Essays on Online Marketplaces

Abstract

This dissertation is three chapters on online marketplaces. The first marketplace that I investigate is online auction platforms, where I investigate the impact that default penalties have on bidding in common-value contexts. Due to the existence of the winners curse and the fact that bidders can observe all the bids after the auction closes, winning bidders have an incentive to decline to pay their bids when their signal about the item's value is not representative of the average bid. I build a structural model of bidding, and estiamte the parameters using data scraped from the internet, and calculate counterfactual default policies. The second marketplace that I analyze is Amazon's, where third-party sellers sometimes compete with Amazon itself in the market for a given product. The impact of Amazon participating as a seller in one of the markets that it hosts has received a lot of theoretical attention, but limited applied analysis. Using a large sample of Amazon products, we leverage differences-in-differences designs to isolate the impact of Amazon participating in such a market and find that Amazon entry drives prices down and increases sales, while not driving third-party sellers out of the market. These results suggest that entry by Amazon is beneficial for customers on its platform, potentially at the expense of third-party sellers. I conclude by formulating a dynamic model to answer the question of how third-party sellers should set prices and invest in their products given the presence of Amazon. This model builds off of the conclusions of the empirical work in Chapter 2, which motivates expanding the static analysis presented there to a dynamic setting to answer important questions about how third-party sellers operate in hybrid marketplaces, how the platforms enforce price discipline, and the welfare consequences thereof.