Title: Factors influencing the living arrangements of elderly in - - PDF document

1

Title: Factors influencing the living arrangements of elderly in South Africa

Authors: Chantal Munthree (chantalmu@statssa.gov.za) and Angela Ngyende

Institution: Statistics South Africa

Abstract Background: The proportion of elderly is growing at a more rapid pace in the developing world. In the last 15 years the proportion of elderly in South Africa has increased (7, 1% in 1996 to 8, 0% in 2011) and will continue to do so given current levels of fertility and mortality (UN, 2015). Projections show that the older population will continue to increase and by 2030 there will be more than six and a half million elderly persons in South Africa (MYPE, 2017). South Africa is ill prepared to cater for the needs of the growing proportion of elderly in the country. Living arrangements are fundamental in old age, particularly during frail years, as it is determining factor on the extent to which elderly access health care and social support. Data and Methods: The census 2011 data as well as the General Household Survey 2005 and 2015 was used to identify the trend in living arrangements in South Africa over time. Multivariate and regression analysis using the GHS 2015 data was used to determine the socio demographic factors influencing the living arrangements of elderly in South Africa. Findings: The proportion of elderly persons living alone has grown between 2005 and 2015, particularly amongst elderly women. More than a half of elderly women (52%), live with 1 or more persons and were shouldering the responsibility of heading their respective households By 2015 a higher proportion of elderly women (26.3%) are household heads catering for the needs of more than 1 person when compared to men (9.2%). A little more than 10% of elderly in South Africa report living alone. The elderly men who by choice or circumstance reside alone are more likely to be Black African with low levels of education, unemployed and accessing a social grant. Elderly women residing alone however are likely 4 time more likely to be white than Black African, older than 64 years of age, having higher education, unemployed and accessing a social grant. Both elderly men and women residing alone are located more so in the Western and Northern Cape. Amongst elderly men living with 1 or more person as head of the household, we find they are more likely to be Black African, married, with lower levels of education, employed, and more likely to be accessing a social grant. Amongst elderly women living with 1 or more person as head of the household the pattern is rather different as we find they are find they are more likely to be Black African, never married/ divorced/ widowed

• f separated with even lower levels of education, unemployed with no access to a social grant.

Recommendations: Home based care and mobile heath social services should target key population among the elderly in South Africa to alleviate the added vulnerability of living alone. Programs and policies should be developed to address the key factors influencing living arrangements of elderly, if such policies are to more effectively address the wellbeing of elderly.

2 Introduction The living arrangements of people vary significantly by age. The focus of this paper is to better understand the living arrangements among people transitioning into older and more vulnerable stages of life. In South Africa the age at which the people are afforded a social pension to support themselves in their old age is 60 years of age. This is also the official age of retirement (National Treasury, 2007).) Apart from a shared experience in the labour market regarding retirement, adults aged 60 and over also display similar physical characteristics and patterns of morbidity and mortality. The elderly are however not a homogenous group thus a secondary purpose of this paper is to elaborate on the diversity of the characteristics and living arrangements of elderly in South Africa. Living arrangements refers to the familial and non–familial relationships of a person to all the other people with whom they usually reside. Living arrangements are fundamental in old age, particularly during frail years, as it is determining factor on the extent to which elderly access health care and social support. Living arrangements of elderly persons continue to be largely influenced by culture, marriage, migration and urbanisation, female labour force participation, and generally, modernisation characterised by individualistic life style (Amoateng, A.Y et al (2007)). This is contrary to living arrangements elderly persons experienced in traditional societies, where they mostly lived with children and families, and were provided all forms of support. One of the implications of the evolving household composition is that the elderly are no longer assured of the family support structure. In South Africa, many elderly are faced with not having the necessary economic means to live independently (May,2003). Furthermore, elderly in South Africa often find themselves bearing the financial and physical responsibility of caring for others in their old age. Understanding the characteristics of elderly living alone as opposed to elderly residing with others can shed light not only on the socio demographic characteristics that influence the living arrangements of the elderly, but also the circumstances under which elderly reside in South Africa. A primary goal of the international Sustainable Development Goals (SDG’s) is to “leave no one behind” and this includes elderly and the vulnerable. To address the African continent’s integrated plan to bring about prosperity, the African Union has placed the inclusion of the elderly as a central focus of Agenda 2063. More locally, the National Development Plan of South Africa is tasked with protecting the growing proportion of elderly in the country, catering to their social, economic, and general well-being. Healthy ageing defined by the World Health Organisation, is the developing and maintenance of functional ability among the elderly. In an effort to better understand whether the elderly in South Africa are indeed provided with an environment that allows for healthy

3 aging, this paper seeks to examine the change on living arrangements among the elderly over a decade, as well as understand the differentials in the living arrangements of elderly in South Africa to draw attention to the plight

• f elderly for policy intervention.

Literature review The world population is reported to be ageing at a far more rapid pace than has ever been experienced (WHO, 2015). The developed world has contributed largely to the global ageing of society with life expectancy reaching upwards of 70’s and 80’s, comprising of approximately 15% of the total population in the developed world (Global ageing, 2015). In contrast, in the developing world, the ageing of populations though at a far lower level than the developed world, is increasing at a more rapid pace than that experienced by the developed world. The forces of ageing in this context, can be attributed to declining fertility and rising life expectancies resulting in significantly higher number of elderly residing in less developing regions (WHO, 2015). The ageing of less developed regions are expected to continue to rise given the dynamics of these forces. According to the United Nations (2015) the percentage change in elderly within the more developed regions is expected to decline from 29.2% between 2000 and 2015, and to 25,6% between 2015-2030. In contrast the percentage change in elderly within the less developed regions increased from 60.3% between 2000 and 2015, to 70.6% between 2015-2030 (United Nations, 2015) . Aging in Africa and Sub Saharan Africa Ageing in Africa is occurring at a far higher rate when compared to other regions of the world (UN, 2015). According to the United Nations (2015), the number of persons in Africa aged 60 and over, increased from 42,4 million in 2005 to 64,4 million by 2015, and is said to triple by 2050 to 220,3 million. Third to Asia and Latin America and the Caribbean, Africa is experiencing the one of the highest growth rates in the proportion of elderly aged 60 and over (UN, 2015). Studies indicate that the proportion of older people living in poverty is higher than the national average, as found among 11 of the 15 countries surveyed. (http://www.global-ageing.eu/agafrica.html). Whilst the developed world has had decades to adjust to the changing proportions and needs of the elderly, Africa’s relatively faster pace of ageing has resulted in the region being ill prepared for the needs of its elderly constituents. There is also limited data on the level and characteristic of elderly in the region as a whole. Though census data, where available, provides some information regarding the characteristics of the elderly, many surveys including the Demographic and health surveys conducted across the globe, exclude elderly age 60 and over when gathering data on heath, reproduction etc. We cannot negate the recent advancement in better understanding the social, economic and physical needs of the elderly in the last decade in the region, however such studies are often limited in their focus. Many studies tend to focus on disabilities and poverty among the elderly neglecting the social circumstances of elderly in the country. Even fewer studies explore factors influencing the plight of elderly. Ageing in South Africa

4 The number of elderly has increased from 2, 8 million in 1996 to 4, 1 million in 2011 (StatsSA, 2015). In the last 15 years (1996-2011), the proportion of elderly in South Africa has increased (7, 1% in 1996 to 8, 0% in 2011) and will continue to do so given current levels of fertility and mortality. As the level of fertility continues to decline in conjunction with declining mortality, it is expected that the proportion elderly over time will grow exceedingly larger

• ver time (UN, 2015). Projections show that the older population will continue to increase and by 2030 there will

be more than six and a half million elderly persons in South Africa (MYPE, 2017). The demographic shift resulting in the change in the age structure takes place in a country with high levels of unemployment, inequality and poverty. Recent trends in poverty and differentials in income inequality indicate that almost half of all elderly living in South Africa were living in poverty (StatsSa, 2017). This does not account for the more vulnerable elderly living in institutions, but is rather only reflective of those living within households. Though there has been a decline in the proportion elderly living in poverty between 2006 (65%) and 2015 (30%), the proportion of elderly living in poverty is significantly higher than that of adults aged 25-64, pointing to financial vulnerability in older age. It is also noteworthy that the proportion of elderly rises significantly in periods of economic shock such as that of the 2008 recession (StatsSA, 2017) In the midst of the turbulent social and economic environment that the majority elderly in South Africa experience, the South African government remains committed to ensuring that all elderly person are able to experience the right to security and dignity. The South African Older Persons Policy adopted in 2006 (Act No.13

• f 2006) provides a framework for providing for the needs of elderly persons. These include the establishment of

the Directorate of Care and Services within the Department of Social Development. The government recognizes that elderly persons are indeed a vulnerable group, given their health and socio-economic circumstances. In South Africa, men and women alike receive a non-contributory pension on turning 60 years of age of R1600 per

• month. This state transfer referred to as the ‘Old Age Pension’ (OAP) fund, is hoped to better the lives of the

elderly who are economically vulnerable. However the complexities of household and social relations often leave many elderly not only caring for themselves but often children and grandchildren. In South Africa numerous studies indicate that elderly, particularly elderly women are often tasked with caring for children with or without the support of parents or family support. The change in the mortality pattern by age over time has had a dire consequence on not only on the lived experience of elderly in the context of AIDS. Apart from the high unemployment rate that results in the state transfer being the only form of income for many families, the impact

• f AIDS, has resulted in many elderly living with the loss of children who were breadwinners whilst burdened with

caring for orphans of AIDS (Maharaj and Munthree, 2013). The responsibilities of care, elderly men and women face, have implication of the health and wellbeing of elderly as well as those they are tasked with caring for. Elderly though often thought of as a burden in developed countries, are an integral part of the generational support system in South Africa. It is evident that the living arrangements of elderly are a consequences of the social and economic experiences of elderly in South Africa.

5 The recently published report on poverty trends in South Africa, indicates that in 2006 more than two thirds of households headed by elderly were in poverty (StatsSA, 2017). Though the incidence of poverty among household heads aged 65 and over has declined from 68% in 2006 to 46% on 2015, poverty incidence remains significantly higher among elderly headed households than households headed by those aged 18-64. Given the vulnerability elderly households face, a better understanding of the factors that influence the living arrangements

• f elderly in South Africa is crucial. The poverty trends report, discusses the financial vulnerability of elderly,

however the there is a mandate to understand the social, physical and emotional demands placed on elderly in South Africa. The growing proportion of elderly in South Africa is indicative not only of a country that is on a demographic and health transition but also alerts us to the changing social, health and financial demands elderly are facing and will face moving forward. South Africa experienced a significant decline in life expectancy at birth between 2002 and 2006, primary due to the impact of AIDS. Expansion of health programmes to prevent mother-to-child transmission and to improve access to antiretroviral treatment (ART) has helped to turn this around, and there has been a marked improvement in life expectancy in South Africa between 2007 and 2017. By 2017, life expectancy at birth is estimated at 61.2 years for males and 66.7 years for females (MYPE, 2017). The expansion of health programmes related to HIV has contributed to a decline in deaths due to communicable diseases such as HIV and TB. This in turn has resulted in a higher proportion of South African deaths being due to non-communicable diseases. Thus the focus now lies on programmes to combat the increase in deaths due to non-communicable diseases such as heart disease, diabetes, etc., particularly among those aged 60 and

• lder (StatsSA, 2017(causes of death report). Policies and programs that focus on extending the lives of people,

particularly in the older ages, should also address the needs of elderly as they progress in age. According to Mckinnon et al (2013) elderly living in Sub-Saharan Africa face an increased risk of depressive symptom when living alone as compared to those living with an adult. A further recommendation of their study was that health and social policies be put in place to mitigate depression, which can often does lead to poor quality of life and increased mortality. The progression to older ages differs by men and women. On average women experience a longer life expectancy in many countries across the globe (World Bank, 2015). In adulthood, 15-50, non-natural causes of death such as hazardous work environments, accidents, injuries which have indeed increased considerably throughout the 20th century affect men more so than women. Non communicable diseases such as alcoholism, smoking etc are also found to be disproportionately higher among men than women. The higher male deaths which can be attributed to more social factors that results in further loss of life to males on average, resulting in an overall lower life expectancy than women. Global life expectancy at birth in 2015 was 71.4 years (73.8 years for females and 69.1 years for males). The size and advantage does vary by region and country. According the most recent population estimates for South Africa, the life expectancy at birth among men is 61,2, lower than that

• f women (66,7) (MYPE, 2017).

6 Variations in age are particularly noticeable by population group in South Africa. Between 2002 and 2017, the proportion of elderly among black Africans increased by 0,5%, among Coloureds by 2,7%; among Indians/Asians by 4,2% and among Whites by 7,3%. Disparities in ageing by province and population group have a historical context that can be traced to fertility, mortality and migration streams over time (MYPE, 2017). Regional variations of ageing exist in South Africa. While almost a quarter of the elderly live in Gauteng, they account for only 7,7% of the total Gauteng population. In contrast, 12% of the elderly reside in Eastern Cape; however, nearly 10% of the Eastern Cape population comprises of elderly persons. Mpumalanga’s population comprises 7% elderly, the lowest among the provinces. These should be factored into future planning and decision-making at provincial level and lower (MYPE, 2017). Living arrangements in South Africa A number of studies globally and even within South Africa have documented the pattern of living arrangements among the elderly, however the paper seeks to determine the socio demographic factors influencing the living arrangements of elderly in South Africa ( UN, 2015; Mba, 2002). Given the pace at which the elderly in South Africa are growing, planning for the needs of this financially and physically vulnerable group is imperative. Numerous studies have confirmed that there is a relationship between the living arrangements of elderly and their psychological and physical wellbeing (Russel, 2007; Zhang and Lui, 2007). Hadi et al (2012), in their study found that older adults who live alone have a lower life satisfaction than those who live in other structures. Similarly in India, Argawal (2012) found that elderly living alone have poorer health than those living with family. Apart from the impact that a living arrangement has on the wellbeing on the elderly, studies in South Africa indicate that elderly, particularly elderly women play a significant role in improving the wellbeing of those that they are living

• with. Maharaj & Munthree (2013) indicate that elderly living with and caring for those infected and affected by HIV

and AIDS in South Africa, report doing so in the context of no financial or physical assistance from family or the

• community. Many studies argue that elderly in South Africa have traditionally been financially taken care of by their

adult children; however, as a result of the HIV/AIDS epidemic there is a complete distortion of intergenerational support (Help Age 2004; Munthali, 2002; Legido-Quigley, 2003). With the limited resources available to elderly men and women, studies in South Africa indicate that the presence of a pensioner significantly reduces household reports of hunger, additionally children are more likely to be enrolled in school when there is an old age pension being accessed within the household (Case& Menendex, 2007). According to Case and Menedex (2007), “social programs targeting the elderly plays a significant role in children’s health and development” (2007, pg1). In South Africa, living arrangements are diverse. Living arrangements among elderly even more so as they transition to an age of changed familial relations, changing economic circumstance and changing physical needs. Schartz et al (2015) in their study of elderly in Agincourt, a rural community in Gauteng, provide a description of the varying living arrangements elderly people in South Africa find themselves. Older people were found to live in

7 large and complex households. Similarly Hoosegood et al (2005) in their study of the northern rural community of KwaZulu-Natal, found that elderly living in multigenerational households were significantly poorer. In an effort to better understand how to create an environment that is age friendly and fosters the functional ability

• f the elderly, an examination of the living arrangement of the elderly including headship is necessary. Household

headship usually refers to someone who bears the economic as well as authoritative responsibility for a household and its members. In South Africa, numerous studies indicate that men, including elderly men are often reported as the head of the household (REF). Given that the age of retirement has only recently changed to allow men to retire as early as women, men were often found working until the age of 65, 5 years more than women. Income and employment equity reports for South Africa indicate that women regardless of age, on average earn less than men, whilst men occupy higher earning positions than women on average (StatsSA, 2015). These reports point to the economic disparities by gender experienced by women at all ages, resulting in vulnerability among in their later life. Economic and social correlates of living arrangement To what extent the living arrangements in South Africa is influenced by culture, social factors, economics, political and demographic developments, the AIDS epidemiological crisis, is a considered a difficult question to answer. It is at best a culmination of these numerous factors over time that have shaped the living arrangement of South Africans, including the elderly. We examine the current characteristics of the elderly in South Africa by their living arrangement, either chosen or by circumstance, to be better understand how policy and programs can be improved to ensure elderly are less vulnerable in their older ages. Previous studies examining the living arrangements of elderly in South Africa using census, indicate that singular households increased from 16.2% in 1996 to 21% in 2001, whilst households with non-related members declined from 5.27% in 1996 to 3.25% in 2001. This pattern indicates that over time the living arrangements in South Africa are becoming more singular and exclusive rather than inclusive (Amoateng et al , 2007). A number of studies indicate that marital status has a direct influence in living arrangements, social support and

• companionship. In South Africa cohabitation without formalised marriage is common, including polygamy, thereby

leading to more complex living arrangements (Sabiti et al, 2008). A change in marital status during old age is known to have a number of socio-economic effects on the elderly. Literature has shown that married people, particularly men, enjoy healthier and longer lives than their unmarried counterparts (Kinsella, 2007). Marriage has been found to increase longevity, more so amongst males than females (Weeks, 2008:193). By 2010 a decree by the minister of finance in South Africa allowed for the inclusion of men between the age 60- 64 according to the new eligibility criteria. By 2017, about 3,4 million older persons aged 60 years and over were receiving the old age grant (http://www.sassa.gov.za/index.php/knowledge-centre/statistical-reports). High levels

• f unemployment and the reduced proportion of working age adults 15-64 relative to the dependent elderly has

fiscal policy implications.

8 Theoretical Focus The concept of household used in this paper refers to a unit that consists of co-residing persons related through, marriage, blood or adoption, and also includes co residing non relative. Between 2002 and 2015, South Africa not only experienced the peak of an HIV and AIDs pandemic, an economic recession, peak in unemployment as well as increased modernization and urbanization among other social, political and economic changes. According to the ‘Resiliency Model of Family Stress, Adjustment, and Adaptation’ (Freidman, 1998; McCubbin & McCubbin, 1993), individuals or families may adapt to ensure the wellbeing of the family and or themselves. The concept of coping, forms part of the adjustment and adaptation. Coping can be defined as a cognitive, affective

• r behavioral effort made by the individual to offset the impact of harm, threat or stress when an automatic

response is not readily available. In this paper, the behaviour of residing within a singular (living alone), living with one other person, and living with more than one person is viewed as a possible coping strategy. Using this theoretical focus, understanding the current factors influencing the living arrangements of elderly men and women in South Africa, will allow for a better understanding of the policies, programs and target populations needed to improve the social and economic wellbeing of elderly people living in South Africa. Data and Methods: Elderly persons are defined as persons aged 60 years and older. Using the census data 2011 and the General Household Survey (GHS) 2005 and 2015, basic descriptive analysis will be used to provide a trend of the socio demographic characteristics of elderly in South Africa between 2005 and 2015. Whilst the census is a complete count of the population, the General household survey is conducted among a selected representative sample of the country which is weighted to a model. As such distributions over time may not follow the trend exactly, given the nature of the census 2011 point. However the two sources of data can be used to confirm or reject the socio demographics and living arrangements of elderly in South Africa over time. Statistical analysis,

including multivariate logistic regression applied to the GHS 2015 data, is used to determine the association between the socio-demographic characteristics of elderly men and women, and their propensity to live alone, live with others as a head of the household and live with others but not at the head of a household. The variables in the analysis include gender, population group, province of residence, age,

marital status, education, and employment status, access to a grant and household size. In determining household headship, respondents were asked to nominate the person they believed to be the breadwinner or person responsible for the household. Results Table 1 examines the change in old age dependency ratios over time. This rise indicates that South Africa is indeed

• ageing. Old age dependency ratio for the purpose of this study is proportion of elderly aged 60 and over relative

to those of working age 15-59. It is clear from table 1 that the old age dependency ratios varies significantly by population group. In an effort to better understand the living arrangements of elderly it is important to examine the

9

• ld age dependency by population group and province. This measure is an indication of the available human

resource of younger working age people to care for their ageing elderly relations. Table 1 indicates that across population groups, Whites have the highest old age dependency ratio of 19.2% among white males and 24.1% among white females, almost 4 times that of Black Africans. With time the old age dependency ratio has increased more so among White and Indian/Asians than Black Africans and Coloureds. The implication of the higher dependency ratios among Whites and Indian/ Asians when compared to Black Africans and Coloureds, is that there is a higher burden of care for elderly amongst Whites and Indian/Asians when compared to Black Africans and Coloureds and this differentiation is increasing over time. A limitation of this measure is that is does not take into consideration the disproportionate levels of financially dependency among elderly and saving by population group, nor does it account for the disproportionate unemployment rate by population group. In South Africa planning in terms of housing, education, health and access to services are done so at a provincial

• level. Table 1 points to the varying levels of old age dependency by province. By 2015, Western Cape has the

highest old age dependency ratio among males (13%), whilst the Northern Cape have the highest dependency ratio among females (18%) Across all provinces the dependency ratio is higher among elderly women when compared to men. The highest difference between male and female old age dependency is recorded in Kwazulu-

• Natal. This may be related to migratory pattern in South Africa that resulted in a number of men leaving KwaZulu-

Natal in search of work and never returning as well as the devastating impact on HIV and AIDS on men in adulthood in the province. Over time we find an increase in old age dependency ratios. This change is not only indicative of the rising life expectancy over time but also the impact of declining fertility in South Africa. Provinces experiencing a higher burden of care are Eastern and Western Cape. It is important to bear in mind these ratios when planning an environment that fosters healthy ageing for elderly aged 60 and over. Table 1: Old Age dependency ratios over time by population group and Province, 2005 -2015 2005 2011 2015 Old age dependency ratio Male Female Male Female Male Female Total Dependency Ratio 8.65 13.06 10.37 15.06 10.09 15.5 Black African 6.77 11.33 8.15 12.91 7.79 13.15 Coloured 8.25 10.45 10.40 13.55 1091 14.31 Indian/Asian 11.19 14.42 13.92 18.78 14.81 19.79 White 21.70 28.62 29.00 34.75 30.29 40.15 Western Cape 8.68 13.35 11.79 14.77 12.96 15.93 Eastern Cape 10.92 17.26 14.05 19.74 11.80 17.59 Norther Cape 8.69 12.62 12.03 15.84 10.97 16.55 Free State 8.69 12.62 10.83 15.77 10.97 16.55 KwaZulu-Natal 9.19 13.39 9.78 14.98 9.31 16.35 North West 8.32 13.51 11.29 15.83 10.38 15.74 Gauteng 7.32 10.44 8.33 11.27 9.48 14.22 Mpumalanga 8.19 10.45 9.29 13.41 8.35 13.32

10 Limpopo 9.19 14.92 11.46 18.32 8.51 15.30 Table 2 shows the percentage distribution of elderly men and women in South Africa by selected socio- demographic characteristics. The proportion elderly men and women in South Africa has increase from 6.8% in 2005 to 8.3% in 2015. Table 2 indicates that the distribution of those age 60 and over by gender varies by

• age. By 2015, a higher proportion of elderly men are found in the age group 60 to 64 (38%) when compared to

elderly women (30%). In contrast a higher proportion of elderly woman (22.3%) are found in the age group 80 and over when compared to men (17%). The contrasting distribution of elderly men and women by age is directly related to the mortality patterns of men relative to women in South Africa. According to the Causes of death report published annually in South Africa, in adulthood 15-50, non-natural causes of death such as hazardous work environments, accidents, injuries which have indeed increased considerably throughout the 20th century affect men more so than women. Non communicable diseases such as alcoholism, smoking and violence are also found to be disproportionately higher among men than women. The higher male deaths which can be attributed to more social factors that results in further loss of life to males on average, results in an overall lower life expectancy than women (StatsSA, 2017b). The disparate distribution of elderly by age and gender is even more significant by population group in South Africa. (See table 2). In South Africa the majority of the overall population is black African (80%). Expectedly the majority of elderly men (61, 8%) and women (67, 6%) age 60 and over are black African (Table 2). Despite the White population ranking 3rd in South Africa, among the population aged 60 and over, Whites are significantly higher than that of Coloureds and Indian/Asians. Also despite Coloured ranking 2nd in terms of overall population numbers, Coloured men and women constitute less than 10% of elderly in the country. The disparate distribution of elderly by population group can be linked to the disparate life expectancy levels experienced by population group in South Africa (MYPE, 2017; StatsSA, 2017). Marital status of individuals is found to be a significant factor influencing the living arrangements of elderly in South Africa. The variable marital status comprises of the following categories, never married, married/ living together, widowed and divorced/separated. Unlike at younger ages, whereby migration due to work or study separated those that were married; elderly are unlikely to find themselves separated from their spouse due to migration for work. The majority of elderly men age 60 and over are either married or living with someone as though they are married (73.4%), whilst the majority of elderly woman are widowed (49.3%). Widowhood is consequence of a partner’s death thus rising life expectancy may be related to the decline in the proportion of widowed among men and women. A higher proportion of elderly women (12.7%) have never married when compared to elderly men (8.1%). Over time the proportion elderly men and women are have never married is

• increasing. Over time there is a growing proportion of elderly men and women who indicate they are divorced or
• separated. This is likely to have occurred prior to the age of 60 and may be a consequence of the growing liberal

views regarding the dissolution of marriage over time in South Africa.

11 The distribution of elderly by province in South Africa is similar to the overall population whereby Gauteng followed by KZN depict the highest proportions of elderly in the population. Due to their relatively smaller population size, the Northern Cape and Free State host the smallest populations of elderly in the country. Table 2 also indicates that over time the proportion of elderly males increased in the provinces of Western Cape, North West and Gauteng, whilst the proportion of elderly females increased in the provinces in Western Cape, Gauteng, Mpumalanga and Northern Cape. The discrepancy in the proportion between Census and that of the GHS is related to vastly different methodologies of a survey and census. Education has generally improved among the elderly over time, with the proportion of elderly with no education declining from 26.4% and 36.1% among men and women respectively in 2005 to 19.9% and 25.9% in 2015. However there is great disparity in the level of higher education achieved among elderly men (9.5%) relative to women (3.9%) depicted in 2015. Though the census confirms the overall pattern of education, there are divergent levels of educational achievement that does not fit the trend well, indicating that there may be an underestimation of higher education in the GHS data. According to table 2, a higher proportion of elderly men (23.7%) indicated they were currently employed when compared to elderly woman (10.8%) in 2015. Consequently a high proportion of elderly woman (75.3%) reported accessing a social grant when compared to men (61.8%) in 2015. Given the proportion of elderly reporting being unemployed and accessing a social grant, the relatively high proportion of elderly men (88.1%) reported themselves as head of household when compared to women (61.4%), provides the context under which this headship occurs. Between 2005 and 2015, the proportion of elderly reporting heading a household has declined despite the increase in the proportion elderly over time. This points to the changing patterns of headship and living arrangements amongst the elderly over time. Table 2: Socio demographic characteristics of elderly (60+) over time by gender in South Africa, Census 2011, GHS 2005, 2015 Characteristics 2005 2011 2015 Male Female Male Female Male Female Proportion elderly 5.4 8.1 6.5 9.4 6.6 9.9 Age 60-64 38.2 32.0 37.1 30.9 38,9 30.4 65-69 27.9 28.1 24.4 22.2 27.0 27.6 70-74 17.8 19.3 17.8 18.2 17.0 19.7 75-79 8.7 10.3 10.0 12.6 9.6 10.8 80+ 7.5 10.3 10.7 16.1 7.5 11.5 Population group Black African 60.2 67.1 62.1 67.9 61.8 67.6 Coloured 8.9 7.6 9.0 8.3 9.7 8.6 Indian/Asian 3.7 2.9 3.8 3.3 4.1 3.2

12 White 27.2 22.4 25.1 20.5 24.4 20.6 Marital Status Never Married 4.9 9.1 10.0 16.1 8.1 12.7 Married/ Living together/ 77.5 30.4 75.9 40.9 73,4 32.4 Widowed 14.6 55.4 10.4 38.3 13.3 49.3 Divorced/Separated/ 3.0 5.0 3.7 4.7 5.2 5.7 Province Western Cape 11.9 11.2 13.6 11.8 14.8 12.3 Eastern Cape 14.5 16.9 14.7 15.8 12.7 13.5 Norther Cape 2.3 2.4 2.5 2.3 2.3 2.5 Free State 5.9 5.6 5.4 5.6 5.5 5.6 KwaZulu-Natal 19.3 20.4 17.4 19.7 16.9 19.8 North West 6.7 6.8 7.6 6.7 7.1 6.6 Gauteng 23.2 18.7 22.1 19.1 26.1 23.0 Mpumalanga 7.0 6.1 6.9 6.8 6.3 6.6 Limpopo 9.2 11.8 9.8 12.2 8.3 10.0 Education None 26.4 36.1 23.6 31.0 19,9 25,9 Primary 21.5 21.2 27.4 28.8 26,6 30,9 Secondary 49.5 41.5 37.7 33.8 44,1 39,3 Higher 2.6 1.3 11.4 6.4 9,5 3,9 *Currently employed Yes n/a n/a 33.7 18.9 23.7 10.8 No n/a n/a 66.3 81.1 76.3 89.2 Accessing a social grant Yes 40.5 61.9 n/a n/a 61,8 75,3 No 59.5 38.1 n/a n/a 38.2 24.7 Head of Household Yes 91.1 62.1 84.0 61.0 88.1 61.4 No 8.9 37.9 16.0 39.0 11.2 38.6 Number 1215610 1956641 1648930 2502828 1724192 2718162 *Unemployed includes discouraged and not economically active Table 3 indicates the percentage distribution of elderly by household headship as well as relationship to household head by gender. The majority of elderly age 60 and over indicate that they are heads of households. A significantly higher proportion of elderly men (88.1%) than women (61.4%) report themselves as head of a household, however over time the proportion of elderly men heading a household is declining from 91.1% in 2005 to 88.1% in 2015. Given the higher level of headship among elderly men a significantly higher proportion of elderly woman report themselves as the spouse to household heads (26.8%) when compared to elderly men (3.4%). A high proportion of elderly women reside with individuals they are related to, either child, grandchild or parent, as well as other relatives, whilst a higher proportion of elderly men report residing with people they are unrelated to (1.2%) when compared to women (0.5%). Table 3 points to the vulnerability of elderly men who are not a head of household, as the majority of such elderly men reside with other relatives (4, 5%) and with those unrelated (1.2%).

13 Table 3: Percentage distribution of South Africa’s elderly (60+) by relationship to the household head over time, 2005-2015 Characteristics 2005 2011 2015 Male Female Male Female Male Female Head 91.1 62.1 83.97 61.01 88.1 61.4 Spouse 3.2 24.9 6.12 24.35 3.4 26.8 Child/Grandchild 0.7 0.5 0.46 0.38 0.8 0.9 Parent/grandparent/in law 1.8 7.3 2.70 7.05 1.3 4.8 Other relatives 3.0 4.9 5.53 6.04 4.5 5.0 Unrelated 0.2 0.4 1.22 1.18 1.2 0.5 Unspecified

• 0.1

0.16 Number 1215610 1956641 1648930 2502828 1724192 2718162 Table 4 provides the living arrangements of elderly from both census 2011 and the 2015 GHS. By 2015, elderly men and women are found to be living predominantly with more than 1 person, and as the household head (52.1% and 41.5% respectively).Whilst elderly men dominate in structure where they are the head i.e. living alone, living with a spouse only, living with more than 1 person and as the head of the household, there are a higher proportion

• f elderly women residing in the more precarious or vulnerable living arrangements i.e. living with someone who is

not the spouse, living in household with more than 1 person and not the head of the household. Table 4: Living arrangements of elderly aged (60+) over time by gender, Census 2011, GHS 2015 Census 2011 GHS 2015 Male Female Male Female Average Household Size * * 4.24 4.42 Living arrangement * * Living alone (hh) * * 11.1 9.7 Living with spouse only * * 26.6 20.7 Living with someone but not spouse * * 1.0 1.9 Living with more than 1 Person (HH) * * 52.1 41.5 Living with more than 1 Person (not HH) * * 9.2 26.3 Number 1 648930 2 502 829 1724192 2718162 *to be completed In an attempt to better understand the living arrangements of elderly key indicators such as headship and average household size is worth analysing. In South Africa the average household size within which elderly reside is declining over time. By 2015 the average household size within which elderly reside is higher among women (4.42) than men (4.24) (table 4). Table 5 examines the living arrangement and headship pattern among elderly men and elderly women by household size. Where elderly men are not the head of the household, they are predominantly found in larger household i.e. 6 or more person residing in the household. Elderly heading household structures containing 7 or more persons are more so headed by elderly men (30.9%) than elderly women (27, 7%), however

14 among household sizes 3 -6 we find that a higher proportion elderly women are reported as the head of the household than men. Table 5: Distribution of elderly (60+) by living arrangement, household size and gender, GHS 2015 Household size Living with others and not HH Living with others and HH Male Female Male Female 3 17.6 21.9 23.1 21.0. 4 15.7 17.7 19.0 19.7 5 13.4 15.5 14.7 17.6 6 15.8 13.2 12.4 13.9 7+ 37.5 31.8 30.9 27.7 158 859 714 288 898 914 1 127 727 Table 6 indicates that the socio demographic characteristics as well as living arrangements of elderly differ by

• age. Given the differential pattern mortality between men and women in South Africa it is not surprising that with

age the proportion of men relative to women declined from 44,8% in the age 60-64 to 29,2% in the age 80 and

• ver whilst the proportion of women continue to increase with age to over 70 percent elderly women in the age

80 and over. In South Africa the differential pattern of mortality is made evident even more so by the percentage distribution of elderly by age and population group. Though the proportion of elderly is highest among black Africans, (Black Africans comprise above 80% of the total population regardless of age), Black Africans as well as the Coloured population show a decline in the proportion elderly with age. In contrast the proportion of the White population, increases significant with age from 20.0% in the age 60-64 to 27.1% in the age 80 and over. Given the higher life expectancy among Whites, elderly Whites are found in disproportionately higher proportions in the age group 70 and over when compared to other population groups. The proportion elderly never married decreased with age among the elderly. This may be related to the lower life expectancy among unmarried persons. Numerous studies point to the protective value of marriage, i.e. higher life expectancy of married men and women relative to their unmarried counterparts, particularly among men (Robards et al, 2012; Farr & Edwards, 1885; Hu & Goldman, 1990). Parallel to this, there is a decline in the proportion of elderly married and or cohabiting and an increased in the proportion elderly reporting being divorced, separated or widowed with age. This may be due to a culmination of factors namely mortality as well as more liberal views and practices regarding dissolution of marriage over time. Marriage and widowhood are inversely related. A consequence of ageing is frailty and ultimately death. As the elderly progress into older ages, they are likely to experience loss of a partner, thus higher proportions of widowhood and lower proportions

• f marriage and living together are found among the oldest old.

In a similar manner the level of education among elderly by age is an indication of the level of education over

• time. With each age cohort, the level of education may be indicative of the experience of the education system

15 and various points in time in the past. The proportion of elderly with no formal education increases with age, indicative of the low levels of education experienced by the majority of South Africans living in apartheid South Africa pre 1994. Similarly among those aged 60-64 the proportion of elderly with secondary and higher education is higher than among those 65 and over, indicative of the improvement in access to education over

• time. Adult basic education (ABE) programs as well as other literacy programs in South Africa are seeking to

address the inequity of the past educational system of South Africa that impacted great on the now elderly population. The proportion of elderly increases with age significantly in the more rural provinces of the Eastern Cape, Northern Cape, Mpumalanga and Limpopo. Among the oldest of the old, those aged 80 and over the Eastern Cape and Limpopo showed far higher proportions of elderly relative to their population size. Table 6 also shows that the proportion of elderly engaged in employment drastically declines with age from 29.3% among those aged 60-64 to 12.7% among those aged 65-69, and finally only 3.5% of elderly aged 80+ are currently employed. In contrast to this the proportion of elderly accessing the social grant increases with age from 59.8% among those ages 60-64, to 73.2% among those aged 65-69. Among those aged 80 and over, 79.7% report accessing the social grant. In South Africa, elderly are afforded a non-contributory pension fund and the data implies that as elderly age, they exit employment and are likely to access this grant. There is a trend in the living arrangement amongst elderly aged 60-79 years of age, however there is a disjuncture in the trend when we reach age 80 and over. There is a general increase in the proportion of elderly living alone and living with a spouse only with age, however among elderly aged 80 and over the proportions declined for these categories. The increased mortality of people over time as well as progression of life i.e., children grow and leave their older parents to study, work, marry as well as start their own families, explain the increase in the proportion elderly living alone or living with a spouse only. The disjuncture at age 80 and over is related to the increased proportion of elderly aged 80 living with more than 1 person and heading a household increasing from 14.8% among those aged 75-79 to 21.9% among those aged 80 and over pointing to the vulnerability of many old men and women in old age. On average, in South Africa, elderly tend to reside in households that are larger, with an average household size of 4. 37 among those aged 60 and over and this does not vary with age. Table 6: Socio-demographic characteristics of elderly (60+) by gender and age in South Africa, GHS 2015 60-64 65-69 70-74 75-79 80+ Proportion elderly 33,7 27.4 18.6 10.3 10.0 Gender Male 44.8 38.3 35.5 36.0 29.2 Female 55.2 61.7 64.5 64.0 70.8 Population group Black African 65.7 67.4 65.5 60.3 63.4 Coloured 10.4 8.7 8.1 9.2 7.1

16 Indian/Asian 3.9 3.5 3.4 4.1 2.4 White 20.0 20.5 23.0 26.4 27.1 Marital Status Never Married 13.9 11.0 10.3 6.9 5.1 Married/ Living together/ 58.4 52.0 43.6 36.4 25.6 Divorced/Separated/Widowed 27.7 37.0 46.0 56.7 69.2 Education None 14.0 18.4 23.9 30.2 40.2 Primary 29.0 31.3 31.7 27.3 21.7 Secondary 47.6 42.0 36.4 35.5 31.7 Higher 6.8 6.6 6.1 4.6 3.8 Don’t know/ other 2.7 1.7 2.0 2.4 2.6 Province Western Cape 14.4 12.8 12.8 13.5 11.2 Eastern Cape 11.6 14.5 14.2 12.1 14.1 Norther Cape 2.3 2.4 2.5 2.2 3.0 Free State 5.9 5.4 5.8 4.6 5.3 KwaZulu-Natal 18.1 19.4 19.4 19.3 17.1 North West 7.2 6.1 7.1 7.0 6.4 Gauteng 26.6 24.1 21.5 23.4 22.3 Mpumalanga 6.4 6.0 7.1 7.5 6.4 Limpopo 7.5 9.3 9.4 10.4 14.2 Accessing a social grant Yes 59.8 73.2 75.5 75.7 79.7 No 40.2 26.8 24.5 24.3 20.3 Currently employed Yes 29.3 12.7 8.4 5.2 3.5 No 70.7 87.3 91.6 94.8 96.5 Average household size 4.37 4.33 4.36 4.37 4.27 Living arrangement Living alone (HH) 8.7 9.8 10.9 13.3 12.2 Living with spouse only 21.7 24.0 24.3 24.8 20.0 Living with someone but not spouse 1.7 1.4 1.1 1.2 2.3 Living with more than 1 Person (HH) 21.7 19.5 17.6 14.8 21.9 Living with more than 1 Person (not HH) 26.1 45.3 46.1 45.8 43.5 Number 1496347 1216235 828243 458675 442855 Table 7 examines the selected living arrangements of elderly by selected social and demographic characteristics

• f elderly in South Africa by gender using the general household survey 2015. Elderly men in younger ages 60-

69 dominate in structures where they are either independent i.e. living alone or are depended on as household

• head. In contrast elderly women in older ages 70 and over dominate in structures where they are either

independent i.e. living alone or are depended on as household head. Elderly black African men dominate in structures where they are independent or depended upon i.e. living alone (75.7%) or living with one or more person has a household head (74.6%). Black African elderly women too are

17 predominantly living with more than one person as a household head (88.2%). Coloured elderly men and women

• ften reside with one or more persons but not as the household head (28.4% and 13.7% respectively). Elderly

Indian/Asians and Whites live predominantly with spouses. Compared to elderly women of other population group’s elderly white woman are 3 times less likely to reside with more than one person as a head of household than their white male counterparts. As expected elderly married or living together as a married couple predominantly reside with a spouse only. In a patriarchal society such as South Africa it is not uncommon for men to dominate as household heads. Less than 3% of elderly women who are married live alone and only 7% of married women report themselves as heads of household when living with more than 1 person. Marital status directly influences the headship of elderly women. Elderly women who are divorced, separated and or widowed predominantly live alone or live with more than one person as a household head whilst elderly men who are divorced, separated and or widowed predominantly live alone. Provincial patterns of living arrangements among the elderly indicates that the more rural provinces of KwaZulu- Natal, Eastern Cape, Mpumalanga and Limpopo have higher proportions of elderly men and women residing with more than one person. In contrast Gauteng has the highest proportion of elderly men and women living in more isolated arrangements such as living alone and living with a spouse only when compared to other living

• arrangements. The Northern Cape and Free State have high proportions of elderly living alone when compared

to other living arrangements. Elderly men and women with no formal education or primary education reside more so with more than one person and often as the household head. The burden of care seems greatest amongst this group, who are predominantly elderly women. Despite the well-known link between education, employment and income, the data indicated that elderly with none or primary education are found to be caring for other as heads of households, whilst elderly women with secondary and higher education are less likely to be caring for others as a head of

• household. Less than 1 percent of elderly women with higher education are found to be living with others as a

household head (0.8%). Elderly men with higher education are found to be living with a spouse only (20%) Elderly men and women living alone are found to have the highest levels of employment along with almost two thirds of these accessing a social grant. This seems to be the mechanisms that allows such elderly to afford an independent living arrangement. Among those living with a spouse only we find an interesting pattern of high employment as well as low access to social grant. This may indicative of a pooling of resources between both elderly spousal partners to afford their living arrangement, without having to access a social grant. A higher proportion of elderly residing with more than one person, access the social grant. Elderly living with more than 1 person also have the lowest levels of employment. Caregiving responsibilities may not afford elderly living in larger households the opportunity to continue with formal employment but rather the dependence on a grant

18 seems a more practical source of income. This is particularly the case for elderly women than elderly men living with more than 1 person. Table 7: Socio demographic characteristics of elderly (60+) by living arrangements and gender in South Africa, GHS 2015 Characteristics Living Alone Living with Spouse Living with more than 1 person Living with more than 1 person (hh) Male Female Male Female Male Female Male Female Proportion elderly 60-64 42.7 18.2 34.5 29.7 39.2 36.8 39.7 29.6 65-69 24.4 27.5 26.2 30.6 25.1 27.6 28.6 26.1 70-74 18.4 20.9 19.0 20.4 14.8 17.1 16.2 20.9 75-79 7.8 17.6 11.7 10.7 10.9 7.1 8.7 11.7 80+ 6.7 15.8 8.6 8.7 10.0 11.4 6.7 11.7 Population group Black African 75.7 48.8 34.7 47.1 50.4 60.3 74.6 88.2 Coloured 5.1 2.9 5.6 5.8 28.4 13.7 9.6 8.1 Indian/Asian 2.2 2.6 5.2 5.1 4.4 5.3 3.9 0.9 White 17.0 45.6 54.4 42.0 16.8 20.7 12.0 2.8 Marital Status Never Married 20.2 15.4 3.4 9.2 34.1 7.4 2.6 16.5 Married/ Living together/ 20.1 2.4 87.4 55.2 43.9 66.9 83.9 6.7 Divorced/Separated/Widowe 59.7 82.2 9.2 35.7 22.0 25.8 13.5 76.8 Province Western Cape 12.4 15.6 19.1 17.0 27.0 15.1 10.9 7.4 Eastern Cape 15.2 12.1 8.0 11.4 6.6 11.3 15.7 16.7 Norther Cape 2.9 3.5 1.9 2.3 2.3 2.7 2.3 2.2 Free State 5.4 6.9 5.8 5.7 5.1 4.9 5.5 5.8 KwaZulu-Natal 14.7 14.0 13.9 14.0 16.5 20.8 19.1 23.5 North West 9.5 6.6 6.5 7.7 5.9 5.7 7.0 6.6 Gauteng 22.9 28.8 35.9 29.7 24.4 22.2 22.2 18.4 Mpumalanga 6.9 6.2 4.2 5.0 4.8 7.7 7.6 6.7 Limpopo 10.2 6.3 4.7 7.0 7.3 9.6 9.8 12.6 Average hh size n/a n/a n/a n/a 6.89 5.79 5.64 5.65 Education None 24.5 16.1 8.2 15.6. 24.3 23.3 19.9 30.2 Primary 31.0 20.8 14.0 20.3 27.8 29.6 31.7 39.8 Secondary 40.4 56.1 55.9 51.7 41.2 41.1 39.3 27.2 Higher 4.2 5.6 20.6 9.8 2.5 3.4 6.1 0.8 Don’t know 0.0 1.4 1.3 2.6 4.1 2.6 3.0 2.0 Accessing a social grant 68.8 61.5 39.1 56.3 71.0 74.7 70.1 88.9 Currently employed 29.5 15.9 29.5 11.5 13.0 9.5 21.5 10.0 Household head n/a n/a 96 49.7 n/a n/a n/a n/a

1912 262633 45865 561436 158858 714288 898916 112772

19 Regression analysis to be finalised. Discussion and Conclusion The results thus far indicate a significant relationship between the selected socio demographics of elderly and their choice or circumstance of living arrangement. The proportion of elderly persons living alone has been growing particularly amongst elderly women. Compared to more developed countries South Africa is ill prepared to cater for the needs of the growing number and proportion of elderly. Elderly men and women are recognised as a vulnerable group in in South Africa, based on the vulnerability of their financial and physical capabilities. Though the non-contributory pension fund does attempt to provide an economic means to assist elderly in ensuring their dignity in older age, the physical and economic assistance for the elderly can also be achieved by improving the living conditions of elderly in South Africa. Living arrangement are known for improving the lives of elderly through, social, economic and physical support. The plight of the elderly persons living alone in the absence of family support system and limited resources makes this group very vulnerable, requiring government

• interventions. Among elderly women, the propensity to live alone increases with age. White elderly women are

more likely to live alone than Black African elderly women. Elderly women living alone were most likely found in the Western Cape and Eastern Cape. However Elderly men and women living alone are found to have the highest levels of employment along with almost two thirds of these accessing a social grant. This seems to be the mechanisms that allows such elderly to afford an independent living arrangement. Similarly Less than 1 percent of elderly women with higher education are found to be living with others as a household head (0.8%). Elderly men with higher education are found to be living with a spouse only (20%). Among those living with a spouse only we find an interesting pattern of high employment as well as low access to social grant. This may indicative of a pooling of resources between both elderly spousal partners to afford their living arrangement, without having to access a social grant. Coloured men and women have significantly higher odds of living with others but not as household head. Black African elderly men are least likely to live with more than 1 person and not head that household. These patterns point to social and cultural prescripts that differ by population group that dictates the living arrangements of Black African elderly men in particular. Despite the well-known link between education, employment and income, the data indicated that elderly with none or primary education are found to be caring for other as heads of households, whilst elderly women with secondary and higher education are less likely to be caring for others as a head of household. Black African men and women are more likely to be living with others as a household head. Elderly women caring for others as heads of households are not only more likely to have lower levels education but also more likely to be unemployed and have no access to a social grant. These results point not only to the disparate burden of care placed upon Black African men and women, but more especially the vulnerability of Black African elderly women residing predominantly in the Western and Eastern Cape, KwaZulu-Natal and Limpopo.

20 Home based care and mobile heath social services should target key population among the elderly in South Africa to alleviate the added vulnerability of caring for others. Unlike in most western countries, in South Africa the vulnerability seems higher amongst elderly caring for other and heading households. Living arrangements are fundamental in old age, particularly during frail years, as it is determining factor on the extent to which elderly access health care and social support. By determining the key factors influencing living arrangements of elderly, programs and policies that impact those particular determinants can be developed to more effectively address the wellbeing of elderly. Living arrangements are known for improving the lives of elderly through, social, economic and physical support. The plight of the elderly persons caring for other in the absence of family and financial support and limited resources makes this group of people vulnerable, requiring government interventions. References Abejo, S (2004), 'Living Arrangements of the Elderly in the Philippines', 9th National Convention on Statistics (NCS) (EDSA Shangri-La Hotel). Aghajanian, et al. (2013), 'Living Arrangement of the Elderly in Iran', Sociology Department Working Paper, Paper 8. Amoateng, A.Y et al (2007) Chapter 3: Living arrangement in South Africa in Families and Households in Post- apartheid South Africa. Human Science research council Argawal S (2012) effect on living arrangements on the health status of elderly in India. Asian Population studies 8: 87-101 Joubert, J. and Bradshaw, D (2006), 'Chapter 15: Population Ageing and Health Challenges in South Africa ', in K Steyn, J Fourie, and N Temple (eds.), Chronic Diseases of Lifestyle in South Africa: 1995 - 2005 (CapeTown: Medical Research Council). Legido- Quigley, H. (2003), 'The South African Old Age Pension: Exploring the role on poverty alleviation in households affected by HIV/AIDS', The 4th International Research Conference on Social Security, “Social Security in a Long-Life Society (Antwerp, Belgium). Munthree, C. and Maharaj, P. (2012) Growing Old in the Era of AIDS: Perspectives and Experiences of Older Men and Women in South Africa, Chapter 6 in Aging and Health in Africa, International Perspective on Ageing 4, Springer Science Martin, L. and Kinsella, K. (1994), 'Research on the demography of aging in developing countries. ‘in L. Martin and S. Preston (eds.), Demography of Aging (Washington, DC.: National Academy Press).

• Mba. C.J (2002), 'Department of Living Arrangements of Lesotho's Elderly Female Population', Journal of

International Women's Studies 3(2), 1-22. May, J (2003) Chronic poverty and old people in South Africa. CPRC working paper 25 February 2003 School of Development Studies, UND. Helpage International. Chronic poverty research centre ISBN Number: 1-904049- 24-9 National Treasury (2007) Social security and retirement reform. Second Discussion paper. National Treasury.

21 Nyambedha, E. , Wandibba, S., and Aagaard-Hansen, J. (2003), 'Retirement lost – the new role of the elderly as caretakers for orphans in western Kenya', Journal of Cross-Cultural Gerontology, 18, 33-52. Posel, D. R. (2001). Who are heads of household, what do they do, and is the concept of headship useful? An analysis of headship in South Africa. Development Southern Africa, 15(5), 651-670. Russel, D (2007) Household composition and psychological wellbeing (dissertation). Florida State University. Rosenhouse, Sandra (1989) 'Identifying the Poor: Is 'Headship' A Useful Concept?' World Bank Living Standards Measurement Study Working Paper No. 58. Washington DC: The International Bank for Reconstruction and

• Development. The World Bank.

Ramashala, M.F. (2002), Living arrangements, poverty and the health of older persons in Africa. (United Nations.

• Dept. of Economic and Social Affairs. Population Division).

Roy, Chaudhuri and (2007), 'Gender differences in living arrangements among older persons in India'. Sabiti, et al. (2007) Chapter 5: Family formation and dissolution patterns. Human Science research council StatsSA (2014) Report-03-01-60 - Census 2011: Profile of older persons in South Africa, 2011. StatsSA. October 2014 StatsSA (2017b) Midyear population estimates 2017. Statistics South Africa TOHME R.A et al (2010), 'Socioeconomic resources and living arrangements of older adults in Lebanon', Who chooses to live alone? (Cambridge University Press 2010). United Nations, Department of Economic and Social Affairs, Population Division (2015). World Population Ageing 2015 http://www.un.org/en/development/desa/population/publications/pdf/ageing/WPA2015_Report.pdf WHO (2015) http://www.who.int/ageing/publications/global_health.pdf .