A Reader’s Questions About Health Insurance

She writes,

how do you define ‘extreme circumstances’? By the cost of the treatment? By the severity of the disease? How and who defines ‘extreme circumstances’?

…Let’s say I get pneumonia. Doctor prescribes antibiotics that cost $100. I decide that the cost of the antibiotics are too high compared to the possible benefit (50/50 chance the antibiotics help) so I hope that enough rest and fluids cure me. Pneumonia gets worse. Now I’m hospitalized and it cost $10,000 a day to be in the hospital. Now my insurance company decides that $10,000 a day is an ‘extreme circumstance.’ How is this scenario more cost effective?

To answer these questions, use some standard insurance company terminology.

1. An insurable event is a rare, unpredictable, costly occurrence. A fire burns down your house. Someone runs into your new car and totals it. What I mean by an “extreme circumstance” is something that constitutes an insurable event. Let me come back to that in a moment.

2. Moral hazard is the failure of an insured person to undertake reasonable steps to reduce risk. Insurance companies address moral hazard through a combination of carrots and sticks. For example, a fire insurance company could offer you a discount on premiums if you have a working smoke alarm. Or they might provide you with a free smoke alarm. Also, because your behavior affects their risk, they might have a rule that says that if your house burns because of negligence (smoking in bed) then they are not liable to pay claims.

The pneumonia patient who decides not to take antibiotics is an illustration of the moral hazard problem. In theory, real health insurance would try to address this. The insurance company might offer a discount to people who obey doctor’s orders. It might offer to pay for the antibiotics. Or it might issue rules warning that it will not pay for hospitalizations that could have been prevented by following a doctor’s reasonable advice.

In any field of insurance, including health insurance, moral hazard is challenging to address. It is hard to list all of the things that insured individuals can do to reduce risk and to specify appropriate carrots and sticks to incent individuals to do those things. So moral hazard will never go away completely, but with well designed insurance contracts it may be reduced to a manageable level.

Now, back to the definition of the “insurable event” in health care. Consider long term care insurance, which pays for a nursing home if you need it. Let us say that a policy covers 10 years, meaning that if the event happens in the next 10 years, the coverage kicks in. If I get this long-term care insurance when I am 50, and I become decrepit at age 59, then the insurance will pay for my long care for the rest of my life. But if I am still independent at age 60, my policy has expired and I would have to get a new one if I want to remain insured. [I am not saying that actual long-term care policies are structured this way. Things are more complicated.]

If you are 40 years old, becoming decrepit any time in the next 10 years is an insurable event, because it occurs so rarely. If you are 70 years old, the chance of having to check into a nursing home some time in the next 10 years is so high that it is not an insurable event. The premiums would have to cost you nearly as much as the nursing home bills.

In dollar terms, I think that the threshold for an event to be insurable needs to be much higher than it is currently in health insurance. People need to get used to saving a lot of money in health savings accounts to deal with short hospitalizations, pregnancies, and the high medical costs of old age, including long term care. It is fair to wonder how practical it is to assign households the responsibility for accumulating this saving. The alternative is to use a government tax-and-transfer scheme. Some combination of both is inevitable, but the relative mix is a subject for discussion. You can think of this is the problem of where to draw the line between health insurance and health charity. How high should a household’s health care spending be as a percentage of income before you give them charity? Keep in mind that in the U.S. average health care spending is 15 percent of income (not nearly that much out of pocket, obviously), so it is mathematically impossible to draw the charity line below 15 percent for everyone.

In health insurance as it is currently designed, the “insurable event” is any expenditure that takes you over the deductible. That is very problematic.

Suppose that you have a chronic illness that is expensive to treat, with an average expense of $25,000 a year. As you apply for health insurance, having high medical expenses next year is not an insurable event. You know that you will have high expenses, and so does the insurance company. Fair premiums would be $25,000 or more, so you would not really benefit by obtaining insurance.

Or suppose that a week before your insurance policy comes up for renewal you get injured in a way that is going to require a sequence of expensive surgeries to fix. Those expensive surgeries are not an insurable event, so your insurance company is not going to renew your policy, and other insurance companies either will not cover you or will charge exorbitant premiums.

We need to change the way we think about insurable events in health care. For the case of the injury, the “insurable event” would be getting into the accident. An insurance adjuster should determine your compensation based on the estimated cost of fixing your body, just as with a car accident an insurance adjuster will determine compensation based on the expected cost of fixing the damage to your car.

It is possible that this “insurance adjuster” model also could be applied to a chronic illness. The “insurable event” would be the point at which you are diagnosed with the illness, and the adjuster would set compensation based on expectations of future treatment costs. However, because chronic illnesses require many years of treatment and the course is unpredictable, this may not be the most workable approach.

Another approach, suggested by John Cochrane (“the Grumpy economist”), is to make the insurable event an increase in your health insurance premiums. Suppose, for example, that you were cruising along paying $1,000 a year in health insurance premiums, and then you get diagnosed with a chronic illness that is expected to cost $25,000 a year. Your current health insurance company, ABC Insurance, is ready to drop you like a hot potato, so you shop for a new policy and you find that the best deal you can get is from XYZ insurance and costs $26,000 a year. If you end up with that policy, you pay $1000 a year, while ABC pays the extra $25,000 a year. That is, your original policy with ABC includes protection against becoming uninsurable!

There is more to this complex topic. I hope this helps.

This entry was posted in Economics of Health Care. Bookmark the permalink.

10 Responses to A Reader’s Questions About Health Insurance

  1. BenK says:

    There is certainly lots more to the topic, including what ‘totaling’ someone would be, but the biggest issue we should address is whether bureaucracy is really the best place for judgments of moral hazard in health to reside.

    There is another option; essentially a jury of your peers. In many cases, even when the government and the markets won’t pay out for someone with big problems, family, who are inclined by long relationship, or neighbors, inclined by sympathetic proximity, or fellow congregants in church, or people with similar conditions, or people with relatives with similar conditions, or professionals for whom the cost of service is lower than market rate, band together to provide charitable care. They make some personal, private judgment about the moral hazards involved, and based on personal inclinations, and from personal resources, provide care.

    This is a very different kind of insurance, but I’m highlighting how it involves an assessment of moral hazard outside of a bureaucratic framework. It is also personalized, specialized, diverse, and distributed. It is compatible with freedom and virtue. It comes with a recommendation, of course, that people act in ways that make them sympathetic – such as prior acts of personal charity and other contributions to the common welfare. It does not reinforce behaviors of ‘bending the rules’ or ‘abusing the system.’

  2. Jon says:

    I used to manage my company’s insurance program, and feel compelled to quibble with your definitions. “An insurable event is a rare, unpredictable, costly occurrence.” Yes, but unpredictable only for the insured. The insurance company knows the likelihood that your new car will be totaled. If it didn’t, that event would not be insurable.
    “Moral hazard is the failure of an insured person to undertake reasonable steps to reduce risk.” No. Moral Hazard is tendency of the insured to engage in riskier behavior, because he is protected from the consequences by the existence of insurance, than he would if he bore all of the risk. The extreme version of moral hazard is when the insured has a positive incentive to incur the loss, e.g. “insurance fires.”
    What we call “health insurance” is partially real insurance, but largely tax avoidance and government subsidization of health care consumption.

  3. Joseph says:

    But is insurance, however one chooses to define it, the problem that people actually want solved? When people talk about health insurance, I don’t think they mean “insurance”, I think they mean “when people are sick, a third party should pay for whatever healthcare services the sick person’s doctor tells them they need”. The third party could be the government or some notionally private entity that is usually highly regulated.

    Libertarians and conservatives think that’s wrong for ethical and practical reasons. They think that people should pay much much more (and depending who you ask all) for the healthcare services the doctor tells them they should consume. In fact, they argue that the reason healthcare costs are so high is due to a third party payer insulating sick people (also known as “healthcare consumers”) from the costs of their healthcare consumption. The theoretical policy is straight forward: force consumers to pay more for healthcare services, which over time will reduce the cost of healthcare services, thereby leading to a virtuous cycle. Which has a strong whiff of the underpants gnome theory of healthcare cost control. But it is possible, right

    What isn’t possible, is selling that politically ever in any way shape or form to anyone who isn’t living in Galt’s gulch.

    • Andrew' says:

      “What isn’t possible, is selling that politically ever in any way shape or form to anyone who isn’t living in Galt’s gulch.”

      Watch me.

      But this kind of s sells itself. All I have to do is sit back and watch democrats install everything. Because when they don’t, their beloved leviathan will go bankrupt.

      We didn’t even have to wait a decade for Obamacare reform to be unavoidable.

  4. R Richard Schweitzer says:

    this form of discourse is a good step toward understanding the differences in:

    Insurance: Transfers and “spreading of RISKS;
    Service Contracts (Healthcare, e.g.): transfers and spreading of COSTS.

  5. collin says:

    People need to get used to saving a lot of money in health savings accounts to deal with short hospitalizations, pregnancies, and the high medical costs of old age, including long term care.

    Maybe we should more free markets for abortions. Or better yet let the insurance cover it! My guess if able to I bet insurance companies would find ways of requiring any woman to take birth control. (Especially if they have one kid.) Or least have more cheap forms of pregrancy. (I do notice the savings accounts heavy Singapore has a really a low birth rate!)

    Also I think the idea for anybody in the bottom 80% to save for long term care at the cost of less consumption today is not good a trade to be honest.

  6. Lord says:

    If everyone should be covered continuously from birth, then unless it is proposed to rely on family histories or genetic testing, all premiums should be equal, varying at most with age. And if the world is deterministic or at most stochastic, there is no choice, no liberty, no moral hazard, though even moral hazard steps would not be choices.

  7. Andrew' says:

    It is fine. We can combine chronic diseases with insurable events. It’s just not very efficient.

    And if we had transportable Healthcare shares or HSAs or Singapore care it wouldn’t be an issue. We are already paying for these things. As unavoidable costs, savings might only come from research breakthroughs.

Comments are closed.