Ageing And Social Security Program Reforms: Canada in International Perspective

Ageing And Social Security Program Reforms: Canada in International Perspective

Jacques Légaré [ * ]



Indexes of ageing for Canada and for UN Economic Commission for Europe countries are estimated and presented. All indexes of population ageing will rise sharply until the middle of the 21st century. This situation will not lead inexorably to apocalyptic consequences for social security systems, but the basis of our social security programs, as well as their management, will need to be seriously modified to handle the real implications of these demographic trends.

In the century between the end of World War II and the year 2050, the total population of the U.N. Economic Commission for Europe (UNECE) countries[ 1 ] studied here is expected to multiply by 1.5, growing from 741,165 million to 1,129,733 million. During the same period, however, the number of people aged 65 years and over will increase three times as quickly, multiplying 4.7 times (Figure 1) and almost 10 times in Canada; inevitably, these changes raise important issues for policy makers.


The ageing of the total population

For our analysis, we group UNECE countries into two broad categories: those with market economies and those with economies in transition.[ 2 ] Both groups will see their crude numbers of elderly people almost double between 2000 and 2050. In regions with market economies, these trends seem to level off and even decrease in some cases, while for those with economies in transition, the proportion of elderly people continues its upward trend, but at a slower pace. Regions that at the present time have younger populations, such as North America and Southern Europe, will see their elderly population more than double— by a factor of 2.5 in Canada. In Western and Northern Europe, as well as the Commonwealth of Independent States (CIS) and Baltic States, numbers will only multiply by about 1.5.

These trends for absolute numbers are reasonably robust as they are based on people who have already been born— those who will be 65 and over in 2050 were born before 1985; variations in mortality and immigration would have only a minor impact on crude numbers. When we consider the percentage of those aged 65 and over in the total population, however, the long-term projections are more risky, as they have to take into account the fertility behaviour of couples who have not yet been born![ 3 ]

Nevertheless, assuming that the U.N. projections are accurate, we estimate that the percentage of those aged 65 and above will also double, moving from about 13 percent now to 14 percent in 2010, 20.5 percent in 2030 and to nearly 25 percent in 2050[ 4 ] (Figure 2). This overall picture obscures the rapid rate of ageing in Europe. Already by 2010, in some countries (Germany, Greece, Italy and Sweden), one fifth of their predicted total population will be aged 65 and over. By 2030, this proportion is expected to reach 25 percent, by which time they will have been joined by Austria, Belgium, the Czech Republic, Finland, Netherlands, Slovenia, Spain and Switzerland. Finally, in the absence of drastic changes in demographic behaviour, more than one third of the population will be 65 and over by the year 2050 in the Czech Republic, Greece, Italy and Spain. During the same period, the elderly population is only expected to reach one fifth of the total population of North America—23.8 percent in Canada. In an era of globalization, such sharp differences in population age structures could have important political, social and economic repercussions.


The oldest old: the most rapidly increasing group

Challenges facing ageing societies cannot be properly studied when old age is taken to begin at 65; old age in the past as well as in the future should be linked to dependency, both in terms of health and of income security. There is no doubt that it is almost impossible to fix a threshold corresponding to dependency. Since the majority of people over 65 are not dependent in most respects, only people aged 80 years and over will be classified here as the oldest old. This group will number nearly one billion in 2050— the fastest growing portion of the elderly population in UNECE countries, for those with market economies as well as for those with economies in transition. For both groups of countries, the oldest old will multiply threefold in the first half of the 21st century, fourfold in Canada, where the population evolved differently as we have already seen. Germany and the CIS, for example, are expecting to have about eight mil-lion people over age 80 in year 2050: this is more than the total population of many UNECE countries!

Of course, this is a long-term outcome, reaching a peak when all the cohorts of the Baby Boom are over 65 and the Baby Bust generations are joining their ranks. When we look at percentages of those aged 80 years and over in the total population, however, the situation is completely different for the two groups of countries. Market economies are expected to have significantly higher proportions of their populations over 80 by the year 2050 than are economies in transition, between 8 percent and 11 percent, as compared with 6 percent and 8 percent (Figure 3). Nevertheless, these calculations are based on predictions of future mortality that may be highly unreliable in countries with economies in transition.[ 5 ]


In 2050, over one tenth of the population of Western Europe as a whole will be aged 80 years and over, a percentage equivalent to that for people aged 65 and over in most UNECE countries very recently.

The elderly: a women’s world

One of the features of the 20th-century mortality decline in the Western world has been the increasing gap between male and female mortality, as mortality has declined more rapidly for women than men. A second important element is that far more men than women died during the two World Wars. Added to these direct mortality effects, is the age difference between spouses; whatever the reason, wives tend to be younger than their husbands. All these factors contribute to the fact that women make up the majority of the elderly population, a situation that seems likely to continue.

If we calculate the proportion of women in the total population (Figure 4), we observe that we are living through a transitional period, with the proportion of elderly women of about 58 percent in 1950 rising over the period and returning by 2050 to a level equivalent to that of the 1950s. In the interval, especially in countries with economies in transition, not to mention the CIS and the Baltic States, the last 50 years have been characterized by important disparities with more than 70 percent of elderly people being female by the 1990s — and most of them not married. Even if the situation is now improving a little, short-term economic and social issues will still have to be faced and adjustments made.


Furthermore, the gender imbalance increases with age, reaching a peak for the age group 90 and over. The majority of women belonging to these older cohorts will have spent little time, if any, in the labour force, a situation potentially creating many problems of income security. Now, and in the year 2050, nearly 75 percent of the oldest old will be women, these percentages being well over 80 percent in countries with economies in transition.

The ageing of the working age population

For the second half of the 20th century, the median age of the working age population hovered around 39, reaching 40 by the end of the century for UNECE countries as a whole (Figure 5). In the near future, this figure will rise rapidly in all regions, levelling off by 2030 in market economies but not in economies in transition. This process will bring the overall median age in year 2050 to about 42 for Canada and 43 for the UNECE countries. This is very close to 42.5 years, the mid-point age of the large age group 20 to 65, a situation that would be observed in a stationary population with all birth cohorts of about the same size and very few deaths before old age.




Ageing indexes

The ageing index most commonly used by demographers and others is the dependency ratio, relating dependants — the population under age 20 and over age 65 (occasionally over age 60) — to the population of working ages. Sometimes, the dependency ratios of the elderly population are compared to the dependency ratios of youngsters. As people intuitively think in terms of the economic burden, such an index can be highly misleading when applied to the older population. There may be an economic cost for supporting retired people,[ 6 ] but it is important to remember that most elderly people are taxpayers and have financial assets, so that, unlike children, they are not necessarily dependent. Even though nowadays we more often use the term “demographic dependency” to avoid this confusion, these indexes are still not very appropriate indicators of change in population age structure. Accordingly, we will now present alternative ageing indices.

To illustrate how the age “pyramid” is becoming more and more an age “cylinder,” we compare the population aged 65 years and over to the population under age 20. During the 1950s there were some 23 elderly for 100 young people; the figure is now closer to 50 — slightly more in countries with market economies, and less in those with economies in transition — and in 2050 it is expected to reach about 110 elderly for 100 youngsters, even in Canada. We will have a cylinder with a top on it.

As mentioned earlier, dependent elderly people are chiefly to be found among those aged 80 years and over, especially in terms of health. The weight of this group in the elderly population as a whole has been increasing for the last 50 years and will accelerate after 2025, mainly in Canada and in the other countries with market economies, when the first Baby Boomers reach that threshold (Figure 6).


The relation of the population aged 80 years and over to those in the age group 50 to 64 years (Figure 7) is probably a more meaningful indicator, as the children of the oldest old belong mainly to this age group. Many policies currently designed to face the problems of population ageing imply a greater reliance on the family than on public services, and some involvement in dealing with their dependent parents is expected from the members of this age group. In 1950, there were only 8 older parents per 100 mature children, and many of the latter still had children at home. Between 1980 and 2020, 100 mature children, mostly at the empty nest phase of the life cycle, may have to take care of 20 elderly parents. By 2050, the number will have more than doubled in countries with market economies and increased by more than 50 percent in countries with economies in transition. In Canada, there would be one oldest-old person for every two mature children. Policies that tend to shift burden from the public sector to the family deserve close scrutiny especially when we know that it is mainly women who will be affected by this shift.


Finally, to enable an understanding of changes in the age structure of the working age population, our ageing index relates the numbers of those in pre-retirement ages (aged 50 to 64 years) to those who are entering the labour force (aged 20 to 34 years) (Figure 8). The graph of this index largely replicates the trends in the mean age of the working-age population (see Figure 5), but this time in a more meaningful way. Presently, there are about 6 older persons for 10 younger persons available for the labour market, but by 2050 it will be one for one in Canada and even more than one for one in UNECE countries as a whole. This movement reflects a pattern observed with an earlier indicator— the upward trend (from 39 to 43 years) of the median age of the working-age population. Of course these are mean values: for some countries or regions the situation envisaged could be catastrophic. For example, the Baltic States could see this index rise to 135 by the year 2050; human resources development and planning could become a nightmare in such a situation.


The impact of demographic determinants of ageing and characteristics of older populations on social security programs

The demographic ageing process we have just outlined was, initially, totally linked to reductions in fertility, not mortality; indeed, in the past, it is mostly infants and children whose numbers increased because of lower death rates. When most premature deaths have been eliminated, mortality rates start diminishing at older ages and have a major impact on the age structure and ageing. And, even if immigration flows slow the demographics of the ageing process, as is clearly illustrated here, for Canada and for North America as a whole, they cannot stop it and only provide a salve to the wound.[ 7 ]

Fertility and PAYG pensions and health schemes

When public social security programs in industrialized countries were subjected to radical restructuring after World War II, Pay-As-You-Go (PAYG) schemes were generally adopted, reflecting assumptions that strong economic and demographic growth would remain the norm. However, economic growth stalled in the early 1970s, and cohort fertility (or family size) has been considerably below replacement level in Canada for more than 20 cohorts, as it has been in most UNECE countries.

PAYG schemes imply that those who pay taxes — mainly workers — in any period assume the costs of health and pension schemes of the same period. These workers tend to be the children of the seniors. As there are — and will continue to be — growing numbers of beneficiaries and declining numbers of workers, the share of total social security benefits coming from PAYG schemes should decrease accordingly if we want to secure solidarity between generations.[ 8 ]

This change should occur at the pace that cohort fertility indexes decline (not period fertility indexes); these rates have been well documented for the past and are relatively easy to predict for the future.

Indeed, when family size—i.e., the cohort total fertility rate— drops by 50 percent over a certain period of time, the proportion of social security benefits coming from PAYG schemes should also drop by 50 percent, if these benefits remain constant. This means that part of the benefits will have to come from other sources, mainly from the beneficiaries themselves, as it is they, and not society, who have decided to have fewer children.[ 9 ]Some may protest that society is then breaking the social contract, but they need to remember that they themselves broke the contract by not having the expected number of children—that is, a sufficient number to ensure replacement levels. It is not our role to assess the merits of such a behaviour, but societies— and seniors—should proceed in an ethical way; this means not putting too heavy a burden on the active population whose numbers are declining as a result of their own parents’ fertility behaviour.

Mortality and pension schemes

As mentioned earlier, current contractions in mortality rates will have an important impact on the number of older persons and on their proportion in the population. Crude numbers can only be on the increase. If we want to keep a fair share between population subgroups in PAYG schemes, we have to keep the ratios stable between workers and retirees and raise the age of access to public social security programs — an age that could be different from the age of retirement from the workplace.

In regards to funded pension schemes, benefits will either have to be reduced or contributions increased to cope with a longer cohort life expectancy; these additional expected years of life, rather than the death rate, should guide decision making, as the death rate will increase in the future due to ageing even if mortality still abates. For example, relatively conservative mortality scenarios project that life expectancy at age 65 in Canada will move from 17 years for the 1901 birth cohort to 20 years for the 1941 birth cohort.[ 10 ] Meanwhile the crude death rate is expected to rise from 7.5 per thousand in 2000 to 12.8 per thousand in 2050. Encouraging early retirement in recent years has only worsened the problem. Even if early retirement is usually penalised actuarially, departure lump sum payments — costly for governments and firms—keep early retirement attractive. Unless very favourable economic conditions prevail, early retirement will have to be seriously questioned, from the point of view of society.

Morbidity and health status

While it is clear that mortality rates have improved considerably, especially at older ages, the question remains as to whether the elderly population is in better health. To ascertain more fully the health status of a population, epidemiologists and demographers have developed morbidity indexes to measure what is usually termed “healthy life expectancy.” For example, the “disability-free life expectancy” is estimated by subtracting those years lived with a disability from “all health status combined” life expectancy. Even if all problems with definitions and data related to this measure have not yet been solved, breakthroughs have been achieved in the last 10 years, due to an international network of researchers — REVES (Réseau Espérance de Vie En Santé).[ 11 ]

The links between mortality and morbidity are not self-evident, and researchers in the field champion one of three possible scenarios.

  • The first one predicts pandemic disabilities at older ages, which means that life expectancy would increase more rapidly than healthy life expectancy.
  • At the opposite extreme, a second scenario forecasts a faster decline in morbidity than in mortality, resulting in fewer years in bad health.
  • Finally, a third scenario predicts what is called a “dynamic equilibrium” between mortality and morbidity.

The experience of the last 30 years in industrialized countries (Canada and many UNECE countries, Japan, Australia and New Zealand) supports the first scenario (Figure 9a): disability-free life expectancy has stagnated, if not worsened, while life expectancy improved greatly, a situation which reflects the worst scenario, namely pandemic disabilities. However, if only severe disabilities are taken into account (Figure 9b), both indices improve. Only trends for women at age 65 are displayed here, but the pattern is similar for men at age 65, and for both women and men at birth. However, some progress has been observed for the most recent past in Canada.[ 12 ]



What about the future? In our opinion, as a society we should not passively accept any of these scenarios, but we should reform the health system so as to make sure that we gain in quality rather than in quantity of life.[ 13 ] In ageing societies, in order to improve quality of life, future health investments should be directed principally at improving home services and fighting chronic diseases rather than fatal diseases, such as cancer and AIDS.

Mortality, morbidity and health costs

If we exclude the costs incurred at birth, it is now acknowledged that the older the person, the higher the per capita health costs. These costs have escalated exponentially in recent decades, at the same time as the ageing process has accelerated. For many people, the link between these two phenomena is evident. However, in-depth studies have clearly demonstrated that, up to now, ageing—or demography—has very little to do with the explosion of health costs:[ 14 ] paradoxically, supply— specifically in terms of technology — was much greater than demand, particularly for the elderly.

The sharp increase in projected numbers and proportions of elderly people raises critical questions about the pace of rising health costs, even if the per capita and per age costs are held constant at their present level. What will be the impact of people living longer on future mortality reductions in terms of ageing? How, by diminishing morbidity levels, can we challenge the spectre of pandemic disabilities at older ages, too plausible a consequence of reduced mortality and increased ageing?[ 15 ]

The fact that health expenditure apparently increases with age is mainly related to costs incurred at the time of dying, whatever the age at which death occurs.[ 16 ]Accordingly, costs for younger elderly persons could eventually decrease, but only if the reduced numbers of deaths resulting from lower mortality is not offset by increased numbers of seniors. To better predict future health costs, therefore, we should put more emphasis on the future numbers of deaths among the elderly than on the increase in the number of elderly themselves. Whatever the mortality scenario used, the former increase is much slower than the latter, thus presenting a less catastrophic, though still relatively serious, situation in terms of future health costs than is usually presented. In Austria, for instance, the numbers of deaths will increase twofold between now and 2030, while the proportion of elderly will increase by five.[ 17 ]

All these considerations apply mainly to health cure (medical and hospital services) costs. For long-term health care (mainly social services) costs, only important gains in terms of morbidity could affect the impact of ageing societies. This has important public health policy implications both for future fundamental and applied bio-medical research, as well as for ethical attitudes when facing death.[ 18 ]

Sex ratios and pension schemes linked to previous labour force participation

When discussing about present and future income security for elderly people, most studies use existing information on pension schemes that are based on the male situation. However, as previously shown, the ageing society of the future will be a largely female world, especially at the oldest ages. To extrapolate the situation more realistically, efforts should be made to understand the real impact that these private and public pension schemes will have on a retirement income based only on previous labour force participation.

There is no doubt proportions of women in the labour force have increased from older to younger cohorts at a great pace; in the early 1980s it was expected that such a change would eventually solve the expected problems of PAYG pension schemes and ageing in general.[ 19 ] However, observed participation rates for Canadian female cohorts have now levelled off during middle age at about 80 percent for almost all these ageing female cohorts (Figure 10). Furthermore, these participation rates veil important differences between men and women in terms of real income, partly because women are paid less and work more often part time. Women’s estimated future income from pension schemes linked to previous labour force participation will thus be proportionally lower.


Accordingly, in terms of equity, basic pension schemes that mostly have a PAYG format should be indexed to take into account the different levels of involvement in paid work. This kind of flexibility will also soon have to be applied to men who increasingly follow unconventional career paths.

Mortality, marital status and support for elderly people

In ageing societies, elderly people are living increasingly autonomous lives. Nowadays, elderly people do not wish to be dependent on their children or their community, at least as long as they are healthy; and even as their health deteriorates, their spouse is generally the best source of support. Thus, the joint survivorship of husband and wife should be optimized. However, with the discrepancy in life expectancies of men and women and with divorce on the increase, recent trends in mortality make this optimum difficult to achieve. The problem of divorce is mitigated by the fact that people continue to prefer living with a partner and often enter a new conjugal union following the breakdown of an earlier one. We should therefore focus our efforts on reducing the gap between male and female mortality (or further reducing it — the trend is already downward for Canada[ 20 ]) by trying to bring male mortality rates at least to the level of female mortality rates, especially in countries with economies in transition.

This could achieve two things: First, it would postpone institutionalization for those with impaired health by encouraging living arrangements in private households. Second, it would minimize some of the drawbacks described in the previous section regarding income security for older women.


Even though the transition toward an ageing society will be less smooth than it might have been had it followed the course predicted by the demographic transition theory, it would be incorrect to talk of an ageing crisis.

Reforms to social security programs should be based on in-depth studies of successive cohorts, as the cohorts entering old age differ widely, both in numbers and characteristics. In Canada, as in many other countries, the steady 20th-century decline in the birth cohort size was interrupted by the Baby Boom, when the number of babies almost doubled. The Baby Bust simply brought the birth cohort size back to its secular trend. Nevertheless, these changes in numbers have implications that must be faced. The fact that these successive cohorts have very different characteristics is also crucial. Older persons of tomorrow will differ from those of today;[ 21 ] women in particular will be far more educated. We can, therefore, expect future older women to be a great deal more autonomous, as they have been throughout their life.

In an era of globalization, however, can we expect industrialized countries to remain indifferent to the fact that some will have growing populations while others will start to decline? Certainly, all these societies will age, but at very different speeds. If we recognize that any major change in fertility behaviour is unrealistic, it is clear that some countries may have to rely on immigration policies that will at least reduce the pace of ageing, while not preventing it. Immigrants can pay the contributions that would have been paid by the babies who were never born.

Some consider the post-war social contract to be of great worth and beyond question. Will the reforms envisaged in our social security programs really threaten it? From a demographic point of view, a mixture of PAYG and funded schemes may prove to be a solution that may save social cohesion against individualism.

For this to be the case, it may be a good strategy to rely on generational accounting, even if it has to be based on a very long-term longitudinal perspective. Any reform should try to measure the implications of its new rules for inter-generational equity. We have certainly benefited from the gender revolution during the second half of the 20th-century, but we must definitively avoid an old-age crisis as we move into the 21st-century.



1a. International trade is of extreme importance to keeping the world economy humming, and all available sea routes are only going to see an increase in traffic. As such an extremely important commodity is metal, and specifically stainless steel. The U.S. General Services Administration lists some uses, problems, and problems. As competition increases and the United States once again becomes a bigger player in the world steel market, shipping metals in a timely manner is vital. One example is stainless steel sheet metal for commercial and residential kitchen backsplashes from Commerce Metals. This enormous market alone uses 10’s of thousands of tons of stainless each year. You can also read more about the invention of stainless steel at Worcester Polytechnic Institute.

*   Jacques Légaré is Emeritus Professor of Demography, Université de Montréal. The author would like to thank Alexandre Genest and Denis Ouellet for their invaluable technical assistance with the data set. This article is part of a larger project on the Demographics of ageing in relation to the social security programs reforms in the UNECE countries. The author would like to acknowledge, with gratitude, the financial support of the Economic Analysis Division of the United Nations Economic Commission for Europe and the Social Sciences and Humanities Research Council of Canada

 1.  UNECE countries, listed in the Annex, refer to member states of the United Nations Economic Commission for Europe, mainly Europe, Canada and United States.

 2.  Contrary to a previous analysis (J. Légaré, The Demographics of Ageing in relation to Social Security Programs Reforms in UNECE countries (Background Paper for the United Nations Economic Commission for Europe Spring Seminar—The Economic Implications of Population Ageing in the ECE Region, 1999)), in this study, countries from the Asian part of former USSR (Eurasia) have been excluded from the UNECE total and from the total for countries in transition. These countries have completely atypical demographic standards when compared to “European” standards.

 3.  Indeed, when we calculate percentages within the total population, we include, for example, children under 15. In 2050, these are the babies born between 2035 and 2050. For the majority of them, their parents are not even born!

 4.  United Nations, World Population Prospects: the 1998 Revision (New York: United Nations, 1999).

 5.  In some of these countries, mortality is on the rise, especially for men of working ages. This has very important consequences especially for the future living standards of many elderly widows. Improving men’s health may prove to be the best investment for preventing poverty among older women.

 6.  International Social Security Association, Restructuring Public Pension Programs (Geneva, 1998).

 7.  M.V. George, F. Nault and A. Romaniuc, “Effects of fertility and international migration on the changing age composition in Canada” in Statistical Journal of the United Nations Economic Commission for Europe, Vol. 8, no.1, (1991), pp. 13–24.

 8.  It is evident that paying for older parents’ health and pension costs is not the same in the case of a couple with 4 children — portraying a situation of the past — as for couples with only 1 child to care for the two of them, which is closer to future reality. We cannot say any longer “everything being equal”!

 9.  H. Le Bras, “Faut-il faire des enfants ou des économies?” in Population âgées et révolution grise: les hommes et les sociétés face à leurs vieillissements (Louvain-la-Neuve: Éditions CIACO, 1990), pp. 791–799.

 10.  R. Bourbeau, J. Légaré and V. Émond, New birth cohort life table for Canada and Quebec, 1801–1991 (Ottawa, Statistics Canada, Demographic Document no 3, 1997).

 11.  J.-M. Robine, I., Romieu, Healthy active ageing: health expectancies at age 65 in the different parts of the world (Montpellier, REVES/INSERM, papers no 318, 1998).

 12.  L. Martel, and A. Bélanger, “An Analysis of the Change in Dependence-Free Life Expectancy in Canada between 1986 and 1996” in Report of the Demographic Situation in Canada, 1998–1999 (Ottawa: Statistics Canada, 1999), pp. 164–186.

 13.  J. Légaré and Y. Carrière, 1999, “Dying healthy or living longer: a society’s choice” in J.-M. Robine et al. (eds.), The paradoxes of longevity (Berlin- Heidelberg- New York: Springer, 1999), pp. 123–134).

 14.  M.L. Barer, R.G. Evans and C. Hertzman “Avalanche or Glacier? Health Care and the Demographic Rhetoric” in Canadian Journal on Aging/La Revue canadienne du vieillissement, Vol. 14, no 2, (1995), pp. 193–224. J.-M., Hourriez, “La consommation médicale à l’horizon 2010” in Économie et Statistique, no 265, (1993), pp. 17–30.

 15.  J. Légaré, M. Rochon and Y. Carrière, 2000, “Évolutions possibles de la mortalité: impacts sur les coûts de santé” in C. Wattelar et J. Duchêne (eds), Le défi de l’incertitude: nouvelles approches en perspectives et prospectives démographiques (Louvain-la-Neuve: Academia-Bruylant et l’Harmattan, 2000), pp. 147–156.

 16.  P. Zweifel, S. Felder and M. Meiers, 1999, “Ageing of Population and Health Care Expenditures: A Red Herring?” in Health Economics, Vol. 8, no. 6 (1999), pp. 485–498

 17.  C. Prinz, Report of the Seminar on Revisiting the Ageing Process, Montreal, Canada, 2–4 October 1996 (IUSSP Newsletter, Vol. 57 (1997), pp. 24–32).

 18.  Op.cit., note 13.

 19.  I.P. Fellegi, “Can we afford an aging society?” in Canadian Economic Observer, Vol.1, no 10, (1988), pp. 4.1–4.34.

 20.  F. Nault, “Narrowing Mortality Gaps, 1978–1995” in Health Reports (Ottawa: Statistics Canada, Catalogue 82–003 (1997), pp.35–41).

 21.  N. Marcil-Gratton and J. Légaré, “Being Old Today and Tomorrow: A Different Proposition,” in Canadian Studies in Population, Vol. 14, no. 2 (1987), pp. 237–241.