David R. Henderson  

Reply to Professor Erik Brynjolfsson

Macroeconomic Theory, Evidence... Friday Night Video: CBS Recogn...

In the comments section of my post yesterday on health insurance, "The Pre-Existing Elephant in the Room," Erik Brynjolfsson, the Schussel Family Professor at the MIT Sloan School of Management, Director of the MIT Center for Digital Business, Chair of the MIT Sloan Management Review, and the Editor of the Information Systems Network, writes:

You say you think people should be responsible for the "costs they create" when they have pre-existing conditions, but aren't most of those due to bad luck?

If so, would you favor a "perfect" market where everyone paid in advance 100% of their expected future healthcare costs? Wouldn't that pretty much miss the whole point of insurance?

On his first question, which is a good one, I don't know the answer. I'm sure some of it is bad luck. A key question is "When did the bad luck happen?" If it happened when the person was in his 20s or later, that's the great advantage of buying health insurance early with a guaranteed renewable clause, a clause that existed in many individual health insurance contracts.

But take the extreme where it was all due to bad luck at an early age. To justify making others pay, you would have to argue that the bad luck should be forcibly put on others who themselves had no role in creating the bad luck. Here's where I balk.

His second question, though, surprised me. The whole point of insurance without a loading fee (which economists sometimes call "actuarially fair insurance") is to have the person pay 100% of his expected future medical care costs. Notice the word "expected." I'm assuming Professor Brynjolfsson is using that word the way we economists use it: the probability of the expense times the amount of the expense. Of course, insurance will always be priced higher than that because of the loading fee, the amount the insurance company charges for running the business--expenses for advertising, offices, claims adjusters, etc.

So, contrary to Professor Brynjolfsson, this doesn't at all miss the whole point of insurance. He seems to be thinking about insurance as a program for the low-risk people to subsidize the high-risk people. But here's what his MIT colleague, economist Richard Zeckhauser, wrote about that in "Insurance," in The Concise Encyclopedia of Economics:

An economist views insurance as being like most other commodities. It obeys the laws of supply and demand, for example. However, it is unlike many other commodities in one important respect: the cost of providing insurance depends on the identity of the purchaser. A year of health insurance for an eighty-year-old costs more to provide than one for a fifty-year-old. It costs more to provide auto insurance to teenagers than to middle-aged people. If a company mistakenly sells health policies to old folks at a price appropriate for young folks, it will assuredly lose money, just as a restaurant will lose if it sells twenty-dollar steak dinners for ten dollars. The restaurant would lure lots of steak eaters. So, too, would the insurance company attract large numbers of older clients. Because of the differential cost of providing coverage, and because customers search for their lowest price, insurance companies go to great pains to set different premiums for different groups, depending on the risks each will impose.

Zeckhauser goes on to point out that the problem of adverse selection arises precisely when insurance companies cannot distinguish the high-risk people from the low-risk people, when the high-risk and low-risk people know what risk category they are in.

COMMENTS (21 to date)
Yancey Ward writes:

Good luck educating people about what insurance actually is. I have long since given up.

Brian writes:

"If so, would you favor a "perfect" market where everyone paid in advance 100% of their expected future healthcare costs? Wouldn't that pretty much miss the whole point of insurance?"

I think maybe someone forgot that the point of insurance isn't for expected costs, but for unexpected costs. I don't expect my car insurance to pay for gas and oil changes. I don't expect my home owners insurance to pay for lawn care or light bulbs. Nor should I expect health insurance to pay for regular check-ups and recurring prescriptions. Unfortunately, most people treat health insurance today like a pre-payment plan rather than something you pay into to help cover outside-of-the-norm, unexpected, and/or catastrophic scenarios. But if that's what someone's talking about, they should at least be aware that what they're describing is not insurance.

Eric Falkenstein writes:

Health care policy is so complicated in large part because it's a Trojan Horse for redistribution. More people find a right to 'health care' makes sense, as opposed to some lump sum of cash, even though all good economists know the latter is the more efficient way of redistributing wealth. The same could be said for education.

Greg Mankiw mentioned a while back, no one would care about health care policy debates if they were all neutral in terms of the effects by income.

John Goodman writes:

The main reason we have problems with pre-existing conditions is that our health insurance system is employer-based and we have to change insurers every time we switch jobs.

John Cochrane has proposed a solution: I call it "change of health status insurance." If my health condition worsens while I am insured with Aetna and then I move to Cigna, Aetna must pay any additional premium needed because of the health condition change.

See: http://healthblog.ncpa.org/rational-health-insurance/

Greg G writes:

Do you favor having government provide police protection, a court system and a common military defense? If so, you will be causing the cost of these things to be "forcibly put" on some people who might feel they don't need or want them and might prefer not to be forced to pay for them.

Do you think that a hospital should be required to treat a child who is bleeding to death after an accident even if the family can't afford to pay? If so, you will "forcibly put" that cost on someone else.

Not everyone assumes that every aspect of the insurance model is ideal for healthcare. Requiring people to participate in paying for some kind of health insurance (like virtually every other modern industrialized country) is an attempt to make health care more insurance-like by reducing the free-rider problem.

Thanks for responding so quickly to my comment. Since you quote Prof. Zeckhauser, let me quote him further in my reply (from one of the his own articles he cites at the end of the article you point to):

"choice in health insurance is a mixed blessing...[There are] losses from having premiums vary with observable health status. Because insurance is contracted for annually, people are denied a valuable form of intertemporal insurance -- the right to buy health coverage at average rates in the future should they get sick today. As the ability to predict future health status increases, the lack of intertemporal insurance will become more problematic."(see http://www.nber.org/papers/w7176) [emphasis added]

In other words, choice can sometimes destroy insurance. As Ken Arrow pointed out over 40 years ago, insurance markets can be tricky and even strong advocates of markets like Ken Arrow understand that such markets can breakdown without regulation.

I agree with you that the best solution is for everyone to buy insurance early, i.e. at birth or before birth (at least before insurance companies can predict the likelihood that you will face high expenses), with a guaranteed renewal clause. Otherwise, as insurance companies get better and better at predicting future health costs of each individual, we will have less and less ability to insure ourselves against bad luck. I think you would agree that insurance against bad luck is very beneficial and eliminating such insurance is harmful. Right?

Some people can't or don't buy insurance at or before birth, so we can a) let them go untreated and perhaps die if they are unlucky enough to have treatable but expensive health care needs, b) treat them via emergency rooms and other care paid for by others, or c) require everyone to buy basic insurance and pay for it if they can at standardized rates where people who have good luck partly insure people who have bad luck.

Most Americans aren't comfortable with option a). Mitt Romney once argued for option c) was the "libertarian solution".

Which do you favor?

Meanwhile, I look forward to your answers to these two questions on my blog:

When the value of information is negative

zdc writes:

"You say you think people should be responsible for the "costs they create" when they have pre-existing conditions, but aren't most of those due to bad luck?"

Absolutely not. A huge portion, if not the majority of hundreds of billions of dollars in annual health care expenditures are for treatment of conditions brought about by poor life-style decisions, willful self-hard (both immediate and long-term), and wanton disregard for personal health and well-being.

Countless governmental and non-governmental reports highlight these astounding costs (and are readily found with the most rudimentary internet searching such that I won't link to them).

For example, let's take a look at the Top 10 causes of death in the US (per the CDC from 2010 data, I use deaths as a data-point because they're a good proxy of diseases, treatment of which is a primary driver of medical spending):

- Heart disease. Sure, there are genetic factors at play, but smoking, poor dietary choices, obesity, and lack of physical activity (all preventable causes of disease) are the major contributory risk factors.

- Cancer. Certainly there are plenty of 'bad luck' diagnoses here, which is probably what most people think of in regards to 'bad luck medical costs'. But many cancer deaths are the sequela of, again, preventable causes of disease. Smoking is a huge contributor to not only lung cancer, but several other cancers. Skin cancer is frequently related to worshiping the sun unprotected for years.

- Chronic lower respiratory disease. This can be multifactorial, but the lions share are the sequela of smoking.

-Stroke. Certainly a 'bad luck' component, but many cases are the sequela of preventable lifestyle choices related to the same factors as those described for heart disease.

- Accidents/trauma. Sure, there are some unfortunate 'bad luck' instances here, but spend any time in an ER, and you'll quickly find that alcohol and drug use are a huge driver of 'accident' and trauma costs.

- Alzheimer's. This one is almost invariably bad luck.

- Diabetes. While type 1 (presumed viral wipeout of pancreas) is bad luck, most are type 2, and most have type 2 from being obese (some don't).

These are just a few examples. Anyone with half a clue with regards to the delivery of medical care in this country could cite numerous more.

Professor Henderson is spot on in his assessment that by preventing insurance companies from being able to price for risk highly associated with certain pre-existing conditions, that insurance market has been hamstrung.

The fact that such an intelligent and highly accomplished economist as Dr. Brynjolfsson is ignorant of the fact that hundreds of billions of dollars in annual medical expenditures are the direct result of personal choices highlights the challenges outside 'experts' with little to no understanding of the actual delivery of healthcare face when expressing their opinions on the topic (and the unintended consequences that arise in situations when they are allowed to shape policy).

Nicholas Weininger writes:

Greg G, I don't know how Prof. Henderson would respond, but to me you're making a good general case for anarchism as an ethical ideal. As you say, because people have very different values, it is always morally problematic to force others to pay for something they don't value as much as you do, even if you consider that thing essential-- whether it be health care or defense.

Now, anarchism has its set of well-rehearsed practical problems. Nonetheless, those of us who find its ethical vision compelling may reasonably seek to minimize the extent to which people are forced to pay for the realization of others' values. I would hope that a hospital facing a child bleeding to death would care for that child regardless of the family's ability to pay. I'd be more than happy to chip in to pay for children in such situations-- and indeed I don't personally object to the (relatively small) portion of my taxes that goes to pay for such things. But I can't in good conscience advocate that others be forced to pay too if there is any practical alternative; their values are not mine to judge and their lives and labors are not mine to use for my ends.

Jay writes:


David wasn't using "Expected costs" like you think he was, please read it again.

David R. Henderson writes:

What Jay said.

Eric Falkenstein writes:

@Byrnjofsynbergerson: "choice can sometimes destroy insurance." It's funny that people very skilled at modeling can find special cases where adverse selection causes markets to break down, but then assumes that putting everyone in a single coercive pool is the efficient solution. The latter has a much higher adverse selection problem and eliminating markets leads to stagnation and queueing. Presumably the government will provide the service just as a private company would. This was the belief of socialists pre-WW2 about productivity, and China and the Eastern Bloc highlighted that it doesn't work that way, you need competition, markets, for productivity growth.

As per buying insurance at birth, that's not renegotiation-proof. At 20 or 30, if I can show I'm of lower risk I should be able to bargain a discount if there are free markets for insurance. In the long run, the birth-insurance-rates would anticipate this the way 30-year mortgages anticipate the pre-payment option for borrowers, that the healthy would leave, making initial insurance the most costly insurance one could buy.

pyroseed13 writes:

David, in a way I think you are sidestepping Erik's question. I agree with you about the purpose insurance should actually serve, but do you propose that we simply do nothing to help those with prexisting conditions? Is there any middle ground here, or does this simply come down to the usual the libertarian complaints about "force?"

Brian writes:

@Jay and David,

I know, I was quoting Brynjolfsson, not Henderson.

David R. Henderson writes:

David, in a way I think you are sidestepping Erik's question.
Actually, in his comment on yesterday's post, Erik asked two questions and I answered them both. I did not sidestep. I also have not answered the questions he asked on his blog. I will get to that but not today. I have a job and deadlines.
I agree with you about the purpose insurance should actually serve, but do you propose that we simply do nothing to help those with prexisting conditions?
Is there any middle ground here, or does this simply come down to the usual the libertarian complaints about "force?"
I'm not sure what the middle is supposed to be between or why you put "force" in quotation marks. I do not want a solution that coerces innocent people. But there are certainly voluntary solutions for people with pre-existing conditions and one obvious one is for you and me to help them. Do you object to helping them?

pyroseed13 writes:

David, thanks for the quick response. Sorry if my tone was somewhat snarky. My worry is that I don't sense that libertarians and conservatives have assuaged liberal's concerns about what would happen if we didn't force insurance companies to cover pre-existing conditions. I agree with much of what you have said about insurance but I think there are alternatives that have been overlooked (see here: http://www.aei.org/article/health/healthcare-reform/covering-pre-existing-conditions-in-a-market-driven-health-system/).

Greg G writes:

@ Nicholas

I was not making a case for anarchism. I was making a case that David's logic here will lead to either anarchism or special pleading.

MingoV writes:

I agree with Dr. Brynjolfsson. Insurance is not about perfectly adjusting premiums to individual risks. Insurance is about a group of people pooling and investing their money so that a group member (policy holder) will not suffer a huge financial loss after a (covered) catastrophic event. Insurers fail in their duty if they categorize, sub-categorize, and sub-sub-categorize risks to create subgroups that are too small to make viable (affordable) pools.

A different misbelief is that insurance should cover trivial and predictable expenses such as doctor office visits. That's not insurance; that's a prepaid service contract. Blending service contracts with tradition insurance almost requires insurers to individualize premiums, because typical health care insurers pay out more to outpatient services than to hospitals. The "service contract" add-on is a primary driver of rapidly escalating health care insurance premiums.

Ak Mike writes:

Mingo, I respectfully disagree. Insurance is a business providing a service. The service should be to protect the customer from catastrophic loss. To obtain this protection, the customer pays, reasonably, the cost of providing this service. Much that is good flows from having people pay the cost of what they buy.

Every other form of insurance charges differentially depending on risk. Fire insurance cost varies with distance from the fire station, presence or absence of hydrants, sprinklers, smoke detectors, etc. Life insurance cost varies with the health and age of the insured. Auto insurance cost varies with the driving record of the insured and the condition of the vehicle. Why should not "health" insurance cost vary with the expected likelihood and magnitude of the insured risk?

If you (and Dr. Brynjolfsson) want to provide public assistance to prevent catastrophic medical expenses from eating into the funds of the medically uninsurable, or public assistance to help pay the premiums of those with high expected costs, please don't confuse that with insurance.

Mark V Anderson writes:

Ak Mike is right and Mingo is wrong (in Mingo's 1st paragraph). If the future is predictable (as it is to the degree the insurance company can sub-categorize levels of risk), then there is no point to insurance. As others have stated, that is redistribution, not insurance.

Jacob A. Geller writes:
"But take the extreme where it was all due to bad luck at an early age. To justify making others pay, you would have to argue that the bad luck should be forcibly put on others who themselves had no role in creating the bad luck. Here's where I balk."
Take a really extreme case -- childhood cancer.

Do you balk at the idea of forcing healthy working-age individuals like myself to pay for a child's cancer treatment?

To what extent is the fact that working-age individuals like myself didn't create said cancer a factor in that equation?

What if I and 51% of my fellow taxpaying Americans preferred to pay for a child's cancer treatment not (merely) by giving our money voluntarily to children/pediatric oncologists chosen by us by hand, but by having the IRS garnish my wages and use the proceeds to reimburse pediatric oncologists -- would that still be "forcible" to you?

If I and 100% of my fellow taxpaying Americans preferred not to pay for a child's cancer treatment at all, would that be coercive of the child in any way? Sure, we didn't literally *give* the child cancer, but haven't we sort of, you know, consigned the child to die, simply because we preferred to spend our money on ourselves?

If none of these are convincing, then what conditions, in your mind, *would* justify forcing society (people like you and me) to pay for treatment of a child via, say, Medicaid, as opposed to purely voluntarily in a free marketplace?

(And no I don't intend for this to be rhetorical... I'd be interested to hear your thoughts, as a libertarian.)

Comments for this entry have been closed
Return to top