International Brief

U.S. Department of Commerce Economics and Statistics Administration BUREAU OF THE CENSUS by Kevin Kinsella, International Programs Center, U.S. Bureau of the Census IB/97-2 Monica Ferreira, HSRC/UCT Centre for Gerontology, University of Cape Town Issued August 1997

Concern about population in Africa African region. Within Southern that large numbers of children con- traditionally has focussed on rela- Africa, South Africa has the highest tinue to be born each year, such tively high rates of fertility and proportions of older population that the majority of the population mortality, expansion of basic health (Table 1), with more than 1 in 8 has been and still is under 20 programs, and, more recently, on persons (5.6 million) aged 50 and years of age. Conversely, the pop- the devastation resulting from the over and nearly 7 percent (2.9 mil- ulation in older age groups remains HIV/AIDS pandemic. Overlooked lion) aged 60 and above. Such ag- a small proportion of the total. The in the face of these pressing issues gregate figures, however, mask the White South African population, on is the fact that most African popula- diversity of aging among population the other hand, already evinces an tions are aging, albeit at slower groups. Figure 1 displays 1997 age structure similar to some of the rates than in much of the develop- population structures by 5-year age world’s more-developed countries. ing world. While gerontology is still groups. The pyramidal shape of After several decades of low fertility in its infancy throughout Africa, the Black population resembles (the 1996 total fertility rate for there is a growing recognition of that of developing countries in the Whites was 1.7 births per woman), population aging—especially in early stages of demographic transi- the largest 5-year age groups South Africa—and the social bene- tion, where fertility and mortality among Whites are between 25 and fits and problems associated with rates are relatively high and popu- 40. More than one-fourth of all this process. To date, awareness lations as a whole are “young.” Whites now are aged 50 or above, has been stimulated largely by The broad pyramid base indicates with nearly 14 percent in the community-based ethnographic studies. The major impediment to Table 1. a fuller understanding of population Percentage of Population in Older Age Groups: 1997 to 2025 aging has been the scarcity of rele- vant, reliable national-level data. In Age group the 1990s, however, researchers in South Africa have sought to re- Region/country Year 50+ 60+ 70+ dress this shortcoming through the Southern Africa 1997...... 12.0 6.2 2.5 use of survey instruments and en- 2010...... 14.1 7.5 3.2 hanced census design and tabula- 2025...... 15.5 9.1 4.2 tion. This report summarizes the Botswana 1997...... 9.7 5.4 2.4 major demographic aspects affect- 2010...... 9.8 5.3 2.5 ing the aging of the South African 2025...... 9.1 5.6 2.8 population and highlights several Lesotho 1997...... 11.7 6.7 2.8 socioeconomic characteristics that 2010...... 12.1 6.5 3.0 2025...... 14.1 7.7 3.3 affect the wellbeing of older per- sons. South Africa 1997...... 13.3 6.8 2.8 2010...... 16.2 8.6 3.6 2025...... 18.0 10.8 5.1 The Southern Africa region has the Zimbabwe 1997...... 8.5 4.3 1.7 continent’s highest percentage of 2010...... 8.6 4.7 2.2 older inhabitants; 6.2 percent of 2025...... 8.4 4.9 2.4 population in 1997 was estimated to be 60 years of age or older, Note: The Southern Africa region comprises Botswana, Lesotho, Namibia, South Africa, slightly more than in the Northern Swaziland, and Zimbabwe. Source: U.S. Bureau of the Census, International Programs Center, International Database. 2

60-and-over category. Correspond- Figure 1. ing figures for Blacks are 11 and 6 Age Structure of South African Population by Group: 1997 percent, respectively, slightly less than for Asians and Coloureds Black (Table 2). 80+ Male 75–79 Female 70–74 The total population growth rate in 65–69 60–64 South Africa is falling rapidly as a 55–59 consequence of declining fertility 50–54 45–49 rates. As more persons live to old- 40–44 er ages, the growth rate of the 35–39 30–34 60-and-over population has come 25–29 to exceed that of the total popula- 20–24 15–19 tion (Figure 2), and the gap will 10–14 widen considerably in the future. 5–9 After the turn of the century, the 0–4 fastest growing population segment 2.5 2.0 1.5 1.0 0.5 0 0.0 0.5 1.0 1.5 2.0 2.5 Millions is likely to be persons aged 70 and over. While age-group percent- ages are useful for appreciating the Coloured/Asian relative weights of different popula- 80+ tion segments, absolute numbers Male 75–79 Female are more important in terms of na- 70–74 65–69 tional population policy and pro- 60–64 gram development. Proportions 55–59 that change only slightly over time 50–54 45–49 may, in fact, deflect attention from 40–44 rapidly growing numbers of older 35–39 30–34 persons and the service require- 25–29 ments generated by such growth. 20–24 15–19 For example, although the percent- 10–14 age aged 50 and over for the Black 5–9 population in Table 2 changes only 0–4 modestly from 1997 to 2010, the 250200 150 100 50 0 0 50 100 150 200 250 absolute growth will be in excess of Thousands 1 million persons.

White 80+ Median population age (the age Male 75–79 Female which divides a population into nu- 70–74 65–69 merically equal parts of younger and 60–64 older persons) is another indicator of 55–59 50–54 the aging process. South Africa cur- 45–49 rently has the highest median 40–44 35–39 age—22 years—in the Southern 30–34 Africa region, and this level is likely to 25–29 increase gradually but steadily 20–24 15–19 through the year 2025. Median age 10–14 varies greatly by population group 5–9 0–4 within South Africa, with a high of 34 years for Whites, 27 years for 250200 150 100 50 0 0 50 100 150 200 250 Thousands Asians, 25 years for Coloureds, and 20 years for Blacks. Source: U.S. Bureau of the Census, International Programs Center, International Database. 3

Table 2. Life Expectancy and Health Percentage of South African Population in Older Age Groups: 1997 to 2025 Overall life expectancy at birth in South Africa in 1997 is estimated Age group to be 54 years for males and 58 years for females. Again, there is Group Year 50+ 60+ 70+ enormous diversity among popula- Asian 1997...... 16.1 7.2 2.4 tion groups (Figure 3). Life expec- 2010...... 22.0 11.1 4.0 tancy for White South African 2025...... 29.7 16.4 7.1 women exceeds that of women in Black 1997...... 11.2 5.6 2.2 some European nations and is 25 2010...... 13.4 6.9 2.9 years higher than for Black South 2025...... 12.9 8.0 3.8 African men. The low levels of life Coloured 1997...... 12.3 5.9 2.2 expectancy at birth for the Black 2010...... 17.7 8.0 3.0 population reflect, to some extent, 2025...... 27.2 14.0 5.1 the growing impact of HIV/AIDS White 1997...... 26.0 14.3 6.7 mortality. The long-term conse- 2010...... 32.5 19.0 8.4 quence of the HIV/AIDS epidemic 2025...... 40.4 25.2 12.6 on population aging remains highly Source: U.S. Bureau of the Census, International Programs Center, International uncertain. HIV/AIDS may suppress Database. urban growth rates in South Africa and in much of Sub-Saharan Afri- Figure 2. ca, and have a selective impact on Average Annual Percentage Growth of Total young and middle-aged adults who and Older Population would be potential supporters of Percent 3.0 and caregivers to the elderly. Over time, however, the projected impact of the epidemic becomes more 2.5 widely distributed across age groups (Way, 1992); there are few- 2.0 Age 70 and over Age 60 and over er children because of pediatric AIDS mortality and reductions in 1.5 numbers of potential mothers, and there are fewer adults overall be- 1.0 cause of reduced infant survival and adult AIDS mortality. 0.5 Total, all ages As is the case in virtually all nations of the world, South African women 0.00 1990 1995 2000 2005 2010 2015 2020 2025 outlive men, regardless of popula- tion group. Except among Blacks, Source: U.S. Bureau of the Census, International Programs Center, International the gender difference in life expec- Database. tancy is greater than in most devel- Figure 3. oping countries. Among persons Years of Life Expectancy at who survive to age 60, women re- Birth and Age 60: 1997 tain their life expectancy advantage Male over men. At older ages, however, At Birth At Age 60 Female years of remaining life expectancy 70 17 for population groups are quite sim- White White 77 21 ilar. 52 15 South Africa is in the midst of an Black Black 55 18 epidemiologic transition from the 59 14 prominence of infectious diseases Coloured Coloured 68 18 to chronic diseases, with different emerging health patterns among 65 15 Asian Asian population groups. The 1990-91 72 18 Multidimensional Survey of Elderly South Africans, conducted by the Source: U.S. Bureau of the Census, International Programs Center, International Database. HSRC/UCT Centre for Gerontology 4 at the University of Cape Town, In prior decades, access to health urban areas, many choose to re- provided the first comprehensive services was difficult for much of turn to their childhood areas. look at various aspects of the the older population because of the Following the elimination of the health of the elderly. The most poor distribution of health facilities, Group Areas Act in 1988, South prevalent self-reported condition af- particularly in rural areas. The Africa has seen a massive move- fecting Blacks, Asians, and post-apartheid government has ment of persons from rural to urban Coloureds was high blood pressure committed itself to the development areas and a proliferation of informal (hypertension). This condition went of a primary health program, one settlements on urban fringes. untreated for 11 percent of both ur- facet of which is to establish prima- From available data, it is not yet ban and rural Blacks. A subse- ry care clinics in underserved quent national household survey clear if older citizens are being ad- areas. The elderly, however, con- versely affected by such movement (Hirschowitz and Orkin, 1995) con- tinue to experience difficulty ac- firmed the importance of hyperten- (e.g., being left behind in rural cessing these (as well as many ur- areas to raise grandchildren), are sion and also found that 70 percent ban) services due to a lack of of the elderly (age 65 and over) participating in and benefitting from transport. And, with the health pro- migration (through better access to had a chronic illness or ongoing gram’s current emphasis strongly health problem. More than half of health and pension systems), or on maternal and child care, there is are adapting in other ways. the survey respondents reported mounting concern that resources having a disability, the most com- are being shifted away from need- Educational levels among older mon being difficulty with vision. ed geriatric care at a time when South Africans remain fairly low in growth in the older population is the 1990s, but will improve as bet- Other epidemiologic evidence accelerating. ter-educated cohorts reach old age. (Bradshaw et al., 1995) suggests Data from the 1991 census reveal that more than one-third of the high mortality risks due to heart Urbanization and Education disease among non-Blacks, with 65-and-over population had no for- very high lung and breast cancer In most countries in Southern Afri- mal educational experience mortality rates seen among the ca, the elderly are more likely than (compared with 12 percent of per- Coloured population. While the im- the total population to reside in ru- sons aged 25-34), while another 18 portance of noncommunicable dis- ral areas. This observation is the percent had completed less than 7 eases clearly is rising, it should be norm in most, if not all, regions of years of schooling. The 1990-91 Multidimensional Survey found that stressed that infectious diseases the world (Kinsella and Taeuber, two-thirds of older (60+) Blacks and remain a prominent cause of 1993), resulting from the migration Asians and half of Coloureds had mortality and morbidity even at of young adults to cities and some- less than 5 years of education. adult ages, as do violence and mo- times from the return migration of Three-quarters of Whites, by con- tor vehicle accidents. older adults from urban areas back to rural homes. In South Africa, trast, had completed 10 or more however, the reverse has been years of school. As in most devel- Because medical insurance has true; data from the 1991 census in- oping countries, older men are been generally unavailable in South dicate that while about 45 percent somewhat more likely than older Africa, most persons reach old age of the entire population (excluding women to have attended school. with few, if any, benefits. In 1990, the former homelands) lived in non- With regard to the ability to read 95 percent of urban and 99 percent urban areas, only 38 percent of the and write, the 1990-91 survey of rural elderly Blacks had no medi- elderly were in nonurban settings. found relatively high literacy rates cal insurance. However, about 90 among the 60-and-over population percent of each group had annual During the apartheid era, South in urban areas—ranging from 61 medical expenses. Many rely upon African policy permitted only male percent among Asians to 100 per- the help of children and relatives to migrants to relocate to urban cent among Whites—but a much defray medical costs. Percentages areas, thus requiring men to leave lower rate (29 percent) among old- of elderly Asians (96 percent) and their wives (sometimes more than er rural Blacks. Coloureds (92 percent) lacking one) and children in rural areas. medical insurance were similarly The remaining rural inhabitants re- Marital Status and high. Sixty-five percent of Whites, lied heavily on remittances from Living Arrangements on the other hand, did have medi- family members working in the city. Because women live longer on av- cal insurance (often via corporate Most older Black labor migrants, erage than do men and tend to schemes), and only 1 percent re- who grew up in tribal communities, marry men older than themselves, ported assistance from family or still attach great significance to their rates of widowhood increase with friends in paying medical expenses ancestral land (Oosthuizen, 1993). age in most countries of the world. (Ferreira et al., 1992). After working much of their lives in South Africa is no exception, and 5 indeed the gender difference is because of the previously-men- Blacks owned their homes. In the striking (Figure 4). Almost half of tioned restrictions on Black female urban setting, 62 percent of elderly all women aged 60 and over are migration to urban areas, polyga- Whites resided in homes owned by widowed compared with only 12 mous families enabled one wife to themselves or their spouse, while percent of elderly men. The abso- visit her spouse in the city while the 72 percent of Asian elderly resided lute number of widows aged 60 other wife (wives) cared for home in dwellings owned by someone and over increased nearly 100,000 and children. else (presumably a child) in the between 1985 and 1991 (to a total household. The South African sur- Survey data for 1990 show that, of 559,000). South Africa also has vey ascertained access to ten except among Whites, more than relatively high percentages of per- household conveniences covering half of all elderly live in households sons 60+ who are never-married or piped water, sanitation, electricity, with three or more co-resident gen- divorced (9 and 8 percent for appliances, and means of commu- erations (Figure 5). Another 30 women and men, respectively), nication. Average access to the percent live in two-generation suggesting at least the possibility ten conveniences was 97 percent households, while fewer than 5 that significant numbers of elderly among Whites, 92 percent for percent live alone. The statistics are without the family support net- Asians, 71 percent for Coloureds, for elderly White South Africans work that is commonly assumed to 47 percent for urban Blacks and 15 look entirely different: 39 percent be prevalent throughout the region. percent for rural Blacks. live alone and another 41 percent In South Africa and several other reside with their spouse in a two- Employment and Subsistence African nations, polygamy can be person household; fewer than 20 an important social factor that typi- percent live in multigenerational Participation in the formal labor cally is not reflected in official data. households. Among very old market fell slightly among older One survey of Zulus in South Afri- Whites aged 85 and over, 7 in men between the mid-1980’s and ca (Moller and Welch, 1990) found 10 live alone. early 1990’s, while the opposite that 10 percent of males were in was true among older women (Fig- polygamous marriages. Such The majority of South African sur- ure 6). Seven in ten men aged arrangements were said to be vey respondents owned their own 55-64 were economically active in beneficial to both women and men; home (or it was owned by their 1991, as were 21 percent of all spouse). Nine out of ten rural men over the age of 64. Corre- Figure 4. sponding levels for women were 29 Percentage Widowed at Older Ages: 1991 percent and 5 percent, respectively. Female Especially for women, these figures may underestimate the true level of Age 55–64 29 economic activity to the extent that Age 65–74 48 informal-sector activities are not ful- Age 75+ 68 ly reflected in census data. Male South Africa is an exception among Age 55–64 5 African countries in its formal eco- Age 65–74 12 nomic support for older citizens. Women aged 60 and over and Age 75+ 23 men aged 65 and over are eligible Source: U.S. Bureau of the Census, International Programs Center, International for a means-tested general social Database. pension of Rand 470 (approximate- ly US$100) per month. In contrast Figure 5. Alone to past levels that varied by race, Living Arrangements of Elderly Spouse only persons of all races are now eligi- South Africans: 1990 Two generations Three or more generations ble to receive this monthly amount. Survey data for 1993 (SALDRU, Asian 1994) indicate that 61 and 68 per- Urban Black cent of age-qualified men and Rural Black women, respectively, receive a so- Coloured cial pension. Nearly 80 percent of White the age-qualified Black population reported receiving a social pension, 0 20406080100although a separate analysis using Percent Source: U.S. Bureau of the Census, International Programs Center, International different data suggests that this lev- Database. el is closer to 90 percent (versus 6

Figure 6. Ferreira, Monica, Frances Lund and Valerie Labor Force Participation Rates: 1985 and 1991 Moller, 1995, “Status Report from South Afri- ca,” Ageing International, Vol. 22, No. 4, pp. Percent economically active 100 16-20. Ferreira, Monica, Valerie Moller, F. R. Prinsloo and L. S. Gillis, 1992, Multidimensional Sur- 80 Male, 1985 vey of Elderly South Africans, 1990-91: Key Female, 1991 Findings, Monograph No. 1, HSRC/UCT Male, 1991 Centre for Gerontology, University of Cape 60 Town. Female, 1985 Hirschowitz, Ros and Mark Orkin, 1995, A 40 National Household Survey of Health Inequal- ities in South Africa, Henry J. Kaiser Family Foundation, Washington, DC. 20 Kinsella, Kevin and Cynthia Taeuber, 1993, An Aging World II, U.S. Bureau of the Census 0 International Population Report, Washington, 20–24 25–34 35–54 55–64 65+ DC. Age group Source: U.S. Bureau of the Census, International Programs Center, International Moller, Valerie and R. Devey, 1995, “Black Database. South African Families with Older Members: Opportunities and Constraints,” Southern Afri- noncontributory old-age pension 19 percent for the White popula- can Journal of Gerontology, Vol. 4, No. 2, pp. scheme have ameliorated some of 3-10. tion; Van der Berg, 1994). the adverse effects. The post- Moller, Valerie and Ayanda Sotshongaye, For many South Africans, the role apartheid Reconstruction and De- 1996, “‘My Family Eat This Money Too’: Pen- sion Sharing and Self-respect among Zulu of the public pension goes beyond velopment Program (RDP) seeks to redress social inequalities by Grandmothers,” Southern African Journal of mere support of the elderly. Pen- Gerontology, Vol. 5, No. 2, pp. 9-19. sion sharing, particularly in Black providing expanded opportunities for education, housing, and em- Moller, Valerie and Gary John Welch, 1990, households, is the norm. One sur- “Polygamy, Economic Security and Well-Be- vey indicated that while only 4.5 ployment, but most of the RDP is ing of Retired Zulu Migrant Workers,” Journal percent of the total population re- focussed on improving the situation of Cross-Cultural Gerontology, No. 5, pp. ceived a pension, 85 percent of all of women and youth; the elderly 205-216. pensioners lived in three-generation are not seen as a priority (Ferreira, Oosthuizen, J.S., 1993, “Patterns of Migration households (Snyman, 1997). Al- Lund and Moller, 1995). Some ob- in Southern Africa with Special Reference to though pensions may enhance the servers fear that resources may be South Africa,” University of Pretoria, Pretoria. economic self-reliance and self-re- diverted from the universal pension Snyman, Ina, 1997, “Pensions: Lifeline of the spect of recipients, pensioners’ own system in order to fund other as- Poor,” In Focus Forum, Vol. 4, No. 4, pp. 8-11. needs may be neglected in the in- pects of the RDP. However, there terests of family welfare (Moller and is a growing public recognition of South Africa Labour and Development Re- Sotshongaye, 1996). Furthermore, the importance of the pension sys- search Unit (SALDRU), 1994, South Africans tem as a social safety net. At the Rich and Poor: Baseline Household Statis- the mere fact of pension receipt tics, University of Cape Town Project for Liv- does not necessarily ensure an ad- same time, the Ministry of Welfare ing Standards and Development, Cape Town. equate standard of living. A major is developing new policies and pro- grams regarding community sup- Van der Berg, Servaas, 1994, “Issues in finding of the 1993 Project of Sta- South African Social Security,” World Bank tistics on Living Standards and De- port services for older citizens. Background Paper, Washington, DC. velopment was that elderly house- The current emphases within South Africa on human rights and policy Way, Peter, 1992, “The Demographic Impact holds (i.e., households with at least of HIV,” Paper presented to the VIII Interna- one elderly member) are poor by reform augur well for greater em- tional Conference on AIDS, July, Amsterdam. most standards. Only 30 percent powerment of the elderly, and for of all surveyed elderly households concrete improvements in basic The International Programs Center were living above the poverty line health care as well as community- (IPC) of the U.S. Bureau of the Census (Moller and Devey, 1995). based comprehensive family care. collects and analyzes population and related statistics from all countries of the world. This brief was produced with the References A Brighter Future? support of the Office of the Demogra- Bradshaw, D., D.E. Bourne, M. Schneider, phy of Aging, U.S. National Institute on Decades of discrimination and dis- and R. Sayed, 1995, “Mortality Patterns of Aging. More detailed information is advantage have negatively affected Chronic Diseases of Lifestyle in South Africa,” available from the International Pro- the health and socioeconomic well- in Jean Fourie and Krisela Steyn, eds., Chronic Diseases of Lifestyle in South Africa, grams Center, Population Division, Bu- being of most South African elderly. Medical Research Council Technical Report reau of the Census, Washington, DC The family support system and a 1995, Cape Town. 20233-8860 USA.