This chapter is concerned with projecting medical care expenditures, especially expenditures on the elderly. I heartily approve of such concern. Within the next two decades, financing health care for the elderly is likely to pose a much greater national problem than “saving social security” (Fuchs 1999a, 1999b). Change in the age distribution of the elderly, the subject of the chapter, is one element in projecting future expenditures. The authors argue that the standard method of projection – based on cross-sectional age-specific expenditures – overstates future expenditure increases for two reasons: (1) improvement in age-specific health status (e.g., lower mortality) leads to lower age-specific expenditures; and (2) end-of-life costs tend to be lower at older ages.

The first point was made by Kenneth Manton (1982); he suggested: “As mortality rates decline at a given age, there would be some compensating decline in the rate of utilization of certain health services (e.g., nursing home care) before that age” (p. 205). I explored this question empirically (Fuchs, 1984) by noting that health care expenditures at any given age are very strongly related to survival status. (The authors' Table 7.9 shows that survival status is a much stronger predictor of expenditures than age per se.)

If age-specific expenditures are adjusted for age-specific survival status, the tendency for expenditures to rise monotonically with age disappears, as can be seen in Table 7-1.1.