With Congress debating ways to reform health care throughout the United States, editorialists railing for or against a single-payer system, and my wife today attending a 4-hour lecture on billing at the hospital where she will be working in a week, my thoughts have been on health care costs. I think there is no question that our current system has gross inefficiencies, with gigantic bureaucracies built up to determine what has been done, what it should cost, who should pay for it, etc. Removing the middleman of private insurers could certainly be worth some significant savings, but turning things over to government generally means more infrastructure, not less. That is a debate I will leave (for now) to those wiser and more knowledgeable than I.
What I would like to discuss today may be an even more controversial position: that what we need is less health care. I do not recall ever hearing anyone else espouse this opinion, but it seems quite logical to me. If we think about it, the relationship between resources spent on health care and the results we get from it (average life expectancy, quality of life, whatever metric you prefer) is a sigmoid function of the type shown below. There is a minimum result level that we will attain even if none of our resources are spent on health care. The results rapidly increase as we allocate a moderate amount of resources toward them, then continue to grow very slowly as more and more resources are used.

Because we value human life and health so highly, it is natural to find ourselves on the asymptote of this curve, doing whatever we can to get the best possible results. If this is the case, then perhaps we could reduce expenditures by 30% and lose only 5% in quality. It may seem that no price should be too high to take care of the health of our citizens, but a simple thought experiment will demonstrate that this is not the case. Suppose that a single person had a very rare disease. A guaranteed cure for this disease is known, but a single dose of it can only be manufactured by the combined labor of 100 people for a year and the destruction of 1000 acres of rainforest in raw materials. No one, except perhaps our unfortunate victim, would advocate this as reasonable.
Knowing where to draw the line is of course quite tricky. It is easy to decide that some treatment is not cost-effective when none of our loved ones are in need of it, but another thing entirely when this is not the case. If we were to do something like this an impartial, evidence-based commission would need to make the decisions. I would guess, however, that the easiest thing to justify axing would be treatments that have been well-studied, are known in the average case to only increase life expectancy by a few months, and require an enormous expenditure per patient.
I am not, of course, proposing that such care should be outlawed, simply that it should not be considered a public good in a possible nationalized system. When insurers refuse to cover the least cost-effective treatments it causes outrage, and perhaps it should if people have been paying premiums under the belief that they will in return be given the best care money can buy. But that assumption must be changed; we could easily, if we valued it highly enough, spend our entire GDP on health care. There comes a point, however, where allocating resources elsewhere would result in a much higher increase in quality of life for society at large.
Perhaps my feelings on this stem from my own preferences. I have health insurance, but not the sort that would help me out in the event of a very serious illness. I want to be taken care of if I have a broken bone, a staphylococcus infection, arthritis, etc. If I were diagnosed with something that is likely to kill me within a few months if left untreated or a few years if I undergo a treatment regimen that will keep me in the hospital and sickly most of the time, I would choose the former in a second. Perhaps I would feel differently if I had children I wanted to see grow up, but my life has been full enough that at this point I feel no need to grasp for more, however unpleasant it might be.