POLITICS DECEMBER 7, 2007
After all three Democratic presidential candidates released their health care proposals this year, a general consensus emerged on who was going to do what. Barack Obama had put forward a substantial plan that would reach a lot of people; Hillary Clinton and John Edwards had put forward even more ambitious plans that would reach even more people. So it was more than a little surprising when, about two weeks ago, the Obama campaign circulated a memo titled “Why the Obama health insurance plan covers more people more rapidly than the Clinton plan does.” And it was even more surprising when the essential arguments in that memo started finding its way into press coverage, most notably a news article in The New York Times called “Clinton Attack on Obama Overlooks Some Realities.”
If you've followed this debate at all, then you know the policy question behind all of this: whether to impose a so-called individual mandate. Clinton and Edwards would require that every single person carry health insurance. Obama would not, although he would make the coverage of children mandatory. When Obama first announced his plan, he suggested he was eschewing a mandate primarily because he wanted to take things a little more slowly--to put his other reforms in place before actually requiring everybody to get insurance. If it turned out he needed a mandate later on, he made clear, then he would do it.
Lately, though, he has been arguing as if his decision on a mandate, considered by some (including me) to be his plan's chief failing, is actually a virtue--and that he, frankly, has a better understanding of this whole problem than his rivals do. As he recently said in Iowa, “Their essential argument is the only way to get everybody covered is if the government forces you to buy health insurance. If you don’t buy it, then you’ll be penalized in some way. ... What I have said repeatedly is that the reason people don’t have health insurance isn’t because they don’t want it, it’s because they can’t afford it.” And while his advisers aren't trying as hard as they once did to suggest that Obama's plan would actually cover close to everybody--an argument that, as I explained earlier this week, the best available evidence contradicts--they have been pushing the line that mandates are inherently problematic, which is why Clinton and Edwards's plans could be expected to do no better than Obama's.
These arguments are resonating both among conservatives who simply don't like the government telling anybody what to do and liberals who worry a mandate simply cannot work in practice. It's true that nobody knows exactly just how much impact a mandate would have. But if it would be wrong to have blind faith in mandates, it would be even more wrong to discount them. The available evidence suggests that a mandate does make a big difference--and that, while it may not guarantee full universal coverage, it offers the best prospects for getting there.
The logic of a mandate begins with understanding exactly why Obama's essential diagnosis of the problem--that it's all about affordability--is wrong. It's certainly true that cost is the single biggest reason 45 million Americans don't have health insurance today. That is why all three Democrats have proposed adopting the same set of strategies to make insurance more affordable: Requiring insurers to sell to everybody, at the same price, regardless of medical condition; creating a public program into which anybody can enroll; cutting down on wasteful treatments that drive up the price of medicine unnecessarily; and offering financial assistance to those people who might still need help paying for coverage. Make those changes--in other words, make insurance affordable--and most of the uninsured will leap at the chance to get coverage.
But not all of them. If you stop there, as Obama's does, some large portion of people--about a third of the uninsured, maybe more, if you believe the projections--will remain uninsured because they will still choose not to get it. Some of these people will make a decision to take their chances on a health catastrophe, figuring insurance isn't worth the price. Another group of people simply won't take the initiative on their own, because, quite simply, they won't give it that much thought. Some will be relatively well-off, some not. In all of these cases, however, the only way to make sure these people get insured is to compel them to do so--in other words, to make it a legal requirement. That's a relatively simple matter if you have a single-payer, government-run system; you just enroll everybody at birth. (That's one more reason why folks like me continue to talk up this option, even though none of the leading candidates have taken to it.) But if you want to provide universal coverage mainly through private insurance--which seems to be where we're heading at the moment--then you have to make people buy it.
Why interfere if people are choosing not to buy insurance and risk the consequences? For one thing, they may not actually grasp the consequences, which in the very worst cases can include bankruptcy or--in the case of those who have to forgo necessary medical care--death. If that sounds paternalistic, then consider that we use a very similar rationale for taking payroll taxes out of people's wages in order to finance Social Security. We know from the days before Social Security that, left completely to their own devices, large numbers of people will make poor financial planning decisions and end up destitute. So, in order to provide everybody with a guaranteed pension, we compel all Americans to pay into the government's retirement system. An individual mandate would rest on the same argument: In order to provide everybody with guaranteed medical coverage, we would compel everybody to pay for it upfront (although, again, we'd also offer financial assistance to those who need it).
It's true the uninsured already have some access to one sort of medical care: emergency care. But this is where those who choose not to get insurance are actually hurting other people, and not just themselves. When the uninsured get treatments through emergency rooms, it's often in the form of charity care--that is, treatment provided at a discount or for free. But when doctors and hospitals provide charity care, they inevitably pass along higher charges to their paying customers. This constitutes a “hidden tax,” as California Governor Arnold Schwarzenegger has called it, on every taxpayer and every private insurance holder. By forcing people to obtain insurance, we would require them simply to pay something for their care--perhaps not enough to offset the whole cost, but at least what they can afford. In other words, everybody would have to share in the responsibility for their own medical expenses.
In reality, people with more money--and in good health--will be doing more than merely financing their own medical care. They'll also be supporting the medical care of others. Indeed, that's the whole point of insurance: to have the healthy, who are relatively large in number, cover the high costs of treatment for the sick, who are relatively few in number. But there's a clear logic for this, too. Universal health insurance offers protection against misfortune, including not just accidents and unforeseen diseases, but also the misfortune of being born with the wrong genes--like the ones that put somebody at higher risk of Alzheimer's, breast cancer, or high cholesterol. Maybe you'll be fortunate never to have such a serious health problem. But don't for a second assume that--if not for some different circumstances, before or after birth--it couldn't have been you.
Still, the most important rationale for a mandate may be a more practical one: It's necessary to keep other reforms from unraveling. If you make insurers sell to everybody, even people with pre-existing conditions, but let people choose whether or not to buy it, people in good health will be more likely to wait until they're sick before buying coverage, figuring there's no point in forking over premiums while their chances of needing care are so low. This will cause a chain reaction. As healthy people opt out, insurance programs will be left dealing with a population of sicker and sicker people. Since insurance relies on contributions from healthy people to offset costs from sick, it will become more expensive--which will cause even more healthy people to opt out. The cycle will repeat over and over again, with the cost of insurance going up and enrollment going down. Wonks call this the “adverse selection death spiral.” And it's hardly theoretical. By the late 20th Century, most of the nation's Blue Cross plans had stopped offering insurance to all comers, regardless of pre-existing condition, because their competitors--who didn't make the same generous offer--had stolen away all the healthy patients.
It's not the theory of mandates so much as the practice of them that worries other critics. This is usually the place where car insurance enters the discussion. Most states require that drivers carry liability insurance. But, in some states, more than a third of all drivers don't have coverage. (If you've ever had an accident with an uninsured motorist, you've learned this the hard way.) If Americans aren't complying with requirements to buy car insurance, the argument goes, why would they comply requirements to buy health insurance?
One reason is that, in general, states haven't put the same effort into making car insurance affordable. Another reason is that car insurance take-up isn't quite as bad as you may have heard, at least not consistently. Georgia set up a system in 2001 to contact drivers right away when insurance lapses, and then impose penalties within short order. Two years later it was up to 98 percent compliance.
Still, automobile insurance is not health insurance. And, unfortunately, we have just one example of a health insurance mandate in the United States from which we can draw inferences: Massachusetts. Governor Mitt Romney signed a law creating an individual mandate in 2006, just before leaving office. Early on, Massachusetts decided to offer 20 percent of uninsured adults exemptions--because, officials said, paying for insurance would be too much of a hardship for them. According to Obama's advisors and other critics, this goes to show how meaningless a mandate really is: Government will never make people buy insurance they cannot afford. The real test of a plan, they argue, is whether it makes insurance more affordable--something Obama's advisers claim they do better than their rivals.
Put aside the fact that Obama's cost-control initiatives don't actually appear to be more comprehensive than his rivals. The real problem is that the lessons from Massachusetts remain ambiguous. The law hasn't even been in place a year--which means, among other things, that the penalties haven't kicked in. What's more, the state had to issue so many exemptions because, thanks in no small part to Romney's opposition to higher taxes, it lacked the money to provide enough subsidies. Both Clinton and Edwards have indicated they would do whatever it takes to make insurance affordable, even if that means raising new revenue.
That's not to say you can't learn anything from automobile insurance or Massachusetts--or, for that matter, from the two European countries with individual mandate schemes (Switzerland and the Netherlands), which both have better than 98 percent coverage. You just have to be careful about how you draw your conclusions. One person who has done so is Sherry Glied, a Columbia University economist. She has co-written what appears to be the most definitive paper on the subject, in a recent issue of the journal Health Affairs. Surveying the evidence, not just from the examples I just described but also other types of government mandates (like childhood vaccinations), she and her co-authors concluded “mandates can be an effective tool in expanding health insurance coverage, but the devil is in the details.”
Those details mainly come down to getting three things right: Making insurance affordable by restructuring the insurance market and offering subsidies (which, again, all three Democrats would do); creating an infrastructure for tracking exactly who has insurance; and imposing real penalties on those who do not comply. Interestingly, the penalties don't have to be onerous, according to Glied. They just have to be sufficient to make people think twice about turning down an opportunity to get coverage. (It's apparently more a matter of getting people to pay attention than changing the cost-benefit calculations.)
I contacted Glied and asked her directly how much difference a mandate would make. After cautioning that she hadn't made any formal projections, she speculated that a well-designed mandate might catch half to two-thirds of the people left out by a purely voluntary scheme. And with automatic enrollment into a plan, she added, the take-up could be higher. (All the candidates have indicated they support automatic enrollment; but without a mandate preventing people from opting out of coverage altogether, Glied said, it wouldn't have the same effect.) An even more optimistic assessment comes from the Lewin Group, a well-respected consulting firm that has its own model for estimating the impact of policy changes. In 2003, as part of a non-profit research project, Lewin made formal projections of ten different proposals for expanding coverage. Two of those had an individual mandate, eight did not. Lewin projected the two plans with mandates would reach nearly all of the 41 million people who were uninsured back then, leaving just 1.6 million without coverage--which, by many people's definition, would count as universal. (A few people will always slip through the cracks.) The plans without mandates did far worse.
Make no mistake, even the Lewin estimate involves some guesswork. According to Jonathan Gruber, a Massachusetts Institute of Technology economist, it's impossible to peg the likely impact of a mandate with even the very rough degree of precision he and other economists use for other estimates, simply because it requires making even more assumptions. But he also stressed that the question isn't whether a mandate will make a difference--it's how much of a difference and how close to full, 100 percent coverage an individual mandate system would get. Nor is Gruber alone in this assessment: It's pretty much conventional wisdom among policy makers. Just yesterday, the Wharton School's Mark Pauly--another highly regarded health economist, but one who tends to embrace more conservative views--co-signed a letter with Gruber and the New America Foundation's Len Nichols affirming that an individual mandate is essential for achieving universal coverage.
If the significance of including a mandate as part of health care reform really came down to nothing more than the numbers, it would not be so important. Obama's plan may not reach as many people as Clinton's or Edwards' would, but it would reach an awful lot of people--and, as such, do an awful lot of good. And, thankfully, Obama continues to say he wants to achieve universal coverage--that is, to make sure everybody has insurance. At least he has the right goal in mind. But achieving universal coverage is as much a political challenge as a policy one. And by talking down a mandate, he makes the political challenge that much harder. Among other things, the moderate Republicans and business groups willing to consider universal coverage will likely demand that a mandate be part of the package, because they see it as a way to reinforce personal responsibility. The insurance industry, meanwhile, will demand it because they think it essential to prevent an adverse selection death spiral. If the idea is to pass universal health care with bipartisan support, then a mandate may be essential.
What does this all mean in the end, in terms of judging Obama's candidacy? Good question. I hope to return to that soon.
JONATHAN COHN is a senior editor at the New Republic and a senior fellow at Demos. He is also the author of Sick: The untold story of America's health care crisis--and the people who pay the price.