This research would study the implications of learning-by-doing in production for theoretical models of industrial organization. The most common technological specification of learning-by-doing is the learning curve hypothesis that a firm's unit cost declines with cumulative output. A first research task is to study the strategic implications of the learning curve in a dynamic model of duopoly. The methodology is to characterize a symmetric Markov perfect equilibrium and evaluate its properties. A second task is to derive implications for antitrust, trade, and patent policies. A third task is to reconsider the foundations of the learning curve hypothesis, studying industrial organization implications of other, more general, learning-by-doing technologies. A final task is to extend these models to investigate the equilibrium timing of production and deliveries when buyers as well as firms behave strategically. This research departs from previous industrial organization research in several respects: (1) the theoretical model is new; (2) the analytical methods are different; and (3) new questions are raised. A longer term goal is to lay foundations for empirical work.
