This column is a collaboration between KHN and
The New Republic.
Are you better off with Medicaid than if you had no insurance at all? The answer seems like a no-brainer: Of course you are. But, for the last few months, a cadre of conservative writers and intellectuals has argued that the program doesn’t actually help beneficiaries and may actually hurt them. To prove their point, they’ve cited a handful of studies in which Medicaid recipients ended up in worse health than people with no coverage whatsoever. According to Medicaid’s critics, this evidence suggests that expanding the program, as the Affordable Care Act would, is a bad idea.
Most social scientists I know reject these arguments. The reason: When you’re comparing people on Medicaid to people with no health insurance, there’s no simple, sure-fire way to account for the underlying differences in the two populations. For example, if you know you suffer from serious medical problems, you’re more likely to sign up for public insurance when it’s available. As a result, the Medicaid population may be fundamentally sicker than the uninsured population — and end up with worse medical outcomes, even if they’re benefiting from the program’s coverage.
The best studies attempt to account for this possibility. And, mostly, they show that people on Medicaid benefit from the program: They have an easier time getting medical care and, as a result, they end up healthier. But even their adjustments aren’t perfect. There simply hasn’t been a way to conduct an experiment that cleanly and randomly separated the Medicaid population from the uninsured population. At least until now.
In 2008, officials in Oregon decided to make Medicaid available to more of the state’s residents. But they appropriated only enough money to add 10,000 people to the rolls — and 90,000 people applied. To cull the list, officials held a lottery, creating the sort of random experiment that would be impossible to create under different circumstances. After all, the only obvious difference between those who ended up on Medicaid and those who didn’t was that 10,000 of them got the right lottery number.
Experts have been watching to see what happens and, on Thursday, they published their initial findings through the National Bureau of Economic Research. Their conclusions? The past studies, like the intuition about Medicaid, appear to have been correct. People with Medicaid really are better off than people without insurance. A lot better off, in fact.
Among the most obvious differences that researchers observed in the two populations was what wonks call “health care utilization” and everybody else calls “getting medical care.” Compared to people who did not get Medicaid, most of whom remained uninsured, the Medicaid beneficiaries were 15 percent more likely to use prescription drugs, 20 percent more likely to get cholesterol monitoring, 35 percent more likely to use outpatient care, and 60 percent more likely (!) to get mammograms. The people on Medicaid also reported that they were healthier because of the program: According to the paper, 25 percent said in surveys that they were in “excellent” health, rather than “fair” or “poor.”
Of course, more health care services don’t always lead to better health. And people don’t always assess their own medical status precisely. The truth is that more objective measures of health probably won’t show big changes, in one direction or the other, for some time. The benefits of taking high blood pressure medicine, for instance, might not show up until many years in the future, when (hopefully) fewer people suffer heart attacks and require expensive medical interventions.
But the study demonstrates clearly, and persuasively, a different benefit of Medicaid: It provides beneficiaries with economic security. The Medicaid population was 40 percent less likely to borrow money or avoid paying other bills because of high medical expenses. The likelihood that unpaid medical bills ended up with a collection agency was also 25 percent lower. Not coincidentally, people on Medicaid were 55 percent more likely to report having a doctor they see regularly and 70 percent more likely to report they had an office or clinic for care.
To be fair, not all of the findings in the new paper make the case for expanding Medicaid easy, at least in the political sense. Emergency room use does not appear to have declined, confounding a promise reformers frequently make, although such a decline might also take time to materialize. And the study found that the total spending on health care for the new Medicaid beneficiaries increased by 25 percent, even though reformers say that expanding health insurance will lead to reduced spending on health care overall.
Then again, that last finding makes sense: These new Medicaid recipients got more medical care and, from the looks of things, they needed it. It was bound to cost more money at the outset. But getting these people into more stable health insurance arrangements can make the health care system as a whole more efficient. And by lowering the overall burden of charity care that doctors and hospitals must provide, expanding Medicaid can create political support for reforms that will effectively pay these providers of care less over the long run. That, more or less, is how the Affordable Care Act is supposed to control health care costs in the long run.
Might future research show this study, too, has flaws? Sure. No research project is foolproof and even the study’s authors caution about extrapolating too much from this one study. But it’s not just this study’s design that makes it unusually significant. It’s the all-star team of scholars that produced it. Among the paper’s authors are Harvard’s Joseph Newhouse, whose analysis of the 1970s Rand “Health Insurance Experiment” is still the gold standard for health policy research, and MIT’s Jonathan Gruber, arguably the nation’s most respected authority on modeling the outcome of health care policy interventions. And although Gruber is a well-known advocate for Medicaid who advised Democrats during the Affordable Care Act debate, one of his co-authors is Harvard’s Katherine Baicker. She served on the Council of Economic Advisers during the Bush Administration.
Oh, did I mention that the paper’s principal co-investigator is Amy Finkelstein, also of MIT. Conservatives frequently cite her research on Medicare as evidence of that program’s limits. I’ve always thought conservatives misinterpret those particular findings, but it speaks to the respect her work commands. And her conclusion about Medicaid seems pretty unambiguous. “Some people wonder whether Medicaid coverage has any effect. The study findings make clear that it does.”
I’m not naïve enough to think this paper will make Medicaid’s most ardent critics rethink their position. But it should.
Jonathan Cohn is a senior editor at The New Republic .