Sam Preston's paper provides an excellent and insightful analysis of past demographic experience and his expectations for the future. Frank Bayo's discussion covers many of the reactions we share, both working at the Office of the Actuary, Social Security Administration. The comments below provide a few additional insights that I hope will be useful.
In his paper, Dr. Preston suggests that future declines in mortality rates may be greater than those projected by the Social Security Administration (SSA). An extrapolation of long-term past rates of decline would support his position. However, a more careful look at past trends, and the reasons behind them, lead us at the SSA to believe that declines will not average as large in the future as in the past several decades.
Between 1900 and 1990, the age-sex-adjusted death rate in the United States population declined at an average annual rate of 1.2 percent. However, it is interesting to note that this period has been made up of several distinct periods with very different experience. From 1900 to 1936, mortality rates declined by only 0.8 percent per year. From 1936 to 1954, they declined by 2.3 percent per year. Little decline was experienced from 1954 to 1968, only 0.2 percent per year, followed by rapid decline from 1968 to 1982, at 2.0 percent per year. Since 1982, mortality rates have again been declining slowly, only 0.4 percent per year.
Precise explanation of this seemingly cyclic pattern of mortality is elusive. The rapid improvement between 1968 and 1982 would seem to be related to the introduction of Medicare, bringing more health care to those with the greatest probabilities of death. Slower declines since 1982 might be assumed to be associated with some of the Medicare reforms that decreased service. However, for these periods, the first rapid and then slow decline in mortality was experienced for all age groups, not just the elderly served by Medicare.
Frank Bayo suggests that future improvement will be harder because we have, in effect, already found the easiest cures and further medical cures will be more difficult. Two additional factors will also contribute to ultimately slower mortality improvement than experienced so far this century. Certainly much of the past improvement is due to our ability to make adequate nutrition and health care available to virtually the whole population. These changes can only be made once. They have contributed greatly to the health and thus the life expectancy of the population this century.
While we will continue to see declines in mortality from these changes, and from trends toward better eating habits by some and a reduction in smoking, for many years into the future, ultimate future gains will require new medical interventions and changes in life style that may be difficult to develop and harder to provide the population at large. Our ability as a nation to afford the new medical technologies for all the people is increasingly in question. Economics will likely restrict our ability to make them available. In fact, the level of current health costs suggests that we cannot even afford the current level of health