May 22, 2011
Are There Sizable Benefits From Extending the Lives of the Old and Frail?
Are There Sizable Benefits From Extending the Lives of the Old and Frail? Becker
The federal government subsidizes basic and applied research on both the quality of life and the length of life. An example of research on quality is the efforts to find implantable artificial kidneys that can replace dialysis for persons with severe kidney disease since dialysis is a limited and very confining procedure. Research into finding cures for breast and prostate cancers exemplifies research on extending the length of life. It might seem clear that research on extending the lives of frail persons over say age 80 does not produce much value to these persons or to society, but economic analysis of the "value of life" shows that this conclusion is far from obvious.
The main criterion used to measure benefits from all kinds of medical advances is how much individuals affected by these advances, and society as a whole, would be willing to pay for them. Clearly, reducing the incidence of Alzheimer's disease, or finding drug substitutes for dialysis, would add great value to individuals who would have been victimized by Alzheimer's or would have been on dialysis. As Posner indicates, however, this does not necessarily imply that putting lots of time and money into developing ways to extend the lives of old people is worthwhile, particularly those who are frail.
On the one hand, the health care delivery systems of the United States and of many other rich countries do not seem able to prevent newly developed expensive drugs, costly surgeries, and other medical interventions from being used to prolong the lives by a tiny amount of very sick elderly persons. It might be better if research spent on these kind of medical innovations went instead to improving the quality of life for other elderly persons, or to helping reduce death rates from cancers, cardiovascular diseases, and other major killers.
On the other hand, even old and frail men and women are often willing to pay a lot for rather small improvements in their expected lifespan. Economists have estimated how much individuals must be paid to be willing to work at jobs with higher death rates, to avoid being involved in deadly traffic accidents, or to take on other life-threatening risks. After manipulating these willingness to pay estimates, the conclusion is that the statistical "value of a life year" to a young or a middle aged American is about $120,000. This is far higher than average earnings in any year- about $45,000 for men who work full time- because the value of life estimate includes the value placed on the many hours not spent at work, and on other sources of utility.
The present value of a reduction in the probability of dying now and at each future age would be worth more to younger persons since older individuals have fewer years of life remaining. But is the value of a "life year" also worth less to the old and perhaps frail? The answer to that is less obvious than may seem (the following analysis is based on work in progress, referred to by Posner, by Kevin Murphy, Tomas Philipson, and myself). Presumably, frail elderly people tend to receive less utility from a year of their current life since their lack of health prevents them from greatly enjoying their leisure time and consumption of different goods. However, the utility cost of any time and money they might spend on prolonging their lives is also lower for them. The fundamental measure of the value of a life year is the ratio of the utility gained to this marginal utility spent on prolonging life. This ratio could even be higher for the old and frail than for healthy younger persons.
In addition, since old and frail people have only a short remaining lifetime, the fear of dying, long recognized as of great importance by novelists and philosophers, is more immediate to them than to younger persons. Their fear of dying could make old people quite willing to spend liberally to delay their deaths even by a few months. Finally, as Posner recognizes, the old would be willing to spend most of their wealth to prolong their lives even by a little bit if they value their lives by a lot more than they value any bequests they would leave to children or others.
Of course, Medicare, Medicaid, and employer health insurance distort many health care decisions in the US since sick individuals mainly pay not their own money, but the monies of taxpayers or of other employees. Medicare, for example, presumably leads to too much spending on the very old because the government ends up financing the great majority of this spending. This is an important determinant of the actual spending on the health of the elderly, but I have argued that many of the old and frail would be willing to spend a lot of their own money to prolong their lives even by a small amount.
Life Extension versus Quality-of-Life Enhancement—Posner
The Medicare program subsidizes medical care for the elderly so heavily as to create serious concern about the fiscal soundness of the federal government. And, as longevity rises, the size of the subsidy rises, and the rise in cost is compounded by the increasing cost of medical technology. Among possible measures that would reduce the rate at which the cost of Medicare is increasing would be means-testing and—the focus of this piece—shifting the balance of subsidized R&D so that more is spent on increasing the quality of life of elderly people and less on extending their (our) lives.
We need to recognize that the public subsidy of medical care for the elderly is not limited to the Medicare program (and Medicaid as well, which provides medigap insurance—insurance against the part of the cost of medical care not covered by Medicare—to millions of Medicare participants who can't afford private medigap insurance), but includes much of the public expenditure on medical R&D, since the elderly are by far the principal beneficiaries of continued advances in medical knowledge and treatments. Although federal expenditures on medical R&D are small relative to Medicare, they have a multiplier effect: every year of life added by advances in medical technology increases the size of the elderly population and hence the cost of the Medicare program.
Thus, life-extending medical research can aggravate the nation's fiscal problems directly (as a major spending program in a political culture that abhors tax increases) and indirectly by increasing the Medicare (and elderly Medicaid) population. Could there be an offset? In National Bureau of Economic Research working papers in 2007 and 2010, Becker and coauthors point out that even elderly, frail people value life extension. The question is how much they value it, and how much we want to subsidize it. A 2006 article by Kevin Murphy and Robert Topel in the Journal of Political Economy entitled "The Value of Health and Longevity" estimates enormous gains to real national income from extending life, but they base their estimates on "value of life" estimates that are inappropriate for this purpose.
What is misleadingly called "value of life" in the economic literature is not that at all; it is a way of estimating optimal expenditures on precaution. Suppose that by observing the behavior of persons engaged in activities or occupations that involve a somewhat elevated level of risk, say a 1 in 10,000 probability of a fatal accident per year, we learn that the average person facing such a risk demands compensation in some form (such as a wage premium) of a shade more than $500 a year, and therefore an expenditure of $500 on a measure that would prevent one such accident would be cost justified. All the analysis tells us is that the value that a person places on avoiding a 1 in 10,000 fatal accident is $500, so if the accident can be prevented for less there is a social gain. But this tells us nothing about the utility that a 25-year-old would derive from knowing that his life expectancy had risen from 80 to 81 or that an 80-year-old would derive from learning that his life expectancy had risen from 83 to 84. Hence it's impossible to determine the optimal level of expenditures on medical care that principally adds years at advanced ages, as distinct from reducing mortality among people who are not yet elderly.
A complication remarked in the Becker paper (as earlier in my book Aging and Old Age) is the nonlinearity of "value of life." The fact that a person would accept a 1 in 10,000 annual probability of death in exchange for $500 doesn't imply that he would accept a 100 percent probability in exchange for $5 million on the theory that he values his life at $5 million and is therefore indifferent between the life and the cash. Dead he would derive no utility from the cash unless he had an unusually generous bequest motive, and rarely would that be strong enough to make him indifferent between life and death. If he has no bequest motive, he will spend all his money on extending his life even for a very short time (unless the additional life has negative utility to him because he expects it to involve great suffering), and let us say that he has $10 million and so will spend it all for a few additional months of life. That does not mean that his value of life is greater than that of a much younger person who "values" his life at only $5 million.
And thanks to Medicare even elderly persons who have a very strong bequest motive have no incentive to economize on medical treatment (at least if they also have medigap insurance). The opportunity cost of medical treatment is zero to them.
So we can't have a clear idea of the welfare gains from extending the life of elderly people. But we can say with reasonable confidence that the welfare of the elderly, and of altruistic members of their families, could be enhanced, without a significant increase in the longevity of the elderly, by redirecting medical research toward diseases or conditions that impair quality of life without necessarily shortening it, or at least without shortening it commensurately. Dementia (which comes in many forms, but Alzheimer's appears to be by far the most common) is the foremost example. It does shorten life somewhat, but on the other hand, as it is largely a function of age, its prevalence is increased by increases in longevity; and dementia is not only psychologically very hard both on the demented and on their families, but also very costly in the amount of care that demented persons require. Blindness, deafness, loss of mobility, and Parkinson's Disease and related degenerative nerve diseases (life shortening, but often the effect on lifespan is less than the effect on quality of life) are other examples of diseases where investing in medical research might yield substantial increases in elderly utility without significantly increasing longevity. Stroke is an example of a medical condition that both reduces longevity and has often dramatic negative effects on quality of life.
Yet the National Institutes of Health expect this year to spend only $18 million on dementia research, $154 million on Parkinson's research, and $337 milliion on stroke research, compared to more than $8 billion on cancer, heart disease, and diabetes research, even though the diabetes "epidemic," while real, is due largely to obesity and bad diet. (Eye disease, however, seems generously funded at $817 million.) Of course the gravity of a disease is not the only factor determining the optimal level of research expenditures; another of at least equal importance is the likelihood that the research will be productive. Optimal allocation of NIH money also requires consideration of the allocation of private research money across diseases. But the enormous expenditures on cancer research have not been very productive, and very promising research programs in dementia (notably research on an Alzheimer's vaccine) are greatly underfunded both publicly and privately. In my view this is a regrettable imbalance.