Our online handbook was written between 2008 and 2011 with the support of the Robert Bosch Foundation. This article reflects the state of knowledge from 2010.
by Frank T. Denton and Byron G. Spencer
National populations increase when the inflows of births and immigrants exceed the outflows of deaths and emigrants, and they decrease when the reverse is true. Substantial changes in any one of those components can have profound effects over the longer term on both the size of a population and its age structure. In the case of Canada the large increase in the number of births associated with the post-World War II baby boom, and the subsequent decrease associated with the baby bust, stand out; so too does the level of immigration, which has been very high by international standards.
The total fertility rate, the number of births per woman over the childbearing years, declined steadily throughout the early decades of the 20th Century, reaching a low of about 2.5 in the late 1930s. It increased sharply in the late 1940s and climbed to nearly four births per woman by the end of the 1950s.
The "baby boom", as it came to be called, lasted for about two decades but by the mid-1960s it was over and the "baby bust" started to set in. Fertility fell below the natural replacement rate of about 2.1 in the early 1970s and has remained well below that level ever since. It was as low as 1.5 early in this century and is now around 1.7.
At its highest (in 1959) the total fertility rate in Canada was 3.9 births per woman, somewhat above the peaks attained in the US and Australia during their baby booms, and far higher than in most countries of Western Europe. Furthermore, the decline from the peak was especially sharp in Canada; the 2008 rate of 1.6 was below that in the US (2.1), Australia (2.0), and much of Europe.
While there has been a great deal of year-to-year fluctuation in the level of immigration, the total number of immigrants has averaged close to 250 thousand per year for the last two decades and has added more than one-half per cent to the population in all but eleven of the 64 years shown in the figure below. By contrast, the rate of emigration has been relatively low and stable. In consequence the proportion of the population that is foreign-born is about 20 per cent today, one of the highest in the world. In what follows we review the growth of the population and the labour force since the middle of the last century and consider what the future might hold to the middle of this one.
The table shows the population at selected intervals from 1951 to 2051; the figures from 2011 are projected. The population today is 2.4 times larger than it was in 1951. Growth was especially rapid before 1966, when the baby boom came to an end. In fact, the population grew almost twice as fast in the 15 years ending in 1966 as it did in the next 15, and the trend has been towards generally slower growth since then.
The growth in the population can be thought of in terms of its components: natural increase (births minus deaths) plus net in-migration (immigration minus emigration). The rapid growth in the 1950s and 1960s resulted mostly from natural increase; the large number of births associated with the baby boom exceeded by a wide margin the smaller number of deaths. In consequence the population was very young. While natural increase accounted for three-quarters of the total increase before 1966, net in-migration has become an increasingly important component of population growth since then.
The average age of the population has been rising steadily since the baby boom ended. Even the very high levels of immigration of recent decades have had only a small impact on the age structure; instead the dominant factor has been the aging of the baby boom generation.
In 1966, when the boomers were young, 42 per cent of the population was under 20 and the median age was 25; that proportion fell to only 23 per cent by 2011, by which date the median age had increased to 40. As the proportion of young people declined the proportion old increased, from 7.6 per cent in 1966 to 14 per cent in 2011, using 65 as the standard marker of "old". Since the decline in the proportion young more than offset the increase in the proportion old, there was growth of the group in the middle. That is reflected in the total dependency ratio, the ratio of the total population to the population 20 to 64 (the "providers"). The ratio peaked at about 2.0 when the baby boom was young and fell sharply as that generation passed the age of 20. It has continued to decrease since, thereby providing the demographic dividend that Canada and other countries experienced by having an unusually large proportion of the population of working age.
Looking to the future, all those born during the baby boom will have passed their 65th birthday by 2031, and by 2051 the few survivors of that generation will be in advanced old age. The projected figures are based on assumptions about the future rates of fertility, mortality, immigration, and emigration. Our assumptions are that the total fertility rate will remain at about 1.7 births per woman, that mortality rates will continue to fall as they have in recent decades (with the result that life expectancy at birth will increase by about 5.8 years for males and 5.2 for females), that immigration will remain at 250,000 per year and that 0.14 per cent of the population will emigrate each year. The result is continued growth in the population: a 27 per cent increase by 2051 compared with 2011. However, the growth occurs at progressively slower rates after 2011, and by the last decade it is only 0.3 per cent per year. Associated with slower growth is continuous aging of the population; by 2051 the median age increases to 46. Again using 65 as the standard age marker, the old will outnumber the young within 20 years, by which time they will account for almost a quarter of the total population.
The projected number of births remains quite stable after 2011, at about 400 thousand per year. The sheer magnitude of the baby boom in Canada is evidenced by the fact that even though the population projected for 2051 is more than twice as large as it was in 1966, the projected number of births is smaller. Net in-migration remains nearly constant. What does increase is the number of deaths: the rapid aging of the baby boom generation results in twice as many deaths in 2051 as in 2011, even though the population grows by only one-quarter and mortality rates continue to fall.
Since the overall population is growing more rapidly than the "provider" component there is a marked increase in the dependency ratio; from its record low level of 1.59 in 2011, it increases to 1.81 by 2031, and 1.87 by 2051. However, even with those increases the ratios remain lower than 1.98, the value reached in 1966 when the baby boomers were young.
The labour force has grown much more rapidly than the population: it was 3.5 times larger in 2011 than in 1951, as compared to the 2.4-fold increase in the size of the population. That was not only because the population of working age grew relatively rapidly but also because labour force participation increased.
The annual rate of growth of the labour force peaked in the 1970s, about two decades after the peak in population growth, as one would expect. By 1981 27 per cent of the labour force was under 25, but the proportion declined to 17 per cent 20 years later as the boom generation aged. The decline was accentuated by the increased enrolment in higher education and the associated later age of labour force entry.
Increases in the participation rates of women were an especially important source of labour force growth. There was a 4-fold gain in the rates for women 25 to 54 between 1951 and 2011, and a 3-fold increase for women 55-64. That far more than offset the reduction in male (and especially older male) participation rates that continued until the mid-1990s. In consequence, the proportion of women in the labour force doubled, from 23 per cent in 1951 to 47 per cent today.
The participation rates of women under 50 have largely stabilized in recent years. However, those of older women have continued to rise and, since the mid-1990s, the decline in participation rates of older men has been reversed. The increases in both the male and female rates for those in late middle age combined with the aging of the baby boom resulted in a 70 per cent gain in the proportion of the labour force 55 and older between 2001 and 2011.
An alternative measure of dependency is the ratio of total population to labour force, rather than to the 20-64 population. Both ratios were highest when the baby boom was young, and both are currently at all-time lows. The decline in the labour-force-based measure is greater because of the increase in participation rates.
The labour force projection assumes that recent trends in participation rates will end by 2021, after which all rates will remain constant. In consequence the labour force grows by only 12 per cent by 2051, or less than half as fast as the population. Thus the labour-force-based dependency ratio rises. It is projected to be about 10 per cent higher by 2031, when the boom generation is in the dependent group, than it was in 2011. Whether that increase is a matter for concern is hotly debated, but we observe that this measure of dependency, as well as the previous one, is projected to be much lower when the baby boomers are in old age than it was when they were in their youth.
While the increase in the total dependency ratio, however measured, is relatively modest, the increase in the old dependency ratio (shown in table "The Labour Force of Canada, Selected Years, 1951-2051" as Pop65+/LF), is much larger and attracts more attention. In the mid-1960s, when the boomers were young, the population that was old was one-fifth as large as the labour force. It is now one-quarter as large, and is projected to be more than half as large within two decades, as the population continues to age. That naturally gives rise to questions about the sustainability of the publicly financed health care and retirement income systems that are now in place.
While population aging is inevitable, a reality that must be faced, the concern related to the anticipated impact on government budgets is often greatly exaggerated. Public expenditures on retirement pensions and on health care for those 65 and older currently account for about one-fifth of all government expenditure in Canada. Thus one-fifth of total expenditure is directed specifically to the one-seventh of the population that is older. If those expenditures varied only with the size of the older population and all other expenditures varied only with the size of the rest of the population, overall government expenditures would increase by 35 per cent between 2011 and 2051, or by only a little more than the 27 per cent increase in the population. Put in those terms, the growth in the tax burden associated with population aging seems manageable. We observe also that the growth would be smaller if participation rates were to increase over time, and retirement were to occur at somewhat older ages. Finally, we note that even a modest increase in productivity growth would fully offset any of the negative consequences that population aging would have on the average standard of living. (For a further discussion of these points, see Denton and Spencer, 2000.) With such considerations in mind, in what follows we focus on overall measures of dependency.
We have reported two measures of overall dependency, both based on the ratio of total population to the "provider" population, with providers, in turn, defined either by age (20 to 64) or by being in the labour force. Here we give further consideration to those two measures and provide five alternative projections to assess the likely future range of values. Each of the alternative projections is designed to assess the consequence of more rapid population growth.
The projections are as follows: A, the standard projection discussed above; B, high immigration (300 thousand per year by 2010, and remains at that level; in the standard projection it remains at 250 thousand); C, high fertility (the total fertility rate increases from 1.7 to the replacement rate of 2.1 by 2016, and remains at that level; in the standard projection it remains at 1.7); D, longer life expectancy (mortality rates at each age continue to fall such that life expectancy in 2051 is 1.7 years higher for males and 1.8 for females than in the standard projection); E, more rapid growth, combines B, C, and D. The same rates of labour force participation are assumed in all projections. Note that each of the alternatives is designed to have a relatively large impact on the size of the population, and hence the potential to affect the dependency measures. Our purpose is to assess how great the impacts would be.
The population-based measure peaked in 1966, just as the leading edge of the baby boom was approaching its 20th birthday. The labour-force-based measure peaked five years earlier because the participation rates of women were already increasing sharply by that time. Both ratios fell until 2006, the only exception being a very modest increase between 1991 and 1996. Both ratios continued to decrease until 2011, when they will be at an all-time low, the first one-fifth below its peak and the other one-third. Both ratios increased through the 1950s and into the 1960s.
All five projections show increases in both dependency measures. Most of the increase takes place by 2031, by which time all those born during the baby boom will have passed their 65th birthdays and, unless current practices change, the vast majority will have retired. With the standard projection the dependency ratio will then be ten or 14 per cent higher by 2031, depending on the measure used, but both will remain well below the levels of the baby boom period. Greater life expectancy (projection D) or even higher immigration (B) would have little impact on the ratios. A return to higher fertility would have a greater impact, but even then the ratios would remain below historical peak levels.
Both the population and labour force will continue to grow. However, slower growth of the population and even slower growth of the labour force are in prospect for the foreseeable future, and both will become older, on average. Furthermore, all growth will depend on continued high levels of immigration; without that both the population and labour force would soon start to decline and both would age even more rapidly.
Although considerable aging of the population is in prospect, we foresee only relatively modest increases in overall dependency ratios. The ratios will become higher as the baby boomers move into old age but they will not reach the levels observed when the baby boom generation was young. While any increase in the population-toprovider ratio suggests a possible decline in the average standard of living, we have suggested that the pressures in that direction are manageable and, in any event, could be offset by increased labour force participation, later retirement, and modest gains in productivity.
Denton, Frank T. / Spencer, Byron G. (2000): Population Aging and Its Economic Costs: A Survey of the Issues and Evidence. In: Canadian Journal on Aging, Vol. 19, Supplement 1. Kingston.
Denton, Frank T. / Spencer, Byron G. (2009): Population Aging, Older Workers, and Canada’s Labour Force. In: Canadian Public Policy, Vol. 35, No. 4. Montréal, Québec.
United Nations Population Division [ed.] (2009): World Population Prospects: The 2008 Revision. http://data.worldbank.org.
last update: December 2010
The future of global migration
How falling fertility rates accelerate development
Will Ethiopia Become a Model for an African Upswing?
What African Agriculture Needs to Achieve
in the Online Handbook Demography