In health economics circles, the word Oregon usually has a negative association – it was the site of one of the greatest health economics failures when it tried to making reimbursement decisions according to strict cost-effectiveness criteria. But all that is about to change. Preliminary estimates of what may prove to be the most important study of the impact of health insurance – at least among low-income populations – on outcomes were released this week from a group of researchers from the NBER, Harvard, and MIT (the NBER working paper is here). I first heard about this study a few years ago and I have been eagerly awaiting the results ever since.
Back in 2008, Oregon had a long waiting list of low-income adults wanting to enroll in its state Medicaid program. Given severely constrained resources it was not able to provide insurance to everyone who wanted it, so it decided to allocate eligibility to enroll into the program by lottery – it randomly assigned insurance eligibility – creating one of the most incredible opportunities to study the impact of health insurance. Period.
Although there have literally been thousands of studies that have compared outcomes between people with insurance and people without almost all of these studies suffer from selection problems: when enrollment is voluntary, people who have insurance are different than those that do not. Finding opportunities to study health insurance is therefore challenging.
One year after the randomization, the authors find:
Being selected through the lottery is associated with a 25 percentage point increase in the probability of having insurance during our study period. This net increase in insurance appears to come entirely through a gross increase in Medicaid coverage, with little evidence of crowd-out of private insurance. Using lottery selection as an instrument for insurance coverage, we find that insurance coverage is associated with a 2.1 percentage point (30 percent) increase in the probability of having a hospital admission, an 8.8 percentage point (15 percent) increase in the probability of taking any prescription drugs, and a 21 percentage point (35 percent) increase in the probability of having an outpatient visit; we are unable to reject the null of no change in emergency room utilization, although the point estimates suggest that such use may have increased. In addition, insurance is associated with three-tenths of a standard deviation increase in reported compliance with recommended preventive care such as mammograms and cholesterol monitoring. Insurance also results in decreased exposure to medical liabilities and out-of-pocket medical expenses, including a 6.4 percentage point (25 percent) decline in the probability of having an unpaid medical bill sent to a collection agency and a 20 percentage point (35 percent) decline in having any out-of-pocket medical expenditures.
So health insurance seems to increase the use of services – notably of both preventive and treatment services – and seems to increase financial risk protection (arguably the main purpose of health insurance). In addition, they find:
Finally, we find that insurance is associated with improvements across the board in our measures of self-reported physical and mental health, averaging two-tenths of a standard deviation improvement. These results appear to reflect improvements in mental health and also at least partly a general sense of improved well being; they may also reflect improvements in objective, physical health, but this is more difficult to determine with the data we now have available.
These early findings thus provide some pretty convincing evidence that – at least among a low-income uninsured population who has expressed interest in having health insurance – the Oregon Medicaid program has provided important health and non-health benefits to those who enroll. Of course the strict generalizability of these findings to other contexts might be somewhat limited in terms of the populations it targeted, the program design in Oregon, and the environment in which it was given – but in many ways many governments are grappling with issues about how to specifically address similar kinds of populations so the findings are likely relevant elsewhere as well.
I was recently talking to a colleague about the impact of user-fees on outcomes and he mentioned that politically it would never be feasible to run an experiment that randomly assigned different user-fee regimes in populations. But I think this experiment is further evidence that you can – and perhaps should – as not only does it allow for amazing research opportunities such as this but it might also be a more fair way to allocate limited resources.
These preliminary results are from one year after the randomization. I am eagerly awaiting the more medium term impact results as well.Share on Facebook