Children in the Shadow of AIDS: Studies of Vulnerable Children and Orphans in Three Provinces in South Africa Michelle Adato, Suneetha Kadiyala, Terence Roopnaraine, Patricia Biermayr-Jenzano and Amy Norman November 2005 Michelle Adato International Food Policy Research Institute 2033 K Street, NW Washington, DC 20006-1002, USA Suneetha Kadiyala International Food Policy Research Institute 2033 K Street, NW Washington, DC 20006-1002, USA Terence Roopnaraine 19B Rajdoot Marg Chanakyapuri New Delhi India Patricia Biermayr-Jenzano Center for Latin America Studies Intercultural Center 484 Georgetown University Washington DC 20057-1026 Amy Norman Department of Geography University of Calgary 2500 University Dr NW Calgary, AB T2N 1N4, Canada ii Abstract This paper examines the experiences of children affected by HIV/AIDS in three provinces of South Africa: Eastern Cape, Western Cape, and KwaZulu-Natal.. By combining the findings of two different studies, the paper analyzes the conditions of children at different stages of impact. It looks at the conditions of vulnerable children living in households with HIV positive members; children at risk of becoming orphans (i.e. children living with HIV positive primary caregivers); and children orphaned after their biological parents have passed away. Using primarily in-depth interviews and observations, the paper contributes toward filling two knowledge gaps—one on children vulnerable to becoming orphans and another on those already orphaned. In Western and Eastern Cape, we interviewed nineteen HIV positive mothers and primary care givers to investigate strategies employed by families in response to HIV/AIDS stresses, focusing on children and examining how parents plan for the future security of their children. Our results emphasize the role of already established patterns of childcare arrangements as primary safety nets in the context of AIDS in South Africa. Children are intimately involved with providing care, support and assistance in treatment to HIV mothers and younger siblings. Our evidence demonstrates the positive impacts of disclosure of HIV status to children and the ability of children to get involved in the care giving and support of their mothers. We conclude that if we are to ensure the future well-being of children, we must first recognize the roles and responsibilities that children are already shouldering and then empower them through a variety of appropriate polices and programs before they enter orphanhood. For mothers who are aware of their HIV positive status, making plans for their children represented a chance to take part in the future growth and development of their children. HIV positive mothers are actively planning for the future of their children, within their limited resources. The plans ranged from organizing future care giving arrangements to preparing wills for inheritance. While many of the women in our study emphasized the desire to save for things like the future educational requirements of their children, meager earnings made this rarely possible. In order to strengthen the roles of mothers, we must understand the dynamics of planning, the challenges HIV positive women face in parenting and the strategies they are utilizing in order to secure their children’s future This paper calls for inclusion of parents in the future policy and planning surrounding the issue of OVCs in South Africa In KwaZulu-Natal, ethnographic research methods were used in 6 localities, with repeated visits, interviewing and observation of 18 households fostering orphans, as well as key informant interviews with individuals involved with community-level interventions to support orphans. The paper also draws on a survey dataset of 1,428 households across KwaZulu-Natal. The qualitative research found that processes of fostering children orphaned by AIDS have articulated with historical patterns of mobility, and with notions of African culture and obligations related to lineage patterns. Family structures are thus far largely coping with the care of orphans, though under the strain of poverty. We found divergence from idealized protocols of patrilineal responsibility because terminally ill mothers are often cared for by their families and children remain in the same household after their mother’s death; and because many children do not maintain links with their fathers and/or fathers’ relatives. Few conflicts were found around iii decisions to take in children, although where they occurred they were related to tensions between the patrilocal residence ideal and the matrilocal status quo; or to efforts to obtain the deceased’s property or access to social grants. The main fostering parents are relatives, primarily grandparents and aunts and uncles. People express a strong ideal of African cultural norms that require that orphans be treated the same as the children of the fostering relatives, and observations largely confirmed this—though we also found cases of discrimination against orphans in some households. Survey data on 333 fostered children also show few differences between orphans and non-orphans in schooling-related indicators, which may be because children are fostered by mainly by close relatives. Key informant interviews suggest that child- headed households face particular problems and risks with respect to food and nutrition, schooling, health, violence, crime, discipline, teen pregnancy, and HIV/AIDS. There is a lack of specialized services to assist them. State grants provide a crucial social safety net for orphans, particularly the Old Age Pension and the Child Support Grant. The survey data show that about 30 percent of fostering households are receiving the Child Support Grant. However, fostering households have a higher likelihood of unsuccessful applications because the applicant was not able to be established as the primary care giver. Only a miniscule percentage of fostering households in the survey are receiving the FCG. The qualitative research found the main reasons for low uptake to be lack of knowledge and assistance with respect to the application process; concern over length of time involved, doubts about success of applications based on observations of widespread failure across one’s social networks; and anticipation or experiences of bureaucratic problems. Apart from grants, most forms of support for orphans were informal. Neighbors and friends help out, but these social networks are made up of poor people with little to share. Material forms of assistance include mainly cash or food donations, loans, clothing, school uniforms and school fees. Creative community-based initiatives included drop-in and community centers offering recreational activities, school performance monitoring, after-school feeding and take-home food, counseling, parenting skills, and assistance with grant applications. These interventions were few and ad-hoc, however, and mainly supported by local contributions, though some had government support. Given the size and rapid acceleration of the orphans crisis in South Africa as elsewhere on the continent, there is an urgent need for a systematic approach to the needs of orphans and fostering households. However, it is also important to find intervention designs that both target orphans and support other vulnerable children, so that policy is non-stigmatizing and fair to other poor children and households. iv Acknowledgements The authors would like to thank the following individuals for their important roles in the research upon which this paper is based: For the Western and Eastern Cape studies: Lungiswa Tsolekile for carrying out the fieldwork along with Amy Norman; Weliswa Binza, Ikhwezi Community Centre and Mbekweni Community Health Project for facilitating the fieldwork logistics; Mickey Chopra and Stuart Gillespie for help in conceptualizing the study and on-going guidance. For the KwaZulu-Natal study: Makhosazana Dlalisa, Wilfred Themba Mbhele, and Themba Mpanza for carrying out the fieldwork; Nina Hunter, Catherine van de Ruit, Myriam Velia, and Julian May for supervision of data collection and processing; Futoshi Yamauchi for analysis of KIDS 3 survey data; Nelly Tioco and Jay Willis for production assistance. The KIDS 3 study was a project of the University of KwaZulu-Natal, the International Food Policy Research Institute, University of Wisconsin, Madison, London School of Hygiene and Tropical Medicine, the Norwegian Institute of Urban and Regional Studies, and the South African Department of Social Development. Financial support for the research came from the Regional Network on HIV/AIDS, Rural Livelihoods and Food Security (http://www.ifpri.org/renewal/), coordinated by the International Food Policy Research Institute; U.K. Department for International Development-South Africa (DfID-SA); the U.S. Agency for International Development (USAID), the Mellon Foundation, the National Research Foundation/Norwegian Research Council, the Government of Japan, and the United Nations World Food Programme. The views expressed in this paper are those of the authors. v Table of Contents 1. Introduction................................................................................................................................ 1 2. Research Sites, Methods, and Data............................................................................................ 7 2.1 The Western Cape and Eastern Cape Studies ...................................................................... 7 2.2 The KwaZulu-Natal Study................................................................................................... 9 3. Preparedness of Children at Risk of Orphanhood in the Context of HIV/AIDS: Findings from Western Cape and Eastern Cape Province ....................................................... 12 3.1 Children’s Mobility and Inter-Household Links................................................................ 12 3.2 Present Roles of Children Living in HIV/AIDS-Affected Households............................. 15 3.3 “Talking” About Future with Children .............................................................................. 21 3.4 The Process of Planning for Children’s Future Welfare and Security............................... 22 4. Living as Orphans: Findings from KwaZulu-Natal Province.................................................. 29 4.1 Who Fosters Orphans? Culture, Agreement, and Conflict in Household Selection......... 29 4.2 Discrimination and Stigma ................................................................................................ 33 4.3 Child-Headed Households ................................................................................................. 39 5. Informal and Formal Support for Orphans .............................................................................. 43 5.1 Informal Support and Formal Interventions for Orphans in Africa................................... 43 5.2 Support for Orphans in South Africa: The Child Support Grant and the Foster Care Grant............................................................................................................... 44 5.3 Other forms of Support for OVCs ..................................................................................... 51 6. Conclusions and Policy Implications....................................................................................... 59 References..................................................................................................................................... 67 Children in the Shadow of AIDS: Studies of Vulnerable Children and Orphans in Three Provinces in South Africa 1. Introduction This paper examines the experiences of children affected by HIV/AIDS in three provinces of South Africa.1 By combining findings from two studies that focus on households at different stages of impact, the paper looks at the conditions of vulnerable children living in homes with HIV positive family members; children at risk of becoming orphans (i.e., children living with HIV positive primary caregivers); and children ‘orphaned.’ It also considers the conditions of other household members affected at each of these stages. The paper is concerned with understanding more about their daily lives, and about informal and formal forms of support available to them at each of these stages. With a population of almost 45 million people and an estimated 5.3 million people infected of HIV, South Africa, has the largest number of individuals living with the HIV virus in a single country. By the end of 2003 the prevalence rate among adults (15-49 years of age) in South Africa was 21.5 percentand among pregnant women 27.9 percent. The rate of infection varies widely among South Africa’s nine provinces. The highest was in KwaZulu-Natal with 37.5 percent and the lowest in Western Cape with 13.1 percent (both provinces are included in our studies).2 Children orphaned by AIDS are found virtually in every country across Sub-Saharan Africa. An estimated 12. 3 million children have lost one or both parents to AIDS in the region (Deininger et al. 2003; Birdthistle, 2003; UNICEF 2004; UNAIDS 2004). Simulations infer that the numbers are still higher and that mortality rates will reach a plateau only by 2020, with expectations of stabilization by the year 2030. The hardest hit country in the region in terms of orphans is South Africa with both the percentage of children who are orphans and the absolute number of children who are orphaned rising dramatically. A recent analysis by UNAIDS/UNICEF/USAID (2004) estimates that in South Africa, out of a total of 17 million children under 17 years of age, 13 percent or 2.2 million are orphans. The number of orphans due to AIDS is estimated to be approximately half of that figure or 1.1 million. Of these children, an estimated 740,000 are maternal orphans; 570,000 are paternal orphans and 290,000 are double orphans.3 Meintjes et al. (2003) estimated future number of orphans from two models: the Actuary Society of South 1 In the Eastern and Western Cape Study, study households had independently confirmed HIV+ status (see section 2). In the Western Cape study, households were not asked, and only in some cases volunteered, whether the orphans deceased parents had died from AIDS. However, we feel it is safe to assume that this was the case in many and probably most of the households, and that the information gathered on orphans in the study households provide insights into the conditions of AIDS orphans in the region. 2 Other important estimates are for women (15-49) living with HIV, at 2.9 million (range: 2.5-3.3 million) and AIDS related deaths (adult and children) at 370,000 (range 270-520 thousand). Over the last four years, consecutively, the rate of HIV infection for young people below the age of 20 has stabilized (Dorrington et. al. 2002; UNAIDS 2004; 2005). 3 Maternal orphans are those children under the age of 15 -or 18 in some contexts- whose mothers have died. Paternal orphans those children whose fathers have died. Double orphans are those children without either living parent (Levine 2001). 2 Africa’s ASSA2000 AIDS and the Demographic model, and the ASSA Orphans model. They found that the number of children orphaned by AIDS is expected to peak at 2.7 million in 2016. 4 With the emergence of HIV/AIDS as Africa’s most prevalent health crisis, the term orphans and vulnerable children (OVCs) is increasingly being used to refer to children made vulnerable by the impacts of HIV/AIDS. Research on OVCs in Africa has primarily focused on conditions of, and interventions to support, orphans. Studies to date highlight the vulnerability of orphans to the HIV virus and impacts of AIDS. Hallman (2004) shows that controlling for wealth and other factors, orphanhood confers added risk for unsafe sexual behaviors. Orphans are likely to be more food-insecure, more malnourished, and less healthy than non-orphans (Ainsworth and Semali 2000; Gertler et al. 2004; Deininger et al. 2003). However, recent studies show that orphans are not more likely to be malnourished than non-orphans (Mason et al. 2005). Many studies have reported the negative effects of HIV/AIDS on children’s schooling (Urassa et al. 1997; Gilborn et al. 2001; Yamano and Jayne 2005; Deininger et al. 2003). These studies show that the impacts of adult death on children is mediated by the type of orphanhood (maternal or paternal), gender and of the child and the initial conditions of the households. Therefore, context specific understanding of pathways through which adult illness and death impacts children becomes crucial. A small body of work focuses on what strategies improve the well-being of children impacted by HIV/AIDS. Deininger et al. (2003) showed that after the adoption of universal primary education in Uganda in 1997, being a foster child no longer impeded access to education. In the absence of proactive policies, however, young foster children’s access to health services worsened over time. School feeding has been seen to keep children in school in Zimbabwe. In a study of food-insecure households, 25 percent of children had dropped out of nonfeeding schools in the previous year, compared to 15 percent where school feeding was in place (SADC FANR 2003). Most recommendations focus on building on existing structures within communities, such as self-help groups, women’s groups, and church groups (Nyambedha et al. 2001). Ainsworth and Semali (2000) in Tanzania found that immunization against measles, oral rehydration therapy, and improved access to health care can disproportionately improve health outcomes among poorest children, and, within that group, particularly among children affected by adult mortality. Relatively fewer studies, however, has focused closely on the experience of children made vulnerable by HIV/AIDS but not yet orphaned5 (Meintjes et al 2004; Giese et al 2003). That is, most of the work referenced above on orphans in Africa focuses primarily on the downstream impact of orphanhood due to adult mortality and very little on the upstream experiences of 4 Conversely, the number of children orphaned for other reasons other than AIDS, is expected to decline which is in line with falling fertility rates. Other estimates predict the rise of maternal orphans by each province in South Africa, which would account to 3-4 percent in the absence of the AIDS epidemic, however, with the impact of AIDS the proportion of children maternally orphaned can rise to 20 percent, with the highest percentage for KwaZulu-Natal, the Free State and Mpumalanga. 5 Although the some of the above studies focus on children in communities with high HIV prevalence rather than orphans per se, focus on children at risk of orphanhood is generally sparse in literature. An exception is Giese et al. (2003), who are concerned with the social needs of children living with sick adults, observing, inter alia, that social workers were not assisting caregivers to plan for their children’s future, nor directing services at children living with sick adults, two clear social needs. 3 children before they are orphaned. Orphaning in the context of HIV/AIDS is a process that begins long before the death of a parent. HIV positive parents, their children and other caregivers face different challenges at different times along the continuum of morbidity, mortality and orphanhood. For children, death of a parent or sibling is an experience that could potentially change their life forever. Often the trauma of death due to terminal illness starts well before the death of a loved one, as is the case with death due to AIDS. It is incurable and the treatment, if available, often not accessible. Therefore, it is often accompanied by fear. In addition, stigma frequently accompanies illness and death related to AIDS. Children as well as adults may remain silent for the fear of being ridiculed or ostracized. This may prevent parents and children accessing resources that may strengthen capacities of children to deal with challenges posed by illness and death of an adult in the family. Another unique aspect of HIV/AIDS is the clustering of death within households. Multiple and serial illness and deaths within families means that children in these households may be overburdened, may be forced to move from home to home, community and school. Children may feel guilt, fear and abandonment. They often shoulder new roles and responsibilities, in addition to their existing ones. Understanding of the trajectory of experiences of children before and after they experience death of their parents in the dynamic context of HIV/AIDS is largely missing from current research and policy perspectives. Also missing is how HIV positive parents anticipate and plan for children’s future. In South Africa, research on the ‘downstream’ dimension of OVCs—orphans—has so far looked statistically at numbers and predicted numbers of orphans (UNICEF 2004; Meintjes et al. 2003; Birdthistle, 2003); impacts on schooling (Case and Ardington 2004; Cichello 2004); interventions (Richter et al. 2004), fosterage patterns (Madhavan, 2004), mobility patterns and migration (Hosegood and Ford 2003); social consequences (Brady 2003) and state policy with particular reference to the Foster Care Grant (Meintjies et al. 2004). But this is still a relatively new field of research in South Africa, compared to that on other dimensions of HIV/AIDS or child poverty, and it is still difficult to draw conclusions about the more broadly-defined conditions and needs of orphans. Some of the implications of this research is somewhat at odds, e.g., around the issue of discrimination against orphans. Our research attempts to shed some new light on their conditions in their changed households, their experience and that of their caretakers. It also focuses on the issue of interventions aimed at assisting OVCs—finding—in contrast to many other parts of Africa—a substantial but widely inaccessible Foster Care Grant, and a near absence of any other interventions by the state or NGOs. In discussing OVC interventions across Africa, Richter (2004) argues that in Africa there is a ‘collective ignorance’ on the impact of AIDS affecting children and families and that information seems to be scattered and available in what they call ‘pockets of local knowledge’. This paper thus contributes toward filling in two knowledge gaps—one on children vulnerable to becoming orphans and another on those already orphaned—using two different data sets of in- depth household interviews from three provinces in South Africa—Eastern Cape, Western Cape, and KwaZulu-Natal. It would have been most desirable to track the same children before and after their experience of adult mortality to understand the trajectory of experiences, but given the background of the two studies (see section 2) this was not possible. Instead we use two cross- sections of children from two different data sets. The first data set primarily focuses on children living with an adult of known HIV positive status (soon to be orphaned); the second data set focuses on children already orphaned. 4 Adult and Child Mobility in the South African Context For both studies, historical and current family structure and mobility patterns are key themes shaping the questions of inquiry and interpretation of the research findings. Apartheid legislation had powerful and long-lasting effects on family structure in South Africa, especially for the African population.6 Under the Influx Control Laws (repealed in 1986 but gradually subsiding), Africans were relocated to “homelands,” with severe restrictions on their ability to travel or find work (Beinart 1994; Jones 1993). Africans living in cities were restricted to townships, small ghettos that generally had inferior housing and public facilities (Jones 1993). Apartheid legislation contributed to a shift to increased complexity in household organization among Africans (Van der Waal 1996; Niehaus 1994; van der Vliet 1991). South African adults were often forced to seek employment and to live apart from their spouses and families. As a result of migratory labor patterns and restrictive housing options, women often became de facto heads of household. Many households came to depend heavily on financial remittances sent in from family members employed elsewhere (Leibbrandt et al. 2000, Posel 2001, Spiegel et al. 1996). In part due to these migratory labor patterns, divorce and non-marital births increased greatly for Africans during the apartheid era (Burman and van der Spuy 1996). Migratory labor patterns meant that one or both of a child's parents were often not present for much of the year, even if they were considered to be current members of the household (Case and Deaton 1998; Siqwana-Ndulo 1998). Parents began to rely increasingly on family members other than spouses for support with the household economy and with raising children (Madhavan 2004; Van der Waal 1996), and rates of child fostering increased (Gordon and Spiegel 1993). Thus, there is a long-history in South Africa, especially among Africans, of children being parented by women (e.g., aunts and grand mothers) if their own mother is absent for periods of time, at least for periods of their lives (Meintjes et al. 2003). Widespread practices of kin fostering; Zulu cultural norms that make out-of-wedlock children part of the maternal family; the frequent unknown status of the father; and widespread work-driven migration leaving children indefinitely with relatives all have led to children experiencing the “death of a social parent” even where biological parents may be alive (Zhuwau 2005). The concepts of “social parenting” vs. “biological parenting” and the implications of this distinction in the context of an orphan crisis in South Africa has thus become a debate in South Africa. While social parenting is not a new concept to South Africa, HIV/AIDS brings with it specific challenges—stigma, discrimination, high costs of medical care, premature death of primary caregivers and clustering of deaths in households. The situation of children in this context can only be understood through a systematic study of household mobility, kinship and the context of fosterage. Maadhavan (2004) makes a useful distinction between voluntary fostering and crisis- fostering. The former refers to informal fosterage that accord with cultural norms about child rearing. Crisis-led fostering pertains to fostering done in response to a crisis—mostly death of biological parents or economic hardship. While crisis-led fostering more accurately describes the situation of children fostered to cope with impacts of HIV/AIDS, this process is articulated with 6 We use the term “African” to refer to black South Africans of African origins. Although this was an apartheid racial classification, it is still generally used when needed to distinguish between groups with distinct cultural and political backgrounds, when such distinction has relevance for the subject at hand. 5 historical patterns of voluntary fostering. The conditions of OVCs in South Africa, and the findings and conclusions of our research on this topic, can only be understood in relation to these processes. Section 2 describes our study sites, research designs, methods and data. Section 3 presents the results of the Western Cape and Eastern Cape studies, focusing on children’s mobility and inter- household linkages; the roles of children in HIV/AIDS affected households; disclosure to children; and planning for children’s welfare and security. Section 4 presents results from the KwaZulu-Natal study, focusing on who fosters orphans; issues and processes around decisions to foster; and experiences of orphans in fostering households. This section also looks at the conditions of child-headed households. Section 5 presents results from the KwaZulu-Natal study that look at interventions; including informal and formal types of support; including cash grants as well as multi-dimensional forms of support. Section 6 considers policy implications and draws some conclusions and recommendations. 6 2. Research Sites, Methods, and Data This paper is based primarily upon two studies using qualitative research methods. Both were part of larger studies that also included panel surveys, thus integrating quantitative and qualitative research methods. The first study, focusing primarily on vulnerable children not yet orphaned, took place in Western Cape Province and Eastern Cape Province, and was part of a larger prevention of mother to child (PMTCT) cohort study. The South African Department of Health supported School of Public Health, University of the Western Cape, the Health Systems Trust (a local NGO) and Medical Research Council to jointly conduct PMTCT cohort study. The second study, focusing on orphans, took place in KwaZulu-Natal Province. This was associated with a panel study on poverty dynamics conducted in 1993, 1998, 2001 and 2004, by the University of KwaZulu-Natal, IFPRI, and several other institutions (see Adato, Carter and May. forthcoming; Carter and May 2001; May et al. 2000). It was conducted with a small sub- sample of the panel survey households, and involved new qualitative modules that had not been previously included. 2.1 The Western Cape and Eastern Cape Studies Sampling Framework The original aim of the study was to build upon a cohort of women who participated in the evaluation of a prevention of a mother-to-child transmission (PMTCT) trial and whose HIV status was already known. The PMTCT trial was initiated in 2003 (Jackson et al. 2004). Recruitment and initial home visits were conducted across the two sites with a total of 400 mother-child pairs identified (of whom 300 are positive and 100 are negative). Our aim was to purposively sample PMTCT mothers who were experiencing HIV/AIDS-related ‘shocks’ which we defined as AIDS-related mortality, prolonged illness, or where children from outside of the home had been taken in due to an AIDS-related death of the parents to study the response strategies households adapt to these shocks. While this was possible in in the Eastern Cape site, in the Western Cape site the majority of the PMTCT women are relatively healthy pregnant women that learnt of their status when they were enrolled in the PMTCT trial and were not yet experiencing any of the above mentioned impacts/shocks related to HIV/AIDS. Therefore, we altered our sampling strategy in Western Cape to enroll HIV positive women, who may or may not be a part of the PMTCT trial, but are aware of their status, and experiencing HIV/AIDS related impacts at the household level as mentioned above. In Western Cape we utilized three sources to select our respondents. The first source was that of Ikhwezi, a local NGO working in Mbekweni to support households who are caring for AIDS-orphans. The second contact was the Mbekweni Community Health Project. The Community Health Workers (CHWs) at the site are aware of the most vulnerable in the community and are in regular contact with many women who are impacted by HIV/AIDS. The last is a Sister at Paarl General Hospital who is also a fieldworker on the PMTCT study. 8 The Study Sites Western Cape Province: Paarl, Mbekweni and the Drakenstein District In Western Cape, the study was conducted in the township of Mbekweni, about 10 minutes outside of the town of Paarl. The majority of people are Xhosa speaking Africans and are unemployed or seasonally employed on nearby grape farms. Paarl is a peri-urban/rural commercial farming area in Western Province with pockets of poverty and poor health. Despite its relative wealth, the socio-economic statistics of the municipality demonstrate immense poverty and extreme deprivation amidst immense wealth (Drakenstein Municipality 2004). The Infant Mortality Rate (IMR) in Paarl is around 40/1000 live births, however the rate in Mbekweni is most likely higher. HIV prevalence rate is fast escalating in Paarl. The District Antenatal Survey of the Provincial Department of Health shows that the HIV prevalence increased from 4.5 percent in 2000 to 15 percent in 2004 (Drakenstein Municipality, 2005). Eastern Cape Province: Umzimkhulu District In Eastern Cape, the study was conducted in Umzimkhulu, which is one of the poorest rural areas of South Africa, the former Transkei homeland in the Eastern Cape. Although part of the Eastern Cape, it is bounded on all sides by KwaZulu-Natal. Due its unique geography, the language most widely spoken in the region is a mixture of Xhosa and Zulu. Employment levels are very low, with only about 12 percent of residents in the municipality employed. There is a marked dependency on government services, pensions and migrant workers, with 38 percent of households reporting no income at all. Only half of the population is literate. Umzimkhulu is one of the previously neglected former Transkei districts with poor health facilities. There are 14 clinics, of which only two are in new, permanent structures. The IMR is 99/1000 live births. The antenatal HIV prevalence rate is 28 percent (Department of Health, 2005). Data Collection In both communities, introductions were made with potential respondents through local contacts. Potential respondents were either phoned or visited with our local informants and told about our study. If they agreed to participate, an appointment time was arranged. In almost all cases the researchers were personally introduced to the participant by local personnel. If they agreed and were comfortable, the researchers would introduce themselves, explain their positions and affiliations, the research they were conducting, and request their formal consent. Our final sample consisted of 25 respondents from 20 households. HIV positive women and/or primary caregivers (PCGs) of orphans were our main respondents. Where circumstances permitted, we also interviewed other household members to better understand response strategies to HIV/AIDS. We also interviewed four key informants in each of the communities. Open ended in-depth interviews guided by a semi structured field guide was the main mode of data collection. Data was analyzed using NVivo qualitative data analysis software. Pseudonyms were used to protect confidentiality. 9 More than half the respondents were interviewed in their homes. When logistics allowed, participants were given a choice between conducting the interviews in their homes or at alternative locations such as the Mbekweni Community Health Project, Ikhwezi, the clinic, or the hospital. In most cases where alternative locations were chosen, it tended to be a question of logistics for the respondent than a matter of stigma or fear for us visiting at home. We found no difference in attitudes or comfort levels between those conducted at homes and other locations. This was most likely due to the fact that the alternate locations were not unfamiliar to the respondents but were places they had visited in the past. For the purpose of this paper analyzing the support to children at risk of orphanhood, we utilize data from mothers and primary caregivers from 19 households that have children below 18 years of age. 2.2 The KwaZulu-Natal Study Sampling Framework Out of 67 clusters7 in the KIDS 3 survey, six were chosen for the qualitative study. This number was kept small in order to allow us to collect data of sufficient depth on four different research themes (orphans, the Child Support Grant, home-based care, and changing household structure and livelihoods) within the study time frame. Data collected in preparation for the KIDS 3 survey in 67 communities was used to select the six qualitative study sites. The clusters selected were to be stratified on the following variables: 1) rural/urban; 2) geographical spread; 3) degree and types of activities (mainly presence but in some cases absence) pertaining to the four research themes. Two of the six clusters were also selected because they had been part of a qualitative study by UKZN/IFPRI in 2001, and we wanted to follow up with some of the same households three years later. Telephone interviews were conducted with local leaders from as many of the survey clusters as possible, obtaining information on variables under (3) above, and a final six clusters selected. Within each of the six selected clusters, six households were selected, with each research theme covered in three of these clusters, and two themes covered per household. This struck a balance between not overloading the fieldworkers and respondents with too many modules per household given the limited timeframe, while also enabling us to capture some interaction between the themes. Because we want to be able to compare quantitative and qualitative data, the first household selection criterion was that they be KIDS 3 survey households; within these, the second criterion was that the household had to contain a pair of conditions to enable data collection on two themes: for example, have a CSG; foster at least one orphan; have a person with a chronic illness. The livelihoods theme enabled the selection of any households. Since these two criteria normally exhausted the supply of households, an additional criterion was not necessary; where it was however, the selection was made purposively. 7 “Clusters” refer to a census units that have been used in the KIDS panel surveys. These are sometimes equivalent to a small locality, and may be sometimes be loosely construed as “communities.” However, given the variation in clusters and the complexity of the term “community” we normally use refer to “clusters.” 10 Study Sites KwaZulu-Natal is South Africa’s most highly populated province, with approximately 20 per cent of the country’s population. Eighty per cent of the population is Zulu-speaking. Nearly half of the population lives in urban areas, but a far larger proportion of the African population is rural-based than is the case country-wide. It is among though not the poorest province, but is argued to have the highest incidence of deprivation in terms of access to services and other measures of well-being (Klausen 1997; Woolard and Leibbrandt 1999). It has the highest incidence of HIV infection among the nine provinces, with an estimated infection rate of 37.5%. In 2005, KwaZulu-Natal was among the two provinces with the highest percentage of children who were maternal orphans, at approximately 10 per cent, and predicted to rise to approximately 20 per cent by 2012 (Meintjes 2004:32-33). Using the criteria described in earlier, the following sites in KwaZulu-Natal were selected for the ethnographic study. In the south, Ndaleni and KwaDunuse; in the north, Mpakama and Kwabrush. The large urban township of Umlazi (L Section) located outside of Durban on the east coast was also selected, as was the urban area of Osizweni near Newcastle in the West of the province. Kwabrush and Mpakama are rural localities, although as in much of South Africa, the rural-urban distinctions are blurred, with KwaDunuse and Ndaleni considered rural but with some urban characteristics. There are variations in infrastructure and services; for example, Ndaleni has relatively good services while KwaDunuse is poorly serviced. Data Collection The first preliminary data was collected from telephone semi-structured interviews with local leaders in the survey clusters (except where they could not be contacted). The main purpose of these interviews was to obtain information for the purpose of stratifying the sample and then selecting six study clusters, but information from 21 clusters with substantial enough information on interventions was analyzed and reported on in section 5 below. As noted above, six clusters were selected for the case study research, but an additional key informant interview was conducted in an additional urban area, KwaMashu near Durban, when no formal or informal interventions supporting orphans could be identified in Umlazi section L, and an interesting intervention identified in KwaMashu. For the six clusters chosen for the main study, permission to conduct the research was obtained from various local leaders. Consent was then obtained from the six case study households selected, after explaining the background and purpose of the study and assuring confidentiality. A ‘modified’ extended case study method was chosen for the study.8 This used household case study techniques including formal and informal interviews conducted during repeated visits to the study household homes, and observation of activity within the household, community, and other points such as welfare offices. Key informant interviews were also conducted with individuals such as teachers, nurses, local authorities, people who run local livelihoods projects, pension committees, and others. Three fieldworkers conducted the research over a six-month 8 The term ‘modified’ was used because the period in the field and the period with each case study household was shorter than is most often the case using an ‘extended case method’ (Burawoy 1998). Also, fairly structured interview guides were used, primarily because of the low level of experience of the fieldworkers. 11 period in 2004, with each person living for two months in each of two communities, staying in the field for a month at a time and alternating communities and periods back in Durban to process field notes. Data was analyzed using NVivo qualitative data analysis software. Pseudonyms were used to protect confidentiality. This paper also draws on data from the KIDS 3 panel survey conducted in KwaZulu-Natal in 2004. This included a community survey of 67 clusters and a household survey of 1,428 households. 12 3. Preparedness of Children at Risk of Orphanhood in the Context of HIV/AIDS: Findings from Western Cape and Eastern Cape Province Preparing for parent’s illness and death is an ongoing and iterative process for both parents and their children. The realities of HIV/AIDS-induced migration, the assumption of new roles and responsibilities due to a parent’s illness, and the disclosure of HIV status of parents present challenges to children. Children must cope with prejudice and stigmatization from others, and with their own anticipatory grief due to their parent’s illness; they may also experience increased anxiety for their own security and welfare. In addition to confronting stigma and discrimination, HIV-positive mothers are often concerned about issues of disclosure to children, and about planning for their future security in the face of severe financial constraints. Understanding more about the processes and experiences involved in preparing for the illness and death of a primary caregiver may help to inform polices and programs so that these experiences are relatively adaptive experiences for both parents and their children. In this section, we begin by examining the roles that children are playing in HIV/AIDS-affected households. We then examine the ways in which mothers prepare children for their impending deaths. Here we discuss the experiences of mothers related to disclosure and how they talk to children about their future. Finally we investigate how mothers plan for their children’s future welfare and security. 3.1 Children’s Mobility and Inter-Household Links Many researchers have investigated household structure and mobility in South Africa. However, there is a dearth of studies on the living arrangements of children in households affected by HIV/AIDS. This section describes the dynamics of mobility of children affected by HIV/AIDS. For the purpose of our research, we asked the mothers/primary caregivers to list the members of their households, recognizing that the terms “household” and “family” are subject to much conceptual confusion, especially in Africa (see Spiegel 1996 for discussion). Our qualitative data on household histories (household with children<18 years of age) show that only a minority of families ever attained long-term residential independence. Among the household histories, nuclear families were both few and ephemeral, a result of the unavoidably dominant pattern of economic migration. In the context of the HIV/AIDS epidemic, the premature death of a partner also rendered the nuclear family unviable. Relationships between households were complex, indicating their mutual interdependence. Many households in our sample included a variety of co-resident relatives, among them orphans fostered formally or informally. Now we turn to the living arrangements of children in our sample households and alternatives sought to ease the burden of childcare in the context of HIV/AIDS. Most children in our households are residing with their mothers, owing to the purposive selection of HIV positive women in order to understand the vulnerabilities of children at risk of 13 orphanhood. Thus, the core members of our sample households are a woman and her children. Fathers of children are noticeably absent from our sample households. Of the nineteen mothers/PCGs considered for this analysis, eight of them reported that the father of the children (and/or their last partner) had died of AIDS. Seven more reported that the father of the children (or their last partner) were living elsewhere. In our sample, only one 16-year-old girl child of a mother in Umzimkhulu lives with her father in another village, as the mother cannot afford to send her to school. Only one household in our sample had a male head of the household (father of the children) living in the household. While absence of partners/father of children in the households is not uncommon either in Western Cape (Spiegel et al. 1996) or Eastern Cape (Siqwana-Ndulo 1998), we find that the frequency of such absence is higher in our sample than in other studies, revealing the deepening of existing vulnerabilities of children due to HIV/AIDS in South Africa. Previous research in South Africa, described in Section One, shows that households headed by women are more likely to migrate or dissolve. Children in female-headed households are significantly more mobile than children in male-headed households. Change in residence of mothers and their children often resulted from marital discord and breakdown, a recurring theme among our sample households. Out of 17 mothers in the sample (excluding one-sibling headed and one grandmother-headed household), 12 mothers were married. The widowed mothers in our sample often reported that marital breakdown began even before the death of their husband. Abandonment was high--most often due to discord related to the HIV status of the partner. Often, too, the relationship with in-laws was affected negatively. Of the 12 married mothers, seven reported having very negative relationships with their in-laws, before or after the death/abandonment of their spouses. Mothers experiencing marital breakdown (and/or death of their husbands) in our sample were thus under pressure to leave their marital homes. Only two previously married mothers continued to reside in their in-laws house/homestead after the deaths of their husbands. One of them had a bad relationship with her in-laws and wished to move out. The other mother had very good relationship with her in-laws and will continue to reside with them. When mothers moved out of their marital homes, children accompanied them. Mothers and their children most often moved to the homes of their mothers/siblings as a response to the burdens of economic hardship and child rearing responsibilities. The following excerpt from a household interview exemplifies some of these issues: In 2002 she began treatment for HIV. Her husband chose not to get treatment. After this break in their decisions things were never the same. Currently he is sick and not going for treatment. He also has not disclosed to his family. In fact, she disclosed for all of them (herself, him, and her child) to his family in the Eastern Cape. She took her husband to court in order to get maintenance from him and now she does collect this money. She is separated from him but not legally and not divorced. She attributes her financial survival to the fact that she can live with her mother and not pay rent. Her sister sometimes watches her youngest child so that she can attend support groups and other meetings. (Neliswa, mother of two children) While moving into maternal homes is a common response strategy, this was not always not the case. The one sibling-headed household in our sample is an illustration of such a case. In 14 addition, half of the mothers in our sample, after parting with their spouses/partners or their mortality, lived by themselves with their children. Interestingly, children of mothers that are not married (there are five of them in our sample), seem to have experienced less mobility. Such mothers in our sample tended to live in their maternal homes and maintain status quo after they separated from their partners. In some cases of paternal death and abandonment, to ease the economic and care giving burdens, children were sent away (children of six mothers in our sample), mostly to matrilateral relatives. These cases of informal fostering of children almost always resulted separation of siblings. The ages of the children that migrated from our sample households ranged from 15 months to 17 years. Children migrated mostly to grandmothers who were often the recipient of old age pension (3 of them). Two mothers sent their children to their maternal aunts. As one key informant noted “they (maternal aunts) will have the most invested in the children- they are the most like their mothers as well”. She has two children (daughters) who are 7 and 2 years old. The fathers of both children are different and neither is very supportive financially. The father of the 7-year old left when the child was only 8 months old. The father of the 2-year old is not supportive although he is supported by a Disability Grant because he is sick with TB (Tuberculosis). Her 7-year old child lives in the Eastern Cape with her aunt. She sent her child away because she couldn’t afford to keep her. It was better to send away her eldest child because she has more needs and it is less expensive to keep her younger child. Both children receive CSG. The aunt of the older child receives the grant for care in the Eastern Cape. (Zoliswa, mother of two children) Sending children to live with relatives is also a mechanism to ease the care giving burden of the HIV-positive mother when she is ill. The response of this mother of four children between the ages of 15 months and 17 years reemphasizes the role children’s migration plays in coping with the challenges posed by HIV/AIDS: The baby moved from the household (at 12 months) to stay with the respondent’s sister when she became very ill and was having difficulties looking after her. Also, because her own children were still attending school, they were unable to assist her and so her sister took the child in. She does plan on fetching her daughter and bringing her back when she is feeling better. Her sister does bring her baby quite often to see her because she asks her to. (Mafu, mother of four children) Fostering is also common among our sample households. Four households in our sample fostered children (excluding one sibling-headed household). Three households received maternal orphans due to AIDS, further reinforcing inter-household links and their mutual interdependence under severe economic strain (Box 1). It is beyond the scope of this study to investigate the adaptations a young and dependent child has to make when they migrate. From our study, we observe that the circumstances of households in our sample require children to accustom themselves to relatively fluid, non- exclusive filial bonds. Relationships with matrilateral kin are of particular importance for children in our sample because these relatives often act as substitute parents. As Van der Waal 15 (1996) notes in a study of the mobility of children in Berlyn “from a young age, children thus learn not to find their physical , social and psychological security exclusively with their parents” (page 46). Box 1 A spirited grandmother’s story Thandiswa, 65 is a pension recipient residing in Umzimkhulu. She lives with her daughter who is self- employed and has 2 children, ages 11 and 8, both of whom are attending school. She has also taken in two of her grandchildren whose mother passed away in 2001. The eldest is a girl, 15 years old, and the youngest is an 11-year old boy, both of whom have also remained in school. There are also 2 more grandchildren whose mother also passed away in 2001, and she has taken in these children as well. The family livelihood subsists of her Old Age Pension, Child Support Grants for the two youngest children, and the earnings she and her daughter receive from selling chips at the local school as well as from their home. She is also supported in small ways by her granddaughter who is now a teacher. From this she has also been able to become a member of a savings club and is one day hoping to have enough money to build a house for all of these children to live in. She would really like to help her granddaughter who is currently matriculating to study further but has no means. Thandiswa is a prime candidate for the Foster Care Grant and is in need of the extra income it would garner. She was told about a grant that is given to people who take care of orphans by somebody that she goes to church with. Indeed, she tried to apply for this FCG. She went to the social workers with the documentation such as the death certificate of the mothers but she was told the children were not eligible as their fathers were still alive. The social workers told her that both parents had to be dead in order for her to get the grant Thandiswa’s story is one of immense resilience in very difficult times. Hers is a story the government of South Africa needs to hear in order to address the complex realities of childcare without the support of men. The rigidity of the system is what will fail Thandiswa’s granddaughter in her efforts to further her education. 3.2 Present Roles of Children Living in HIV/AIDS-Affected Households The various roles and responsibilities children bear in and outside of the home in resource-poor settings is well recognized. In South Africa, the 1999 Survey of Activities of Young people (SAYP) states that 45 percent of all South African children are involved in some form of ‘work’ if a broad definition of work is used (Bray 2003; Statistics South Africa 2001). Fetching water and wood was the most common ‘economic activity’ (62 percent) followed by help with farming (15 percent) and ‘unpaid domestic work’, meaning household work by children living in households without parents or grandparents (7 percent). About 12 percent of South African children are engaged in domestic chores seven or more hours per week. A trend in increasing engagement with age emerges from this survey. Using Time Use Survey (TUS) , the Cape Area Panel Study (CAPS), and ethnographic methods, Bray (2003) investigates concerns around the levels of household chores by children and implications for their well-being. The study shows heterogeneity between boys and girls across different age categories and between urban and rural areas. 16 In South Africa, examining participation of children in work requires viewing their work as a function of their roles as members of households and as an opportunity to learn new skills required in adulthood. However, concern over work in childhood is increasing the light of HIV/AIDS. There is some evidence from South Africa that children living in communities where AIDS is exacerbating chronic poverty are undertaking a larger share of household work and missing school (Booysen and Arntz 2002; Giese et al. 2003). Our qualitative study adds to understanding of the experiences and roles of children in HIV/AIDS-affected households. For this analysis, we include data from only those mothers/PCGs that have children eight to 18 years of age residing in the household. The rationale for this is that by the age of 8, children are often lending support to their families whether in the form of emotional support, care giving, or through financial assistance. Of 11 households in Paarl, nine were characterized as such, while four of six in Umzimkhulu included children 8 years of age or older. Children as Caregivers: Care, Support, and Assistance in Treatment Our interviews reemphasize the continued role of children in performing routine chores such as those mentioned above. Almost all of the mothers/PCGs with older children residing in the household mentioned their children performing domestic duties. This was especially the case for female children as they were depended upon for the care of younger children. An assessment of care giving activities performed by children is important to deepen our understanding of roles children perform in response to rising rates of AIDS-related illness and death in the household. Bray (2003) finds that far fewer African and colored children engage in care giving activities than other domestic tasks. She finds that in Masiphumele, where HIV prevelance is high, “female relatives living close by or in other parts of Cape Town, and/or neighbors are their first port of call” in times of need for childcare. While female relatives were an important port of call, our study also shows that as HIV-positive primary caregivers attend support groups, apply for grants, and visit clinics and hospitals, older children, especially girls, are increasingly taking on the primary care giving responsibilities of their younger siblings. The experience of the mothers in Box 2 and Box 3 below demonstrates how the traditional role of care giving by older children is assuming a heightened importance in the era of HIV/AIDS. Box 2 Children providing care and support to their HIV positive mothers/PCGs Lolo is a mother of eight children—six sons and two daughters. Their ages range from 22 years to 3 years. Her 22 year old son lives with his uncle. Her 15 year old daughter is in standard 4 and lives with her paternal granny in Eastern Cape. The three-year old boys are twins and are HIV positive. Her mother passed away in 2003. Her mother used to assist her as well. She would get remittances from her from the Eastern Cape. She does have a sister and brother but they do not assist her at all. Her sole source of income is the Child Support Grant that the twins get. Her eldest daughter (19 years old) is also very helpful with other children. If she were to get sick or if anything happened to her, she thinks that her daughter would be able to look after her other children. There have already been days where she has done this. 17 A domestic role unique to HIV/AIDS-affected households is that of providing treatment support. Our study shows that both boys and girls are increasingly playing important roles in helping their mothers follow treatment regimens (Box 3). Such support is critical in adherence to antiretroviral (ARV) regimens (Murray et al. 2004; Simoni et al. 2002). Non-adherence allows the virus to resume rapid replication resulting in multidrug resistant strains that can endanger health of the individual on ARVs as well as the health of the public should transmission to others occur. Box 3 My daughter acts as my buddy Rose’s husband passed away in 2002 from AIDS. She has two sons (ages 11 and 4) and a daughter (age 13). She works as a domestic worker. Her youngest child receives CSG. She is also a recipient of Disability Grant. When her husband died, her mother came to live with her. She is not emotionally supportive or in any other way. If she needed someone to watch the children it would be her eldest child and not her mother. She and her daughter are very close. Her daughter knows about her status, found out while listening to a conversation between her mother and her brother-in-law. The daughter cried a lot when she first overheard the conversation. But now, she and her daughter are very close. Her daughter often reminds her to take her tablets. She comes home from work and is often very tired and will fall asleep. Her daughter wakes her up after work and reminds her—she acts as her buddy. Perhaps one of the most important areas where boys and girls are providing their HIV-positive caregivers with support is that of emotional comfort and love. No matter the context, an HIV - positive status is a difficult reality. In South Africa stigma remains a barrier to support. Treatment is not yet a viable option for many resource-poor people, despite the rollout of ARVs. Emotional support and comfort remain the most important factors for stability, optimism, and overall positive living of those living with HIV/AIDS. Mothers/PCGs in our sample found children to be a critical source of emotional support, making it easier for them to face the daily tribulations of being HIV positive—as Lolo states, “The fact that my children are handling my [HIV] status well makes it easier for me to deal with stigma”. Such support has may have important implications for longevity of mothers/PCGs. There is evidence that perceived support from family members positively influences health outcomes of HIV positive men in the United States (Serovich et al. 2000; Kalichman et al. 2003). The stress of giving care to a terminally ill person is well recognized. Adolescent children could be especially under stress. Adolescence is a time of rapid change—both physically and emotionally. Children now not only have to deal with their individual changes, but also take on the onus of supporting adults. Neliswa’s experience is a window into the strain adolescents in these households may be going through. Neliswa is a mother of two in Paarl. She disclosed her status to her 18 year old son. Although her son seems to have accepted her status, she and her son have strained relationship. They have not spoken much to each other lately. She states, “I do not attribute this to my status but to normal teenage problems.” Like other studies, children in our sample are not the primary or secondary sources of monetary support. Only one mother in our sample cited her son to be a potential source of support. This 18 could be due to high adult unemployment rates in South Africa that prevents young people from finding work. In contrast to other literature investigating the impact of HIV/AIDS on children’s well-being in South Africa (Booysen and Arntz 2002), we find that very few children in our sample dropped out of school. None of the mothers/PCGs reported their children missing school due to economic activity. The most important reported constraint on schooling is school fees. Our findings indicate that mothers/ PCGs prioritize children’s education over domestic chores and economic activity. 3.2 Experiences of Disclosure to Children The increasing care giving responsibilities being placed on children’s shoulders presents HIV- positive parents with the immediate dilemma of whether to disclose their HIV status to their children. Previous work from developed countries in this area suggests that women are most likely to disclose to their close friends, mothers and romantic partners (Armistead, Morse, Forehand, Morse, and Clark 1999; Simoni, Demas, Mason, Drossman and Davis 2000), finding disclosure to their children to be the most difficult (Semple et al. 1993; Tompkins, Henker,Whalen, Axelrod and Comer 1999). In these studies less than one-third of the mothers disclosed their status to their children. Disclosure of HIV status of parents to children is critical for preparing children for impending parental illness and death. Evidence shows that open discussions about illness and anticipated loss have been shown in some studies to lead to more fulfilling family relationships (e.g., Brown and Powell-Cope 1993) and children’s misconceptions about their mother’s health will likely persist if such discussions do not occur (Siegel and Gorey 1994). Disclosure is crucial for children to be effective agents of response and yet has been largely ignored in the literature from Africa. This section illustrates some of the experiences of mothers in disclosing their status to their children, and the importance it may play in child preparation and well-being. Understanding more about factors that are associated with mothers’ disclosure to their children may foster efforts to guide disclosure decision-making, as well as increase the likelihood that when disclosure occurs it is a relatively adaptive experience for mothers and their children. Out of a total of nineteen mothers/PCGs interviewed, fourteen have children that are eight years of age or older living with the mother or elsewhere.9 Some six mothers/ PCGs, (less than half) disclosed their status to their children. Our findings show that disclosure ranged from being a process to a single important event. But in almost all the cases it involved a lengthy mental and emotional preparation. For some women, disclosing to children came relatively easy. Lolo, the mother of six children, two of who are HIV positive states “It took me a year to disclose. All my children know about my status now. I disclosed to them alone and the children accepted it”. Lolo’s experience was not atypical in our sample. Even those mothers that stated that they could disclose to their children without much anguish took approximately a year to do so. For others, although they had 9 This analysis includes mothers that have children living elsewhere. 19 a desire to be open and honest with their children, the process represented a more difficult challenge. Two mothers also expressed concern and confusion over the appropriate age for children to know about the positive status of their parents: She has not yet disclosed to her 8-yr old. She doesn’t think he will be old enough to understand, he is not mature enough. (Pule, mother of two children) She says that she will definitely talk to her son one day about her status and about HIV/AIDS in general. She has been thinking herself about when to disclose to him and feels that he is very clever and she will probably due it when he is 6 or 7. (Tumi, Mother of one child) Our study confirms results from previous studies, albeit from the developed countries, that HIV infected parents are more likely to disclose to children who are older (Weiner et al. 1996; Rotheram Borus et al. 1997; Shaffer et al. 2001; DeMatteo et al. 2002). This could be because mothers/PCGs may feel that older children may be better able to understand the implications of their illness. These studies also show that parents are more likely to disclose to female children. This cannot be confirmed from our data, given the small sample when disaggregated by age and gender of the child. Further exacerbating the difficulties of disclosure is the logistical fact that for a large proportion of South African families, children are not residing with their parents. The question of proximity therefore became a factor in disclosing to children, while once again no support existed to help parents overcome this predicament: Her children do not yet know of her status. She thinks it is possible though that her mother has told them. She hasn’t really thought of telling them, in large part because they are not around each other (Bulelwa, Mother of children age 10 and 15 years who live with their grandmother in Eastern Cape). The following case (Box 4) highlights the complexity of disclosure to children, especially to youth who are already knowledgeable about HIV/AIDS. It is also an example of the startegies HIV-positive people utilize in order to ‘feel out’ how their disclosure would theoretically play out in their closest and most important relationships. Previous research from elsewhere shows that disclosure of HIV status may not often bring anticipated relief. These studies from Western countries have reported a significant increase in child behavior problems and a decrease in mother-child relationship quality from pre- to post- disclosure (Shaffer et al. 2001; Rotheram Borus et al. 1997). In stark contrast, in our sample, where mothers disclosed their status they did not report any deterioration of their relationship with their children. Where they were able to discuss their illness with them, the children were largely supportive. Further, in some cases the children took on key roles in the support network of the HIV-positive respondent. A prime example of this was in the sibling-headed household where the young man’s mother had spoken to him prior to her death about her illness and a time when she would pass away (Box 5). 20 Box 4 Experiences of disclosure: ‘Feeling out’ disclosure to children Bongi, 40 years old, is a mother of a 17- year old daughter. She now lives with her boyfriend. Her daughter lives with Bongi’s mother at a homestead close by. At this homestead are also her sister-in- law and their three children. This homestead has a few huts and also an 8-room house where they all live. The father of her living daughter lives in Durban and is married to someone else. He helps in small ways but Bongi says it is not enough, especially to help raise a teenager. He gives her R100/month. In 2003, Bongi gave birth to a baby daughter. At this point she still had good relations with the father of the baby, man different from the father of her teenaged daughter. The baby fell sick when her father was away on job training. Bongi phoned the father of the baby to tell him that she and baby were being tested for HIV. After that phone call she never heard from him again. When she got her results that they were positive she never told him. She never heard from the father of the child again, even at the funeral for the baby. Bongi disclosed her HIV status to her mother and her sister-in-law. She has not told her current boyfriend or her teenage daughter. Her daughter has bad asthma and so she is afraid that if she told her she might get sick. She also does not want her to worry about her or be devastated by her status. However, there was a time recently where she told her daughter that she was HIV positive and then said that she was joking. Her daughter told her that even if she was it would be okay because there are many other people who are positive and it is just like any other disease. This gave her hope that one day she would tell her, but she is not ready to do that yet. Box 5 A well prepared child is a well adjusted child His mother organized this [inheritance of the house they now live in ] before she passed away. She put the house in his name so that there would be no problems when she passed away. There had been uncles who were interested in the property but there was no grabbing because she organized inheritance before her death. He did not see this as a difficult process for his mother. He himself has now opened a Funeral Policy because he is receiving enough financial support that he has some left over each month to save. Before his mother passed away she spoke with him about a time when she would not be around. She told him that times would be difficult and that he would have to deal with it on his own. In fact, this made things easier for him because he knew what was to come. His mother also empowered him with HIV education. Because of this emotional/educational support, he feels he has the strength to support others/his peers. He is now very open about the subject of HIV/AIDS. (Head of a sibling headed household, 21 years old. His mother passed away when he was 18 yrs) 21 In another case, the relationship between a mother, her child, and her HIV status is clearly highlighted. Not only was the more informed child better able to negotiate the illness of his mother, but in being prepared, he was able to become a key support to her: She has also disclosed to her 17-year old son. He is a very supportive son and will even help her with things like medications. He understands what is happening because he learns about it at school. At times she thinks he knows even more about it than she does. She worries though about what it will be like one day when he goes away to school because he has been so supportive of her. (Mafu, Mother of four children) Our research shows that there are many positive implications of disclosure. While all positive mothers may have yearned to tell their children, the process was not easy. An important facilitating factor aiding disclosure to children was external counseling and support. Where mothers accessed counseling and support, they were better prepared to disclose to their children. Gugu’s experience highlights the role counseling played in disclosure and acceptance by children: She has disclosed to her two eldest children and it was not difficult for her to tell them. Both her children have gone for counseling and both children can now counsel others. Some mothers (e.g., Rose from Paarl), expressed a desire to access counseling to disclose to their children. The actions of the activist NGO TAC (Treatment Action Campaign) in bringing HIV and AIDS much more explicitly into the public discourse seems to have emboldened some women to discuss previously taboo subjects with their children. When he is old enough she will talk to him about sex. Part of her inspiration is TAC, which always says to “call a spade a spade.” She wants to be open with him and tell him about condoms, etc. Things need to be given a name and spoken about (Tumi, mother of a three year old child). 3.3 “Talking” About Future with Children Talking to children about their future after the impending illness/death is a critical part of preparing them to deal with the trauma of parental death and for their future roles and responsibilities. Our hypothesis was that mothers who disclosed their status to children are more likely to discuss their children’s future and security with them. While no conclusions can be drawn from our extremely small and purposively drawn sample, it is interesting to note that no noticeable pattern exists between disclosures to children and talking to them about their future in our study. Bulelwa, a mother of a 10 and 15 year old hasn’t disclosed to her children yet. She thinks that her 15-year old son would be capable of looking after the other child and himself. However, she says, “I have not yet spoken to him about this. But one day I will get the guts to talk to him about it.” Disturbingly, six out of eight mothers/PCGs that disclosed to children eight years or older, did not explicitly discuss the future with their children either. As for the orphans in four of our sample households, only one respondent of a sibling-headed household mentioned their mother explicitly talking to them about their future before her death. We could not obtain such information for the rest of the orphans fostered in three other households in our sample. 22 We find that factors that hinder disclosure to children—age of the child, child not living in the same household as the mother, children having difficulties pertaining to adolescence are also the most important factors that hinder discussing the future with children. One important assumption in the discourse on the impact of HIV/AIDS on children is that siblings in South Africa tend in any case to live in different households , and that this should not be seen as a unique negative impact of HIV/AIDS on children. The two mothers (one was the mother of a now sibling-headed household) who had discussed the future with their children, expressed a desire for siblings to live together after the death of their mothers. Powerful examples below of how mothers discuss the future with their children, and the benefits of such preparation (Box 5) clearly indicate the need to facilitate such discussions between parents and children: She feels that her children should stay together and should not be separated. Her house will be a family house. She does discuss issues such as preparation of a will and other legal issues. She feels that “When God calls you, you should be ready” (Gugu, Mother of four children. The eldest is 21 years old and other children are below 16 years of age.) 3.4 The Process of Planning for Children’s Future Welfare and Security For parents who face worsening illness and impending death, one of the most painful realizations is the eventual inability to care for their children and seeing them grow to maturity. Previous work in Western countries shows that children are at a greater risk of long-term negative outcomes if their parents fail to make custody arrangements before they die (Simoni et al. 2000). Children bereaved by sudden, unexpected parental loss demonstrate more negative outcomes than children who are prepared for such a loss, and the legal complications are greater (West et al. 1991; Rotheram-Borus et al. 1997). Future planning for the children can create a peace of mind for parents by assuring them that the children will be cared for according to the parents' wishes concerning their future. There are, however, many barriers to this sort of planning. Parents may often be reluctant to initiate such planning because of a sense of guilt, denial of the seriousness of the illness, or fear that others may learn about the diagnosis (American Academy of Pediatrics Committee on Pediatric AIDS 1999). They may lack the information, skills, financial resources and tools to plan for their children’s future. While there are previous studies on stresses that mothers face, comparatively little is known about how HIV-positive mothers/PCGs think about and plan the future of their children, or about the resources women draw upon for the security of their children once they pass away, especially in Africa (Marcenko and Samost 1999; Rotheram- Borus et al. 2004). Our study attempts to fill this gap in understanding in a South African context. In this section, we report findings from our interviews with seventeen mothers/PCGs who are HIV-positive, one sibling head of a sibling-headed household and one grandmother of a grandmother-headed household. From the interviews, we found that all of the mothers were thinking about the future and in the majority of cases were carrying out some level of planning for their children. Concurring with 23 the findings of Marcenko and Samost’s (1999) study of HIV positive mothers in Philadelphia, we found that plans of HIV-positive mothers in our study ranged from organizing future care giving arrangements to preparing wills for inheritance. Respondents also identified very specific barriers to their planning initiatives, and to the hopes they had for their children once they had passed. Discussions during this section of our interviews were imbued with sadness, tears, and in some cases fear. Care giving Arrangements The most important type of planning identified by the mothers interviewed was that of planning for the future care giving of children. As Bongi stated, “Mothers will usually identify family members in advance who could care for their children and sometimes even speak with them about this. It is usually family members because they are the ones who will care for the children as their mother’s did- raise them as their own.” Our research found that identification of family members as potential caregivers was a common trend amongst HIV-positive mothers (Box 6). Identified family members ranged from the respondents’ mothers and siblings, to their older children who mothers hoped would be able to cope without external adult support once they had passed on. Box 6 Going to Live with Your ‘Other Mother’ Mafu, 36, is a mother of four and lives with her husband in Umzimkhulu. In 2003 she tested positive for HIV after the birth of her daughter. Recently she and her baby were diagnosed with TB and are currently being treated. The first person she disclosed to was her sister in law who has been incredibly supportive of her emotionally as well as through care giving and frequent visits. When she became very ill both of her sisters offered to take her baby to alleviate the burden on her, and the child is currently in the care of her sister who lives closer to her so that they can visit. When she was especially sick she spoke to her sister about caring for her children when she passes away and her sister agreed. The three youngest children will go to live with her and she has even spoken to them about it. She has said that “there will come a time when you will go and live with your other mother.” She says that “even when you ask the children who their other mother is, they know.” Mafu trusted both of her sisters to watch over her infant daughter and was in a position to ask either to take in her children when she passes away. This story attests to the strong ties between families, and especially women. Children are aware of their extended families as their most intimate support networks, to the point where children not only have one mother, but can claim to have multiple mother figures ready to watch over them. Grandmothers were the most often identified persons for the role of future caregiver. The dynamic and essential roles played by grandmothers can be seen in a variety of ways. First, in the three cases where mothers were not sure what would happen to their children, the relationships with their mothers were extremely strained. This was very distressing to the mothers. They broke down into tears during their interviews. She has made no plans. She says that even if she fell very sick she would still look after her daughter. She doesn’t think her mother would look after her daughter at all even if 24 she passed away. She would probably take the child to the baby’s father if she were extremely sick because she would not know what else to do (Molatela, mother of a 2 year old who lives on her mother’s homestead in Umzimkulu. Her mother lives in the city.) In many cases grandmothers are already playing significant roles in child-rearing (as discussed in the section on children’s mobility) as well as in supporting their HIV-positive daughters emotionally and materially. Six of eighteen mothers indicated that their children would be living with their grandmothers, thus demonstrating the strong ties between children and their maternal grandmothers. Only one mother who is currently living with in-laws stated that the child would continue to live with her paternal grandmother once she had passed on. As one key informant working with orphans described, “in most cases, before the parents pass away they send the children to the home they are going to be living in” so that the transition is relatively painless for the child. Because traditional Xhosa culture has assigned a significant care giving role to grandmothers, this transition is generally seen to be less traumatic for children. An excellent example of the active participation of elder women in sustaining the strength of family networks comes from our Paarl site. In this case we had the opportunity to interview not only the young HIV-positive woman, but also her mother in order to better understand how networks respond to the impact of HIV/AIDS at the household level. In this case the young mother had been abandoned by her husband, leaving her and her young daughter to fend for themselves. At this time, the respondent’s mother had taken them in to live with her. When we asked the mother about future planning for her daughter, she described the following: She finds it difficult for herself to do this [planning] because she does not have the financial means to plan certain things. Right now, no one in her immediate family could look after her daughter. She has told them that should something happen to her today she thinks the best place for her daughter would be the orphanage. Her mother wouldn’t really have the strength or capacity to do it and she doesn’t know what would happen to her. She thinks that she would be best treated at an orphanage (Pule, mother of two children). In terms of planning, the young woman described a last-resort reality whereby there would be no one left to care for the child, forcing her to live in an orphanage without any ties to her extended family. However, what is so interesting about this case is what we learned when we spoke with her mother, the 59-year old grandmother of the child. This grandmother had been a very supportive mother emotionally and was also acting as a caregiver to her granddaughter while her daughter served as a coordinator for the local TAC office. She told us that although it was very difficult for her at first, she says, “when you discover your child is positive, you should treat them as you treated them before you found out.” In terms of the future well-being of the child, this grandmother had her own idea of what she was capable of and the role she would play: If anything happens to my daughter I would help and care for my granddaughter. If I was too old to do this then my other daughter would be the ideal person to care of her and would raise her (Pumza, mother of an HIV-positive mother). This case attests to the fact that not only do mothers plan for their children, but in some cases the extended family may be doing so as well. Further, while some women may feel that children would be a burden on elderly grandmothers, grandmothers themselves may feel up to the 25 challenge. There is more research required on how elderly women themselves interpret their roles in the era of HIV/AIDS, and how this can be bolstered as part of an effective response to the crisis of future care giving of children. When grandmothers were not present, available or capable, sisters to the mothers were often the next choice. Five mothers indicated that their children would be living with aunts. Three mothers in our sample are already hosting maternal orphans who are their nephews or nieces, demonstrating how common this practice is. For example, 35-year old Nomaindia thinks that her sister would be the most appropriate person to take care of her two year old daughter. But, she says, “I haven’t yet spoken to her about it but will [do so] when the time comes”. The following examples display some of the thought processes of the women interviewed: She thinks that her sister will be the ideal person to look after her child. Her sister is 21 and already looks after her child. She loves him and knows that she would take good care of him. She has asked her already and her sister has agreed to look after him. However, she does this in a more general manner. For example, she will say, “one day, if anything were to happen to me…” She is planning though to disclose to her in June. Her sister does ask why she talks about death a lot but she hasn’t yet told her. They have a very good relationship though so she is confident she will be supportive. (Tumi, mother of one child) While the extended family continues to be the first line of defense for most mothers and their children in the era of HIV/AIDS, three mothers identified their eldest children as potential caregivers for their youngest in the future: In another case from Paarl, Gugu, a mother of four stated that “her children should stay together and should not be separated”. In this case the respondent felt that her children would be able to survive once she passed on because they would be protected and able to reside in what would be their family house. While there are some cases where children were deemed capable of looking after themselves and each other, this was only evident in Paarl where fewer households in general have remained intact. In most cases, mothers expressed a preference to have another adult take over the care giving responsibilities. Apart from the three mothers who had poor relationships with their own mothers, all of the mothers in the sample trusted the identified future caregivers to care for their children. Significantly, none expressed fear of abuse, neglect or property grabbing. This has profound implications for the positive living of the mother as well as the future well-being of HIV/AIDS affected children. Ziyekele’s story is an illustration of this. Ziyekele, a widow and a mother of a two-year-old daughter, lives with her mother-in-law and sister-in-law on their homestead. Ziyekele is currently trying to access her husband’s retrenchment funds. Is she gets this money, her daughter will inherit it. She says, “if anything were to happen to me, my mother-in-law will take care of my daughter. I am not worried that my mother-in-law will try to access any of [the money]. I trust her.” 26 Our research also demonstrates the incredibly strong ties that exist between women in South African extended family networks. Most of our respondents were confident that the future caregivers would do their best to provide for their children. Planning for Inheritance and Will Preparation Most of the mothers in our sample either rent or live with their parents/in-laws. Often they are not in a position to pass on assets to their children. However, one important thing to note is concept of ‘family house’ in South African Xhosa culture. For many Xhosa families, the deed of the house is in a family name, and not in the name of one specific person. In cases where some women were living with family, their death would not necessarily entail the necessity of formulating a will, because through tradition her children would be guaranteed the right to remain. However, in other cases, women formulated wills in order to guarantee the rights of their children to reside on the property and to ensure that the house remained in the family. This was the case for two of the households interviewed (Box 7). Box 7 The Family house For two women in Paarl, the question of inheritance and where their children would reside once they passed away was settled by the Xhosa concept of a Family house. Vuyo, 40, is a mother of one child, but is also fostering five of her deceased sisters children. They are currently living in what was her parents’ home. Her parents died 11 years ago and left it to her and her sisters as a family house. For many years she has lived in this house with her two sisters until they both passed away. In terms of the house and inheritance she says that the house is a family house and will always remain so. “The house belongs to everyone.” She does plan on going through a process to make sure that the house remains in the family and she can leave it the children of the family. It was their grandparents house and they have a right to remain living there. Gugu, 38, is a mother of four and lives with three of them in a house that she owns. When she passes away she would like the house to become a family house. She feels that her children should stay together and should not be separated [this is something that her son does not like to discuss]. Her house will be a family house. She has already prepared a will and has hired a lawyer to handle things. While she is not necessarily afraid that others in the family will try and ‘grab’ the property from her child, she is working now while she is healthy to ensure that they will always have a home. Financial Planning Overwhelmingly, the most commonly cited constraint on mothers’ financial planning for the future was poverty. More than half of the mothers in our sample had not yet started to save for their children’s financial future. The common occurrence of abandonment by husbands/fathers of children had exacerbated the vulnerability of these households and made saving impossible. Indeed, in only one family in our sample did the father of the children reside in the same household. However, in this case the father also described the constraints of poverty in his efforts to plan for his children’s future. 27 Despite such poverty and the lack of stable incomes, several examples illustrate that mothers are engaging proactively, within their constrained economic circumstances, in providing financial security for their children. Mothers typically started saving whatever they could and often said “although it is not a lot, it is something.” (Box 8) Box 8 Working to guarantee future financial security for children Tumi, 30 is a widowed mother of a four-year old son who currently lives with her mother and grandmother in the Eastern Cape. She resides in Mbekweni, a township near Paarl where she supports herself through the Disability Grant. She has matriculated and worked in the past for the municipality, but currently is volunteering for the local branch of the Treatment Action Campaign (TAC). She educates others, including her father about HIV/AIDS in order to alleviate stigma and pass on the information is learning. While she thinks that many women do not plan, she has been planning in a variety of ways for her own child. Right now she is putting R50 per month into an account for him. When her husband died she did not receive anything. There was struggle for assets and the property that she shared with her husband but after a period of time her mother told her not to bother anymore and to give up fighting. However, her in-laws have told her that they have set aside money for her child and that when he turns 21 they will pay for his Varsity. She has concerns though about whether or not this will materialize and would like to gain support on how to take legal action to make sure this is there for him. Further, in two examples where the father of the children was alive, two mothers in our sample were seeking child maintenance support through legal channels. In both cases their efforts had been successful. Tuleka, mother of three says, “[the father of her children] is actually paying her some money right now per month (R700 total) but has been trying to stop payments and so it is a constant struggle”. Agency of mothers to provide for children, even in the most difficult circumstances, can be found in Ziyekele and Thandiswa’s stories. Both of them are trying to work with the employers of their deceased family members to gain access to their savings. For many of our respondents, government grants were the only stable sources of income, and were utilized directly to support the children of the households. In some cases, mothers made sure that their children would receive the CSG even after they passed away, making it easier for the subsequent caregivers to take care of the child. The twins’ mother was sick for a long time and during that time she applied for the CSG even though she was weak at times. This has made it easy for grandmother to take care of her grandchildren (Thandiswa, grandmother taking care of maternal orphans). Funerals are often expensive and can drain family resources. Membership in a burial society or having a funeral policy spares the family from resource drain. Thandiswa’s experience with her daughters’ planning illustrates the point. She states “My daughters planned for their children. Both my daughters belonged to Burial Societies, thus it was easy for them when they passed away”. 28 While most mothers expressed the desire to either belong to a burial society or to buy a funeral policy, most could not afford to do so. Only six of nineteen mothers/primary caregivers in our sample either had a funeral policy or belonged to a burial society. Only one mother in our sample had directly included her child on her funeral policy. 4. Living as Orphans: Findings from KwaZulu-Natal Province This section turns to the study of households after children have become orphaned, looking at where children go, who fosters them and why, and experiences and conditions of orphans within new household structures. As emphasized throughout this paper, household mobility patterns over many decades in South Africa have left children living without biological parents, so that the ‘adoption’ of ‘social orphans’ is a long-established practice in the country. Madhavan (2004) elaborates on patterns of foster care before and after the onset of HIV/AIDS, and debates whether existing living arrangements are just a continuum (of previous patterns) or a new development responding to the pandemic. Fosterage has always been present in African families; however, fostering patterns have been exacerbated by the devastation of family structures from AIDS. Madhavan distinguishes between two basic patterns of fosterage: voluntary and crisis-led fostering. Voluntary fosterage pertains to arrangements between biological and foster caregivers to raise the child. Crisis-led fostering occurs in response to the death of a biological parent or the occurrence of a major shock or hardship, as is the case of the huge number of children orphaned by AIDS. While social parenting is not a new concept in South Africa, children orphaned by AIDS have often experienced (as section 3 shows) stress and trauma from prolonged caretaking for sick adults, living through premature death of biological parents, clustering of deaths in households, and prolonged and exacerbated periods of poverty. They may be dealing