Is Medicaid a bad idea?
That's what Tyler Cowen asks after reading a working paper from economists Amy Finkelstein, Nathaniel Hendren, and Erzo Luttmer. The paper is highly technical, but its headline finding is that a dollar in Medicaid spending doesn't lead to anything near a dollar in value for Medicaid beneficiaries.
It finds this two ways. First, if you test whether Medicaid users would pay the price of their Medicaid coverage if the government took it away, they say no. This makes some sense, as they are very poor, and one reason they have Medicaid is they don't have the money to buy insurance on their own.
The more interesting finding is the second one: if you look at where Medicaid's money goes, less of it than you might think goes to covering the uninsured — and more goes to paying back the people who are already covering the uninsured.
But the paper's conclusions are more complicated than that makes them sound. Read closely — and depending on your values — the paper either isn't saying anything very new, or it's arguing something quite radical: if Medicaid is overrated, it might be because all health insurance is overrated.
The main finding: the uninsured are kinda-insured
There's one fact that drives pretty much every other finding in Finkelstein's new paper: the poor and uninsured only pay about 20 percent of the cost of their medical care.
Who pays the rest? Finkelstein and her co-authors don't know. There's good evidence that a lot of it is borne by hospitals, who in turn pass it on to the government, the insured, and so on. But it's also likely that some of these costs are being borne by churches, charities, family members, etc.
Whoever is paying these bills, the result is that a lot of Medicaid spending ends up offsetting this pre-Medicaid spending. "We estimate that a substantial portion of the government's Medicaid spending — $0.6 on the dollar — represents a transfer to the providers of this implicit insurance," the researchers write.
It's perhaps easiest to explain this through example. Imagine John breaks his leg. If John is uninsured, his brother, Mike, pays for his medical care. But if John has Medicaid, then the government pays for his care.
John got medical care either way. So in this case, Medicaid's money actually didn't go to John so much as it went to his brother, because it was his brother who actually would have ended up paying the tab.
This doesn't always happen, of course — a lot of people without insurance simply go without necessary care, or they go into debt to get the care they need — but it happens a lot. It happens enough to drive pretty much every other finding in Finkelstein's paper. It's the main reason a dollar in Medicaid coverage isn't worth anything near a dollar in health care to Medicaid's beneficiaries.
What the paper doesn't tell us
First, the paper can't answer whether there are gains from giving people actual Medicaid insurance rather than leaving them to whatever patchwork, uncertain system of care they're using now. That is to say, it doesn't even try to estimate how much it's worth to be able to see a doctor when you need one, as opposed to when the situation is so dire you simply rush to the ER; it doesn't know how to value the long-term health benefits of stable care or the differences in the kind of care that the insured and the uninsured get; it has no formula for weighing what it means for John to be able to get treatment without begging his brother to lend him cash.
Second, there is real cost — in anxiety and terror, as well as in money — to families scrambling to come up with the money to pay for heart medication or chemotherapy. There's real cost to parents who need to beg their local church group to help pay for their child's medicine. How do we value the relief a family gets — both emotional and financial — of knowing a child can get the medical care he or she needs? This study can't measure that.
Third, the study can't test the value we, as a society, place on everyone being able to go to the doctor when they need medical care. As an example, Social Security offsets a certain amount of support children used to provide for their parents. So a dollar in Social Security is not worth a full dollar to Social Security's beneficiaries, because it partially replaces support they would have gotten anyway. But as a society, we've decided it's really, really important for the elderly to have guaranteed income, and we are willing to pay the cost of that guarantee.
Medicaid raises the same question. If we want to have a society where people can get the medical care they need, then Medicaid makes sense. If not, then perhaps it doesn't.
Why does anyone care about some random Medicaid paper?
To understand why this particular paper matters, there's a bit of academic context worth knowing. Its lead author, MIT economist Amy Finkelstein, is perhaps the most celebrated health economist in the country right now, having recently won the prestigious Clark Medal. And part of Finkelstein's celebrity comes from a massive experiment that underlies this paper, too.
The experiment had grim origins: Some years back, Oregon scrounged up the funds to add 10,000 residents to the state’s Medicaid program. The problem? There were 90,000 residents who qualified. So the state held a lottery, because that's apparently how we make life-and-death decisions in the United States of America.
The upside, though, was the lottery created something that health researchers had wanted for decades: a way to isolate the effects of health insurance. A gaggle of health economists, led by Finkelstein and Harvard's Katherine Baicker, convinced Oregon to let them follow a randomly selected group of the lottery's winners and losers.
That may not sound like a big deal, but in health economics, it was an earthquake: there had never been a randomized controlled study comparing the uninsured with the insured, because giving some people health insurance and withholding it from others just to see what happens is, well, kind of a horrible thing to do.
The study tracked two years' worth of health data, and came to some surprising findings. Medicaid worked as health insurance: The enrollees got more medical care (including preventive care) and were protected from catastrophic health costs.
This might seem obvious, but it wasn't: There was real concern, prior to this study, that Medicaid paid doctors so little that the insurance itself was worthless because providers wouldn't take it.
But that wasn't the case. "People who gained access to Medicaid did use more health care," Baicker told me. "We can eliminate the story that Medicaid is so lousy you can’t get in to see a doctor."
What was harder to find was the effect of all that health care on actual health. People said they were healthier, and, interestingly, they reported much lower levels of depression, but there was no significant change in indicators like blood pressure or cholesterol, and the study didn't last long enough — or have enough enrollees — to tell us much about more serious conditions or overall death rates.
Moreover, as good as the study was, the researchers were limited to two years of data, among one particular population, in one particular geographic area. We have no idea what the results would be if the researchers followed, say, 20,000 Texans for 15 years.
Even so, Finkelstein's new paper is, to a large degree, based on the results of that previous study: If the initial experiment was an effort to try to figure out what Medicaid did, the new paper tries to extrapolate those results to show what Medicaid is worth.
What is health insurance actually worth?
Recall what the initial experiment, limited though it was, found: Medicaid worked perfectly well as health insurance. The problem was the health care that the health insurance purchased didn't seem to do anyone much good.
In a traditional economics framework, that's not actually a problem: health insurance exists to insulate people from catastrophic health costs. So long as it's doing that, it's working.
But no one actually feels that way about health insurance. Indeed, health insurance isn't really insurance anymore — not in the way that fire insurance or car insurance is. People buy health insurance so they can use more health care. People don't buy car insurance so they can get into more accidents.
Health insurance is supposed to buy us health care, which is supposed to buy us health — that's why we pay so much for it, and why many Americans think that everyone should have it, regardless of whether they can afford it.
It isn't clear, though, that health insurance is buying us nearly as much health as we hope. America's health-care costs are far higher than any other nation, but we're not any healthier. We know that between 15 and 30 percent of the care we get is wasted, and we know that when people go to the doctor, there's a damn good chance they're not going to get the treatments they're supposed to get.
The reason the Oregon Medicaid experiment was so interesting is that it was the first methodologically sound study to truly try to test whether health insurance actually leads to health improvements, and while its results were by no means definitive, they definitely weren't encouraging, either. The implications of that are big: in 2014, the average employer-provided health plan cost more than $16,000 — and the truth is, we really don't know how much healthier it made anyone.
That said, we know from the medical literature that a lot of health care really does improve health. We know that statins control cholesterol and chemotherapy improves cancer survival and bones heal better when they're skillfully set. So we know that much medical care works. And we know that health insurance improves access to medical care. But that's about where our knowledge ends.
This uncertainty is sometimes taken as an argument that we should cut subsidies for health insurance for the poor. Usually, the people making this point have extremely generous employer-provided coverage that they would never think of giving up — it's a gamble they are only willing to make with someone else's care.
I think a sounder conclusion is that we, as a society, often get less value out of health insurance than we think, and that that might be particularly true for people who have weak ties to the medical system (which would include people who have long been uninsured) or particular difficulties getting care (like language or transportation barriers).
That isn't to say health insurance is useless, or that medical care doesn't help. But we're probably paying too much and getting too little, and now that we're a lot closer to a world where every American who wants health insurance can afford it, we should be focusing on making sure that all that health insurance we're buying is actually delivering the health we're expecting.
All of which is to say, the next great health-care campaign shouldn't be about cutting cost. It should be about increasing value.