The fundamental problems with health insurance

27 July, 2009 (19:34) | Public Policy | By: david

After watching countless TV ads for life insurance as a child, I finally went to my parents and asked “Do you have life insurance?” They assured me that they did. So I asked the natural follow-up: “Do I have life insurance?” They laughed. “Why would we need life insurance for you?” I was hurt. You see, I had misunderstood. The ads I was watching said “buy life insurance, because you might die.” So I reasoned that buying life insurance was a way to prevent death, and I was sad to find out that my life wasn’t worth the premiums.

I understand the concept of insurance better now than I did then, but the concept of health insurance is in some ways just as baffling to me today as life insurance was then. And I worry that what’s getting overlooked in the current debate about our health care system is just how peculiar the notion of health insurance is.

Think about all the things we normally insure and the hazards we normally insure them against. Usually we insure tangible items: houses, cars, boats, jewelry, etc. And we insure them against very unlikely events: fire, flood, theft, fraud, and so on. This kind of property insurance makes a lot of sense for everyone involved. The big winners, obviously, are people who make claims. They get back far more from the system than they put in. People who never make claims also win, though, because even though they are effectively throwing their money away, they are only out a fraction of what they might otherwise have lost, and they got to sleep better at night. Insurers win because they have very smart people calculating the odds of these unlikely events happening and how much each policy is worth. (For example, if a ring gets stolen, they have to pay whatever that ring was appraised for – not for “a new ring.”) So insurers have a pretty good idea of how much they are likely to pay out each year. Occasionally they encounter insurance fraud, but for the most part they can be assured that their policy holders will take reasonable precautions to prevent the losses from occurring in the first place.

Life and disability insurance are a little trickier because no one can objectively calculate the replacement value for a life or a career. To correct for this, life and disability insurance policies are written with defined benefits. For example, there might be a fixed amount awarded in case of accidental death, and someone who becomes disabled gets awarded a percentage of his or her previous salary. Again, there are instances of fraud (and murder), but generally speaking people would much prefer never to collect on these policies.

Now think about health insurance, and note how radically different “health” is from life, disability, and property:

  • Everyone gets sick or injured eventually.
  • Defined benefits are usually not part of the equation – certainly not as explicitly as they are with other types of insurance. Some health insurance policies have lifetime benefit limits, but no policies tell you in advance exactly how much will get paid out for every conceivable malady. The general proposition came to be “if you get sick or hurt, we’ll pay for you to get well.” (With the advent of managed care, that got amended: “And we’ll also pay to keep you from getting sick in the first place.”)
  • Most people intentionally behave in ways that increase their need to “cash in” on their insurance policies. These behaviors range from things we generally frown upon, like smoking and overeating, to things that are commonly accepted in everyday life, like playing sports or driving a car, to things that are essential to society, like working in dangerous jobs or getting pregnant.

So the very building blocks of what makes the casualty insurance system work – unlikely hazards, defined benefits, and motivation by the insured to avoid claims – are all absent in the health insurance system.

Now add to that the fact that the health care services purchasing market is like no other and you have a real mess.

I was curious as to how this obviously silly idea of health insurance came about so I did a little poking around. There’s clearly enough here for a doctoral dissertation so I won’t attempt to provide “the answer” here. But if you’re interested in the history of health insurance, here are a few links to get you started.

Any health care system that has “insurance” as its foundation is doomed to fail. The only solution as I see it is to define some package of health care services that every resident of this country is entitled to regardless of income, assets, age, medical condition, or any other factor. This package would include preventive measures like annual physicals and vaccinations but not gym memberships. It would include a reasonable standard of care for chronic conditions but not the latest, greatest, and most expensive treatments (even if those treatments were proven to prolong life). It would provide crutches but not million-dollar prosthetics. Most people, of course, would not want their care to limited to such a bare-bones package, and we could let the private sector figure out a way to provide supplemental benefits, like quicker access to a wider range of physicians, more expensive operations and drugs, private hospital rooms, and so on. Ideally the supplemental benefit system would expose the true costs of health care to the consumers and give them an incentive to spend conservatively.

In a system like this, rich people would live longer and more comfortably than poor people. Could we abide such a result? We do now.

Share

Comments

Comment from Michael Pahre
Time July 28, 2009 at 11:02 am

My understanding is that your proposal for core, guaranteed, nation-wide health services with optional, privately-purchased supplemental packages is similar to that in France.

Comment from Thomas Moran
Time July 29, 2009 at 3:47 pm

My understanding is that any insurance is designed to mitigate risk. Insurance can be applied wherever risk is widely shared but where a relatively small percentage of the population experiences the adverse event or condition. Properly constructed there should be no winners and/or losers among the insured population. Those who don’t experience the adverse event are better off for that, and those who do are compensated just to the extent of their loss. This is the model, the actual implementation may, of course, depart somewhat; where it departs widely fraud is typically involved.
This understanding should be applicable to all forms of insurance. homeowners and auto insurance are cited as paradigm cases. The other cases come under this understanding as well:
In the case of life insurance what is insured against is the premature loss of one’s life. A certain small percent of the population lose their lives early. The payout is not set at a given value because there is no way to properly value life. The premium is set based on the desired payout vs. the likelihood that the person will die within the premium cycle. The desired payout is determined by the insured based on the desire to provide for heirs, etc.
Likewise disability insurance insures against the relatively small probability that a person will not be able to earn their expected income over the course of their working years due to some defined cause, sickness, injury, etc. The premium is set by the likelihood of disability and the expected payout, often based on the salary of the individual.
Health insurance may differ from this model in the end of life period, where a much higher percentage of the target population is subject to adverse events or conditions. So it may not make sense to speak of health insurance for people above a certain age, but rather to speak of providing health care coverage for that population. Restricting the discussion to the population under 65 years of age, say, it seems that the standard insurance model can be applied.
If health insurance departs from this model, it is not because health care and health risks differ from other risks. It has to do with the marketing of health insurance and health care in our society.
As Harold Luft points out in his book ‘Total Cure’ the routine parts of health care do not conform to the insurance model. If everyone should visit the doctor periodically and if everyone should have screening for various medical conditions and indicators, then it makes no sense to insure for these. It might make sense to incentivize them, but that would not be an insurance function.
What can properly be insured medically in Luft’s terms are ‘acute episode’, which is to say ‘hospital care’ and ‘chronic illness’, such as diabetes. In these cases the classic insurance model obtains: a small percentage of the population actually experiencing the adverse effect. Insurers can adjust rates for members of the population whose behavior increases their risk, just as is done for auto insurance. There is a further condition that an individual experiencing a chronic condition not be dropped from the insurance rolls., this is part of the set of assumptions required in setting up the risk pool.
Chronic conditions that reach epidemic proportions, as it is sometimes suggested diabetes and obesity are doing, may then move outside the domain of insurable risks and into the domain of public health risks. This seems to me to be a sensible way to think about such cases.

Comment from david
Time July 29, 2009 at 4:50 pm

Thomas, we are in agreement about most things, but I think that the days of acute episodes and chronic illnesses affecting only a small percentage of the population under 65 are long gone. How many people our age do you know who have made it even this far without encountering either? Injuries and illnesses are common today that couldn’t even be diagnosed 50 years ago, much less treated.

As for risk adjusting, that’s a completely different but equally troubling kettle of fish. Do we want our monthly health care payments to be based on the DNA samples we provide?

Write a comment