Blog
The Oregon Medicaid experiment, explained for students
5 min read, by the Student Research Institute
In 2008 Oregon had money to add about 10,000 low-income adults to Medicaid and about 90,000 people who wanted in. The state ran a lottery. That accident of budgeting produced the only randomized study of what health insurance does for adults in the United States.1
Why a lottery matters
People who have insurance differ from people who do not, in income, health, and a hundred things nobody measures. Compare the two groups and you cannot tell what insurance did from what those differences did. A lottery breaks the link. The winners and losers were the same kind of people on the day of the draw, so any gap that opened afterward came from the coverage.
What the first year showed
Winners used more health care: more doctor visits, more prescriptions, more hospital stays. They were far less likely to have medical debt sent to collections or to borrow money to pay a medical bill. They reported better health and less depression.1
What the second year showed
A team then examined about 12,000 people in person, drawing blood and measuring blood pressure. Medicaid cut the share of people who screened positive for depression from 30.0 percent to 20.9 percent, and it nearly eliminated catastrophic medical spending. It did not produce a significant change in measured blood pressure, cholesterol, or blood sugar over the two years.2
Source: A drop of 9.2 percentage points. Financial strain also fell. Measured blood pressure, cholesterol, and blood sugar did not change significantly.2
How to read that
Both sides of the insurance debate quote this study, and both are quoting it correctly. Coverage protected people financially and improved mental health. It did not, in two years, move the physical measures that clinicians watch. The honest summary is the whole sentence, not half of it.
Why students should know it
The Oregon study is a model for what a health economics project can look like: a clear question, a source of random variation, public data, and results reported whether or not they flatter anyone. The papers are in our journal article library, and the data sources page lists where similar public data lives.
Sources
- Finkelstein A, Taubman S, Wright B, et al. The Oregon Health Insurance Experiment: evidence from the first year. The Quarterly Journal of Economics. 2012;127(3):1057-1106. doi.org
- Baicker K, Taubman SL, Allen HL, et al. The Oregon experiment: effects of Medicaid on clinical outcomes. The New England Journal of Medicine. 2013;368(18):1713-1722. doi.org
Sources checked September 21, 2026. See every source we cite.
For companies and foundations
Fund a student researcher
Your gift pays for rent, a stipend, books, or a journal fee. It goes to a student who has the position and lacks the money to take it.