17. Growing numbers of old people

The aging of Western populations during this century was brought about by the great demographic transition in which both fertility and mortality fell sharply. Mortality was reduced most importantly in infancy and childhood but considerably also in youth and middle age with the result that increasing numbers of people lived to retirement age. These swelling numbers ensured that, though mortality in old age declined only slowly if at all, more people reached high ages than before.

The increase in the numbers of oldest-old was, however, quite moderate until a new factor came into play in the post-1950 era: a decisive decline in age-specific mortality even at oldest ages, a development as unprecedented as it was unexpected. Not only were more people reaching ages 65 and 80 but those who did were surviving considerably longer than their predecessors. This caused an acceleration in aging and a situation where the growth rate is the faster, the higher the age. The annual rate of increase of the oldest-old contrasted sharply with that of the population as a whole (absolute numbers given in Annex Table 16)

	Western Europe (12 countries)	Eastern Europe (5 countries)	Japan
Ages 80 and over:
1960 to 1970	2.4%	2.4%	3.7%
1970 to 1980	2.7 %	2.4 %	5.3 %
1980 to 1990	3.7 %	2.3%	6.2 %

Total population:
1960 to 1990	0.6%	0.1%	0.9%

In all three regions, the growth of the old age population has many times as fast as that of the general population leading to a rapid process of extreme aging. In Eastern Europe, the growth was caused mainly by past decline in death rates at younger ages which allowed constantly growing numbers to survive to retirement age and beyond but, lacking the additional impetus of declining old age mortality, the growth did not accelerate. In Western Europe and still more Japan, this new phenomenon led to considerable acceleration.

A wide diversity can be observed between individual countries. This is shown in Table 34 for 28 countries (absolute numbers in Annex Table 15), including some where data quality is not very good but considered sufficient to give an approximate idea of the overall growth. Japan is far ahead of the field with a 30-year growth factor of 4.42 while in most countries it varies between 2 and 3. The development of the growth rates over time reflects trends in population size, mortality decline and past historical events. A major factor to be taken into account are war losses which, once suffered, will mark the cohorts in question until their extinction.

Figure 34, based on data in Annex Table 13, demonstrates how the increase of the male population at specific ages was in recent years affected by casualties in World War I. At the present time and during the next 10 - 12 years, the number of men to reach age 80 is reduced in countries which participated in World War II. In addition, the cohorts of both sexes are affected by temporary fluctuations in the birth rate during crisis periods.

The proportion of oldest-old in total population has naturally increased substantially in the same 30-year period (see Table 35). In the 25 countries, for which data are available for both 1960 and 1990, the proportion of oldest-old has doubled from 1.42 to 3.00 percent and everyone of them had in 1990 a higher percentage of oldest-old than the highest observed in 1960.

The diverging trends and the past history have caused wide-spread changes in the ranking of countries. Nevertheless, France, Norway and Sweden have remained in the group of top five. East and West Germany, both aging but at different speeds, have moved in opposite directions in ranking. Japan and Finland, the last two in 1960, have moved much higher. Countries outside of Europe are among the least aged.

Besides growing rapidly in size, the oldest-old population itself is aging. For the group of thirteen countries as a whole, the following aging indicators in successive age groups describe the development.

The growth factor has been ever higher at successive ages, and always higher for women that men. The proportion of those in the next higher age group has also regularly grown and has been higher for women than men. The mean age of persons 105 years or older grew very slightly and, being based on small numbers, was marginally higher for men.

The age pyramid in absolute numbers in Figure 35 demonstrates the magnitude of the increase in numbers, very substantial, particularly among women, up to ages around 90. The highest ages require a presentation in larger scale (in the upper right corner) to show the explosive growth of the centenarian population.

The following is an assessment of the contribution of the decline in oldest-old mortality after 1960 to the growing numbers of oldest-old in each country. For the purpose, the actual cohorts as they reached age 80 were projected up to 1 January 1990 according to the national life table of 1950-60 (in case of the Netherlands, 1960-70), available in Annex Table 2, and then compared with the actual population of 1 January 1990. Divided by the population on 1 January 1960, this gives two sets of growth factors: projected and actual. The results are given in Annex Table 16 by country, age group and sex and illustrated in Figure 36 with bar charts in which the lower blue part shows the growth factor as projected, i.e. if mortality had remained unchanged at the 1950-60 level. The entire bar shows the actual growth factor, and the red part the contribution of the decline of oldest-old mortality.

In most cases, the existing momentum would have increased the old age population by about 50 percent, or, at most, doubled it. In most countries, furthermore, the lower part of the bar does not increase with age, indicating that the cumulative nature of survival by age was still largely absent in the 1950s. In several countries, shown among the first in Figure 36, the subsequent mortality decline was so slight as to have relatively little effect. In the majority of the countries, however, its effect was decisive. Without it, the oldest-old population would have grown only modestly, and the phenomenon of faster growth at older ages would have been largely absent. As it was, the oldest groups grew in many countries by a factor of 10 or 20.

Most spectacular are the growth factors of the centenarians. VAUPEL and JEUNE (1994) found in a study of Scandinavian cohorts born between 1860 and 1890 that the contribution of mortality decline above age 80 to the increase in the number of centenarians varied as follows between groups of cohorts:

amounting thus to above two thirds of it. The remainder was mostly due to mortality decline at younger ages leaving only a marginal importance to the increase in births.

With our material we cannot trace cohorts from birth and so separate the two last-mentioned components but we can show the contribution of mortality decline above age 80 since 1960. Table 36 shows the results regarding centenarians in 1990 in ten countries with reliable information. The methodology is different and our data apply to all living centenarians instead of those who reached age 100, and consequently, the figures are not strictly comparable. They measure, however, essentially the same thing, and our data corroborate the findings of Vaupel and Jeune that the increase in the number of centenarians is overwhelmingly due to the recent decline in the mortality of the oldest-old: an average of 78 percent with a range from 63 to 88. This decline increased the growth factor from a projected 2.9 to the actual 9.7 and thus the growth portion (where the original population is excluded) from 1.9 to 8.7.

Finally, it needs to be re-emphasized that although the relative growth has been most rapid at the highest ages, the great bulk of the population increase among the oldest-old - 85 to 90 percent - has taken place at ages between 80 and 90 years. The spectacular surge in the number of centenarians is a salient feature of the recent development and as such of great significance but very small in terms of population.

	Growth factor	Percent aged 90 and over
	1960 - 1990	1960	1990

Males	2.3	5.0	7.6
Females	2.9	6.7	11.6

	Growth factor	Percent aged 100 and over
	1960 - 1990	1960	1990

Males	3.4	0.4	1.0
Females	5.0	0.7	1.5

	Growth factor	Percent aged 105 and over
	1960 - 1990	1960	1990

Males	8.6	2.5	3.6
Females	11.1	2.6	4.1

	Growth factor	Mean age
	1960 - 1990	1960	1990

Males	12.4	106.1	106.5
Females	17.2	106.0	106.4

Denmark	65 - 68 percent
Norway	51 - 65 "
Sweden	65 - 75 "