The Oregon Medicaid Experiment and New Possibilities for Quasi-Experiments

Jonathan Adler has already noted the new article in Science using data from Oregon’s Medicaid experiment. It found that expanding Medicaid led to a 41% increase in visits to the emergency room, including increases in visits for conditions that would be better treated in a primary care setting. Here is the abstract:

In 2008, Oregon initiated a limited expansion of a Medicaid program for uninsured, low-income adults, drawing names from a waiting list by lottery. This lottery created a rare opportunity to study the effects of Medicaid coverage using a randomized controlled design. Using the randomization provided by the lottery and emergency-department records from Portland-area hospitals, we study the emergency-department use of about 25,000 lottery participants over approximately 18 months after the lottery. We find that Medicaid coverage significantly increases overall emergency use by 0.41 visits per person, or 40 percent relative to an average of 1.02 visits per person in the control group. We find increases in emergency-department visits across a broad range of types of visits, conditions, and subgroups, including increases in visits for conditions that may be most readily treatable in primary care settings.

Like Adler, Sabrina Tavernise at the New York Times explains one reason this study might be important:

Supporters of President Obama’s health care law had predicted that expanding insurance coverage for the poor would reduce costly emergency room visits as people sought care from primary care doctors. But a rigorous new study conducted in Oregon has flipped that assumption on its head, finding that the newly insured actually went to the emergency room more often.

As some have noted, greater use of the emergency room is not necessarily a bad thing; one would want to look at health outcomes, as well as the modest increase in costs.

I wanted here to applaud the research design. It is rare that any social welfare agency does real randomized experiments, even though that is by far the best way to determine how programs actually work. The Oregon project is likely to generate lots of wonderful papers. I’ve often lobbied the law schools I’ve taught at to randomize access to some of their programs. In the rare times they have done so, very interesting results have emerged.

The rollout of the Affordable Care Act actually offers some great opportunities for quasi-experiments. For example, in some states Medicaid eligibility was expanded to include those making about $11,000-$16,000, while in other states it was not. Though the two groups of states are not perfectly matched, the opportunities for comparison are good.

As another example, before the Affordable Care Act, there were functioning individual health insurance markets in Guam and the US Virgin Islands, but after the Affordable Care Act there appear to be no companies offering private individual policies in those territories. In part that is because the individual mandate does not apply to US Territories while most of the rest of the Act does. This result already offers some support for the idea that the mandate is necessary if pre-existing conditions are required to be covered. One could study the effects on individuals in Guam and the Virgin Islands before and after the Act, and between the territories that had functioning individual markets before the Act (eg, Guam) and those that did not (eg, Puerto Rico).