13. Sex ratio and sex differential of mortality

The greater longevity of the human female is most clearly manifest in the life of the elderly and translates virtually without exception into a sex ratio of mortality (male rate per female rate) higher than unity. With increasing age, however, this ratio likewise almost universally -declines and this has often led to the mistaken conclusion that at high ages death rates for the two sexes tend to converge and may eventually intersect.

        Actually, the declining ratio is nothing more than an inevitable result of arithmetic: at young ages the ratio may reach, and often does, values of 2 or 3 but such ratios become increasingly unlikely and finally impossible as mortality reaches higher levels in old age. If, instead of the ratio, we look at the differential of the rates, the picture changes completely: instead of converging, the rates maintain their distance.

        These two structural features of old age mortality - sex ratio declining but sex differential fairly constant with advancing age - have persisted through the recent momentous development which has been much more beneficial to women than men.

         An examination of the sex ratio based on age-standardized death rates at ages 80-99 in the 28 countries with acceptable data (Annex Table 10) confirms, first of all, that the males are subject to considerably higher mortality than the females and that this difference has lately been increasing. The male excess mortality which at these ages in 1960-64 varied between countries from 10 to 33 percent, had by 1985-89 grown to an excess of 24 to 48 percent while the mean had grown from 20 to 36 percent.

        The highest levels of excess male mortality in old age are now found in Northwestern Europe and Australia as shown in Table 15 and the lowest generally in Eastern and Central Europe. The very low ratio for Spain is probably due to age overstatement by males whose death rate therefore has a downward bias.

The increase in the sex ratio during this last quarter century, also shown in Table 15, has been remarkably sharp in the Low Countries and Scandinavia where improvement in the mortality of old men has been slight. Although this has not been the case in Switzerland and Finland, the gains were in these countries much larger for women. At the other extreme, the sex ratio has increased only moderately in the former East bloc. The apparent decline in Spain is most likely an artifact. The development is illustrated by age group in Figure 9 which is composed of only the most reliable data. In it an aggregate of eleven Western European countries is compared with a group of three Eastern European countries and with Japan. The actual ratios are given in Annex Table 11.

        In both European groups the decline of the sex ratio of mortality with advancing age was very moderate in the 1950s, after which the gradually increasing female advantage has translated into ever steeper slopes as the upper end of the life span has not kept pace with the rapid changes among octogenarians. In Japan the slope was very straight and stable for three decades and approached the European pattern only in the 1980s when it also confirmed the existence of a mortality difference between male and female centenarians.

        When, instead of sex ratio, we look at the sex differential in Figure 10, we note its nearly constant size by age. In the European curves there is a tendency for a slow increase of the differential over age which belies the assumption of convergence. As the female advantage has grown over time, the curves have been moving to higher levels while roughly maintaining their slowly ascending form.

Updated by V. Castanova, 1 March 1999