2.3  Trends in OECD averages

This section averages across the entire OECD, to establish the major trends. Cross-country differences are examined in the following section.

Movements in population size and age structure are driven by movements in fertility, mortality, and migration. Figure 1 shows fertility and mortality rates. Fertility is measured using the “total fertility rate”. This is the sum of (single-year) age-specific fertility rates; it has an intuitive interpretation as the number of births the average woman would have over her life time if prevailing age-specific fertility rates were to prevail indefinitely. Average fertility has clearly been falling since the 1960s. Fertility rates during the 1950s and 1960s, the height of the baby boom, were in fact higher than they had been for several decades, except in the recently developed countries (RDCs) and Japan. Even in the RDCs, however, the post-war declines in infant and child mortality meant that “net” fertility was higher than previously. The UN median variant projection shown in Figure 1 assumes that average fertility will rise somewhat in coming decades.

Mortality is measured in Figure 1 by life expectancy at birth. As is apparent in the figure, the UN assumes that gains in life expectancy will be smaller in future decades than they have been in past decades. There are as yet no signs of a slow-down, however, and demographers are increasingly arguing that the UN is underestimating future life expectancies (Vaupel 1997).

Figure 1 – Estimates and projections of fertility and mortality levels for the OECD

Source: Calculated from data from the UN Population Division’s World Population Prospects online database.

Figure 2 – Estimates and projections for the age structure of the OECD

Source: Calculated from data from the UN Population Division’s World Population Prospects online database.

Unfortunately, data on total migration into and out of the OECD as a whole are scarce. The UN’s Population Prospects online database does show, however, that net migration into “developed regions” in 2000 was about 1.8 migrants per thousand population. Migration on this scale is sufficient to materially affect population growth rates, but not age structure.[3]

Past and projected population sizes and growth rates for the whole of the OECD are shown in the bottom row of Table 1. The OECD as a whole grew steadily over the period 1950-2000, but the UN predicts that in-migration and future improvements in mortality will only partly compensate for lower birth rates, so that population growth will slow markedly in the future.

Figure 2 shows trends in age structure. Note that the working-age population is defined here, and throughout the paper, as ages 20-64, and not 15-64, 15+, or 20+ as is often done. Most 15-19 year olds in OECD countries are now in education rather than work, so an increase in the size of this age group is more accurately represented as an increase in the “dependent” population than the “working age” population. An analogous argument applies to people aged 65 and over.

The age groups used here and in most other charts in this paper are very broad, to simplify comparisons. This means, however, that only the most pronounced structural changes are visible.

Source: Calculated from data from the UN Population Division’s World Population Prospects online database.

The proportional shares of the three age groups in Figure 2 show the imprint of the baby boom and subsequent baby bust.[4] The proportional share of the 0-19 age group increased, mainly at the expense of age group 20-64, during the baby boom decades of the 1950s and 1960s. From the 1970s, however, the proportional share of the age group 20-64 increased again as the baby boom cohorts began to turn 20. As long as the baby boom cohorts remain in the working ages, the proportional share of 20-64 age group will remain high. From about 2020, however, when the baby boom cohorts start to reach age 65, the proportional share of the 20-64 age group is projected to fall, and the share of the 65+ age group is projected to rise. The low fertility rates experienced since the 1980s have kept the proportional share of the 0-19 group small.

The population waves induced by the baby boom and bust have been occurring simultaneously with a rise in the proportional share of older people, resulting from continuing life expectancy gains. Given that the UN may have underestimated the potential for continued increases in life expectancy, the proportional share of the aged may well turn out to be higher than Figure 2 suggests.