More than a week after some very smart health economists released findings from a study of Medicaid in Oregon, policy experts, politicians, and pundits are still arguing over exactly what the study showed—and how, if at all, it should change what we think about making health insurance more available to the poor.
That’s good. The study should force all of us, on the left and the right, to think more carefully about how we write and talk about health care reform. And I’ve read plenty of writers doing just that.1 But the idea that this study all but blows away the case for the Medicaid expansion, as some Obamacare critics have suggested, still strikes me as wrong-headed.2 And that’s because these critics make two unfair claims: that the study “proves” Medicaid doesn’t improve health and that financial security, which Medicaid clearly provided its Oregon beneficiaries, was an afterthought in the health care reform debate.
A refresher, for those of you who decided not to read the wonky set over the last ten days or so: A few years ago, when Oregon officials had to allocate 10,000 Medicaid slots from among about 90,000 applicants, they decided to use a lottery. That created two groups, randomly selected: One with an opportunity to enroll in Medicaid, one without the opportunity. Past studies of Medicaid suffered because those studies didn’t have such random assortment between people who were on the program and those who were not. People who enrolled on Medicaid tended to be poorer and sicker; researchers would adjust for that fact, but the adjustments might not have been correct. The Oregon Medicaid lottery made it possible for researchers to make a cleaner comparison of the two groups.
The headline from the study (at least on the right) is that the Medicaid beneficiaries didn't see statistically significant improvements in physical health. That's true. But it's important to understand what exactly that means. The folks on Medicaid appeared more likely to have lower blood pressure, cholesterol, and (among diabetics) blood sugar levels. But the improvement was smaller than the researchers had expected to see. And the researchers couldn't even be sure that improvement was real, because the sample size wouldn't allow them to pinpoint differences in health of that magnitude. Instead, they could give only a range of possibilities. At best, the Medicaid recipients had "clinically significant" improvements in physical health. At worst, they had no improvement at all, or perhaps even ended up in worse health.
What should we make of that? It's hard to know precisely without knowing what the researchers really should have expected in the first place. Katherine Baicker, a Harvard economist and a lead author of the paper, has said the Oregon experiment didn't produce the kind of physical health improvements researchers had found in previous studies, including the famous RAND health insurance experiment and some studies of MediCal (California's Medicaid program) during the 1980s. Aaron Carroll, a pediatrician and health care services researcher at Indiana University, thinks the comparison is flawed—that, in effect, the researchers were looking for a level of improvement that prior studies haven't really showed. I spent much of the weekend trying to sort out the particulars of this argument, via some lengthy e-mail exchanges with several of the researchers as well as some outside researchers. And I'm still not sure who’s right. Maybe they'll settle it soon.
In the meantime, there’s a point on which all the researchers agree. The Oregon study can't disprove that Medicaid produced physical health benefits, because it can't pinpoint the results with enough precision. Meanwhile, lots of other studies (not to mention intuition) suggest that people on Medicaid do end up healthier. Given this evidence, I would say the best bet is that Medicaid has at least some positive effect on health. If the authors are right about the expectations—and, for the sake of argument, let's assume they are—then the Oregon findings should make us less optimistic about the magnitude of improvement or less confident that the improvement is real. But we should still think that, all else equal, people who go onto Medicaid will end up healthier. The question would be how much.
Should Obamacare's defenders be careful about predicting that simply giving people Medicaid will yield large improvements in health? I think so. Here it’s useful to think about what might have actually been happening in Oregon. A good place to start would be with an interview with Mary Carson, an Oregonian who actually ended up on Medicaid. In the interview, conducted by Jordan Rau of Kaiser Health News, Carson explains that getting healthy is not easy. You have to go to the doctor, take your medication, change your lifestyle, and, by the way, hope the medication works. (Sometimes it takes a while to figure out which drug works, and at what level, etc.) Medicaid, in other words, is a necessary but not sufficient precondition for better health.
But the best way to think about this is the way that Bloomberg's Josh Barro did: The problem with Medicaid isn’t Medicaid. It’s our health insurance system, which is lousy at getting people good care. If that's true, then expanding Medicaid make more sense if it's happening alongside other reforms designed to make health care better. And that's pretty much what is happening. Obamacare isn't just giving people insurance. It’s also trying to reorganize the way people get health care, by, among other things, getting more people into “medical homes” that make precisely the kind of improvements the researchers hoped to find in Oregon. I wouldn't pretend that Obamacare is going to complete this transformation overnight—or anytime soon—anymore than I'd pretend it's going to make sure every American has insurance on January 1, 2014. But I would argue it's pushing the entire health care system in the right direction and that, over time, more coverage will translate into better health.3
The glass-half-empty version of that argument is that government will never get health care right, making the whole Medicaid exercise pointless. This is what many conservatives believe. But Medicaid clearly delivered other benefits for the enrolled Oregonians, starting with mental health. The presence of large improvements in mental health were just as surprising to the researchers as the lack of large improvements in physical health. Better mental health might be partly the result of a placebo effect, or the fact that people who don’t worry about medical bills have one less reason to be depressed. Whatever the reason, it's still "a big deal," as Adrianna McIntyre wrote at the blog Project Milennial. Not only might better mental health save some money, given the costs to society of lost productivity and activity. It should also save lives.
The findings on financial security were just as significant. Medicaid effectively wiped out catastrophic medical expenses among the poor—and improved financial security in other ways. Obamacare critics have ignored or dismissed this finding, saying that financial security wasn’t a primary reason for health care reform. I suggest they go back and read the speech that Harry Truman gave in 1948, or the one Lyndon Johnson gave in 1965, or the one Bill Clinton gave in 1993, or the one President Obama gave in 2009. Financial security—specifically, the need to protect Americans from crushing medical bills—is the dominant theme in each of those speeches.
To be clear, better health was a big theme of these speeches, too. But to ignore the financial security benefits, as well as the mental health benefits, would be as wrong as dismissing the findings on physical health as irrelevant. Jonathan Gruber, the MIT economist who was also a co-author of the Oregon study, told Talking Points Memo that “I would view this study as somewhat weakening the argument for universal coverage based on health improvements ... and greatly strengthening the argument based on financial security and mental well being.”
In fairness, there’s a more complicated question here: Whether, all things equal, Medicaid is the best possible way to provide the benefits it does—or, at least, the best of the available alternatives. That’s the question that such Obamacare critics as Ross Douthat and Megan McArdle raised—and one I’ll address in a subsequent item.
Jonathan Cohn is a senior editor at The New Republic. Follow him on Twitter at @CitizenCohn.
Barro would actually disagree with this part, I think—but that's a topic for my next column.