*Professor, University of the Philippines Population Institute, Quezon City Metro manila, the Philippines*

1. Introduction

   In mid-1995, the world's elderly population (herein defined as persons 60 years and older) was estimated at 542.7 million, nearly equaling Africa's population in 1985. One in eleven of the earth's inhabitants is at least 60 years of age. Fifty-five per cent of the world's elderly are women, underscoring their lower levels of mortality as compared to men.

   Changes in age distributions, which are occurring as a consequence of declines in fertility and mortality levels, have led to two distinct trends. One is the ageing of population, the other is its juvenation. However, according to the United Nations (1988), the most dominant trend from the mid-1980s to the first half of the 21st century, in both the more and the less developed regions, will be the ageing of populations. In many low fertility and low mortality countries, natural increase can no longer sustain population growth and such countries may expect a decline of their population size. At low fertility levels, the improvement of mortality, particularly at old age, will be a major factor in the growth and ageing of populations. In the Third World, population ageing is primarily a result of the diminution in the proportions at the younger ages, a result of falling fertility. As both fertility and mortality levels continue to descend to much lower levels in the less developed regions, their age distributions will very closely resemble those of the more developed regions in 1950.

   Over half (52 per cent) of the world's senior citizens dwell in Asia; four in every fifteen are concentrated in Eastern Asia1, one in six inhabit South-Central Asia, about one in fifteen live in Southeast Asia and the remainder reside in Western Asia. Less than two in five (39 per cent) of the world's elderly are to be found in the more developed regions. Of this proportion, 64 per cent are located in Europe, 22 per cent live in North America, 12 per cent reside in Japan and the remaining 2 per cent in Australia and New Zealand.

   The ageing and feminization of the world population is clearly evident in table 1 where the projected male population for Europe, aged at least 60 years in 1995, makes up about one-sixth (16 per cent) of the total male population whereas the corresponding proportion of women comprises two-ninths (22 per cent). Thirty years later, the European elderly male population is forecasted to reach almost a quarter (24 per cent), while their female counterparts are expected to account for three-tenths (30 per cent) of all women in the European region. The corresponding proportions in 2050 are anticipated to reach 28 and 34 per cent, respectively. Eastern Asia, which reports one-ninth of its female population as elderly, will be a close second to North America's elderly female populations in 55 years' time (27 vs. 29 per cent). The fractions of male elderly in Eastern Asia are foreseen to rise from one-eleventh (9 per cent) in mid-1995 to nearly a quarter (24 per cent) by the mid-21st century.

   A significant fraction of this expansion will be due to greater female longevity at the very old ages. In 1995, as shown in table 2, European men aged 80 and older constitute close to 12 per cent of the male elderly population; women aged 80 and over formed 19 per cent of the female elderly population. By 2050, with improved life expectancies, one in five elderly European men are foreseen to be at least 80 years while the analogous proportion for women is anticipated to be two in seven. Parallel proportions for Eastern Asia reveal that the very old men and women make up and over 11 per cent of the 1995 elderly population, respectively. By the middle of the coming century, the corresponding percentages are contemplated to be 15 and 22 per cent.

   The following sections will describe the rapid greying of Asia and Oceania, concentrating on the changes in age structure using relevant indicators; the sex ratios, life expectancy, urban-rural residence and marital status of the elderly, and the support ratios which measure the potential stress that shifting age structures place on a society. The indicators were calculated using data from the United Nations publications: The Age and Sex Distribution of the World Populations: The 1994 Revision (1994), The World Population Prospects: The 1994 Revision (1995) and the special issue on Population Ageing and the Situation of Elderly Persons of the Demographic Yearbook(1993). Supporting data, where appropriate, were obtained from cross country studies on ageing undertaken by the Economic and Social Commission for Asia and the Pacific (ESCAP) in connection with its projects on Emerging Issues of the Ageing of Population(1989) and on Local-Level Policy Development to Deal with the Consequences of Population Ageing(1994). In this chapter, Western Asia is excluded in the discussion except for comparative purposes and only those countries in the Pacific area for which data are available in the United Nations publications are included.

   Across Asia, populations are ageing rapidly. In some of the countries, notably China, population ageing will proceed much faster than comparable socio-economic growth, an unprecedented development in world history. In fact, ageing, like transformations in birth and death rates, does not take much time any more. While it took France some 140 years to double the proportion of its population, 60 year sand older, to 18 per cent in 1976, only 45 years will pass in China before improved health care and low fertility allow the percentage of the elderly to increase from 6 per cent in 1970 to 14 per cent in 2015.

   The transition from high to low levels of fertility and mortality induces profound alterations in age structure. As decreases in mortality usually precede similar declines in fertility and since the diminution in mortality at elevated levels tends to be particularly salient for infants and children, the age structure becomes younger at the demographic transition's early stages. When significant decreases in fertility are initiated, the age structure trend changes direction and the population starts to age. Continued fertility decline and extensive mortality decreases at middle and older ages contribute to further population ageing.

   The accompanying figure 1 shows the trend in the proportions of the elderly between 1950 and 2050 for Asia and Oceania. Although these two regions are at different stages of the demographic transition, their age structures are following and are projected to follow the course of the transition. It is evident that in 1950, the latter's percentage of persons 60+ surpassed the former's by about 5 percentage points. A century later, it is foreseen that the proportion of elderly Asian women will exceed that of their male counterparts in Oceania but will still lag behind their Pacific sisters by 3 percentage points.

    In mid-1995, as recorded in table 3, Japan (22 per cent), Australia and New Zealand (17 per cent each) and Hong Kong (16 per cent) had the largest proportions of women elderly in the region under review. On the other hand, while Japan reported the highest proportion of elderly men (18 per cent), Australia, New Zealand and Hong Kong had around 14 per cent of their male population aged 60. Singapore and the Republic of Korea both reported 11 per cent of their female population to be 60 years or older while the analogous proportion for China was 10 per cent. The lowest female fractions were those of Vanuatu, East Timor and the Solomon Islands - about 4 per cent. The countries with the lowest male elderly ratios were Cambodia and Samoa (3 per cent). The rest of the countries had percentages ranging from 5 to 9 per cent for women and 4 to 9 per cent for men. By 2050, two in five Japanese women will be at least 60 years of age; the corresponding fraction for Japanese men will be one in three. In contrast, only one in nine women and one in ten men in Afghanistan are foreseen to belong to the elderly group. Among the subregions, South-Central Asia will lag behind its mates in Eastern and Southeastern Asia and Oceania. Table 3 also reveals that with very few exceptions, the percentges of women elderly surpass those of men.

    The absolute increase in the number of elderly is another way of examining population ageing and is the most relevant to policy planners and service providers. Among the countries, China is projected to have the largest absolute increase--nearly 160 million--in the number of persons aged 60 and over between 1995 and 2025. This increase represents more than half (52 per cent) of the addition to the country's total population during the 30-year period. Similarly, the increment to India's elderly population in the same period is estimated to be about 106 million, over two-ninths (23 per cent) of the increase in the country's total population between 1995 and 2025.

    The relative increases of the population 60+ in the coming decades as presented in table 4, manifest an accelerating trend, peaking during the period 2010-2030 and falling rapidly thereafter. During the twenty-year period, 1990-2010, Asia's senior citizens are projected to grow by 70 per cent, doubling in the next twenty years before plummeting down to about 50 per cent. Oceania's increase is more moderate - about 50 per cent in the period 1990-2010, advancing to 66 per cent during 2010-2030, and subsiding to 28 per cent in 2030-2050.

    Notable deviations from this trend are Japan where growth is projected to stall and even turn negative after 2010-2030; Hong Kong's increase is expected to become negative during the period 2030-2050, while Singapore is anticipated to evince a status quo during the same period. Other slow growers are the Republic of Korea, Australia and New Zealand.

    Worth noting too are the lower percentage increases reported for women 60+ as compared to men 60+ in Eastern Asia and in Australia-New Zealand during the period 1990-2010. Nowhere else in the two regions under review is a similar pattern evident.

3.2 The oldest-old

     The most common way to look at population ageing is to consider changes in the proportions of the elderly as just discussed. Another is to examine the relative growth of age groups over time. The age composition of the elderly population is being altered because of the tendency of the older age groups within the elderly population to expand more rapidly. Countries which have undergone a rapid fertility transition in the recent past are ageing the fastest compared to those nations where the transition proceeded at a more moderate pace.

     The ageing process currently observed and projected for developed countries (and for developing countries in the coming century) is characterized by a disproportionate growth of the very old and of widows due to higher male mortality. Important age differentials are uncovered when the population 60 years and older is divided into three groups, those aged 60-69 years, those 70-79 and those 80 years and older. In 1950, persons in the 'young-old' category (60-69 years of age) numbered 65 million in Asia and Oceania, comprising 68 per cent of their total elderly population. Forty-five years later, this group is estimated to exceed 175 million and is projected to swell to 615 million in 2050. For the 'oldest-olds' (80 years and older), the increase will be much faster. The population of this group which was 4.5 million (5 per cent of all elderly) in 1950, is estimated to be close to 25 million in mid-1995 and is anticipated to reach 192 million (16 per cent of the population 60+) by the year 2050. By 2050, the 'young-olds' will only account for one half of the total elderly population, a decline of 18 percentage points in a century.

     The large increase in the 'oldest-old' is evident in both developed and developing regions, and among women and men. Of the total population 80 years and older in 1950, 38 per cent came from the developing world. By 1995, this proportion rose to 44 per cent; in 2050, it is expected to approach 72 per cent. The parallel proportions for Asia and Oceania were 33 per cent in 1950, 40 per cent in 1995 and 58 per cent in 2050. Of the world's women 80 years and older in 1950, 36 per cent came from the less developed regions; a proportion which grew to 40 per cent in 1995. Fifty-five years thereafter, the figure is foreseen to reach 70 per cent. The corresponding fractions for Asia and Oceania are 32, 37 and 57 per cent. These figures demonstrate that Latin America and Africa will be catching up with their Asian and Pacific counterparts by the 21st century. Notable is the fact that the proportions of the very old women are almost identical to those for the total population.

     Table 5 discloses that for the most part, in the subregions of Asia and Oceania, the percentage increases for women 80 years and older, for certain subregions during the next two decades, exceeded those for men for the periods, 1950-1970 and 1970-1990. From then on, the reverse will take place with the exception of South-Central and Southeastern Asia and the small island countries of the Pacific. By 2030-2050, the proportional increases among those who are at least 80 years of age will be smallest for women and men in Australia and New Zealand.

3.3 Population 70+ compared to those 60-69

     Still another way of indicating the fast growth of the older age groups is to express the population 70+ relative to those aged 60-69 as listed in table 6. For Asia as a whole in 1970, the numbers of women and men in their 70s or older were over half those in their 60s. By 1995, the ratio had risen to 62 per cent. Thirty years hence, it is anticipated to reach 70 per cent. By the mid-21st century, the population 70+ is foreseen to almost equal that of the 60-69 year-olds. Oceania's ratios were much higher - 74 per cent in 1970, 97 per cent in 1995, dropping to 94 in 2025, but reaching a high of 130 per cent in 2050.

     In 1970, the lowest percentages of the populations 70 years and older compared to those in their 60s were reported by Samoa (25 per cent), Afghanistan (37 per cent), Maldives and the Lao People's Democratic Republic (40 per cent). The highest proportions were reported by the Solomon Islands (80 per cent), Australia (78 per cent) and New Zealand (76 per cent). In East Asia, the fractions ranged from a low of 44 per cent for Hong Kong to a high of 65 per cent for Japan. In South-Central Asia, the highest proportion recorded was that of Pakistan (61 per cent), the lowest, Afghanistan (37 per cent). For Southeastern Asia, the percentages varied from a low of 40 per cent in the Lao People's Democratic Republic to a high of 63 per cent in the Philippines. In Melanesia, the difference between the lowest (Papua New Guinea) and the highest proportions (Solomon Islands) was 35 percentage points.

     By the middle of the coming century, the numbers of those who will be at least 70 years of age will equal or even double the number of those in their 60s. This will be true for the Eastern Asian countries, with the exception of Mongolia, for Brunei Darussalam, Singapore and Thailand in Southeastern Asia, and for Australia, New Zealand, New Caledonia and Guam in Oceania.

3.4 Populations under 15 and over 60

     The size and proportion of the population under 15 years of age are more directly affected by recent fertility changes than are other age groups. However, it is the change in old age mortality that directly influences the population aged 60 years and older more than any other age group. Also, the variation in the proportion of children in a population is to some extent related to the shift in the proportion of the elderly as well as to other age groups.

     The comparison of the proportions of elderly to those of the child population indicates which age group is dominant and whether the trend is one of juvenation or ageing. The trend disclosed in table 7 is definitely one of ageing. At the turn of the century, Japan is expected to be the first Asian nation to have its elderly population exceed its child population; although as early as 1995, the proportions of Japanese female elderly already had this distinction. By 2025, Japan is projected to be joined by Hong Kong, the Republic of Korea, Singapore, Australia and New Zealand. A quarter of a century later, all the rest of the East Asian countries, India, Sri Lanka, Brunei Darussalam, Indonesia, Malaysia, Thailand, Fiji, New Caledonia, Guam, French Polynesia and Samoa are anticipated to join the list.

3.5 Median ages

     When a population ages, the share of the older people in the population increases while the share of children and youth decreases, resulting in a rise in the median age. As disclosed in table 8, Asia as a whole had a median age of 24.6 years in mid-1995 compared to the world's 25.3 years. Among the Asian sub-regions for the same period, Eastern Asia was found to have the highest median, 28.3 years, while Western Asia had the lowest, 21.2 years. Oceania's median age was 29.8 years with the Australia-New Zealand sub-region topping this by 3.5 years. By 2030, Asia's median age is projected to climb to 33.6 years and that of Oceania to 35.8 years. A score of years thereafter, the median ages of Asia and Oceania are forecast to rise further to 37.7 and 38.5 years, respectively.

     Among the Asian countries, Japan ranked first with a median age of 39.3 years in 1995. In second place was Hong Kong with 34.1 years. Bringing up the rear were the Maldives, 16.7 years; the Lao People's Democratic Republic, 17.6 years; and Pakistan, 17.9 years. Among the countries in the Pacific, Australia and New Zealand headed the list with median ages of 33.6 and 32.1 years, respectively. At the bottom was Samoa with a median age of 16.3 years.

     The trend projected in the coming 35 years for Eastern and Southeastern Asia is for steadily mounting median ages, with the pace decelerating in the following two decades. To illustrate, Hong Kong with its 1995 median age of 34.1 years is anticipated to have half of its population below the age of 50.6 years in 2030, an annual average increment of 0.47 per year. In 2050, its median age is forecast to be 53 years representing a yearly average increase of only 0.12 years. Indonesia had a median age of 23.1 years in 1995, which went up to 34.1 years in 2030 and to 37.7 years in 2050. These levels imply annual average expansions of 0.31 and 0.18 years, respectively. On the other hand, the projected median ages for South-Central Asia and the small island countries of the Pacific exhibit a continuous upward trend. A case in point is that of Nepal whose 1995 median age was 18.6 years. By 2030, the median is foreseen to attain 26.9 years or an annual average growth of 0.24 years. In 2050, the median is expected to reach a level of 33.8 years, indicating an addition of 0.34 years per annum. Papua New Guinea's median age in 1995 was 20 years, projected to move up to 27.7 years in 2030, representing a yearly average increment of 0.22. By 2050, the median age is forecast to rise to 34.1 years or an annual average addition of 0.32 years.

4. Other Indicators

4.1 Sex ratios for the elderly and the total population

     Sex ratios of the elderly reflect the differential mortality and migration that have occurred over the lifetimes of cohorts. Although more boys than girls are born worldwide, men tend to have higher mortality rates than women at all ages, which accounts for imbalances in sex ratios at older ages. As has been stated earlier, in most countries of the world the ageing of the population implies a greater increase in the number of elderly women as compared to elderly men. Thus, ageing societies are increasingly becoming disproportionately female.

     Sex ratios among the elderly are projected to decline in the coming decades as seen in table 9. In general, elderly women are expected to outnumber elderly men in 2025 and in 2050, but to a lesser extent than in 1995. The Islamic Republic of Iran, Maldives and Pakistan, which had elderly sex ratios exceeding 100 in 1995, as a consequence of past high maternal mortality, will witness the reverse in the next century.

     In contrast, the majority of Asian and Pacific populations as a whole are forecast to continue having sex ratios favouring men. Among all the countries, Cambodia has the lowest sex ratio (93), but is projected to achieve a gender balance in the decades to come. Japan's sex ratio of 96.4 in 1995 is predicted to remain almost stationary through the year 2050. On the other hand, Cambodia, whose projected sex ratio was 93 in 1995, is expected to achieve a gender balance in the coming century. The same is true for Viet Nam.

4.2 Life expectancy among the elderly

     Around the middle of this century, life expectancy at age 60 in the countries of Asia and Oceania, with available data, was of the order of 12-14 years, being generally 1 to 2 years longer for women than for men. By 1990, life expectancy at age 60 for women in the countries shown in figure 2 varied from 17 years to 24 years. The range for men was 15 years to 20 years.

     These ranges mask considerable differences among countries. A Japanese woman who reached 60 years of age in 1990 could expect to live a further 24 years, on average, compared with women of the same age in the Philippines or Malaysia whose life expectancy would be around 18 years. Men's life expectancy is uniformly lower than women's due primarily to the effect of the former's less healthy lifestyles during adulthood compared with the latter. Male life expectancy at age 60 varied from about 15.4 years in the Philippines to 20 years in Japan.

     Moreover, as demonstrated in the 1991 United Nations Demographic Yearbook (table 3), there were some quite significant changes in life expectancy at age 60 over the period 1960-1990 compared with the average gains for women and men mentioned earlier. Perhaps the most outstanding example is Japan where life expectancy at age 60 for women had increased by 7 years, or almost 40 per cent, since 1960. For men, the corresponding increase was 5.3 years, or 35 per cent. Among men in this region, the most impressive increase in life expectancy at age 60, was observed in Australia (2.9 years).

4.3 Urban-rural residence

     Between 1980 and 1990, most of the elderly population in Asia and Oceania were to be found in rural areas. According to the special issue of the United Nations Demographic Yearbook on the elderly, the exceptions were Hong Kong, Japan, Brunei Darussalam, Islamic Republic of Iran, Australia, New Caledonia and New Zealand.

     Data for the 1980s revealed that only one in twenty elderly lived in Nepal's towns and cities. The comparable proportion was 13-14 per cent for Bangladesh and Afghanistan. Around one-third of the elderly in the Philippines and Malaysia were urban residents while the corresponding fractions in Australia and Hong Kong hovered around 90 per cent.

     As illustrated in figure 3, from just over one-half (Mongolia) to about nine-tenths (Vanuatu) of the elderly resided in villages around 1990. In contrast, only one in nine of New Zealand's elderly and less than three-tenths of Japanese senior citizens were rural dwellers.

4.4 Marital status

     The distribution of marital status is a function of historical process that is embedded in cohorts as they reach the older ages and indeed, preceding cohorts that are already aged. In this respect, marriage, remarriage, divorce and widowhood play a vital role. The current trends reflect past developments.

     While data on marital status have improved in terms of age detail at the older ages, there are still serious difficulties in obtaining time-series information for many countries, particularly the developing ones. Categorization of marital status is subject to considerable differences between countries in the manner in which separation is treated. In addition, statistics on consensual unions may be variously recorded and reported. This is a matter of emerging concern in many countries, not only in the developing but in the developed countries as well.

     Data on marital status of the population 60 years and older around 1990 for selected Asian and Pacific countries show a large gender difference (see figure 4). Noteworthy is the majority of male elderly in these countries (with Japan leading the group) who are still married while from nearly two-fifths to close to three-fifths of their women counterparts in Vanuatu and Indonesia, respectively, are already widowed. In fact, two to four times as many women as men were widowed, due in part to the greater longevity of the women and to wider social acceptance of remarriage among widowers than among widows. Fewer women than men had never married except in Japan and Vietnam. One in six of New Caledonia's male elderly were still single; the analogous proportion for women was one in eight.

     The general picture obtained from figures 3 and 4 is supported by multi-country studies conducted by the Association of Southeast Asian Nations (ASEAN) in the 1980s, the ESCAP studies of ageing in Malaysia (Arshat and others, 1989), Republic of Korea (Choe, 1989) and Sri Lanka (Perera, 1989) and the more recent (1994) ESCAP Project on Local-Level Policy Development to Deal with the Consequences of Population Ageing in Bangladesh, India, Indonesia, Pakistan, Thailand and Viet Nam.

4.5 Support ratios

     Policy makers are concerned with the relationship between changing age structures and varying demands on social services and support structures. An important illustration of this relationship was the post-Second World War baby boom in the developed countries. This posed a problem for the educational system in the 1950s when large numbers of students began to exceed the capacity of existing schools and teachers. However, when the successively smaller birth cohorts of the mid-1960s attained school age, the educational system was faced with progressively smaller entering classes. Consequently, school services were reduced; many schools were closed or converted to other uses.

table 10      As a population ages, supply and demand issues also are created. The elderly place different demands on support systems than the non-elderly. Expanding numbers of elderly may strain pension schemes, housing supply and health care systems. Growing numbers of elderly also imply rising health care costs, particularly as people survive to older ages and require more expensive medical treatment.

     Approximate measures of potential 'stress' that shifting age structures may place on a society are called support (or dependency) ratios. The total support ratio is commonly defined as the number of persons under age 15 and over age 60 per 100 persons aged 15 to 59, the so-called productive ages. The elderly support ratio is defined as persons 60 and older per 100 aged 15 to 59. There are inherent imprecisions within these indicators, not the least of which is the fact that many people over age 60 are still economically productive. Likewise, not everyone in the 'productive' age group is economically active. Official retirement ages vary between 55-65 for women and 60-65 for men, although many individuals continue to work beyond official retirement age. Despite such measurement problems, support ratios can be used broadly to exemplify the changing pressures that a generally dependent population component might exert on society's tax base.

     With major fertility declines, Asia's elderly support ratios which averaged 13 in 1995 are expected to reach 22 in the year 2025 and 36 by 2050. But it is in East Asia where the aged dependency is projected to double as early as the first quarter of the coming century. For example, table 10 demonstrates that Hong Kong's working age population in 2025 is projected to bear the burden of 63 persons aged 60+ compared to 22 in 1995; the comparable figure for Japan will be 59 elderly persons as against 32, thirty years earlier. In Singapore, the population of working age will have to support 51 elderly persons in addition to children, an increment of 36 dependents since 1995. Nowhere else is the burden posed by the elderly as heavy. The numbers of dependent elderly in the year 2025 are foreseen to vary between 7 in Samoa to 41 in Australia. By mid-21st century, the burden placed by the elderly on the working age population is forecast to range from 22 in the Solomon Islands to 48 in the Republic of Korea.

5. Social Implications of Population Ageing

5.5 The family and household

     Age structure changes have an impact on the family as an institution and on the conditions of women. As a population ages, the size and structural type of the family is altered. Likewise, households which are the basic units within which employment, consumption and savings decisions are taken, would be both smaller in size and older in structure. Throughout this century, the average household size has been shrinking in most countries of the world. Much of this change has been due to a mounting proportion of one-person households. Myers (1992) reports great increases, more than double in some cases, in the proportion of older persons living alone in most developed countries in the past 30 years. In England, for example, the proportion had grown from 7 to 17 per cent among men and from 17 to 45 per cent of older women between 1951 and 1981. In developing countries, average household size has in some cases expanded, following the juvenation of the population resulting from gains in infant and child survival. But this picture is expected to alter in the coming decades as the developing countries follow the trend of the more developed regions.

     The nature of families is linked intimately with household structure or co-residency patterns. The number of generations included in the household and the length of time during which several generations live together seems inevitable as a result of demographic ageing (Kono, 1988; United Nations, 1994). Among Japanese households, 54 per cent included three generations, which denoted that the elderly were normally taken care of by their child or their child's spouse (Kono, 1994). This trend could endanger the extended family's survival by placing intolerable burdens on the younger generation.

     From the family's viewpoint, the lengthening of joint survival is an important aspect of the demographic transition. As individual survival is prolonged, joint survival lengthens even more than proportionately. Moreover, the age difference between spouses contracts as do those between parent and child and between siblings. By extending joint survival, low mortality raises the proportion of one- and three- generation families relative to two-generation families (Ryder, 1988).

     The marital status of older persons is an aspect of family structure that deeply affects their living arrangements, support systems and individual well-being. In some cases, support provided by an older women to her husband may, on one hand, leave her physically and financially drained and lacking any support of her own when he dies. On the other hand, unbroken husband-wife families provide a continuity of the marital bond established through the life course. Thus, they constitute a multiple support system for spouses in terms of emotional, financial and social exchanges. The provision of care in coping with ill-health from chronic diseases and functional limitations becomes increasingly important at older ages. To a large extent, marital status also determines the living arrangements of older persons. While independent living arrangements generally characterize the situation for married couples in most developed world societies, it is far from universal in other regions of the world. Finally, it is frequently reported that married persons tend to enjoy higher levels of survival, mental health, use of health services, social participation and satisfaction compared with older persons who never married (Myers, 1994).

5.6 Conditions of women

     Ageing has led to a decline in the numbers of men relative to women due to the latter's advantage in mortality at all ages, leading to a gender gap in life expectancy. While the world as a whole had about 80 elderly men per 100 elderly women in mid-1995, the developing regions were estimated to have 90 men per 100 women for the same period, a situation echoed in the Asian region. This imbalance widens with age so that those at the oldest ages and with the utmost need for care, are far more likely to be women.

     One of the most important developments in the twentieth century is the extension of women's economic activity. In many countries, however, the women 60 years or older have considerably lower incomes than elderly men primarily due to smaller pensions. Divorced women suffer the greatest economic hardships because they are generally not covered by the pension schemes of their former husbands.

     In general, as a reaction to growing opportunities in the economy, women in ageing populations had adopted strategies which implicitly assumed that their major care giving responsibilities were to children. As more and more women joined the labour market, its opportunities assumed greater importance to them. In particular, the more recent expansion in women's education made younger women workers relatively more useful to their employers than older women. Hence, in some countries, older and less-educated women have commenced increasingly to care for the offspring of their employed daughters and/or daughters-in-law.

     In many countries, the elderly look to women of childbearing age to minister to them. As greater numbers of women join the labour force, their time must perforce be taken up not only by bearing and rearing children, but also by catering to the demands of the growing number of elderly in the households. At the same time, the caregivers will be approaching the age when they themselves will be joining the cohort of senior citizens (Concepcion, 1994).

     If the family is to remain the locus of care, the logic of population ageing requires both women and men to be involved in this activity. Many women are starting to resist the presumption that they should take on total responsibility for family and household care. The choice might have to be made in the near future between establishing institutions to replace the family in care giving or having men assume a growing share in the burden of caring for their children and ageing parents.

6.1 Labour supply

     The increase in the old age dependency ratio as evident in both the developed and less developed countries is expected to accelerate in the early part of the 21st century. Many industrialized countries, including Japan, have experienced a rise in the average age of employed persons. The number of young economically active persons has shrunk while the working populations aged 35-54 years and those 55 and over have expanded.

     In the developed economies, the expected growth of the economically-active population and rapid technological change might create problems in maintaining full employment until the turn of the century. Thereafter, the ageing of the labour force would lead to lower overall unemployment, with relatively more unemployment among older workers. Firms might substitute capital for human resources, modernizing their plant and equipment or redeploying production to countries with ample low-cost workers.

     The ageing of the labour force may significantly transform the wage and employment structures of firms. The growing number of older workers might narrow wage and salary differentials to the benefit of younger workers and might also induce a loosening in the rigidity of tenure arrangements.

     Job opportunities for older persons are determined by the relative supply of, and demand for, older workers. Industrial structure is one of the principal factors governing the labour market prospects of older persons (Durand, 1975). In self-employment and the agricultural sector, older workers find it easier to phase into retirement by switching job assignments and by reducing hours. Self-employment and employment in family-related activities also tend to be greater in the rural sectors. Thus, countries that are less urbanized and have a larger agricultural sector are expected to have higher labour force participation among older persons (Chen and Jones, 1989; Arshat and others, 1989; Choe, 1989; Perera, 1989).

     Labour force participation rates among the elderly in the countries comprising the Association of Southeast Asian Nations (ASEAN) remained quite high at ages above the official retirement age (Chen and Jones, 1989). This serves to underscore the point that only a minor fraction of the ASEAN populations except Singapore, are in wage and salaried employment affected by compulsory retirement ages. In self-employment, economic needs force a lot of people to continue working as long as they are able to, albeit sometimes at a reduced pace. In agriculture, where the elderly workers were concentrated, the pace of work can readily be adapted to the skills of the elderly, who in some instances may help out only in the busy seasons.

     In the developing countries, the increasing dependency burden of the ageing population would be more than offset by the decreasing dependency burden of young people during the early stages of ageing. However, the total dependency ratio is likely to rise rapidly in the early part of the coming century. Industrialization, accompanied by the mechanization of agriculture is likely to reduce the opportunities of the elderly for self-support. Increasing urbanization and social mobility, smaller families, and the entry of married women into paid employment would leave them less time to care for elderly parents.

6.2 Savings and investment

     In the developed countries, the elderly did not stop saving in old age, but actually continued to increase their net worth. In the case of Japan, Kono (1988) observed that the effect of a low official age of retirement in conjunction with a high life expectancy had contributed to a high savings rate.

     Demographic ageing might affect rates of capital formation. If a parental generation expected greater life expectancy, they might reduce the net transfer of capital to the next generation, thus slowing the rate of economic growth. In India, the property owned by parents was generally transferred to sons at the death of the household head. The common view was that lengthening life expectancy of the aged parent would reduce incentives for the young to remain on the farm. A contrary view supported by the evidence was that experience was considered important in India and that the presence of an elder in the household increases productivity. Thus, sons were less likely to be absent, the older the household head (United Nations, 1994). On the other hand, if parents expected a longer retirement period, they may have a greater incentive to invest in the human capital of their children as a source of old-age support. Furthermore, demographic ageing is accompanied by smaller family size so that if the total amount of the intergenerational transfer was constant, there would be a larger net transfer per child (Tabah, 1988).

6.3 Intergenerational transfers

     It has been generally assumed that the persistence of the extended family and multi-generational living arrangements denotes that the elderly in the developing world are well cared for by their family members. Older persons transfer property to their children who assume responsibility for household tasks and care for their ageing relatives. The studies in Pakistan (Afzal, 1994) and Thailand (Phalakornkule, 1994) undertaken under ESCAP auspices have shown, however, that elderly parents were central contributors to the economic support of the households. The contrary was found in Indonesia (Wirakartakusumah, 1994) where only 22 per cent of elderly who lived with their children or grandchildren contributed to household expenses. The sources of support differed by country with Indonesia reporting other sources apart from salaries; Pakistan, salaries and/or savings as well as other sources; and Thailand, income from work and relatives. These studies indicate that support of the elderly by their children at home is delayed until later ages than are being measured by the broad age category of 60+.

     Income from one's own economic activity and from pensions and provident funds were more important for men than for women. The percentage of male respondents who cited pensions or provident funds ranged from 5 per cent in Thailand to 6 per cent in Malaysia. There were no comparable figures for the Philippines (Chen and Jones, 1989). The majority of elders in Sri Lanka had no definite source of income after retirement. Only one in nine elderly received a pension with five times more urban dwellers receiving pensions than rural residents.

     Remittances from family members who migrate for better economic opportunities can be an important source of inter-generational redistribution of income within the family. This is particularly true in more agricultural economies. The strength of the migrants' motivation to return or to bring their families to join them often determines the degree to which these remittances may be viewed as investment for the migrants future return home (often after retirement). Alternatively, migrants may send money for philanthropic reasons or as part of an implicit contract to contribute to their parents' well-being, without any intention of returning to take advantage of their investment in the family. But upon retiring, they may return as the pension they receive is insufficient for subsistence in the city environment (Hugo, 1988; Lucas and Stark, 1984).

6.4 Pension schemes

     As populations age, the share of old age pensions in total public expenditure will increase significantly. In part, this will reflect the decline in the public expenditure requirements of children. Population ageing, by increasing the cost of public old age pensions, makes necessary either a reduction in benefits, an increase in contribution rates, financing from other sources or some combination of these policies. In the event that contribution rates are increased, the diminished burden of child care should give families a greater capacity to absorb the tax increase while maintaining their real disposable income per capita.

     The negative economic effects on pension plans of population ageing could be significantly reduced by policies designed to increase the activity rates of physically- and mentally-able older persons. Eliminating compulsory retirement and raising the age of eligibility from 60 to 65 years could reduce the expenditures of some social retirement systems by half.

     In most developing countries, pension systems are virtually non-existent, covering only a small portion of the population such as government workers. Even then, many are simply provident funds providing lump-sum payments upon retirement which are inadequate and quickly spent. Most older workers, then, work to much more advanced ages than is the case for the more developed regions. On retirement, such elderly return to live with one or more children, depending almost entirely upon family support. A case in point is that of the Philippines where coverage under the government pension scheme was limited to only 8 per cent (Domingo, 1989). About 15 per cent of Malaysian men enjoyed employment benefits as contrasted with 5 per cent for Malaysian women (Arshat and others, 1989). The number of Korean elders receiving public welfare is very low (Choe, 1989). The reason is that the government-level social welfare system in not well established in the Republic of Korea. The majority of elders in Sri Lanka had no definite source of income after retirement. Only one in 9 elderly received a pension, with 39 per cent of the urban dwellers receiving pensions as contrasted with 8 per cent of rural residents (Perera, 1989). Thus, public pension plans might have in egalitarian implications for the transfer of resources between different social classes of the same generation as well as the transfer of resources between generations.

    The ageing of populations is a worldwide phenomenon, an ineluctable consequence of decreasing fertility and lengthening life expectancy. The population 60 years and older is enlarging both in relative and in absolute numbers. As the population age structure rectangularizes, intergenerational and intercohort relations need to be re-appraised and altered.

    The structure and relations of families are being modified by these fertility and mortality changes. Greater numbers of generations live together in families for much longer time periods. New relationships are evolving quickly within these kin groups. Simultaneously, the change in the acceptability of alternative family patterns and the variation in social and economic status between generations within the larger society have contributed to alterations in the family.

    The earlier sections underscored several demographic aspects of population ageing that deserve attention. One, is that as a consequence of falling fertility, the size of the population of non-working-age relative to the size of the working-age population is expected to swell in the developing regions within the next half century. Two, since the sex ratio tends to be lower at older ages, population ageing implies a greater increment in the number of elderly women than in that of elderly men. Three, owing to the tendency of the older age groups to increase more rapidly, older persons will constitute increasingly higher proportions of the population 60 years and older. Four, population ageing and the more rapid population growth in the developing regions than in the developed regions will lead to more and more fractions of the world's elderly concentrated in the developing countries.

    The analysis of fertility and mortality effects on age structure leads to several important policy implications. First, even if fertility and mortality were to remain constant at current levels, the ageing of populations will move forward in the decades to come. Demographic history has already built in an age-cohort structure that makes population ageing almost unavoidable in the future. Second, although the populations of developing countries are considerably younger than those of the developed countries, it is foreseen that the former will follow the course of population ageing experienced by the latter. In addition, the pace of population ageing in Africa, Asia, Latin America and Oceania will surpass that of the remaining major regions of the world. Finally, it should be stressed that until nearly the middle of the 21st century, the growth in the size of the population 60 years and older will be determined solely by mortality variations because these persons have already been born.

    The considerable numbers of elderly in the Asian and Pacific sub-regions will require a vast effort on their part, and on the part of families, governments and private organizations, to ensure that the elderly can be as useful and as productive as possible.

REFERENCES

Afzal, M. 1994. Local Level Policy Development on Consequences of Ageing: Country Case Study in Pakistan. Asian Population Studies Series No. 131-D. Bangkok, Thailand: Economic and Social Commission for Asia and the Pacific.

Arshat, H., P.C. Tan and N.P. Tey. 1989. The Ageing of Population in Malaysia. Asian Population Studies Series No. 96. Bangkok, Thailand: Economic and Social Commission for Asia and the Pacific.

Chen, A.J. and G.W. Jones 1989. Ageing in ASEAN: Its Socio-Economic Consequences. Singapore: Institute of Southeast Asian Studies.

Choe, E.H. 1989. Population Ageing in the Republic of Korea. Asian Population Studies Series No. 97 . Bangkok, Thailand: Economic and Social Commission for Asia and the Pacific.

Concepcion, M.B. 1994. 'Implications of increasing roles of women for the provision of elderly care'. In Ageing and the Family. United Nations publication (Sales No.E.94.XIII.4).

Domingo, L.J. 1989. Status of research on and programmes for the elderly in the Philippines. Paper presented at the Workshop on Research on Aging in the Developing World, XIV International Congress of Gerontology, Acapulco, Mexico, June.

Durand, J.D. 1975. The Labor Force in Economic Development. Princeton, N.J.: Princeton University Press.

Hugo, G. 1988. 'The changing urban situation in Southeast Asia and Australia: some implications for the elderly.' In Aging and Urbanization. United Nations publication (Sales No. E.91.XIII.12).

Kono, S. 1988. 'Social consequences of changing family and household structure associated with an aging population'. In Economic and Social Implications of Population Aging. United Nations ST/ESA/SER.R/85.

1994. 'Ageing and the family in the developed countries and areas of Asia: continuities and transitions'. In Ageing and the Family.. United Nations publication (Sales No.E.94.XIII.4).

Lucas, R.E. and O. Stark. 1984. Motivations to remit. Migration and Development Program Discussion Paper, No. 10. Cambridge, Massachusetts: Harvard University.

Myers, G.C. 1992. Demographic aging and family support for older persons. In Family Support for the Elderly: The International Experience. H. Kendig, A. Hashimoto and L. Choppard (eds.). Oxford: Oxford University Press.

1994. 'Marital-status dynamics at older ages'. In Ageing and the Family. United Nations publication (Sales No. E.94.XIII.4).

Perera, P.D.A. 1989. Emerging Issues of Population Ageing in Sri Lanka. Asian Population Studies Series No. 98. Bangkok, Thailand: Economic and Social Commission for Asia and the Pacific.

Phalakornkule, S. 1994. Country Policy Development for the Elderly at the Local Level in Thailand.. Asian Population Studies Series No. 131-E. Bangkok, Thailand: Economic and Social Commission for Asia and the Pacific.

Tabah, L. 1988. 'The economic and social consequences of demographic aging'. In Economic and Social Implications of Population Aging. United Nations publication ST/ESA/SER.R/85

United Nations. 1988. 'Global trends and prospects of aging population structures'. In Economic and Social Implications of Population Aging. United Nations publication ST/ESA/SER.R/85.

1994. Ageing and the Family. United Nations publication (Sales No. E.94.XIII.4).

Wirakartakusumah, M.D. 1994. Local Level Policy Development on Consequences of Ageing: Indonesia Case. Asian Population Studies Series No. 131-C. Bangkok, Thailand: Economic and Social Commission for Asia and the Pacific.