This dissertation examines learning-by-doing in the commercial aircraft industry with the goal of increasing our understanding of both learning itself and its implications to optimal policy.
Chapter 1 introduces the topic of learning and discusses previous progress in both the empirical and theoretical literatures.
Chapter 2 introduces a new cost dataset for a commercial aircraft firm and uses this data to analyze the dynamics of learning in commercial aircraft production. This dataset is found to be inconsistent with the simple learning hypothesis. Instead, strong support is found for the hypothesis of organizational forgetting, a more general learning model where unit costs are similarly dependent on a firm's past production experience, but where that experience depreciates over time. Additionally, it is found that some, but not all, of a firm's production experience transfers from one generation of an aircraft to the next.
In chapter 3, I formulate, estimate, and solve a multi-agent dynamic model of the commercial aircraft industry in which firms are differentiated in their products and cost structure, and where entry, exit, prices, and quantity sold are endogenously determined in a dynamic equilibrium. I utilize the cost model of aircraft production developed in chapter 2 and a discrete choice model of demand, both estimated econometrically using industry data. It is found that the dynamic model predicts industry price levels and movements well and closely matches many aspects of the observed industry dynamics. Additionally, it is found that the organizational forgetting hypothesis is not only critical to explaining industry cost data, but is also helpful in explaining the strategic behavior of aircraft producers, especially the exit of mature products from the market. Finally, the model is used to evaluate the welfare implications of an anti-trust policy that places a restriction on industry concentration and it is found that such a policy would be welfare reducing.