INTRODUCTION Brazil is experiencing a deep demographic change with significant transformations in its age structure. The main determinant of these changes has been fertility, whose decline is one of the more intense observed among the most populous countries in the world in modern times. The Brazilian Total Fertility Rate dropped from 5.8 children per woman, in the mid 60’s (Carvalho, 1978) to 2.3 children per woman during the second half of the 90’s (IBGE, 2001). It will probably reach the replacement level at the beginning of the first decade of this century and would continue to decline at moderate rates afterwards (IBGE, 1997). The Brazilian population growth was about 3.0% per year before the onset of the fertility decline, i.e., during the sixties. By the end of the 20th century, however – during the period 1991/2000–, the population growth averaged as much as 1.6% per year. Available projections agree with the persistence of decline in the population growth rate in the short and long run; it would go from 1.0% per year, between 2000 and 2010, to 0.3% per year, between 2045 and 2050 (IBGE, 1997; United Nations, 2001). As a result of these changes, it is estimated that, only between 1980 and 2000, approximately, 35 million children had not been born in Brazil. Had the 1980’ fertility rates remained constant, at least another 35 million would be born between 2000 and 2010. As for the elders, defined here as those aged 65 or more, around 3.7 million were added to them during the period 1980 and 2000. Between 2000 and 2010, a time interval corresponding to half of the previous period, the absolute number of new elder should be almost the same figure, that is 3.3 million people, thus signalling the acceleration of the changing process. The paramount structural changes in the Brazilian age pattern is better seen by considering its age composition over the 100-year interval between 1950 and 2050. According to the United Nations (United Nations, 1998), the percentage of the elderly population, initially below 3%, will reach 18% at the end of the period. The deep reduction in the proportion of the youngsters, i.e., those aged 15 years or less, will be of equally great magnitude. It was 41.6% in 1950, 28.8% in 2000, and will represent only 19.9% in 2050. In the long run children and the aged will have approximately the same weight in the total population. Consequently, Brazil will experience one of the most intense processes of population ageing among the most populous countries in the world during this period. Such considerable variations in a short time will have important repercussions on the Brazilian society, the type of social development programs to be implemented and the quality of life of the population. This ageing process occurs in a society that despite its fast modernisation simultaneously shows one of the world's most unequal income distributions, a significant number of people living below the poverty line and experiencing very low economic growth rates and high unemployment rates since the 80’s. This paper outlines a general view of the Brazilian demographic ageing process during the period between 1950-2050. It strengthens the evolution of this process and the consequent changes in the generations’ interrelationships through the demographic dependency ratios. Some of its most specific characteristics as feminisation of the ageing process and the survivorship of the elder are also emphasised. The international context is considered for better understanding of the process.[2] CHANGES IN THE BRAZILIAN AGE STRUCTURE The Brazilian population, that represents approximately one third of the Latin-American population, grew from 54 million to 170 million during 1950 and 2000, being estimated that it will reach 244 million in 2050[3]. In the five final decades of the 20th century the population aged less than 15 years grew from 22 to 50 millions and would remain around this figure with small oscillations, up to 2050. The population above 65 years old on the other hand, increased from 1.6 million in 1950, to 8.7 million in 2000, and, it will probably reach 42 million in 2050. Thus, while it is expected the young population twofold, the elderly would grow approximately 26 times in these 100 years.[4] A synthesis of this change is shown in Figure 1. The chronological sequence of the age and sex distribution is initiated having a typical young format pyramid that remained virtually constant up to 1970, becoming practically rectangular at the end of the period (2050). This series points out a process toward a relatively quick new stabilisation when compared to the demographic evolution had by currently developed countries.[5] Figure 1 Brazil – Relative age and sex distribution (%) – 1950, 1970, 1990, 2000, 2025, 2050 Source: United Nations (1999) Graphs 2 and 3 illustrate, in detail, the different variation by age, in absolute and relative terms of the Brazilian population age structure. Figure 2 highlights the most significant variation in volume during the first 25 years, when, as it was seen in the case of the population pyramids, the variation in the structure was almost non-existent. The young population contingent (less than 15) stabilises around 1985; inside this large age group, the children up to 5 years old decline in absolute terms afterwards. The maximum volume of youngsters, slightly above 51 million people, corresponds to 1990; absolute figures decrease in 2000, tending to stabilise numerically afterwards. Since the 60’s, when the levels of national fertility started to decline, the absolute increment in the successive cohorts had been lesser each time until 1985. This trend remained in the following decade, where the absolute variations – positive or negative – are cyclical, tending to stagnate around a null asymptote value. The inertial effect of this destabilisation, initiated with the fertility decrease around the 70’s, which appeared in the 90’s in terms of the absolute number of births, will remain over the future decades with the corresponding variations in the social and economic demand that each age group generates. This aspect will be brought up again later. Graphic 3 – Brazil – Accumulated Population until Indicated Age – 1950-2050 Source: United Nations (1999) Graph 3, with the proportional age distribution of the total population, shows the transformation mentioned above. The curves that suggest a moving net, evidence, proportionally, more accentuated changes among the young population (left side of the Figure) after 1965 –beginning of the destabilisation– and until approximately 2005, when it would tend swiftly toward an new quasi–stabable population. Graph 3 - Brazil - Age Distribution of Total Population - 1950 -2050 Source: United Nations (1999). INTERRELATIONSHIPS BETWEEN AGE GROUPS Table 1 quantifies the transformations of the Brazilian age structure. It includes measures of the age structure according to large age groups and the values of the mean and median ages. Before analysing the relationship between the large age groups, it is worth to observe the variation in the average values of the age distribution, particularly the median age. It eloquently shows the speed of change, since it duplicates over the period of reference. While at the beginning of the period, 50% of the population was less than 20 years old, in 2050, such proportion would correspond to those up to 38 years old. In the Latin-American context, this increase would be smaller than the cases of Mexico, Colombia and Cuba only; in any case, it would be larger than the foreseen for Latin America, as a whole, and, even, for Asia.[6] Table 1 - Brazil – Total population and proportional distribution by large age groups, Dependency Ratio and its components, and Mean and Median Age – 1950-2050 Years Total Population Relative age distribution (%) Dependency Ratios (%) Mean Age Median Age 00-14 15-64 65 + TDR(a) CDR(b) ADR(c) (1) (2) (3) (4) (5) (6) (7) (8) (9) 1950 53.977 41,6 55,5 3,0 80,3 74,9 5,4 23,5 19,2 1960 72.757 43,3 53,5 3,3 87,1 80,9 6,1 23,2 18,6 1970 96.021 42,4 54,0 3,7 85,3 78,5 6,8 23,6 18,6 1980 121.673 38,1 57,8 4,2 73,1 65,9 7,2 24,7 20,3 1990 147.939 34,8 60,9 4,3 64,1 57,0 7,1 26,2 22,7 2000 170.116 28,8 66,0 5,1 51,4 43,7 7,8 28,7 25,9 2010 190.876 25,3 68,4 6,3 46,2 37,0 9,2 31,3 29,4 2020 209.734 23,3 68,2 8,5 46,7 34,2 12,5 33,8 32,9 2030 225.161 21,5 66,6 11,8 50,0 32,3 17,7 36,2 35,5 2040 236.541 20,6 65,0 14,5 53,9 31,7 22,2 37,9 37,0 2050 244.228 20,1 62,6 17,3 59,6 32,0 27,6 39,0 38,2 Source: United Nations (1999) As a first approach to changes in the intergenerational relationships it can be seen that, due to ageing process, there are substantial modifications in the demographic dependency ratio. Although it is a purely demographic measure, since it takes the age as the only variable, in highly indicative of important intergenerational social and economic relationships that will happen. Both, the Child Dependency Ratio (CDR - ratio of population below 15 to population 15 to 64) and Aged Dependency Ratio (ADR - ratio of population above 64 to population 15 to 64), although with different evolution, determine the declining trend of the Total Dependency Ratio (TDR - ratio of population below 15 and population above 64 to population 15 to 64) that was practically constant and at high levels (around 80-90%) up to 1970. It starts to decrease exclusively due to the proportionally higher reduction of the youngsters (column 2). Thus the total-dependency ratio would reach its minimum values during the period 2010 to 2020, due to fast reduction of the child-dependency ratio (CDR - column 6). The aged-dependency ratio component (ADR - column 4), that remains almost non-expressive until, say, 2010, will present significant increases after 2020/30, due to the increase of the elders. The significant changes in the path of the dependency ratios over the period 1950-2050 are plotted in Figure 4. Figure 4 - Brazil – Total, Child, Aged Dependency Ratios - 1995-2050 Source: Table 1. Summarising, for a long while the dependency ratio should remain descending below historical levels previous to the beginning of the age structure changes. In this transition the child-dependency will decline quickly while the aged-dependency increase – slowly at the beginning, and faster afterwards. This combination will certainly bring about important implications for demand and allocation of public and private resources among both youngsters and elders. As to the youngsters, such drastic variation in the ratios, inclusive in absolute number, has important repercussions in the investments needs. It certainly creates favourable chances to the improvement of the educational and health level, for example, that would guarantee, in the future, a better quality of life for the Brazilian population. The period that will extend approximately until 2025 was characterised as a “window of opportunity” by Carvalho and Wong (1995) or ‘demographic bonus’ related to similar impressive structural changes in other underdeveloped settings. This relatively optimistic approach rely in the assumption that it is a period where the quantitative pressures imposed by the child-dependency attenuates, while quantitative demands of the elder are not yet growing expressively[7]. Thus, a favourable conjuncture draws round favouring both: Expanding coverage and improve quality of socio-economic resources to attend young population. Time for rearrangements to adequately face the challenges imposed by the foreseen growth of the elderly population. BRAZILIAN AGING POPULATION IN THE INTERNATIONAL CONTEXT Transformations mentioned in the previous lines show, above all, the ageing process of the Brazilian population. This process is analysed with more detail next considering its volume and intensity in the international context. Intensity is measured comparing the Ageing Index[8] since its evolution throughout the time is directly related to the population ageing process. The Ageing Index is easily understood and highly sensitive to the variations in the age distribution caused by the fertility decline which is the case for Brazil. Table 2 portrays the probable evolution of the aged population, its relative weight and the Ageing Index in Brazil. It also includes world figures classified according to the United Nations concept of more and less developed countries and some selected regions[9]. The data points to the existence of distinct worlds when referring to the aged population: one of them, compounded by the less developed countries (LDC), in which the number of elders is large but has relatively small weight. The other formed by the more developed countries, in which the absolute size of the elderly population is small but their relative participation in the total of the population is high. In the LDC, the elders were approximately 250 million in 2000 but they corresponded to only 5.1% of the total population. In the more developed countries, this contingent reached 170 million representing, however, 14.4% of the total population. While, in 1950, the aged population of the LDC was practically equal to the more developed (60.6 and 64.2 million, respectively), in 2050 it would be almost four times larger in the LDC (1.2 billion against 300 million). It should be notice, in addition, the speedy ageing process of those regions that had initiated recently the demographic transition, which is the case for Latin America and the South–eastern Asia. Taking 1950 as ground for comparison, these regions would present a faster ageing process since its aged population may be 20-fold over this 100-year time interval. The Ageing Index, that in 1950 presented values below 10 (i.e., one elder for every 10 youngsters) for both regions, would be nearly 10-fold in 2050 when it will be more than 80 elders per 100 youngsters. Despite this huge ageing process, however, the large difference regarded Europe would remain; in Western Europe, for instance, youngsters and elders almost equal (Ageing Index = 93,5) already at the beginning of the 21st century; the ancients, however, would be significantly more than the youngsters afterwards. In sum, the ageing process between 1950 and 2050 measured trough the variation of the Ageing Index would be faster in the LDC than in the more developed countries: the Ageing Index would increase in 617% and 486%, respectively.[10] Table 2 – Brazil and Selected Regions – Population above 65 years old, Percent of Elderly and Ageing Index - 1950-2050 Brazil and Selected Regions 1950 1975 2000 2025 2050 Population above 65 years old (in thousands) Brazil 1.604 4.247 8.709 21.919 42.243 Less developed countries(*) 66.644 118.027 248.221 561.749 1.163.054 More developed countries (*) 64.202 116.029 171.069 253.912 299.249 Latin America and Caribbean 6.178 14.061 28.080 67.472 135.666 South–eastern Asia 6.735 10.691 24.371 58.100 131192 Western Europe 14.373 23.513 29.151 41.749 46.918 Percent of elderly in the total population Brazil 3.0 3.9 5.1 10.1 17.8 Less developed countries(*) 3,9 3,9 5,1 8,5 15,0 More developed countries (*) 7,9 10,7 14,4 20,9 25,9 Latin America and Caribbean 3,7 4,4 5,4 9,7 16,8 South-eastern Asia 3,7 3,3 4,7 8,5 16,7 Western Europe 10,2 13,9 15,9 22,7 27,5 Ageing Index (per cent) Brazil 7,2 9,7 18,1 46,8 92,9 Less developed countries(*) 10,3 9,4 15,7 34,1 73,9 More developed countries (*) 28,9 44,2 79,1 133,1 169,3 Latin America and Caribbean 9,2 10,6 17,2 41,1 83,7 South–eastern Asia 9,4 7,7 15,0 37,0 85,2 Western Europe 43,8 61,0 93,5 152,3 185,8 Source: United Nations (1999) (*) See previous footnote Regarding Brazil and compared to Latin America and Caribbean, Table 2 shows important findings between 2000 and 2050: The country will present similar proportional growth in the size of the elderly population (around 385%); notwithstanding, a larger relative increase in the proportion of the elderly related to the corresponding total population (249% against 211%, for Latin America) and, an even bigger relative increase in the Ageing Index (313% against 286% for Latin America). This apparent contradiction on similar proportional growth in the size of the elderly population and a bigger relative increase in the Ageing Index is due to an inertial effect. In fact, that the absolute number of the elders in 2050 will be less affected by the time of the onset and speed of fertility decline in the Region and in Brazil than the other two measures. The data points to a faster ageing process in the Brazilian population than in the Region, as a whole. Additionally, by accepting the United Nations’ projections, Ageing Index in Brazil (93) would still be much lower than in Western Europe (186) and in South–eastern Asia (115) in 2050. As a matter of fact it is impressive the ageing process to be experienced by the latter between 2025 and 2050, when its Ageing Index would change from 37 to 85 elders per 100 youngsters. Finally, according to the United Nations projections’ (United Nations, 1999), the Brazilian ageing process, measured by the evolution of the Ageing Index would be among the most intense seen in the period 1950-2050. These data characterise Brazil among the 35 more populous countries in the world with the 4th most intense ageing process, following the Republic of Korea, Thailand and Japan. In this one hundred-year period the Brazilian Ageing Index would be 12-fold, while the corresponding to Republic of Korea would be 21-fold, 19-fold in Thailand, and around 16-fold in Japan. (See Table A2). Graph 5 – Brazil and Selected Countries – Variation of the Ageing Index – 1950-2050 Index: (1950 = 100) Source: Table A2. Figure 5 displays the evolution between 1950 and 2050 of the Ageing Index corresponding to Brazil. It also includes the Republic of Korea, Japan, Spain, United States and the Democratic Republic of Congo (first and last countries corresponding to the most intense and slowest ageing processes among the 35 most populous countries in 2000, as shown in Table A2 in Annex). The intense ageing process of the Brazilian population contrasts with the one observed among the countries with already old age structure and where the transition to low fertility levels started well before and took a longer time interval. Consequently, in these countries the institutional adjustments and the social and economic transformations needed to assist an aged population also had a longer time interval. SOME CHARACTERISTICS OF THE BRAZILIAN AGING PROCESS Evidence of the relatively fast ageing process that the country might experience has been given already; this item analyses some more specific characteristics as the sex differential and, briefly the age groups inside the elder population.[11] Feminisation Ageing happens differently by being man or woman. Although the sex ratio (SR) at birth favours men, females have lower mortality levels by age vis–a–vis the males; thus there are more women than men from, say, age 20 onwards. Usually lower sex ratios are located at older ages, where the cumulative effect of higher male mortality is more intense. Because of the probable mortality trends, an increasing feminisation in the Brazilian ageing process is expected. In fact, it is presumed that, at least in the long run, improvements in the Brazilian life expectancy will happen keeping constant the current and significant mortality differentials by sex. Data referring to Brazil (Table 3) show the probable evolution of the sex differences among the elders, over the period 1950-2050. The already higher proportion of elderly women existing in 1950 (almost 25% more women than men) should increase substantially (there would be 40% more women than men would by 2050). This would result in an increment of their participation in the elderly population of 55,5%, in 1950, to 58,4%, in 2050, thus extending the distance between sexes of the Ageing Index. Additionally, two aspects must be considered when dealing with the feminisation in the expected ageing process. In the first place, this sub–population would probably have more improvements in their mortality levels, above the projected ones, perhaps, that will enlarge sex differentials. Thus, certainly the data above are not completely disclosing the degree of feminisation that the Brazilian aged population might reach. Table 3 – Brazil – Elderly Population, Sex Ratio, Proportion of Total Population and Ageing Index by Sex – 1950-2050 Years Elderly Population (in thousand) Sex Ratio (%) Proportion of Total Population (%) Ageing Index (per cent) Men Women Men Women Men Women 1950 715 891 80,2 2,7 3,3 6,3 8,0 1960 1.068 1.315 81,2 2,9 3,6 6,7 8,4 1970 1.614 1.918 84,2 3,4 4,0 7,9 9,5 1980 2.378 2.677 88,7 3,9 4,4 10,2 11,6 1990 2.886 3.505 82,5 3,9 4,7 11,1 13,8 2000 3.790 4.919 77,0 4,5 5,7 15,2 20,4 2010 5.094 6.893 73,9 5,4 7,1 20,7 29,0 2020 7.509 10.345 72,7 7,3 9,7 30,1 43,1 2030 11.105 15.476 71,8 10,1 13,4 44,8 65,1 2040 14.131 20.052 70,4 12,3 16,5 56,8 84,2 2050 17.560 24.683 71,2 14,8 19,7 70,1 103,0 Source: United Nations (1999). In the second place, it must be emphasised, however, that the striking unbalance between elderly men and women is not a Brazilian particularity. Indeed, the sex ratios shown in Table 4 indicate that Brazil is currently close to the Latin American average despite the fact that –always according to the United Nations' projections– it is among the lower ratios of the world[12]. Notice that among the more developed countries, after reaching a minimum sex ratio in the 70s (62,9, in 1975), they will probably have a persistent, yet slow return to higher sex ratio (74.1 in 2050).[13] Table 4 - Brazil and Selected Regions - Elderly Sex Ratio - 1950-2050. Brazil and Selected Regions 1950 1975 2000 2025 2050 Brazil 80.3 86.9 77.0 72.1 71.1 Less developed countries(*) 80,3 84,9 84,1 83,4 82,7 More developed countries (*) 71,0 62,9 65,4 72,4 74,1 Latin America and Caribbean 84,7 85,4 78,4 75,7 76,2 South-eastern Asia 83,7 82,2 78,9 78,4 79,8 Western Europe 74,6 62,3 64,9 75,6 76,6 Source: United Nations (1999). Ageing among the elderly The effect of the fertility decline, combined with survivorship improvements among the elders tend to benefit the eldest population, that, on turn, are responsible for higher growth rates of the elderly population together. Thus growth among the “old–old” i.e., above the age 80, at least during a certain period, will have the highest growth rate. Being female survivorship much higher than male at these extreme ages of life, the feminisation of the “old–old” is, to a large extent, due to the survivorship of the eldest age groups. Table 5 presents the elders above the ages of 65, 70 and 80, the respective relative participation and growth rates for the period 2000-2050. Between 2000 and 2050 the size of the elderly population would be 5-fold, and the older the age group, the greater the increase rate would be. The volume of people aged 70 or more would be nearly 7-fold; yet, this increment should be less than among the “old–old’ that would be nearly 8 times and would extend its participation to more than 20% of the aged population. Finally, the high growth rates, above of 4,0% for the periods after 2020 in the ages above 70, will be product of the combined effect of the high fertility of the past and the improvements in the mortality levels that is expected for these ages. Table 5 – Brazil – Elderly Population by Age Groups – 2000-2050 (in thousand) a) Age groups (cumulated) Years 65 or more 70 or more 80 or more Total Proportion (%) Total Proportion (%) Total Proportion (%) 2000 8.709 100,0 5.371 61,7 1.219 14,0 2010 11.987 100,0 7.507 62,6 1.865 15,6 2020 17.854 100,0 10.798 60,5 2.763 15,5 2030 26.581 100,0 16.664 62,7 4.179 15,7 2040 34.183 100,0 23.081 67,5 6.606 19,3 2050 42.243 100,0 28.916 68,5 9.263 21,9 b) Average Annual Growth Rate (%) Interval 65 and above 70 and above 80 and above 2000/10 3,19 3,35 4,25 2010/20 3,98 3,64 3,93 2020/30 3,98 4,34 4,14 2030/40 2,52 3,26 4,58 2040/50 2,12 2,25 3,38 Source: United Nations (1999) THE SOCIAL CONDITIONS OF THE AGED IN BRAZIL One of the first assemble evidences about the social-demographic situation of the elders in Brazil was elaborated by the Seade Foundation (1990) that centred its investigations on the São Paulo State elderly population. Several researches have been released since then and those presented at National Meetings of the Brazilian Association of Population Studies (ABEP) are of particular importance[14]. Camarano (1999) organised the most recent publication focusing the elder at national level tracing a general view of recent social conditions of the aged, based on secondary data and field research. A review of the elder life conditions allow us to infer that, generally speaking, the elders have better life conditions than the rest of the population. They have higher income (either wages, pensions, saving returns, etc); higher proportions own the houses they live in and contribute significantly to the family budget. The latter applies particularly to elder man and to his work income, i. e. salaries (Conceição, 1998; Wajnmam et alii, 1999). In economic crisis and huge unemployment context, the financial contribution of the elder to the family support is fundamental. Furthermore, there are evidences of expressive increase in income level of the elder during recent periods, to which much would have contributed the welfare state. Expressive increments would have occurred between women, mainly due to the extension of the welfare state coverage in rural areas. It was also the women who experienced major improvements in the literate and enrolment ratios. Continuation of the elder as important contributor to family support, in the future, is questioned, when the current difficulties for generating and cumulating wealth among young adults are considered. (Camarano et alii, 1999; Camarano, El Ghaouri, 1999). It is known that those difficult are caused by conjunctural and structural reasons, among them, the high unemployment rates, the growth of the ‘non–formal’ work market that prevails in the Brazilian economy and the global process of destroying unskilled jobs, very often suitable for the younger labour force. The importance of the governmental transfers to the elderly population via welfare state is the decisive element for maintaining a reasonable standard of living at the final stage of life. Turra (2000) shows that through welfare state, an enormous part of public resources is transferred to the population; i. e., payments made to pensioners and retired, in one way or another, are passed through to children and relatives of different ages. Altogether, certainly, the most benefited are the aged, that compare to the younger population, receive higher volumes of governmental transfers. In the public sphere, the rural Welfare State constitutes a significant element of social protection to the aged; it contributes to expand their family income, which, as a rule, are generally very low. Rural welfare state has also represented an important channel social integration for women. A household survey carried out in the two most populous regions of Brazil (Northeast and Southeast), found that more than 60% of the retired and pensioners were women (Delgado and Cardoso Jr, 1999). Although this can be reflect of the sex composition of the elderly, where, as has seen, the women predominate, it is worth no note that in the rural area, 87% of the total income of women over 60 years old come from welfare state. As for men, this percentage is of 55%[15]. Other data also show that the welfare state constitutes an important source in small cities and villages where the welfare benefits constitute a significant fraction of the monetary resources that circulate there. The welfare state benefit and its importance for the family income constitute a central element so that, in the Northeast, the direct intergenerational transfers of income go from the aged to their children. This diverges from results from São Paulo, where the stream has the expected opposite direction, as Saad (1999) demonstrates. The author shows that, in the Northeast, the co-residence between generations are much more due to the adult children's needs than of the aged parents' (Saad, 1996). Ferreira (2001) draw similar conclusions related to the peripheral areas of the Metropolitan City of Belo Horizonte. In the same sense and based on secondary data, Camarano et alii (1999) find expressive proportions of young adults and children sharing the household headed by an elder. At the aggregate level, Turra (2000) points out that, the intergenerational stream flows from the older to the younger in Brazil, opposing Caldwell's (1976) suggestion. He builds up further elements for explaining the persistence of the fertility decline in Brazil by showing that bringing up children is expensive for parents, in net terms. Camarano (1999) also affirms, based on the studies congregated in her publication, that in Brazilian families, the intergenerational transfers have bi-directional flows, particularly due to the economic crises that have affected more the young population. Barros et alii (1999) find that the percentage of poor people tends to be less among the elder than among the rest of the population, which, to a large extent, is explained by their earnings from the welfare state and their higher capability of saving. The authors also find that the average income of the elder is above the per capita income of most of the households, particularly among the poor ones. Thus, besides the fact that the elder do not represents an increase in the economic dependency ratio, they contribute to mitigate poverty among the non-aged. Facing the fast Brazilian ageing process, two important questions come out: will it be possible to continue with this relatively positive diagnoses for the elder that, furthermore, transfers wealth to the younger generations? Will the Welfare state have – with such an important social role– conditions to continue to do so in the long run? CONCLUSIONS The fast and intense fertility decline in Brazil does not have equivalent to any historic experiences in the developed countries. In conjunction with the equally fast reduction in the country's population growth rates, it is experiencing a deep age pattern transformation in a short period of time. While showing one of the fastest demographic ageing processes among the most populous countries in the world, Brazil is facing a great challenge that goes beyond demographic matters. The key issue is how to bring together economic development and reduction of both the abyssal social differences and high levels of poverty that stigmatise the country, preserving, at least, the current life conditions. During the period of time, when the decrease of the young population is not yet compensated by increases in the elderly population, Brazil experiments a demographic bonus, represented by a declining total dependency ratio, that will be further extended without, however, reaching previous historical levels. In this change, the dependency eminently young at the beginning will inexorable age, changing the nature of the social demands by pressuring health system and, particularly, welfare state. Public policies' efforts in order to equate the Brazilian demographic transition and the needs that this peculiar population growth originate are still at the incipient phase; there is no evidence, however, that the elders are currently in worse life conditions than the remaining social groups. On the contrary, the governmental transfers are favourable to them and the rural welfare state, for example, constitutes a clear example. Furthermore, there are evidences that the intergenerational income streams still occur from the elder to the younger and that, among the poor, the situation of the elders is rather favourable related the rest of the population. Given the fast ageing process, however, Brazilian society has many challenges to confront in order to guarantee quality of life for the elder. When formulating social security policies special attention must be given, for instance, to the gender issue, since women are on average, in disadvantage vis–a–vis men. Although they are responsible for most of the non-paid household/family work, women have, in general, lower participation rates in the formal job market, are usually found among the lower income stratum and very often do not have social security. Finally, although the combination of demographic bonus and the appropriate welfare state and social support policy for the elder would much contribute to guarantee adequate quality of life to the elder, there are other important issues to consider: the increasing number of elderly population, in both relative and absolute terms. the lack of resources and permanent competitive social demands, typical of an underdeveloped society. the evidence that expressive segments inside the elderly population are above the average standard of living These aspects bring about the importance of accurately identify inside the elderly population the less privileged groups, and consequently the most vulnerable. They should constitute the public policies' preferential target face to the chronic scarce resources and imminent increase of ageing. References BARROS, Ricardo P. de; MENDONÇA, Roseane; SANTOS, Daniel. (1999) Incidência e natureza da pobreza entre idosos no Brasil. In: CAMARANO, Ana Amélia (Org.) Muito além dos 60: os novos idosos brasileiros. Rio de Janeiro: IPEA, 1999, p.221-249. CALDWEL, John C. (1976) Toward a restatement of demographic theory. Population and Development Review, v.2, n.3/4, p.321-366, Sep./Dec.. CAMARANO, Ana Amélia; BELTRÃO, Kaizô, I.; PASCOM, Ana Roberta P.; MEDEIROS, Marcelo; GOLDANI, Ana Maria.(1999) Como vive o idoso brasileiro? In: CAMARANO, Ana Amélia (Org.) Muito além dos 60: os novos idosos brasileiros. Rio de Janeiro: IPEA, p.19-71. CAMARANO, Ana Amélia; El GHAOURI, Solange K.(1999) Idosos brasileiros: que dependência é esta? In: CAMARANO, Ana Amélia (Org.) Muito além dos 60: os novos idosos brasileiros. Rio de Janeiro: IPEA, p.281-304. CARVALHO, José Alberto M. de; WONG, Laura R.(1995) A window of opportunity: some demographic and socio-economic implications of the rapid fertility decline in Brazil. Belo Horizonte: CEDEPLAR/UFMG. CONCEIÇÃO, Maria Cristina G. da. (1998) Jefaturas, tipos de ingressos y estrutura de los hogares con indivíduos mayores de 60 anos em Brasil. In: ENCONTRO NACIONAL DE ESTUDOS POPULACIONAIS, 11, 1998. Caxambu. Anais...ABEP: Belo Horizonte, p.2311-2339. DELGADO, Guilherme C.; CARDOSO Jr., José Celso.(1999) O idosos e a previdência rural no Brasil: a experiência recente de universalização. In: CAMARANO, Ana Amélia (Org.) Muito além dos 60: os novos idosos brasileiros. Rio de Janeiro: IPEA, 1999, p.319-343. FERREIRA, Frederico P. M. (2001). Estrutura Domiciliar e Localização: um estudo dos domicílios com idosos em Belo Horizonte - 1991. Belo Horizonte: Cedeplar/UFMG, Tese de Doutorado. IBGE.(2001) Sumário dos Indicadores Sociais. Rio de Janeiro: IBGE. IBGE.(1997) Projeção da População do Brasil para o período 1980-2020. Rio de Janeiro: IBGE-DEPIS, mimeo. MOREIRA, Morvan de M.(1997) Envelhecimento da População Brasileira. Belo Horizonte: CEDEPLAR-UFMG, Tese de Doutorado. MOREIRA, Morvan de M.(1998) Envelhecimento da população brasileira: intensidade, feminização e dependência. Revista Brasileira de Estudos Populacionais, Campinas, v.15, n.1, jan./jun., p.79-93. PRESSAT, Roland.(1970) El análisis demográfico. Havana, Instituto del Libro. SAAD, Paulo M.(1999) Transferência de apoio entre gerações no Brasil: um estudo para São Paulo e Fortaleza. In: CAMARANO, Ana Amélia (Org.) Muito além dos 60: os novos idosos brasileiros. Rio de Janeiro: IPEA, p.251-280. SAAD, Paulo M.(1996) Living arrangements of the elderly and the family in Northeast and Southeast Brazil, 1980. IN: ABEP. ENCONTRO NACIONAL DE ESTUDOS POPULACIONAIS, 9, Caxambu. Anais...Belo Horizonte: ABEP. SEADE- FUNDAÇÃO SISTEMA ESTADUAL DE ANÁLISE DE DADOS.(1990) O idoso na Grande São Paulo. São Paulo: SEADE. TURRA, Cássio M.(2000) Contabilidade das gerações: riqueza, sistema de transferências e conseqüências de mudanças no padrão demográfico brasileiro. Belo Horizonte, CEDEPLAR/UFMG (Dissertação de Mestrado). UNITED NATIONS.(1999) The Sex and Age Distribution of Populations: the 1998 revision. New York. UNITED NATIONS.(2001) World population prospects: the 2000 revision – highlights. (www.um.org.esa/population/wpp2000.htm) WAJNMAN, Simone; OLIVEIRA, Ana Maria H. C. de; OLIVEIRA, Elzira Lúcia.(1999) A atividade econômica dos idosos no Brasil. In: CAMARANO, Ana Amélia (Org.) Muito além dos 60: os novos idosos brasileiros. Rio de Janeiro: IPEA, p.181-220. WOOD Ch. H., CARVALHO, J.A.M. and ANDRADE, F.C.D. (2000): Notas acerca das categorias de cor dos Censos e sobre a classificação subjetiva de cor no Brasil: 1980/01. In: ENCONTRO NACIONAL DE ESTUDOS POPULACIONAIS, 12, 2000. Caxambu. Anais...ABEP: Belo Horizonte Table A1: Brazil – Population by Age and Sex - 1950, 2000 e 2050 Age Group 1950 2000 2050 Total Men Women Total Men Women Total Men Women Total 53.975 26.776 27.199 170.115 84.003 86.112 244.230 118.653 125.577 00-05 8.984 4.533 4.451 15.994 8.142 7.852 16.374 8.373 8.001 05-10 7.013 3.523 3.490 16.123 8.196 7.927 16.330 8.346 7.984 10-15 6.436 3.230 3.206 16.960 8.614 8.346 16.282 8.314 7.968 15-20 5.405 2.711 2.694 17.389 8.812 8.577 16.169 8.247 7.922 20-25 5.016 2.515 2.501 16.525 8.263 8.262 16.011 8.148 7.863 25-30 4.127 2.068 2.059 14.601 7.232 7.369 15.912 8.071 7.841 30-35 3.445 1.718 1.727 13.650 6.736 6.914 15.887 8.024 7.863 35-40 3.071 1.525 1.546 13.039 6.400 6.639 15.854 7.961 7.893 40-45 2.592 1.277 1.315 10.977 5.349 5.628 15.543 7.748 7.795 45-50 2.082 1.010 1.072 9.103 4.404 4.699 15.050 7.437 7.613 50-55 1.816 856 960 7.121 3.412 3.709 14.452 7.058 7.394 55-60 1.362 633 729 5.399 2.551 2.848 14.079 6.765 7.314 60-65 1.022 461 561 4.526 2.098 2.428 14.042 6.599 7.443 65-70 723 330 393 3.338 1.511 1.827 13.327 6.070 7.257 70-75 463 206 257 2.570 1.130 1.440 11.335 4.884 6.451 75-80 267 116 151 1.582 670 912 8.318 3.340 4.978 80-85 153 63 90 790 320 470 5.341 1.987 3.354 85-90 326 124 202 2.832 951 1.881 90-95 88 31 57 916 277 639 95-100 14 4 10 163 47 116 100 + 1 - 1 11 4 7 Source: United Nations (1999) Table A2 – More Populous Countries in 2000 – Ageing Index - 1950-2050 Countries Ageing Index 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 Spain* 26,9 29,9 35,0 40,2 71,3 117,0 138,5 174,3 239,0 293,1 311,4 Italy* 31,4 37,6 44,3 59,0 96,6 127,5 162,2 208,1 250,5 288,1 290,8 Japan * 13,9 19,0 29,4 38,4 65,1 115,2 145,5 187,8 202,9 216,3 229,6 Germany * 41,9 54,1 59,0 84,4 93,0 105,5 149,5 165,8 194,1 218,5 212,1 Ukraine 27,9 29,3 36,6 55,8 56,6 79,6 106,3 117,3 150,2 168,3 195,9 Russian F. 21,4 21,0 28,9 47,2 43,7 68,9 82,2 98,0 136,1 146,4 174,5 Poland 17,8 17,2 30,6 41,7 40,0 61,8 76,9 103,5 135,5 143,1 167,6 Korea 7,3 7,9 7,8 11,2 19,4 31,3 46,9 69,1 108,4 139,9 154,3 U Kingdom * 48,0 50,2 53,3 72,1 82,1 85,2 102,3 119,5 138,3 156,0 154,1 France * 50,1 44,1 51,8 62,7 69,1 85,2 95,3 116,6 136,7 152,7 152,6 Canada* 25,9 22,4 26,0 41,4 54,1 67,4 86,3 108,6 133,2 142,0 140,0 China 13,4 12,4 10,8 13,4 20,1 27,6 39,8 60,2 90,4 130,2 138,5 Thailand 7,1 6,1 6,5 8,8 13,6 23,1 34,5 51,0 81,6 116,5 137,4 United States* 30,6 29,8 34,7 49,7 56,5 58,3 69,1 90,5 115,7 124,9 127,1 Myanmar 8,6 8,3 9,0 10,1 11,4 17,3 22,4 30,5 50,4 75,6 96,9 Turkey 8,6 8,6 10,7 12,0 12,2 20,6 25,6 36,7 53,8 74,2 96,7 Mexico 10,6 10,2 9,2 8,4 10,3 14,2 20,8 32,1 49,3 74,2 96,6 Argentina 13,8 18,0 23,7 26,6 29,2 35,0 39,6 49,5 60,4 71,8 90,4 Vietnam 11,4 10,9 9,8 11,3 12,4 16,0 20,1 25,5 43,8 68,3 88,0 Brazil 7,2 7,6 8,7 10,9 12,4 17,7 24,8 36,5 54,8 70,2 86,2 Indonesia 10,1 8,3 7,2 8,2 10,8 15,4 22,3 30,3 44,7 65,3 83,0 Colombia 7,4 6,7 7,4 9,3 11,9 14,4 19,0 29,6 46,4 64,9 80,2 Iran 13,7 9,3 7,5 7,4 7,8 12,1 16,8 21,3 34,6 54,6 80,2 India 8,6 8,6 9,0 10,5 11,9 14,9 20,7 30,4 43,4 60,3 77,3 Algeria 10,9 8,8 8,6 8,4 8,5 10,3 12,0 20,2 34,0 49,2 72,1 Egypt 7,5 7,7 10,4 10,1 9,6 11,6 15,3 25,4 37,5 51,0 71,8 Philippines 8,2 6,8 5,9 6,7 8,6 9,9 13,5 22,9 35,7 49,4 70,7 Bangladesh 9,6 9,1 7,7 7,5 7,1 9,1 12,1 19,1 29,2 43,3 69,0 Pakistan 14,1 9,5 6,8 6,5 6,8 7,6 9,4 14,0 24,6 36,5 49,3 Kenya 9,9 9,0 8,2 6,7 6,1 6,9 6,7 8,5 14,3 23,3 43,2 South Africa 9,3 9,4 9,3 8,5 8,8 10,2 11,6 15,1 20,8 28,0 40,2 Nigeria 5,2 5,6 5,5 5,7 6,1 7,1 8,2 9,1 12,5 20,7 31,8 U.R. Tanzania 4,8 5,0 5,1 4,9 5,4 5,7 6,1 6,8 9,4 16,5 26,8 D.R. Congo 8,7 6,6 6,3 6,1 6,1 5,8 6,1 6,6 8,3 12,9 21,7 Ethiopia 6,8 5,7 5,6 6,6 6,5 6,2 6,1 6,6 8,4 12,8 21,2 Source: United Nations, 1999 * Richer developed countries [1] Director of the Department of Population Studies of the Joaquim Nabuco Foundation - Institute of Social Research and Lecturer at the Federal University of Pernambuco. The author thanks Jose Alberto Magno de Carvalho and Laura R. Wong for their reading and suggestions and takes entire responsibility for the remaining deficiencies of the paper. [2] In order to facilitate comparisons, estimates published by the United Nations were used in this paper: The World Population Prospects – The 1998 revision (United Nations, 1999) and World Population Prospects – The 2000 Revision (2001); the latter available on line (www.un.org/esa/population/wpp2000.htm). Unless otherwise expressed, all data are from the medium variant of the mentioned reports. [3] Table A1, in the Annex, shows the sex and age distribution of the Brazilian population for the years 1950, 2000 and 2050 [4] Due to the inertial effect of the age structure and the relatively narrow interval of the future fertility and mortality variations, we believe that the possibility of these forecasts to fail shall be relatively small. [5] See as an evidence of this comment, the French evolution of the age structure for the period 1775 to 1959 made by Pressat (1970). [6] Cuba has a population dynamic very different from the rest of Latin America. Notice that, already in 1950, the Cuban median age was above of the Latin–American average: 23.0 and 20.1 years respectively. For 2050, the corresponding values would be 43.1 and 37.5 (United Nations, 1999). In the Asian context, however, Thailand and Republic of Korea stand out, with an increase greater than 22 years in the median age during the period. [7] This profile, demographically advantageous will be present in other the underdeveloped countries experiencing similar demographic changes. This is the case of some Latin American countries, such as Mexico and Colombia. It is also applied to Asian countries like Indonesia, Republic of Korea and Thailand, among others. [8] The Aging Index is the ratio of the population over 65 years old to the population below 15 years old. [9] According to The 1996 General Assembly of the United Nations the categories include the following countries/regions: More developed: Northern America, Japan, Europe and Australia–New Zealand. Less developed: Africa, Latin America and the Caribbean, Asia (excluding Japan) and Melanesia, Micronesia and Polynesia. Notice that this category excludes least developed countries (38 in Africa, 9 in Asia, 1 in Latin America and 5 in Oceania). [10] Notice that the developed countries had certainly initiated this process prior to 1950; furthermore, for this date its Ageing Index was almost three times higher than the corresponding to the LCD. [11] It would have been important to characterize this process considering ethnic characteristics, but the census data concerning this variable are unusable due to obvious errors. Data deficiency is evidenced, for example, in the sex ratio over 100 in ages above 30 for the black/brown population. Inconsistency is evidenced, also, in the increase of the size of brown cohorts as we go further on time. (A clear example of this, is the brown population of age 10-14, in the 1960 census, whose volume was of 1.343 thousand children; in 1980, being at the age 30-34, the cohort size was 1.355 thousand. Finally, in 1990, when they would be at the ages 40-44, the volume extended to 1.485 thousand). For “migration” movements between the diverse ethnic groups, see Wood et al. (2000). [12] This is due, to a large extent, to the significant differences by sex peculiar to the Brazilian mortality, which, in addition, is expected to remain. Sex difference in the Brazilian life expectancy oscillates between 7,5–7,8 years favoring women for the beginning of the 21st century (Pronex/Cedeplar, 1999; CELADE, 2001). [13]The low values of the SR among the aged in the most developed countries, more specifically those located in Western Europe, are explained, in part, by the world wars. [14] See especially the available survey in Moreira (1997) [15] Informe da Previdência Social (Informs of the Social Welfare)- 2001.v. 3, p. 4.