Arnold Kling  

Is Private Health Insurance Impossible?

PRINT
Buying Habits, Race, and Incom... Tranquility for A Dollar a Day...

Mark Thoma channels Paul Krugman.


It's nice to have Paul Krugman discuss a question that has been addressed repeatedly at this site, market failure in the provision of health and social insurance due to moral hazard and adverse selection

This moral-hazard-and-adverse selection story is one of those ideas that strikes economists as so clever that it simply must be true. Tim Harford tells the same tale in his Undercover Economist book, which is going to make it difficult for me to give a favorable review.

Mark Pauly and Bradley Herring have dared to do somethin that Krugman, Harford, Thoma, and similar "experts" have not done. Pauly and Herring examined the way insurance companies actually behave! It turns out that they don't adversely select. Individual health insurance is expensive because of poor scale economies--there are very few consumers in the individual market--and not because of the factors near and dear to the theoreticians.


Rather than focus on fears of risk-segmentation, the authors suggest that political energy should be directed towards the non-group market's major problems: high loading costs, lack of persistence in purchasing, and mistaken choices.

There is no direct evidence that private health insurance markets break down. Instead, the empirical argument made against private health insurance boils down to the usual litany that the U.S. spends more on health care than other countries, yet our longevity is worse than that in many countries that spend less. That is an interesting fact, but it is just as true of the over-65 population in the U.S., which has the national health insurance that Krugman thinks all of us need, as it is for the under-65 population, which Krugman sees as the victim of the cruelty of private health insurance.

On Monday, Wednesday, and Friday, when Krugman talks about health care, he tells us that Medicare works great. On Tuesday and Thursday, when he argues that Social Security's shortfall is relatively small, he tells us that Medicare is the real fiscal time bomb. His Tuesday-Thursday outlook on Medicare is more realistic.


Comments and Sharing





COMMENTS (28 to date)
Lab_Frog writes:

Frog thinks a better solution is to guarantee every citizen small amount of money for health-care per year that is not savable (essentially a health-care voucher designed to encourage preventive healthcare) and a total amount per citizen (some amount in some year dollars, for example $ 30,000 in birth year dollars) in government funding for general healthcare (generally consumed towards the end of life, but can also be used to pay for mandatory private health insurance during times of financial hardship). This would encourage preventive healthcare, discourage over-consumption as ultimate funds are limited, and competition between insurance firms would limit “predatory pricing.” However the administrative costs may outweigh the benefits of such a program, especially if the government makes the program too complex or does not communicate how much funding remains and what the intent of the funding is.
From Lab Frog’s Blog: “Frog thinks that a good way to encourage more preventive healthcare would be to reallocate some government hospital funding to health vouchers, say $300 per person bi-annually, which if not used, do not carry over to the next year. Preventive healthcare would than be “free” and the number of people who got their teeth cleaned twice a year and got an annual physical would go up, thus in the long-run decreasing demand for corrective healthcare, which is more expensive.”

Is there a model with solid emperical evidence to explain why adverse selection and moral hazard does not occur in private health insurance?

Alberich der Zwerg writes:

Arnold,

When is your book on healthcare going to be released?

AdZ

Joe Stafura writes:

In addition to mediocre performance and the false free market nature of health care the load on small businesses is a true drag on entreprenuership.

Most of the workers are young when you start a new tech business but in a short period entering the 30's and having a family raises the premiums to the point where you are reducing the quality of the benefits.

The sad fact is that we ( other than health care ) businesses are supporting large pharmas and insurances companies that make large profits on a few people along with a government system that treats the rest.

Free market models in health care and education are non-starters but with everyone in Washington so fond of the campaign contributions there will be a market for books defending this foolish concept.

David J. Balan writes:

What is the basis for the assertion that "it is just as true of the over-65 population in the U.S...as it is for the under-65 population?"

L writes:

Krugman does an interesting slight of hand, promising at the beginning of the article to talk about markets in health care, and concluding at the end that markets in health insurance are not viable.

Most countries subsidize medicine so heavily that the difference between government insurance and subsidized medicine probably isn't important, but it's a pity that no one talks about markets in heatlh care.

Robb Lutton writes:

Mr. Kling,
You cite this study approvingly, but doesn't it raise a problem for you? You say that, instead of theorizing, they study what really happens. Isn't that a big slap in the face for a self described neoclassical economics professor? Isn't that what you are all about? That is to say...theorizing and stuff?

dsquared writes:

Two points:

That is an interesting fact, but it is just as true of the over-65 population in the U.S., which has the national health insurance that Krugman thinks all of us need, as it is for the under-65 population, which Krugman sees as the victim of the cruelty of private health insurance.

This would only be a knock-down argument if a person's standard of health after age 65 were independent of the healthcare that person had received before age 65, which is clearly false.

and second, the study linked makes a fundamental misunderstanding about adverse selection in insurance markets. They are analysing the degree of price discrimination in American health insurance markets. But the classical problem of "adverse selection", in the sense in which Krugman uses the term, is one of private information on the part of the buyer. If the insurer is able to make use of the information in order to discriminate by price, it clearly isn't private to the buyer and there can be no question of adverse selection.

What you would expect to see in an insurance market where adverse selection is a big problem would be precisely what you do see in American health insurance markets; big risk pools, high deductibles and co-pays, and lots of people who are buying less insurance than they would otherwise wish to, as part of the cost of raising the deductible high enough to induce the (unknown to the insurer) bad risks to self-select out of the insurance market on the basis of their private education. It seems to me that Krugman has this one about right, or at least this study doesn't prove he doesn't.

Bill Stepp writes:

The reason there are few consumers in the individual market is that government restrictions on the provision of health insurance have driven carriers out of the market. Guardian, the last of the old indemnity insurers, exited the health insurance market altogether in the early 1990s.
As for moral hazard, it exists because government subsidizes risk-taking activities.
Krugman really doesn't have a clue about medical insurance, or much else for that matter.

Tim Harford writes:

Alas! Well, I hope you enjoy the other nine chapters... Look forward to reading and learning from your book.

Dan writes:

"People holding individual insurance whose anticipated medical expenses are twice the average pay premiums only about 20 to 40 percent higher than other individual insurance customers."

Hmmm. I'd like to see the regressions for that data, but that sounds like a strongly positive correlation to me.

Let's say an average insurance plan is $12,000 per annum. 20-40% more is $1200-2400 -- and this is not a serious problem for workers with jobs so poor that they are not offered health insurance, how?

I am not an idealogue on this issue, but the Pauly-Herring article does little to erode the case for single payer care. Even the excerpt you quoted supports it:
"Rather than focus on fears of risk-segmentation, the authors suggest that political energy should be directed towards the non-group market's major problems: high loading costs, lack of persistence in purchasing, and mistaken choices. "

Government mandated care solves both persistence in purchasing and "mistaken choices" (maybe these should be referred to as ineffecient choices?). And we all know that costs would be reduced if this group stopped taxing our emergency rooms for non-emergent care.

Randy writes:

Want to reduce the cost of health care? Then allow competition. Allow the practice of medicine without a license. Remove restrictions on drug production and distribution.

Every other solution I have ever read will eventually fail, because health care is a monopoly. Throwing money at it, from any direction, will simply drive up the price.

What we have now is people who cannot afford health care simply self medicating. I expect to see an underground market develop soon. Not that it would be a bad thing.

Perhaps you think that only certified specialists can safely practice medicine. Sure, and only a certified mechanic can change my oil.

Lord writes:

Generally insurance company prices are determined by the class of the policy holder, and the number of these classes has mushroomed significantly to increase price discrimination. If significant costs are incurred, premiums rise sharply, which diminishes the utility of insurance. That insurance companies flee risk is without doubt. Look at all the low probability, low risk insurance offered, e.g. accident insurance, and all the high probability, high risk insurance that is avoided, flood, etc.

Eye Doc writes:


One could postulate that the longevity of our elderly population has been adversely affected by what occurred when they were younger and had their health care delivered by a non-socialized system. I think that's a very reasonable arguemnt, but it does not in anyway mean that Krugman is correct. I don't think he is. I think he's completely clueless.

Looking at what a disaster socialized health care has become in Canada, the UK etc. you would think would turn off most thinking beings to adopting that sort of a system, but I guess that pretty much leaves Krugman out.

If you want lower cost health care you need to lower malpractice costs and the practice of defensive medicine by adopting medical malpractice reforms such as capping non-economic damages, and you need to privatize Medicare. Most of the cost of Medicare is bureacracy. All you need to do to lower Medicare costs is to give vouchers to Medicare patients so that they can buy their own health insurance, and get the federal government out of the health care business. Same goes for Medicaid as well.

Captain Video writes:

I have a number of chronic health problems and have been trying to obtain long-term health insurance. Don't try to tell me that insurance companies do not adversely select. If an item of research shows that they do not adversely select, my suspicion is that they must be doing something wrong.
In the physical sciences empirical results are not accepted as demonstated until they have been independently reproduced by other researchers acting independently. Has the research you are quoting been independently repeated by other researchers, or is it what physical scientists would call a "nonrepeatable result?"

Captain Video writes:

"Looking at what a disaster socialized health care has become in Canada, the UK "

Even ideologue like Margaret Thatcher did not dare abolish the National Health Insurance system in Britain because that would have been a political disaster for her. If these programs are such disasters, why do not political parties in those countries try to win elections by promosing to abolish them?

Captain Video writes:

"Want to reduce the cost of health care? Then allow competition. Allow the practice of medicine without a license. Remove restrictions on drug production and distribution."

"Everything is most efficient in this most efficient of all possible worlds!" (as long as the government does not interfere) Thank you Dr. Pangloss!

Captain Video writes:

"obtain long-term health insurance"

OOps, I meant long-term care insurance.

if a person's standard of health after age 65 were independent of the healthcare that person had received before age 65, which is clearly false.

Highly doubtful. After age 65 health is far more likely to be the result of things like previous diet, alcohol and tobacco use, and heredity than the health care one received at earlier ages.

Not only is it likely that the outcomes of over-65s will be strongly affected by the care they've received earlier in their lives, the fact remains that the health care delivery system that they use is largely shaped by the private insurance that the rest of the population uses.

A better comparison than Medicare would be to compare longevity etc in the VA hospital system, where the system itself is relatively independent of private insurance.

Randy writes:

And thank you Captain Voltaire :)

But seriously, how will you provide universal health coverage without driving up demand, and therefore price? How will you stop medical technology from advancing, and thus raising the price?

Price increases will defeat your best plans, and so healthcare will always exist in tiers - a system for those who can afford the best, and a system for those who can only afford the basics.

So create a system that provides the basics. Better yet, just let them, and people will create it for you.

Steve writes:

Arnold,
Although I find Pauly's and Herring's findings intriguing, the assertion that adverse selection does not occur in the health care market seems like wishful thinking? Your explanation of "poor scale economies" does not seem to adaquately explain the wide delta between premiums payed by employers with large shared risk pools and small business and individuals. I am not a fan of Krugman or his solutions. But adverse selection is certainly rational. I have always assumed it is emperically verifiable. Maybe not. But whatever the cause, the fact that those who need health insurance the most have the toughest time getting is a problem.
Steve

Jim Glass writes:
dsquared wrote:

... the study linked makes a fundamental misunderstanding about adverse selection in insurance markets. They are analysing the degree of price discrimination in American health insurance markets. But the classical problem of "adverse selection", in the sense in which Krugman uses the term, is one of private information on the part of the buyer.

If the insurer is able to make use of the information in order to discriminate by price, it clearly isn't private to the buyer and there can be no question of adverse selection.

Hmmm...
Krugman wrote:

...good insurance is hard to come by, because private markets for health insurance suffer from a severe case of the economic problem known as "adverse selection"...

That's why insurance companies don't offer a standard health insurance policy, available to anyone willing to buy it. Instead, they devote a lot of effort and money to screening applicants, selling insurance only to those considered unlikely to have high costs, while rejecting those with pre-existing conditions or other indicators of high future expenses.

I think dsquared is correct -- Krugman self-refutes (yet again)

Gary writes:

I think the free market economist are forgetting the progress of genetic testing., Even if private health insurance is possible now without severe government regulation it sure will be impossible once genetic testing becomes even more sophisticated. After all, it would be absurd for a private insurance company not to test for ones potential chronic diseases, make an actuarial calculation as to how much those diseases are likely to cost and charge you accordingly. The process would be fair--people would pay what they are likely to cost the company--but the process, would I think, be unconscionable. Ergo, even if Krugman isn't right yet, he will be right very shortly. Once an insurance company has sufficient information to make the calculation, they will obviously do so--and at that point private health insurance becomes impossibly expensive for those who lost the gentic lottery absent state regulation to require say age based policies with no preconditions--something I am not averse to but one that an insurance company would not be happy about precisely because of the adverse selection problem. You either must require everyone who can to buy health insurance which has no precodition testing and then the state must also pick up the tab for the rest of the population or the system must inevitably fall apart.

George Weinberg writes:

I think the problem is one of the hidden premise. Krugman obviously believes that everyone is entitled to cheap, high-quality medical care. The essence of the "adverse selection" argument is that those who are currently healthy and unlikely to beome sick will not willingly subsidize those already sick or who know they are more likely to become sick in the near future. This isn't really an economic argument, and of course will be unpersausive to those who believe there is no such right.

The "moral hazard" argument, which is economic, is of course an argument against any mandatory natiobal health insurance scheme: if everyone is forced into one big insurance pool, this will tend to encourage reckless behavior, since people will be (at least to some extent) shiting the cost of their recklessness onto others.

Tino writes:

"You say that, instead of theorizing, they study what really happens. Isn't that a big slap in the face for a self described neoclassical economics professor? Isn't that what you are all about? That is to say...theorizing and stuff?"

A tragic, complete but common missunderstanding of neoclassical economics.

The methodological premiss of neoclassical economics, in it's simple form, is simplified ("unrealistic") assumption that explain empirical phenomenon.


Read Milton Friedmans essay on possitive economics, but more importantly read his book on the consumption function and their monitary history of the united states.

aa2 writes:

When the doctors are in control of how many doctors are allowed to practice they will always choose less then the market demand. That is simple human nature that any philosophy student should understand.

So they have restricted the number of practioners to less then demand, as every profession in the history of the world that has the power to control its own numbers.

Thus there is less practioners then demand. In a free market only those who can pay the most will get care. And in a socialist system like in Canada, where doctors are even more aggressive in limiting supply, healthcare will be rationed.

Same situation with housing. 86% of America is owned by the government and off limits. On the rest of the land especially in the high density areas there is an anti-development attitude. So most projects like big apartment buildings aren't allowed. Yet at the same time there is natural population growth and immigration.

So housing is a simple bidding war where the poorest group gets left out. And everyone else pays a ludicrous amount of their income for property.

In all cases simply get rid of the restrictions and the problem disappears.

rvman writes:

Since when are 'adverse selection' and 'moral hazard' involving the INSURERS the issue in health insurance markets? The adverse selection involved is among the buyers, not the sellers. Low risk people won't buy a 'standard' offer, creating a higher risk pool, and thus raising prices, thus driving out moderate risks, raising the risk of the pool, raising prices, resulting in higher prices and lower coverage than is efficient. Only high risk and highly risk-averse low-risk people will buy. The economically efficient solution would be for insurers to do exactly what that study says doesn't happen - segmenting of the market by risk level. That there are not multiple offers from the same company based on perceived health screams "not everyone is getting reasonable offers".

As for moral hazard, that is people taking risks because they have insurance. Not studied here. I doubt that it is a big factor, relative to adverse selection.

Your thesis about using using Medicare population as a control to study the health effects of universal health care breaks down when one looks at longitudinal health studies. It turns out that the effects of poor health care imputs early in life frequently do not express until late in life. Thus the spotty health care imputs endemic in the United States from prenatal and continuing into childhood and adulthood can, and frequently does, express later in life. Thus is the unquantified poorer health and longevity you point to in the aged is likely the result of deprivations of adequate health care earlier in life. Poorer outcomes among that population "that has socialized medicine" is thus not caused by socialized medicine in old age as you imply, but rather the lack of it earlier in life.

Comments for this entry have been closed
Return to top