Essays on the Analysis of Dynamic Games

Abstract

This dissertation seeks to combine ideas from literature in Machine Learning and the econometric analysis of games, and contributes to the analysis of dynamic competition in the context of high dimensional covariates. Chapter 1 studies new entry and mergers in the U.S. airlines industry and explores how the incentives of legacy carriers to accommodate new entry change when they merge and whether low cost carriers are sensitive to these changes when making entry decisions. We estimate an explicitly network-wide, strategic and dynamic model of airline competition, and find evidence that Southwest was more likely to enter markets where, from Delta and Northwest's perspective, the expected value of committing aircraft capacity, relative to other markets, fell the most post-merger. Chapter 2 develops a method for deriving policy function improvements for a single agent in high dimensional Markov dynamic games. We derive a one-step improvement policy over any given benchmark policy, and the one-step improvement policy can in turn be improved upon until a suitable stopping rule is met. Chapter 3 applies the method proposed in Chapter 2 to solve for policy function improvements in a high-dimensional entry game similar to that studied by Holmes (2011). The game has a state variable vector with an average cardinality of 10^42. We find that our algorithm results in a nearly 300 percent improvement in expected profits as compared to a benchmark strategy.