Essays on the econometric analysis of treatment assignment rules and altruistic preferences
OA Version
Citation
Abstract
This dissertation has two main themes: treatment assignment rules and altruistic preferences. The first two chapters are about comparing different treatment assignment rules using observational data. The third chapter studies how altruistic preferences are affected by markets and incentives.
In Chapter 1, I develop a theoretical framework to compare different treatment assignment rules. A treatment assignment rule is a mapping from observed characteristics to binary treatment status. The welfare difference between two given treatment assignment rules is not point identified in general when data are obtained from an observational study or a randomized experiment with imperfect compliance. I characterize the sharp identified region of the welfare difference and obtain bounds under various assumptions on the unobservables with and without instrumental variables. I conduct estimation and inference of the bounds using orthogonalized moment conditions to deal with the presence of infinite-dimensional nuisance parameters.
In Chapter 2, I apply the method I proposed in Chapter 1 to examine two applications in economics. First, I study the problem of assigning individuals to job training programs. I calculate the welfare differences between different hypothetical policies using experimental data from the National Job Training Partnership Act Study. Second, I apply the method to study public health insurance policies. Specifically, I calculate the welfare impact of Medicaid expansion using data from the Oregon Health Insurance Experiment.
Chapter 3 (joint with Ching-to Albert Ma and Daniel Wiesen) studies how altruistic preferences are changed by markets and incentives using a laboratory experiment. Subjects are asked to choose health care qualities for hypothetical patients in monopoly, duopoly, and quadropoly. Prices, costs, and patient benefits are experimental incentive parameters. We combine a theoretical model of strategic interaction with a nonparametric estimation method and find that markets tend to reduce altruism.
Description
License
Attribution-NonCommercial-NoDerivatives 4.0 International