Abstract
In this paper, the possibility of predicting future health care costs from predicted changes in age structure and age-specific nlortality is explored, based on data for the Swedish population. A linear relationship between age-specific mortality and agespecific health care costs is established for 1997. By combining this relationship with predictions of the future age structure and the future age-specific mortality rates, the Swedish health care costs in 2010 and 2030 are predicted. In order to test the validity of the method, the same methodology is applied retrospectively to data from 1985 in order to predict the health care costs in 1997. The results show that the method gives an underestimation of the actual costs. This should come as no surprise, since international research has shown that the age structure plays a relatively insignificant role for the level of health care expenditures. The most important factor for explaining differences in health care expenditures over time, and across countries, is the level of GDP per capita.