The Atlantic Home

Multiple readers want to know why US health care costs so much.  A complicated and fascinating question on which agreement across ideological lines is extremely limited.


Proponents of national healthcare prefer to blame the administrative costs of private health insurance, the bad incentives of private health insurers which cut corners on preventative care, greedy providers (drug firms especially), and excess reliance on emergency rooms.

Perhaps predictibly, I find this unconvincing.  Since private insurance covers about 36% of health care expenditures, it's hard to imagine how the savings could possibly be that big.  And indeed, it seems that the net cost of private insurance accounts for about 7.3% of healthcare spending, which itself is about 16% of GDP--about 3-4 percentage points higher than our nearest competitors.  So we might save a little over 1% of GDP if we got rid of those parasites--but it wouldn't make up the difference between us and high-rent systems like France and Switzerland, much less lower cost providers like Britain.

And as Arnold Kling points out, it's not that easy to get rid of the private insurers:

Suppose we were to abolish private health insurance tomorrow and put everybody on Medicare. Here are the things that would happen.

1. You could get rid of everybody in the private health insurance industry. That would save at most 15 percent of health care spending.

2. There would no longer be a private-sector benchmark for Medicare to use in pricing. You would need to hire a lot of those former private health insurance folks to work for Medicare to figure out what every medical service is worth, to negotiate prices with doctors, hospitals, and so on.

3. Medicare would have to process more claims, which means you would have to hire back some more of those former insurance workers.

After (1)-(3), costs might be less than the existing system. By a small amount.

Likewise, drug industry profits were a tiny portion of healthcare spending.  If we eliminated all profits--hell, double it to account for the much maligned "marketing costs"--we'd possibly push down expenses by another 0.5% of GDP.  But as with insurance administrative costs, it's more complicated than that.  Without profits, no one would do R&D--and on net, drugs save us money, because they often replace expensive procedures.  Years of statins are still cheaper than one hospital stay for a heart attack or stroke.  And about half the marketing cost is free samples, which most people would agree are a good and useful device for letting patients see whether a drug works for them.

As for emergency rooms, there's not all that much evidence that lack of insurance is the primary issue there.  First of all, it seems to be a worldwide problem.  Second of all, a recent report from the state of Massachussetts indicates that insurance is not the driving factor:

First of all, inappropriate -- or non-urgent -- use of the Emergency Room was not limited to uninsured populations. It showed up across the board. People covered by private insurance, Medicaid and Medicare were just as likely to use the ER for non-urgent care as people without health insurance. About 20% of all ER visits by privately insured and Medicare patients were for non-urgent purposes. About 24% of all ER visits by Medicaid beneficiaries and people without any insurance were for non-urgent purposes.

Reading between the lines, poor people are more likely to use the emergency room for non-urgent care than those in upper income quintiles; presumably, this has more to do with their chaotic lives, lack of control over their work schedules, or planning skills than the payment problem, since Medicaid patients show the same usage patterns as the uninsured.

Finally, preventive care.  It is true that if people with certain chronic diseases do everything they're supposed to--not just visit the doctor but comply with their treatment regimens--it probably saves money in the long run.  But this is a complicated problem.  Compliance with treatment regimens often fails not because the patient doesn't go to the doctor, but because the patient doesn't check their blood sugar multiple times a day, take their inhaled steroids and pills at the scheduled intervals, cut salt and fat out of their diet, or what have you.  Also, people who don't die of diabetes now may just die of diabetes a little later, or of something else massively expensive, like cancer.  Preventive care may be an excellent way of enhancing net social welfare.  But the evidence that it saves money on the whole is not really there.

So why does American health care cost so much?  Conservatives like to blame out-of-control spending.  And they have a point; the healthcare benefits of extra spending don't show up in the best controlled studies, like the Rand experiment that randomly assigned people to either a sort of HMO or a fee-for-service plan.  Spending on retirees varies wildly by Medicare region, but health doesn't.  And so on.

But again, this cannot explain all of it.  Healthcare costs are exploding around the world; in most places, healthcare is a vastly more important ministry than defense, and the favorite political activity is arguing about how to tweak the medical system.  So why?

To my mind, the real answer is threefold:

1)  We pay more for our medical services.  But though the pharma industry is important, the real action is in wages.  Our medical personnel cost vastly more than their counterparts abroad in almost every category.

2)  We consume more services.  Americans get shiny new facilities--my British colleagues once derisively commented that American hospitals are "like hotels".  American hospitals don't have open wards for almost anyone.  They staff at very high levels.  Doctors conduct an inordinate amount of tests.  We use an expensive machine rather than watchful waiting.  And often, those expensive machines catch conditions that never would have turned into anything, which we then treat.  Natasha Richardson probably would have lived if she'd had an accident here, because doctors would have done a cat scan, and there would have been a Medevac helicopter available.  That's tens, maybe hundreds of thousands of dollars to save a single life.

3)  There are inefficiencies.  I don't mean "compared to other systems"--every system has some screwed-up illogicality that costs it money and makes patients worse off.  But compared to what we could have.  For example, Medicare pays for procedures, not wellness, which means that there's a chronic undersupply of geriatricians, because the specialty isn't particularly well paid even though the nation's largest healthcare provider is specifically designed for old people.  This is madness.  But every real-world system that has attempted to pay physicians for wellness has ended up giving up in disgust.

So how much scope is there for reducing our costs, relative to the rest of the world?  Some, obviously, though it's not clear that this would actually be a net benefit to either us and the world, since the iatrogenic effect would probably be to wipe out most industry research into new drugs and medical devices.  But not really that much, for both political and practical reasons. 

Politically, state health care systems have so far proven unable to control labor costs--indeed, the health care unions are some of the most powerful political forces in most states, while the AMA has dominated the Medicare reimbursement schedule.  There's no evidence that is going to change any time soon.  Politically, also, conservatives have got to face the fact that we are not going to stop providing health care to people who are in dire need, and that this will undercut any attempt to move towards a fully private model. 

Practically, we have to pay healthcare workers a lot because we have to pay everyone a lot; in a rich country, wages for healthcare workers are high.  And measuring healthcare productivity is really insanely difficult, which makes it very hard to figure out what's worth spending money on. As long as Americans don't want to sacrifice access to procedures--and they don't--there's just not much room for decreasing costs.

That doesn't bother me that much.  The mindless trend extrapolation about how much we'll spend on health care in the future elides the point that we'll be much richer in the future; why shouldn't we spend all that extra income on healthcare?  Your ancestors spent 2/3 of their daily income on food.  Now you spend about 15-20%.  But spending much more of your income on clothes and housing doesn't mean that you're starving; it means that you're so rich, you only spend a small fraction of your income on food.  When I look around at our incredibly bountiful economy, I don't see any obvious lack that we're creating by spending ever more of our income on leading longer, healthier lives.