With tumbling fertility rates, the fraction of children in the population decreases in 2010 and 2050, when compared with historical averages. On average over the past 60 years, children comprised 25-33 percent of developed countries' populations and 41-48 percent of emerging countries' populations (excluding Russia, where the demographic profiles more resemble developed economies, even if their life expectancy does not). In 2010, these fractions range from 18-27 percent in developed countries and from 27-40 percent in emerging countries (again, excepting Russia). In 2050, they decrease further, to an average of 21.6 percent in developed countries, and the same in emerging countries.
Among senior citizens, we observe the opposite effect. As life expectancies have increased over the last century—and will likely continue to increase in the coming decades—the fractions of senior citizens in the populations soared. Their averages from 1950-2010 are roughly 10-15 percent in developed countries and Russia, and a scant 4-5 percent in the remaining emerging countries; by 2010, they ranged between 13-23 percent in the developed countries plus Russia, and between 5-8 percent in emerging countries. Continuing the trend, in 2050, these numbers jump to 21-36 percent and 14-25 percent in developed and emerging countries, respectively. Note that the demographic profiles in emerging countries in 2050 will be very similar to those of the developed world today.
To get a hint of how severely out-of-sample these numbers are, notice in Figure 4 that the fraction of senior citizens in the United States in 2050 will be more than six standard deviations away from its 1950-2010 historical average ([21.2 percent-10.9 percent]/1.7 percent). Other countries will experience a similar magnitude of transition in future years; in some cases, more extreme.
The most interesting effects occur in the working-age populations, which we define as spanning the age group 20-64. We see an increase in most countries' factions of working-age adults when comparing 2010 with both 1950 and the historical averages (1950-2010). By 2050, the trends revert and the fractions start to decline. India and Brazil are the only two exceptions because their fractions continue to rise after 2010, but China and Russia show the same pattern as developed countries. In some cases, these changes from 2010 to 2050 are impressive; the fraction goes from 63 to 53 percent in Canada, from 59 to 46 percent in Japan, from 61 to 48 percent in Italy, and from 61 to 50 percent in Germany.
This erosion in the work force, as a share of the population, is exactly Japan's circumstance today, and for at least the next three decades, regardless of what happens to fertility rates in the interim. Even if Japan's fertility rate were to soar past replacement rates tomorrow, births today are not part of the work force for roughly another 20 years; so the profile of seniors, relative to the working-age population, is more or less preordained for the coming 20 to 40 years. Western Europe faces a like circumstance, albeit a bit less daunting than Japan's.
For those readers who currently think that the demographic head winds that we're already seeing in Japan and Europe won't get much worse, we borrow an expression from vaudeville: "You ain't seen nothin' yet." At some future date, however, births must approximately match deaths.13 This stage is likely to be reached—globally—around the middle of this century. Especially for countries that experience fertility rates below the replacement level of 2.1 children per woman, Phase III will be a daunting transition to the new steady state.