Medicaid Fails Patients at an Astronomical Cost
July 10, 2017
With healthcare reform sputtering in the Senate, warring lawmakers are endlessly arguing over what’s broken and how to fix the problems. The debate over reforming Medicaid is particularly heated, with narratives from all sides engulfing America’s predominant low-income insurer. In diagnosing the state of Medicaid, the differing assessments of the right and the left make for an interesting Rorschach Test. To program advocates on the left, Medicaid serves as an effective insurer for poor and near-poor Americans, shielding them from the medical misfortunes of life. Many critics on the right, however, suggest that Medicaid is just another malfunctioning, ineffective federal program. At an annual price tag of nearly $600 billion, taxpayers need to know if Medicaid is a worthwhile investment in its present form.
Despite the claims of many defenders of the status-quo, most signs point to the program’s colossal failure. In assessing the impact of any government policy or program, economists prefer to use a randomized control study. In this type of analysis, the program of choice is applied “randomly” to a subset of people within a broader population, while the rest of the population sits on the sidelines. Ideally, a random assignment ensures that the program recipients are identical in attributes to the population not assigned the “treatment.”
This, of course, is nearly impossible to conduct on a large scale, because, well, there are many problems ethical and otherwise with randomly assigning a government program to certain individuals. But in 2008, Oregon responded to a limited federal supply of Medicaid expansion dollars by setting up a randomized lottery and commissioning economists to study the health results for the selected individuals. After two years they published their findings of the health impacts of Medicaid, and to the dismay of many, no significant physical health benefits of Medicaid were found. Sure, rates of self-reported health shot up, but two-thirds of this effect seemed to occur near the onset of the program and may be tied to the idea of having insurance coverage rather than actual care.
While many of the right declared victory in the intellectual battle over Medicaid, the left prepared a counter-crusade. And, as a recent column by Ezra Klein shows, there’s plenty of evidence seemingly contradicting the Oregon study that ought to be examined. The problem, though, is that the analyses don’t live up to the randomized “gold standard” study cited by policy gurus on the right. Take, for example, a study led by Harvard Professor Benjamin Sommer comparing health outcomes in states that expanded Medicaid verses states that did not. While the headline figures suggest that Medicaid significantly improved health outcomes, the individual state-by-state comparison shows a significant decrease in mortality only for New York (NY) State.
And while the Empire State did generously expand Medicaid in the early 2000s period targeted by the study, many other factors besides Medicaid could have led to NY health outcomes relative to other states. New York City, for example, saw sharper decreases in particulate matter emissions in the same time period than many other cities, and ramped up anti-smoking campaigns. One way to see if New York’s health results are the result of Medicaid expansion or other factors is to examine whether or not decreases in death rates were concentrated amongst low-income folks. But as Sommers and his research team note, changes in death rates are observed for high-income individuals and senior citizens unable to take advantage of Medicaid expansion. Clearly, then, something else was driving the results.
Other studies purporting to show the benefits of Medicaid expansion similarly fail to pass muster. Another study, led by Indiana University Professor Kosali Simon, also attempts to compare health outcomes in Medicaid expansion states and non-expansion states. Strangely, statistically significant health effects were only found for low-income childless adults, despite a large expansion effort targeted toward low-income parents. This bizarre result suggests either that families are incapable of benefiting from Medicaid, or other factors are behind the selective increases in health.
These studies will inevitably be compounded by other studies claiming to show similar Medicaid benefits. But, given the existing stock of research, taxpayers are right to doubt the effectiveness of this half-a-trillion-dollar entitlement program. If proponents are right, Medicaid improves health in the straightforward way of providing low-income patients with increased access to physicians and treatments for maladies. Why, then, are findings so state-specific and non-applicable to huge swaths of beneficiaries? Medicaid’s defenders will need to produce better research if they want more of taxpayers’ hard-earned money.