Dynamic Equilibrium in the US Prescription Drug Market After Patent Expiration

In this paper I formulate and estimate an empirical dynamic oligopoly model that incorporates consumer learning, consumer heterogeneity and forward-looking firms. I apply this model to explain the evolution of pharmaceutical markets after patent expiration, and to address its related policy issues. I develop a practical method to estimate the parameters of the model that does not require finding instruments or solving the equilibrium model. Using this new method and a data set detailing the evolution of prices and market shares for 31 chemical entities from 1984-1990, I estimate the distribution of consumer preferences that determine how consumers evaluate risks, perceived attribute levels, and prices when choosing among brand-name originals and generics. I also design and program a backward induction algorithm to numerically solve the equilibrium model with different market structures. This computer program, together with the estimated preference parameter values and calibrated cost parameter values, are used to select a market structure that best fits the data, and the selected market structure is applied to analyze the firms' strategic behavior. According to the preliminary estimate results based on data from two markets, flurazepam and temazepam, I find that learning plays an important role in explaining the slow increase in market share for generic drugs. I also demonstrate that consumer heterogeneity has the potential to explain the pricing pattern that brand-name prices increase in response to generic entry. Finally, the model will be used to quantify the benefits of introducing generic drugs and to simulate the impact of various new public policies, including restricting the price for brand-name originals below some arbitrary level, reducing the average approval time for marketing generic drugs, and changing the cost of obtaining such approval.