Abstract:
In 2008, Oregon policymakers offered low-income and uninsured residents access to health insurance called the Oregon Health Plan Standard through a lottery rationing device. Studies done in conjunction with the Oregon Health Insurance Experiment to examine subsequent outcomes use an instrument variable design where assignment to the treatment group (selection in the lottery) serves as a proxy for actual treatment (enrollment in health insurance). Consequently, previous research assumes that the entire treatment group act as “compliers” and apply for enrollment. This thesis thus examines the likelihood that an individual selected via the lottery system applied to the Oregon Health Plan across demographic variables such as racial and ethnic identification, gender, income strata, educational attainment, employment status, age, household size, and geography. This thesis finds statistically significant evidence to support that take-up of the Oregon Health Plan Standard is non-random across demographic traits.