Perhaps, in this moment, it is useful to refer to a chart showing changes in age-specific mortality for males and females separately. Figure 8 indicates age-specific patterns of mortality for 1921-1925, 1960 and 1996, separately, over several decades. Notice that this chart is based on the semi-logarithmic graph. From this chart, some salient features may be pointed out.
(1) The decline in infant and child mortality from 1921-25 to 1996 is very significant.
(2) What may be called "tuberculosis lump" appeared in 1921-25 at ages between 15 and 25 have virtually disappeared among females in 1990. The recurrence of a lump or bump around age 20 among males in 1996 should be considered to be caused by external cause of death, notably attributable to motor-bike accidents.
(3) On the other hand, however, the mortality improvements are not particularly phenomenal at advanced ages, though the nature of semi-logarithmic graph might disguise some appreciable changes.
These charts may suggest that increases in life expectancy occurred in the postwar years were largely attributable to the rapid declines in infant and child mortality. This pattern of trends is also observed widely in present-day developing countries since the mortality can be more easily reduced in the young age groups rather than in the elderly. Mortality in childhood is caused mainly by infectious and parasitic diseases which can be more easily controlled by the anti-biotics and modern medical technology. On the other hand, however, mortality in the old ages is not easily controlled even by modern medical technology since diseases characteristic of old-ages are of attritional and degenerative nature of human organs. Hence, if other things being equal, the remarkable reductions in mortality in the infancy and childhood have an immediate effect of expanding the bottom part of the population pyramid, thus youthening or rejuvenating population, rather than ageing population.
Table 6 indicates the rate of contribution of improvement in mortality in each age group to the lengthening life expectancy in Japan. According to these tables, for example between 1955 and 1960, much contribution has been noted among the young ages under 15 years toward an increase in the life expectancy. There about 65 percent are attributable to mortality reductions in this young age group. On the other hand, however, in recent years, say 1985-1990, the contribution in the improvement in infancy and childhood has become considerably outshined by the contributions in adult and old-age mortality. When the effect of mortality declines among the elderly becomes more pronounced than among the youth population, then, it would be inevitable that the population ageing becomes an important political issue which demographers could not clearly envisage.
