Medicaid Expansions 2014: What Can Be Learned From the Oregon Health Insurance Experiment?

Background and Purpose: In 2008, a group of uninsured low-income adults in Oregon were selected by lottery for the chance to apply for a Medicaid-expansion program. This natural experiment provides the unique opportunity to assess the effects of expanding access to public health insurance on the health care use, finances, and health of low-income adults using a randomized controlled design. This is the first U.S. study to use the gold standard of research to evaluate the impact of insuring the uninsured and is poised to inform the Medicaid-expansion efforts scheduled to begin in 2014.

Methods: This analysis relies on both primary and secondary data sources to evaluate healthcare access, quality, financial strain, and self-reported health. We mailed a baseline and 12-month follow-up survey to almost 60,000 individuals who signed up for the lottery, approximately half who were selected (treatment) and half who were not (controls), with a weighted response rate of 50%. Administrative data, not subject to self-report or survey response bias, included hospital discharge data, credit reports, mortality records, and Medicaid administrative data. To account for imperfect program take-up, we produced two types of equations for each outcome measure: 1) Ordinary Least Squares (OLS) modeled the difference between treatment and control groups, preserving the integrity of randomization; and 2) an instrumental variable approach, Two-stage Least Squares (2SLS) used the difference in insurance coverage between the two groups attributable to the lottery (the “first-stage”) to estimate the treatment effect if every person who won the lottery had gained coverage.

Results: About one year after enrollment we found that insurance coverage is associated with improved access and quality of health care, reduced financial strain, and big improvements in self-reported physical and mental health.  Access: Medicaid coverage increases the likelihood of using outpatient care by 35%, prescription drugs by 15%, and being admitted to the hospital by 30% but does not initially appear to have an effect on use of emergency rooms. This increased utilization translates into about a 25% increase in annual health care spending.  Quality: Medicaid increases the probability individuals report having a regular source of primary care by 70% and the likelihood of having a particular doctor they see by 55%. Insurance also increases the use of preventive care, such as mammograms (60%) and cholesterol monitoring (20%).  Financial Strain: Medicaid coverage decreases the probability of having to borrow money or skip paying other bills to pay for health care by 40%, and the probability of having an unpaid medical bill sent to a collection agency by 25%. Health: Medicaid coverage increases the probability that people report themselves in good to excellent health (compared with fair or poor health) by 25% and increases the probability of not being depressed by 10%.

Conclusions and Implications:  Our initial findings suggest that public coverage expansions are more cost-effective than previous non-experimental studies have indicated. Gaining insurance had a significant positive impact on the lives of individuals across all measurement domains.