Sustainable Development Banner

April 2001

Population ageing in developing societies:
How urgent are the issues?

By Alain Marcoux
Population Programme Service (SDWP)
FAO Women and Population Division

Population ageing - the increase in the proportion of "older people", i.e. persons aged 60 and over, in the population - has become a prominent topic for studies on the implications of demographic change. Ageing has emerged as a global phenomenon in the wake of the now virtually universal decline in fertility and, to a lesser extent, of increases in life expectancy. The theme is of immediate concern in developed countries, where ageing is already well advanced and will continue, with serious consequences on the economy of pension schemes. It is also gaining importance in developing regions, where a number of countries have started worrying about the medium- or long-term implications of the ongoing or incipient fertility decline for their age structures.

Current demographic trends

Charts 1 to 4 illustrate, at regional level, expected changes in the proportion of total population that will be found in the 60-plus age group according to United Nations population projections.1 The differences between regions are considerable at the present time: a 15-point gap between the percentages for the least developed countries and the more developed regions in 2000. Those differences should indeed grow larger, as the gap would reach 20 points in 2050. Sub-regional prospects are the following:

Table 1 places individual countries with respect to these broad trends. We chose to assess the period during which the proportion of persons aged 60 and above in a given country is expected to reach 10 percent, which is half the current average for the more developed regions and one third of the level expected around 2050 for these same regions. Even this relatively low percentage should generally not be reached very soon:

In conclusion, population ageing of the kind that raises serious and much discussed economic and social issues in the more developed countries still is a distant prospect in most developing countries - sometimes indeed a very distant one.3 Population ageing is an inevitable prospect, but its degree of urgency varies widely. In examining its implications for public policy and action by the civil society, one must at least distinguish a typology of country situations and assess the specific needs of each type.

Table 1. Distribution of developing countries by ten-year period during which the proportion aged 60 and above surpassed or is projected to surpass 10 percent

Implications of ageing

The economics of ageing have been analysed mainly in developed countries. In the very different context of developing regions - large informal sectors, flexibility of labour participation patterns, large non-monetarized economy, lack of institutionalized pension systems etc. - analyses so far have been less detailed. Attention has focused principally on the "negative" consequences of ageing, such as (i) problems regarding economic support for older people who no longer participate in the labour force, or do so with a low productivity; and (ii) problems regarding elderly health care, i.e. the financing of facilities and services and their adaptation to changing needs.4

But the economics of ageing in developing countries must be examined in the context of broader demographic changes, of which ageing is only one aspect. What is needed is a balanced, comprehensive view of the implications of those changes. During the demographic transition that developing countries now are undergoing, the decline of fertility - which is the main factor in ageing - causes not only an increase in proportions of older people, but also, conversely, a reduction in the proportions of younger people. Of particular interest is the proportion of people under 15 years of age, as these are also dependant. So we must place projected ageing within the context of overall age dependency.

Age dependency: trends and transitions

For those concerned with the implications of demographic change for development, an important aspect of age structures lies in the consequences of that change from the viewpoint of dependency. Large numbers of young or/and older people tend to burden the economy as the consumption needs of economically non-productive members of the society reduce the overall capacity for saving and investing: hence the tradition of calculating an "age dependency ratio", i.e. the quotient of the size of population aged under 15 and over 59 by that of the population aged 15-59 years, usually expressed in percent (children and older people per 100 adults).5

The overall age dependency ratio therefore can be looked at as the sum of two components, i.e. (a) the quotient of the size of population aged under 15 by that of the population aged 15-59, or juvenile dependency ratio, and (b) the quotient of the size of population over 59 by that of the population aged 15-59, or old-age dependency ratio. Now, as fertility decline proceeds, population ageing does raise the latter ratio, but at the same time the former ratio declines. What becomes of the overall age dependency ratio depends on the relative speeds on change in these two components.

In recent decades, the most significant change in age structures in developing countries has been the reduction in the proportions of young people, due to fertility declines: the proportion aged 0-14 has been declining in all the developing regions since 1970-19756. It will continue to decline, and the resulting reduction in numbers will be roughly as large as the increase in numbers of older people. Table 2 summarizes the expected changes.

Charts 5 to 13 illustrate the projected changes in the two components of the age dependency ratio and the outcome in terms of the overall ratio.

In Africa, two main situations will occur. In the Eastern, Middle and Western sub-regions, the overall ratio (chart 7) is expected to keep declining throughout the period and bottom out towards 2045-2050, much along the pattern we have seen for the least developed countries, but starting from higher values of the ratio. Charts 5 and 6 show that this trend is driven almost entirely by the decline in juvenile dependency: the old-age ratio is small and will rise only slowly. In Northern and Southern Africa, age dependency is much lower now and also declining, but will bottom out sooner - especially in Northern Africa, where old-age dependency will rise significantly faster than in the rest of the region.

In Asia (charts 8 to 10), the situations are rather diverse and will remain so, but the relative positions of the sub-regions will be reversed at the end of the period: Eastern Asia, which has the lowest overall ratio in 2000 - because its fertility transition is well advanced - will have the highest in 2050 because of more advanced ageing; Western and South-central Asia, which have the highest ratios, will then have the lowest for the opposite reasons. Western Asia exhibits different trends than any other sub-region: because both the projected fertility decline and subsequent ageing would be very slow, the overall dependency ratio has a very flat profile. The overall ratio for Eastern Asia, by contrast, would follow a path very similar to that of the more developed regions, bottoming out only shortly after the latter and then rising at a similar pace (but at a lower level of the ratio).

In Latin America and the Caribbean (charts 11 to 13) the sub-regional paths are quite similar. Juvenile dependency will have achieved most of its projected decline during the first half of the period surveyed; around the same time, the rise in old-age dependency will start accelerating due to ageing. Overall, total dependency will start increasing again around 2020 in all the sub-regions.

A "window of opportunity"

Therefore, as demographic transition proceeds, the juvenile dependency ratio decreases while the old-age ratio increases. This increase is at first slower than the said decrease; the result is a decline in the overall dependency ratio, which is important in a development perspective. In effect, during this decline, countries benefit from a reduction in the relative burden of the younger generations in terms of consumption and investment needs, while the burden of the elderly has not grown enough yet to offset that reduction.

The reduction in the total burden of dependants per person in the active age groups, in turn, opens opportunities to invest in economic and human development. This is why the period during which the age dependency ratio declines, enabling to better invest in national resources, has been described as a "window of opportunity"7. How large will that "window" be? Charts 7, 10 and 13 illustrate it at the sub-regional level. Table 3 gives answers at the level of single countries, classifying those by approximate date when the overall age dependency ratio reached or is expected to reach its minimum:8

Table 3. Distribution of developing countries by estimate or projection year in which the age dependency ratio reached or is projected to reach its minimum



Asia and Pacific

Latin America and the Caribbean

Before 2000


Qatar - Singapore




Korea Rep

Puerto Rico



Sri Lanka - Thailand - United Arab Emirates

Cuba - Trinidad & Tobago


Mauritius - Tunisia

Armenia - Azerbaijan - Bahrain - Brunei Darussalam - China - Georgia - Israel - Kazakhstan - Kuwait - Myanmar

Barbados - Brazil - Chile - Costa Rica - Guyana



Korea DPR - Lebanon - Malaysia - Mongolia - Turkey - Viet Nam

Bahamas - Colombia - Dominican Republic - Jamaica - Panama - Suriname


Egypt - Morocco

Fiji - India - Indonesia - Iran - Kyrgyzstan - Turkmenistan

Belize - Ecuador - Mexico - Peru - Uruguay - Venezuela


Algeria - Libya

Bangladesh - Philippines - Uzbekistan

Argentina - El Salvador


Cape Verde - Kenya - Zimbabwe

Samoa - Syria - Tajikistan



Botswana - Comoros - Côte d'Ivoire - South Africa

East Timor - Maldives - Nepal - Pakistan - Papua New Guinea

Honduras - Nicaragua


Rwanda - Sudan - Swaziland

Cambodia - Iraq - Jordan - Saudi Arabia - Solomon Islands - Vanuatu

Bolivia - Guatemala - Haiti - Paraguay


Benin - Burundi - Cameroon - Central African Rep - Chad - Djibouti - Eritrea - Gabon - Ghana - Guinea - Lesotho - Madagascar - Mauritania - Namibia - Nigeria - Senegal - Tanzania - Togo - Uganda - Zambia



2050 and after

Angola - Burkina Faso - Congo - DR of the Congo - Equatorial Guinea - Ethiopia - Gambia - Guinea Bissau - Liberia - Malawi - Mali - Mozambique - Niger - Sierra Leone - Somalia

Afghanistan - Bhutan - Oman - Yemen


Source: elaboration of data from United Nations, World population prospects. The 1998 revision, New York, 1999.

During the transition towards increasing dependency ratios, the composition of the dependant population will shift from a majority of young to a majority of older people. This obviously implies changing needs and therefore requires adaptations in social investment programmes as well as in the management strategies of both governments and private corporations. For instance, the overall costs of education for the society obviously decline during this process. On the other hand, the additional health costs of ageing are undeniable, largely due to the contrast between costly curative methods required by old-age pathologies and less costly preventive strategies (immunizations etc.) that have much effect on infant and child mortality.

In 50 of the 129 countries considered in this paper, the number of older people is expected to surpass that of persons under age 15 during the 2000-2050 period. Table 4 lists those countries and indicates when the reversal would take place.

Table 4. Developing countries where the number of older people is projected to surpass the number of persons under age 15 before 2050, by period when this would occur



Asia and Pacific

Latin America and the Caribbean






Georgia - Korea Rep - Singapore




Armenia - Azerbaijan - Bahrain - China - Korea DPR - Qatar - Thailand - United Arab Emirates

Puerto Rico - Trinidad & Tobago



Brunei Darussalam - Myanmar - Sri Lanka - Turkey

Bahamas - Chile - Guyana - Jamaica - Mexico - Panama - Suriname - Uruguay


Egypt - Morocco - Tunisia

Fiji - India - Indonesia - Iran - Kazakhstan - Kuwait - Kyrgyzstan - Lebanon - Malaysia - Mongolia - Viet Nam

Argentina - Belize - Brazil - Colombia - Costa Rica - Ecuador - Peru - Venezuela

Source: elaboration of data from United Nations, World population prospects. The 1998 revision, New York, 1999.

Ageing policies in different demographic contexts

At the simplest level of analysis, concern for ageing is merely a function of the proportion of older people in a population. Where that proportion is low (as e.g. in sub-Saharan Africa, or in the least developed countries) the magnitude of related issues - support for the elderly, health care - is limited. (This does not mean that those issues are easy to handle, because a poor society will have difficulties tackling even a modest per caput burden.)

When the proportion of older people starts increasing, it is important to start acting upon the related issues in a progressive fashion, lest the magnitude of the delayed tasks should become unmanageable. The rate of increase is also important; that rate is usually small at first, but as fertility decline accelerates, the growth in the proportion of older people will also accelerate. That acceleration can easily be anticipated - even though not precisely forecasted, of course - and public strategies need to take it into account.

The overall demographic context matters to the possibilities of responding to the challenge of ageing; in particular, a diminishing proportion of children enables reallocating resources from the needs of these to the needs of the increasing elderly population. At the present time, one concern is that in the countries who are rather advanced in the said transition (those in the upper rows of Table 4), not enough has always been done to take advantage of the situation by adequately investing in the future, in particular in human resource development.

Countries facing that hurdle in the short or medium run need to assess the needs of the older people, in particular their basic needs - food, health, housing etc. This should be compared with available resources: the nature of livelihoods and the extent of family support, taking into account trends in living arrangements and changes in household structures; the possible need to supplement the role of family networks and of the civil society at large; the prospects for further involvement of the elderly in the labour force and the desirability/feasibility of such an involvement.

Countries with more lead time at their disposal (those in the lower rows of Table 4) have the possibility of planning well ahead for the necessary adaptations. In this perspective, they should define long-term strategies to partly reorient public investment efforts as well as training programmes (in particular from mass education to health services); set up public mechanisms for welfare where feasible; and foster or assist with the development of targeted initiatives and institutions in the civil society.

In-country differentials and levels of intervention

While policies addressing the consequences of overall ageing at the country level must first be defined in general terms based on national trends, it may often be the case that ageing proceeds at a markedly unequal pace in different parts of a country. Accordingly, related issues are more urgent in some areas - or ethnic groups, or socioeconomic categories - than in others, suggesting in-country priorities.

For instance, both fertility and mortality tend to be lower in urban areas of developing countries, as the typical "demographic transition" - i.e. the move from high to low levels of these two variables - usually sets in sooner in urban societies. In the absence of other influences on the age structure, therefore, ageing would be generally more advanced in urban areas. In reality, however, urban populations are not necessarily more "aged" than rural ones. The reason lies in another factor, namely rural-to-urban migration. That migration usually is age-selective: it is mostly the adult age groups that supply migrants. Therefore, the populations left behind have higher proportions of older people.10 The age structures thus modified can be much skewed towards an "aged" pattern ;at the 1984 census of Tunisia, for instance, the ten most "aged" districts all were rural ones, exhibiting proportions of older people about fifty percent higher than the national average and old-age dependency ratios about double the national one.11

In this context, issues related to rural ageing will often need to be addressed as a priority, even in countries where the process of overall ageing will take many more decades to display all its effects.12 In any event, a different measure of urgency between urban and rural ageing issues is bound to mean priority to different types of interventions, because the form taken by those issues also differs significantly: urban and rural support networks differ in their breadth and functions; health needs and facilities differ; so do housing conditions and constraints, social institutions and so on.

Finally, attention to patterns of ageing and their consequences at a disaggregated level leads to highlight the issue of the appropriate levels and kinds of intervention required by the various issues identified. Different settings are characterized not only by different ageing patterns, but also by different economic and socio-cultural conditions relevant to the evolution of livelihood systems. Those conditions in turn affect both the nature and extent of ageing-related issues. It is therefore useful to consider separately a 'macro' level of intervention - concerned e.g. with government budget orientations and their long-term shifts, or with institutional policies - and a local-level policy development more concerned with how to foster successful adaptations by the families and communities, based on the specific forms of ageing issues in the given setting as well as on the socio-cultural characteristics of the local population that will influence its perception of those issues, its capacity to deal with those and the manner in which it will do so.13


1United Nations: World population prospects. The 1998 revision, New York, 1999. The medium variant has been used in all cases. The elements of novelty contained in the 2000 Revision of the UN projections, which appeared at the end of February 2001, do not significantly affect the contents of this note.
2"Eastern Asia" here refers to the developing part of the sub-region (Japan excluded).
3Inequalities in the transition to ageing are reflected in the varying degrees of attention given to its economic and social implications in different regions. Asian countries have been studying these issues and cooperating in the ageing area for a long time, as shown e.g. by the numerous technical papers and meetings promoted by the United Nations Economic and Social Commision for Asia and the Pacific (ESCAP). By contrast - and understandably - African countries have not yet undertaken a similar effort. Latin American countries as a group are in an intermediate position, although a number of them have tackled the ageing issue independently.
4This note will not dwell on the developmental implications of ageing. A forthcoming paper by the FAO Population Programme Service will deal with these aspects (see "
Rural population ageing in poorer countries: possible implications for rural development").
5These age groups do not necessarily correspond to economically active and inactive categories, because societies with large proportions of children or/and older people often require these to work. Nevertheless, the age groups used in calculating the ratio should not be modified for those societies, because the definition is based on a desirable state of affairs - barring child work as well as protracted use of older people's labour. "Actual" dependency should be measured by the economic dependency ratio, i.e. the quotient of the non-economically-active by the economically active population.
6However, that decline has been extremely slow in sub-Saharan Africa so far.
7This consequence of the transition in demographic structures is also sometimes referred to as the "demographic bonus".
8We have considered projection years only (the years in 0 or 5); the expected low point of the curve would normally be located within 2.5 years from the projection year showing the lowest ratio.
9A new factor affecting age structures in a growing number of countries is the HIV/AIDS pandemic, which contributes to raise the proportion of elderly because it mainly affects adult age groups. This may modify the pattern of urban-rural differences in ageing, if the incidence of the disease differs between rural and urban populations.
10They also tend to have higher proportions of children, but (a) adult migrants may take their children with them; and (b) in populations where only one parent normally emigrates, this pattern of migration contributes to lower fertility in the medium term, lowering the proportion of children - and contributing to "natural" ageing (as opposed to migration-induced ageing).
11Marcoux, Alain: "Ageing rural populations in the developing countries: patterns, causes and implications", in United Nations: Ageing and the family, New York, 1994, pp. 144-148.
12On the specificities of rural ageing, see "Rural population ageing in poorer countries: possible implications for rural development" (cf. endnote 4 above).
13For instance, the important work done so far in the Asia region has opportunely placed emphasis on local-level action (see in particular the ESCAP project Local Level Policy Development to Deal with the Consequences of Population Ageing and the publications to which it has given rise).

SD Homepage Back to Top FAO Homepage