For decades there was this debate about Medicaid, the health insurance program for the poor.
On one side were people making what seems like the straightforward argument: People who get Medicaid fare better than people who don't.
On the other side were those making the contrarian argument. They argued that there is already a safety net for the poor and the uninsured, and that Medicaid's reimbursement rates are so low that most doctors don't see Medicaid patients anyway.
The debate was perennial and unresolvable. You couldn't simply compare people with Medicaid to those without, because the two groups had different characteristics.
To truly answer the question, you would need to take a big group of people, and randomly divide them into two groups. One group would get Medicaid, and the other wouldn't. But doing that would be unethical.
Then, a few years back, Oregon announced that it had 10,000 new slots in its Medicaid program. Far more than 10,000 people wanted to enroll. So the state held a lottery.
The state of Oregon didn't mean to be creating the perfect experiment to test whether Medicaid actually helps people. But that's what it was doing.
Katherine Baicker, a health economist at Harvard, followed people who entered the Oregon Medicaid lottery — and she compared the outcomes between those who were given Medicaid, and those who were denied.
On today's show, we talk to her about what she learned — and we hear what the findings mean for the debate over health care in America.
KATHERINE BAICKER: It's natural to think, you know, health insurance must be good for your health. That's why people want health insurance. But, in fact, the data's not so clear. And there's more controversy in the public debate about this than you might think.
ALEX BLUMBERG, HOST:
Hello, and welcome to PLANET MONEY. I'm Alex Blumberg. And that was Katherine Baicker you heard at the top. She is a professor of health economics, the Harvard School of Public Health. And today on the show, Katherine and I will be talking about this big study of hers that did more to settle the debate that she mentioned than any study in decades. And this study of hers might not have even happened if a certain state government hadn't held a very unusual lottery. It's all coming up.
(SOUNDBITE OF SONG, "AMERICA")
IMAGINE DRAGONS: (Singing) Is this just an illusion that I made inside my head to get me by? Twenty years in debt...
BLUMBERG: For decades, in policy circles, there's been this debate about Medicaid - an unresolvable debate. Now, this debate has taken a particular significance because of the Supreme Court's upcoming ruling on the fate of the health care bill - you know, the big health care bill that the Obama administration passed.
On the one side, were people making what seems like the straight-forward argument - that expanding access to Medicaid helps the people who get access to Medicaid. In other words, Medicaid does what it's supposed to do. It makes you healthier and helps pay for your health care. But there is this other side making a contrarian argument about Medicaid - the argument that when you give people access to Medicaid, you actually hurt them. You make them sicker.
Now, this contrarian argument goes like this, Medicaid reimbursement rates are so low that a lot of doctors don't even see Medicaid patients. They won't see them. And so even though you're on Medicaid, it doesn't actually give you more access to health care. And what's more, it might hurt you because when you get on Medicaid, it removes you from this sort of hidden safety net that already exists.
So the contrarians argue that there are lots of charities and religious organizations already providing free or low-cost health services to the poor. And so when you put poor people on Medicaid, it takes them out of a non-governmental system that's already working and puts them into a dysfunctional government system that doesn't work.
Now, this debate has raged for decades with each side pointing to data and studies to bolster their arguments. But the problem has always been none of the data is clean. None of the studies are conclusive because none of the studies have what you need to make a definitive claim about something. They don't have a control group.
You can't do a big study where you take a big group of poor people and give half of them access to health care and half of them, you don't. That is not seen as ethical. And this is the way things were until very recently, when the state of Oregon, utterly by accident, delivered researchers in this field the control group that they'd been looking for for so long. Here is Katherine Baicker.
BAICKER: What happened in Oregon was that they had limited budget to enroll new populations in their Medicaid program for low-income adults. So in 2008, they had enough money to enroll 10,000 new people. But they knew that there were many more than 10,000 people who would like to gain access to this program. So they decided that the only fair thing to do would be to have a lottery from among people on the waiting list.
BLUMBERG: In the end, about 90,000 people signed up for those 10,000 spots which meant you got 80,000 poor people randomly denied access to health care. Now, if you're in the camp that thought Medicaid was good, this is a tragedy. If you're in the camp that thinks Medicaid is bad, it's a blessing in disguise. But in either case, it is the perfect control group that Katherine Baicker had been waiting for her entire career.
Her and her team got in touch with Oregon and went about designing a study. They sent out surveys to tens of thousands of people in both groups, Medicaid and non-Medicaid, asking all sorts of detailed health questions as well as questions about other stuff - financial situation, mood, psychological state. They also got hospital discharge records and credit reports on both groups.
And after analyzing the data, they came out with some pretty definitive results. And I will tell you right now, this is not one of those cases where the counterintuitive narrative is the true one. It turns out getting access to Medicaid, it definitely seemed to help.
BAICKER: The probability of having an unpaid medical bill sent to a collection declined by 25 percent for the group that was insured relative to the group that wasn't. So 25 percent fewer of the people in that group had a bill sent to collection. And that's a substantial reduction. And it was concentrated, as you'd expect, in medical bills. You were much less likely to have a bad medical bill sent to collection.
That's clearly very important for individuals access to credit and housing and jobs and all of that. It's also important for the health care system because bills that are sent to collection are almost never collected upon. So the ultimate cost of that bad medical debt is being borne by providers and by the other people they care for and might shift costs to if that occurs, not just the individuals who weren't able to pay the debt, although they clearly bear a substantial cost, as well.
So that's the formal access to credit. It also showed up in the informal credit or the things that don't show up in your credit report. The probability of having to borrow money or skip payment on other bills declined by 40 percent.
BLUMBERG: Wow, wow.
BAICKER: And this is a huge change in people's financial well-being. And we did this series of interviews and focus groups with several hundred people from this study to try to get a sense of what it really meant for their lives and well-being. And a story that we heard over and over again was that people really hated being uninsured because of the constant stress of knowing that they had bills that they were unable to pay.
It dissuaded them from seeing a doctor because, in the past, they hadn't been able to pay their bills. People talked about feeling like second-class citizens. One woman said that she wasn't raised not to pay her bills, that this was very upsetting to her much more broadly than the dollar amount on the medical bill might have suggested.
So this reduction in financial strain, I think, is an underappreciated aspect of what health insurance is supposed to do. It's not just supposed to gain you access to health care; it's supposed to prevent financial ruin if something bad happens to your health.
BLUMBERG: Right. What kinds of things do the people who are enrolled say?
BAICKER: They said that it changed so much about their interaction with the health care system that they could now - people said that they were much more likely to have a doctor who was their doctor - go to a doctor's office instead of a clinic to have care that they described as excellent or very good.
All of these measures of access to care mapped into people's reported experiences with the health care system - that they could now find a doctor who would treat them, that they could get follow-up care if the doctor recommended something. They could actually go do it. People reported not splitting their pills or taking their medication more regularly.
All of those things are consistent with, you know, what we saw as increases in use of preventive care. We saw mammograms increase by 60 percent. Cholesterol monitoring increased by 20 percent. These are big increases in primary care use. And this maps to the way that people were able to interact with providers once they had a way to pay for the care they were getting.
BLUMBERG: OK. So you saw these - there was, sort of, a very different reality, basically. Any other big changes that came through in that first initial wave of analysis?
BAICKER: Well, people reported their health to be much better once they were insured. We asked people things like, would you describe your health as excellent, very good, good, fair or poor? And the probability that they reported themselves in good, very good or excellent health increased by 25 percent. The probability of screening positive for depression went down by 10 percent. The probability that they said that their report - that their health was getting worse over the last six months went down by 40 percent. So these are major improvements in self-reported health.
But I want to insert a caveat on how to interpret those. Those improvements in self-reported health manifested themselves within a couple of months of people being insured, even though increases in utilization took longer to occur. So you do have to wonder what exactly people are telling you about.
Are they telling you just about their physical health, even though you've asked them about their physical health? Or are they also telling you about their overall well-being - that their, you know, stress from debt or not having access to care or being worried about whether they're going to get their medication next month - all of that probably played into people's answers about their health as much as their actual physical health did.
But that said, you know, as a policymaker, the question is, what do you care about? Do you just care about things like people's blood pressure? Or do you also care about depression? Which, I would think most policymakers do. But then one step further, what about generalized strain and, you know, well - how it affects people's well-being more broadly cast?
BLUMBERG: So let me just ask you, to what extent already has your study answered some of these questions that were out there?
BAICKER: I think there was a little something in our study for everyone to hate, which is a sign of a good study. The - it put to rest two very different arguments about the effect of expanding Medicaid to low-income adults, which is exactly the population we're seeing covered under the Affordable Care Act's Medicaid expansions.
On the one hand, some people said Medicaid is terrible insurance. It doesn't pay providers enough. People who gain access to Medicaid get a card. But they don't really get increased access to high-quality health care, so it's not going to improve their health by much. And our study says, no, Medicaid actually improves people's well-being along all these different dimensions in a measurable way.
On the other extreme, some people said, well, expanding Medicaid will actually save money. People will stop going to the emergency department. They'll stop going to the hospital. They'll get preventive care. It'll improve their health by so much that it actually pays for itself. And we should expand Medicaid because it'll actually save us money. And our study says, no, it costs money when you expand Medicaid. People consumed about 25 percent more health care. So it's not free. Although, it looks like it increases health care use by less than you would naively think if you just used the existing non-experimental studies.
And that leaves policymakers in the tough position of having to weigh real costs against substantial benefits. And economics can't tell you what the right answer is. You have to weigh the costs against the benefits and think about the alternative uses of the money and how much you value the improvement and well-being and what the opportunity costs of the money spent on health care is. But I think we eliminated the line of argument that Medicaid doesn't do anything to improve people's well-being. And we eliminated the line of argument that Medicaid is so effective that it's actually free.
BLUMBERG: Now, the so effect of that it's actually free that - is that not true even over the long term? Because my understanding was that, you know, it's sort of like there's - the part of that argument is that you pay a little bit in preventative care, and it will prevent sort of the costlier things from occurring down the line. So you get your blood pressure checked regularly. You don't have a very expensive, you know, coronary bypass surgery later or whatever.
BAICKER: No that's...
BLUMBERG: Yeah, is that not true?
BAICKER: That's a very good point in that our study, so far, has only told you about the first year of experience and soon will tell you about the second year of experience. You could still make an argument - well, 10 years out, you're actually going to save money. We don't have any data that supports that, but we can't refute that either.
The only thing I would note was that, at least, in the first year, we saw an increase in hospitalizations, not a decrease in hospitalizations. And the increase in hospitalizations was concentrated or showed up the most in planned hospitalizations, not admissions through the emergency department. But, rather, people going to the hospital for a scheduled procedure where their doctor had found something that needed to be corrected in the hospital, and they went to the hospital and had that procedure done.
That might - the long-run cost implications of that are not so clear. But I would argue against measuring the success of a health insurance expansion by whether or not it saves money in the sense that we don't say for food stamps - food stamps, they save money. We think, food stamps, they give people food. And that's what they're - that's the purpose. Health insurance, it gives people access to health care and that improves their health.
BAICKER: I think that the right metric shouldn't be, does it save money? The right metric should be, is however much it costs warranted by the benefits that it generates? Now, certainly, if it were free and it produced any benefits, it would then, obviously, be worth it. But I would argue even if it doesn't save money, if it produces enough health, it will be worthwhile.
BLUMBERG: There is a second part to this study. In addition to questionnaires and financial records, Baicker and her team also gathered actual blood samples from study participants. They've collected other forms of medical data, as well. And this data is being analyzed right now. The results should be complete within the year. And when they are, they'll be able to compare the two groups across a whole range of indicators - blood pressure, cholesterol level, a whole bunch of stuff. And this will be the most scientific answer yet to this deceptively simple question, does Medicaid make you sicker or healthier?
(SOUNDBITE OF IMAGINE DRAGON SONG, "AMERICA")
BLUMBERG: As always, we would love to hear your thoughts on today's show. You can reach us at firstname.lastname@example.org or find us on Facebook, Twitter or on our Tumblr. I'm Alex Blumberg. Thanks for listening
(SOUNDBITE OF SONG, "AMERICA")
IMAGINE DRAGONS: (Singing) Believe in you. Rise To the top of the world, America. America, don't you cry. Lift me up. Give me strength to press on. Rise to the top of the world, America. America, America, don't you cry. Transcript provided by NPR, Copyright NPR.