| I explore the effect of incentives on the decisions of patients and physicians in the market for prescription drugs. To study the static and dynamic effects of incentives empirically, I develop estimation approaches that can accommodate features of the drug choice setting, including the large choice set and long time horizon for treatment, that complicate the application of existing tools.;In the first chapter, I focus on the interaction between physician and patient incentives and the treatment choice. The division of responsibilities among patients, providers, and insurers in the market for medical care creates scope for agency costs and moral hazard in the physician's treatment recommendation. I model the origin of these costs and measure the power of insurer-designed incentives to limit them, focusing both on incentives aimed at the "demand-side" interaction of the patient and physician and the "supply-side" interaction of the physician and insurer.;In the second chapter, I study demand for experience goods---products whose quality consumers learn through experimentation. I develop a dynamic framework for analyzing demand that interacts pricing and promotion with this learning process, and accommodates settings in which agents search over large choice sets. I employ this approach to study the antidepressant market, testing whether insurer policies can improve the quality of drug-patient matches and overall health.;Finally, in the third chapter, written jointly with Eduardo Morales, we provide a framework for estimating the structural parameters of single agent models using moment inequalities. This framework permits us to perform inference on model parameters that are computationally impossible or very costly to estimate using maximum likelihood. Specifically, we detail the assumptions needed to generate moment inequalities for single agent dynamic models and conduct inference on the structural parameters without computing the value function for any individual or time period. We outline a simple computational algorithm to estimate the boundary points of an identified set defined by moment inequalities and to conduct inference. |
