers to marry studs. In addition to the vertical types, the utility functions for agents also depend on an idiosyncratic matching shock. I use this shock to capture an element of horizontal differentiation. In particular, the existence of this shock implies that even though two studs go on a date, a marriage may not result if the idiosyncratic shock is bad. Externalities enter the model in two ways, the first being through the proportion of studs on the market, which all agents care about. The second way is through competition for studs, who can be seen as rival goods. I find that when the monopolist has the ability to observe types and split them into separate platforms, it will choose to do so to maximize profits. In the case where the monopolist cannot separate the agents, I solve for a steady-state equilibrium and study the variation in optimal per-interaction prices used by the monopolist for different types of distributions for the idiosyncratic shock. When types are observable and the monopolist operates as a social planner, I characterize parameter ranges for which the firm charges distortionary (non-zero) levels of per-interaction prices to correct for externalities on the platform.