Public HealthTargetingMixed

Oregon Medicaid Lottery

Harvard University / MIT / NBER · Oregon, USA · 2008

Summary

The Oregon Medicaid Lottery is one of the most debated experiments in US health policy. Its finding that Medicaid coverage produced no statistically significant improvement in physical health measures at 2 years — blood pressure, cholesterol, blood glucose — was widely cited as evidence that insurance coverage doesn't improve health. But the finding was more nuanced: recipients used substantially more health care, were dramatically less likely to face financial catastrophe from medical bills, and showed large reductions in depression. The physical health null result most likely reflects the short follow-up period (2 years is insufficient for most chronic disease improvements) and the fact that 60% of uninsured adults already had some prior care access. The debate over the Oregon results reveals as much about how we measure health policy success as about what Medicaid actually does.

Research question

"Does expanding Medicaid coverage to low-income uninsured adults improve physical health, mental health, and financial security?"

Methodology

Intervention

Oregon had resources to expand Medicaid to only 10,000 additional adults on the waiting list of 90,000. Rather than using an arbitrary selection rule, the state held a lottery: 35,169 individuals were selected by lottery and 74% enrolled in Medicaid. Researchers used lottery selection as a natural experiment to estimate the causal effect of Medicaid coverage vs. remaining uninsured. Outcomes measured at 2 years.

Assignment

Lottery-based natural experiment; 89,824 names on the Oregon Health Plan waiting list; 35,169 lottery winners, ~30,000 enrolled in Medicaid; outcomes compared by lottery selection (intent-to-treat) and by Medicaid enrollment (instrumental variables)

Sample size

~12,000 individuals surveyed at 2-year follow-up; 6,387 received in-person clinical assessments

Primary outcome

Physical health outcomes (blood pressure, cholesterol, blood glucose, glycated hemoglobin); mental health (depression screening score); financial hardship (out-of-pocket spending, catastrophic medical expenses)

Effect estimate

Physical health: no significant effects on blood pressure, cholesterol, or diabetes management at 2 years. Mental health: significantly lower rate of positive depression screening (−9.2 pp). Financial security: significantly lower catastrophic medical expenses (−25%), lower medical debt, lower probability of borrowing money for medical care. Healthcare use: significantly higher (primary care, hospitalizations, prescriptions).

Decision

Results were among the most debated in health economics; interpreted by conservatives as evidence Medicaid is not worth its cost; interpreted by proponents as showing Medicaid protects against financial catastrophe and depression even if it doesn't improve physical health markers in 2 years; findings influenced the ACA debates, state waiver discussions, and Medicaid work requirements debate; ongoing follow-up tracking longer-term mortality and health outcomes

View source report →

Result

Mixed

Evidence strength

Strong

Randomized controlled trial with large sample.

Replication status

Open for replication

Institution

Harvard University / MIT / NBER

Location

Oregon, USA

Year

2008

Policy area

Public Health

Mechanism

Targeting

Could your jurisdiction test this?

This finding has not yet been independently replicated. A replication in a new context — a different city, country, or population — is one of the highest-value contributions to the evidence base. If your team could host a replication, we can help with study design, sample-size estimation, and analysis.

Propose a replication →What does "open for replication" mean?

← Back to registry What the evidence shows →