UC Berkeley

This dissertation investigates the role of firms in international trade using the recently available comprehensive datasets on international trade transactions of firms in different countries. The first chapter investigates empirically second-degree price discrimination in business-to-business dealings. I use highly detailed transaction-level Colombian imports data to document the presence of quantity discounts. These data helps me identify firms on both ends of the transaction. I document a robust negative relationship between prices and quantities of the same good purchased from a given seller by different Colombian buyers. This relationship is not an artifact of the measurement error or the scope for the quality differentiation of particular goods. I also show that this result is not driven by the length of the relationship and between a given buyer and a given seller or the number of transactions between a given buyer and a given seller per year. Finally, the negative relationship between prices and quantities holds when I condition my estimation on the measure of buyers' and sellers' market power. I find that, on average, a 10\% increase in the quantity purchased reduces the price charged per unit by 2\% and conclude that this relationship is most likely consistent with price discrimination rather than alternative explanations

To rationalize this empirical fact, the second chapter of this dissertation develops a tractable theoretical framework that embeds nonlinear pricing (second-degree price discrimination) into a standard international trade model and characterizes optimal policies. The model can be conveniently aggregated and yields a standard gravity equation of international trade flows. However, the model has several important implications that are absent from the traditional trade models. I show that welfare losses from second-degree price discrimination can be quite substantial. I also characterize optimal trade policy from the perspective of a social planner in a small open economy. I prove that optimal tariffs are higher when firms use non-linear prices as compared to standard models. In addition, if the policymaker sets tariffs that are optimal under linear pricing, but firms use second-degree price discrimination, this will lead to significant welfare losses.

The third chapter focuses on the the importance of the extensive margin (the number of firms exporting) for trade flows, which is highlighted by the Melitz model of international trade. Using the World Bank's \emph{Exporter Dynamics Database} featuring firm-level exports from 50 countries, it documents that around 50\% of variation in exports does occur on the extensive margin --- a quantitative victory for the Melitz framework. The remaining 50\% on the intensive margin (exports per exporting firm) contradicts a special case of Melitz with Pareto-distributed firm productivity, which has become a tractable benchmark. This benchmark model predicts that, conditional on the fixed costs of exporting, \emph{all} variation in exports across trading partners will occur on the extensive margin. Combining Melitz with lognormally-distributed firm productivity and firm-destination fixed trade costs can explain the intensive margin seen in the EDD data. In the EDD, the importance of the intensive margin rises steadily when going from the smallest to largest exporting firms across source countries, as is also predicted by the Melitz model with lognormally-distributed productivity.