Kevin Drum just wrote a post on the Bush administration's idea for health care reform: Health Savings Accounts. I wanted to expand on what he said, since HSAs are a Very Bad Idea, and it's worth knowing why.
The basic idea behind health savings accounts is simple. You get a health insurance policy that is, ideally, cheaper, but has a much higher deductible. In Kevin's example, the deductible is $2,000; he suggests that such a policy would be $2,000 cheaper, but that's wrong. CNN reports:
"According to the Kaiser Family Foundation, premiums for an employer-sponsored family plan averaged $9,068 in 2003, with workers kicking in $2,412. The premiums on a high-deductible plan will run you 20 to 40 percent less, estimates Herschman."
If you have such a policy, you or your employer can deposit money in a tax-sheltered account. The idea is that this account will cover some or all of the health care costs you run up before you hit the deductible. Since you will have to pay for the first few thousand dollars of your health care costs, the proponents of HSAs argue, you will be motivated to be a good consumer, and use no more health care than you need. In this way, health care spending will be driven down.
Sounds good, right? A nice, market-oriented solution to a serious problem. (And the idea that liberals are hostile to market-oriented solutions, in general, hasn't been true for at least fifteen years.) However, this is just one more illustration of Hilzoy's First Rule of Policy Wonkery: Never, ever rely on slogans instead of looking at the details*.
First, under the Bush administration proposals, employer contributions to HSAs are optional. That means that the account you supposedly get to use to pay your medical bills below the deductible might not exist. You could put the money you save into it, but as noted above, those savings will not cover the whole difference in the deductible. If they don't, you'll just pay more.
Second, HSAs will not, in fact, lower health care spending overall. The best study on the issue concludes that "health spending would change by +1% to -2%." One reason, as Kevin notes, is this:
"the vast bulk of healthcare dollars are spent on people who are extremely sick and quickly blow past even a large deductible anyway. Since HSAs don't affect that spending at all, it means that, at best, their effect on the total cost of healthcare is probably pretty negligible."
It's worth being clear about this. To that end, I have created my first ever chart using Excel (I am so proud), from data found here (see Exhibit 1.11):
[UPDATE: OK, I blew my chart. Darn. Along the y axis, it should say: percent of health care spending. The idea being that the 1% of the population who spend most on health care account for 22.3% of all health care spending, and so on.]
The top 50% of health care spenders will probably breeze right by their deductible, and will thus be completely impervious to HSAs' incentives to save money. Any health care savings will have to be gained from the bottom 50% (or from people who will be in the top 50% but don't yet realize it.) But the bottom 50% accounts for only 3.4% of health care spending. This means that this proposal leaves the overwhelming majority of health care spending absolutely untouched.
More problems below the fold.
Third problem: the health care market is just about the last place you'd expect an idea like this to produce better health care decisions. For one thing, consumers often have absolutely no idea of what they will pay for services, especially non-routine services. Hospitals, for instance, rarely tell you which services you can expect to receive while (say) having your appendix removed. And they almost never tell you how much you will pay for these services.
One reason for this is that how much you pay depends a lot on your insurer. Different insurers negotiate different pay scales with medical service providers; and so there is no handy price list for consumers to survey. (A persistent unfairness in the health care system: the insurers who cover large companies tend to get much better rates. This means not only that many small businesses pay more for health insurance, but also that their workers pay more for the very same health care if they haven't reached their deductible. Small businesses tend to employ less prosperous people. Of course, the uninsured haven't negotiated any reduction, so they pay sums undreamt of by people with insurance, for the same health care.)
Besides the lack of information on price, it's very hard to get good information on quality. From an article in TNR:
"Within the federal government, the Centers for Medicare and Medicaid Services recently began publishing information on hospital quality, in the hopes of steering beneficiaries to the best providers. But, while a few basic criteria are helpful in drawing narrow judgments--generally speaking, you want to get a surgical procedure at a hospital that does a large quantity of them--others can be grossly misleading. Does a high mortality rate suggest a hospital has lousy staff? Or does it mean the hospital simply takes on the most difficult cases?"
So we are hoping that giving consumers incentives to save money will produce savings even though those consumers will have to make decisions without good information on either price or quality. This is not, it seems to me, a recipe for success. And that's even without considering the fact that a non-trivial amount of health care decisions are made in emergencies, when detailed comparison shopping is really not an option.
Fourth, whether you're concerned with helping the sick or with saving money over the long term, you want to achieve savings by having consumers spend money more wisely, not by having them skimp on things like annual checkups, pap smears, and the like, which can catch diseases early, thereby avoiding both a lot of needless suffering and a lot of needless expense. But HSAs give people incentives to skimp across the board.
This is especially true when you consider that most people do not consciously decide to spend excessively on health care. They buy what they need and think they can afford; they don't just decide to have a colonoscopy for kicks. The rich will probably continue to spend money as before, even if they have to pay more of their health care costs themselves. The poor, however, will probably save money on checkups and preventive care, especially since, if their employers don't contribute to their HSAs, the money will come entirely out of their own pockets. Possibly some savings might come from the middle class, but it's not at all clear that those savings would offset the higher costs incurred as a result of poorer people skipping checkups and preventive care.
Finally, there's the truly ghastly problem of adverse selection: HSAs will draw young, healthy people out of traditional health insurance, since they are the ones who can best afford to take the risk. This will make the premiums for traditional health insurance skyrocket, and this, in turn, will mean that people who are already sick will find their costs driven up, even if they remain with their original plans. Last time I wrote about this, I tried to explain the phenomenon this way:
"Insurance is all about spreading risk among large numbers of people. Premiums basically reflect the amount of risk per capita: when everyone ends up needing big payouts, premiums are high; when most people don't, they are low. Bush's proposal creates enormous incentives for young, healthy people to opt out of normal medical insurance and into a system of catastrophic coverage plus health savings accounts. The poor, old, and sick, by contrast, wouldn't be able to afford to. But if all the young, healthy people opt out, then the premiums for those who remain -- the poor, old, and sick, mostly -- will skyrocket. At that point, the costs for those businesses who still offer normal health insurance also skyrocket, and those businesses will (quite understandably) try to move to a system in which they offer only catastrophic coverage, which is much cheaper. And what this means is: Bush's proposal has the potential to break our current system of employer-provided health care, leaving healthy and wealthy people, as well as corporations, better off, but at an enormous cost to the poor and the sick."
The upside, supposedly, is that this is part of the "ownership society", and will encourage personal responsibility. But I can't imagine why breaking the existing system of health insurance encourages our "ownership" of anything besides our medical bills. And if you own enough of them, you will stop owning a lot of other things -- your house, your car, your savings -- in very short order.
Moreover, insuring yourself against some hazard is a way of taking responsibility for dealing with it. And it's exactly the right way to deal with large and unforeseeable costs: the amount of health care costs you can insure yourself against is always larger than the amount you can prepare for by saving up money, and of course you are also less likely to end up dead or needlessly disabled. Why anyone thinks that in this case responsibility and insurance are at odds is a mystery to me.
So here's the deal: Health Savings Accounts will achieve minimal savings at best; they will introduce more market incentives into just about the last place you'd expect those incentives to work well; and they will break the existing health insurance system, leaving the old, the poor, and the sick to pay more, and the young, rich, and healthy to pay less.
Advocating Health Savings Accounts is a triumph of ideology over serious thought. They are deeply unfair, and we should oppose them.
*Hilzoy's Second Rule: there are some problems to which game theory cannot usefully be applied. There are, of course, more.