| This thesis combines three essays on the industrial organization of medical care. The first two essays consider the restrictions imposed by health insurers on their enrollees' choice of hospitals. The network of hospitals offered by each insurance plan affects consumer welfare, hospital profits and the incentives faced by hospitals to invest in new capacity, technology and quality. The first essay uses a three-step econometric model to predict consumer preferences over health insurers conditional on the hospitals they offer. The results indicate that consumers place a positive and significant weight on their expected utility from the hospital network when choosing plans.; The second essay uses the demand estimates as one input to a model of supply. A simple profit maximization model explains roughly half the observed contracts between insurers and hospitals. A generalization of the model demonstrates an additional effect: hospitals that do not need to contract with all plans to secure demand (for example, hospitals that are capacity constrained under a limited or selective network) can force insurers to compete for contracts. Some plans may exclude these hospitals in equilibrium. I estimate the expected division of profits between insurers and different types of hospitals using data on insurers' choices of network. Hospitals that have merged to form systems, and those that are expected to be capacity constrained, capture significantly higher markups than other providers. I show that these high markups imply a negative incentive for hospitals to invest to remove capacity constraints despite a significant benefit to consumers from the investment. The results demonstrate one route by which hospitals may modify their characteristics in order to increase their bargaining power.; The third essay develops a methodology to analyze multiple-agent discrete choice problems where there may be both endogenous regressors and multiple equilibria. This approach is used to estimate plan and hospital profits in the second essay. The identifying assumption is that firms expect their choices to lead to higher returns than those from their alternative options. The approach is a modified method of moments algorithm where the moment conditions hold as inequalities at the true value of the parameter vector. |
