Vulnerabilities and impacts among the Warwick Junction grannies

Careful study of the AIDS/HIV epidemic's effects on the traders and the conditions that have shaped their vulnerabilities points to five overarching struggles in their lives (see Table 1):

1. 1. Financial burdens: The grandmothers were supporting growing numbers of children and grandchildren and losing income-earning family members. They were struggling with increasing medical, food and funeral expenses, and the costs of taking time away from paid work to care for the sick.

2. 2. Emotional stresses: These women had to cope with the anguish of losing their children and the immense pressures of caring for their families. Many were finding it extremely difficult to talk about sex with younger family members, which compounded their emotional stresses.

3. 3. High unemployment levels: In the province of KwaZulu-Natal, unemployment has reached 29.9 per cent (Statistics South Africa 2006). The rising level affected these women in various ways: family members were not working, more and more customers could not pay reasonable prices, and they themselves were forced to work ‘informally’. Street trading is a matter of survival, but working long hours hindered them from providing their families with what they regarded as proper care.

4. 4. Displacement: Most of the grandmothers migrated between the urban area where they worked and a rural area where they supported and looked after the sick. Taxing emotionally, the stresses of displacement were compounded by their ‘informal’ work, which attracted little respect. In many ways, they were treated as and felt expendable. The discourse of ‘informality’, which pervaded their existence, assumes that their insecurity is somehow temporary. It also de-legitimises their urban work and residence, which manifested itself in a number of ways, including being denied health-care services because they did not have ‘proper addresses’ in Durban.

5. 5. Untreated personal illnesses: All the women had chronic, debilitating and largely untreated illnesses, such as hypertension, asthma, diabetes and arthritis, which made it even more difficult for them to care for others. They worried about who would take care of them if they could no longer work, and what would happen to their families when they died.

TABLE 1. Impacts and vulnerabilities and collective responses in the grandmothers' words

While research on grandmothers' vulnerabilities has been limited, some studies of AIDS-affected families and communities elsewhere have shown a similar suite of financial and emotional impacts among older women (e.g. Schatz Reference Schatz2005; Oburu and Palmerus Reference Oburu and Palmerus2005); these have been most extensively documented by HelpAge International (Mall Reference Mall2005; HelpAge International 2006). Less attention has been paid, however, to the compounding and converging factors that shape grandmothers' vulnerabilities (e.g. unemployment, displacement and chronic illnesses), which clearly came to light in Warwick Junction.

A worsening situation

The cumulative effects of AIDS, unemployment, displacement and chronic illness were exemplified in the grandmothers' changing household structures. Alongside the repeated in-depth interviews, the traders' ‘household maps’ were drawn at six monthly intervals (in December 2004, June 2005, December 2005, and June 2006). In December 2004, the traders were also asked to describe their ‘households’ at Christmas 2002. The term ‘household’ was referred to by the isiZulu word umdeni, meaning a self-defined social network of shared responsibilities and entitlements, not one necessarily confined to one location or limited to kin (Hosegood and Solarsh Reference Hosegood and Solarsh2004). The older women defined their ‘households’ as the people for whom they were responsible. The maps showed a clear trend, of an ever-greater number of people being supported on ever-shrinking incomes – a worsening situation.

One story, of a seamstress who supported 19 people on the equivalent of three US dollars per day illustrated the circumstances of many. The changes over time in her household are depicted in Figures 1 through 3. Over three-and-a-half-years, the woman's household changed from 15 to 19 members; the number of income earners declined from three to one; her income also diminished because of her chronic illness. She was not working in June 2005 or June 2006 because of diabetes and hypertension, which meant that her family had virtually no income. She lost four children, three of whom she suspected had AIDS. She took on responsibility for two orphans and provided personal care for six people, all of whom were seriously ill. It is worth reiterating that her story was by no means exceptional among the traders at Warwick Junction. Indeed, all the household maps revealed a consistent pattern of a growing reliance on grandmothers, a trend that has been shown elsewhere (e.g. Ogunmefun and Schatz Reference Ogunmefun and Schatz2006).

Figure 1. Seamstress household map, December 2002.

Figure 2. Seamstress household map, December 2004.

Figure 3. Seamstress household map, December 2005.

The identified general pattern and trend imply neither a uniformity in people's experiences nor a predictable sequential downward spiral. The household maps of the woman who sold seeds showed new income-earners moving into her network after one six-monthly interval, which meant that fewer people were being supported by more earners, suggesting at least some improvement in economic security. Over the two years, however, the seed seller, the seamstress and every one of the older women in the study reported larger ‘households’ but a net loss of income-earners. The household maps also revealed some limitations in common research constructs. All the grandmothers defined their ‘households’ as large, fluid and multi-locational networks of people for whom they felt responsible. The people in the seamstress's household, for instance, frequently moved between three residences (in the township, in the city and in the rural area). She was affected if any member fell ill or died. To assess how AIDS affected her, one needs to understand the dynamics of her social web. The maps therefore showed not only a worsening situation for grandmothers, but also how conventional conceptions of ‘households’ as comprising only members that reside at least four nights of a week in the same house (as used in most longitudinal research), are inadequate for tracking the complex effects of AIDS.Footnote 5

Responses and responsibilities beyond the family

Not only were the grandmothers taking on ever more responsibilities in their families, they were also active in their communities, for example by devising collective responses to the epidemic's stresses. Most were involved in group savings schemes that assist AIDS-affected families with food and funeral expenses, or prayer groups that address the emotional and spiritual traumas associated with witnessing so much illness and death (see Table 1). The findings also corroborated research findings from other parts of southern Africa, which has shown grandmothers taking on responsibilities that extend well beyond their families, and playing central roles in AIDS-related home-based care initiatives, income generation projects, support groups and child-care initiatives (Chazan Reference Chazan2006; Mall Reference Mall2005; Ogunmenfun and Schatz Reference Ogunmefun and Schatz2006). Many older women derive support and purpose from such community involvement, but community initiatives like ‘home-based care’ can also mean that they become engaged in additional, labourious and usually unpaid care-giving (Marais Reference Marais2005).

Overall, the epidemic had had considerable financial and emotional impacts on the grandmothers in Warwick Junction, which were compounded by myriad other stresses in their lives. Cumulatively, their burdens appear to be worsening, although not in any sequential or single way. As strains built up, they tried to adapt. They assumed ever-greater responsibilities for the growing needs of their families and communities.

Assumption 1: AIDS is fundamentally changing grandmothers' roles in South Africa and beyond. Counter view: The situation is as old as it is new

As stories of grandmothers affected by AIDS reach national and international audiences, one commonly held assumption is that the epidemic is fundamentally changing their family situations and roles – from women who were once cared for by their adult children, to women who are becoming mothers anew to the next generation. As the narrative goes, if it were not for AIDS, grandmothers in high-prevalence African countries would be ‘enjoying a restful retirement’ (Reuters 2006). Grandmothers are viewed as ‘Africa's newest mothers’ and their roles are understood to be transforming: ‘the changing roles of grandparents, or surrogate parenthood, as a result of the epidemics, for example, may result in changed relationships between the oldest and youngest generations’ (UN 2004). This is not entirely accurate. The subtlety of the situation was aptly described by one of the Warwick Junction grandmothers:

The historical context and social conditions must be taken into account if we are to understand how AIDS is transforming the lives of grandmothers. South Africa has a long history of labour migration; for many decades, young adults (especially men) have worked away from their homes in factories and mines (Marais Reference Marais2005). Grandmothers customarily stayed behind to care for their grandchildren, and were dependent on remittances from their adult children. They were in a situation of reciprocal exchanges rather than ‘restful retirement’: grandmothers raised children, young children assisted with house and farm work, and absent adult children provided income (Madhavan Reference Madhavan2004). As Marais (Reference Marais2005) has pointed out, AIDS has undermined the traditional reciprocal care arrangements. Migrant parents who fall ill and die can no longer provide remittances to the grandparents caring for their children, and young adults themselves may move back home to be nursed. Older people, especially poor women, are thus compelled to assume both productive and child-raising duties (Marais Reference Marais2005). This was shown by the changing household maps in Warwick Junction (see Upton 2005 for similar evidence from Botswana).

Grandmothers in this region are facing a situation that is both old and new. Moreover, the ‘oldness’ of the situation is actually shaping the ‘new’ burdens: powerful and deep-seated gender norms, roles and identities uphold historical care arrangements and underlie the mounting pressures on grandmothers. Participants in the Warwick Junction research, for instance, described their situations as a facet of their ‘duties’ and ‘roles’ as women and grandmothers; they view it as ‘natural’ for them, amidst the escalating epidemic, to assume additional family responsibilities. Likewise, in the Agincourt District in rural South Africa, grandmothers did not perceive their increasing responsibilities as a ‘burden’, but rather viewed themselves as ‘bound’ to the wellbeing of their children and grandchildren (Schatz Reference Schatz2005).

Assumption 2: AIDS is an exceptional force in southern Africa; it is the primary stress in grandmothers' lives. Counter view: That AIDS is exceptional is a myth

The second closely related assumption is that in high-prevalence African communities, AIDS is the primary stress in most families and therefore in most grandmothers' lives. As one well-known advocate remarked, ‘all of us who live in privileged western societies experience death from time to time, but in much of Africa that's all people know. Their lives consist of attending funerals … every family in every country … has suffered a loss in the carnage of AIDS’ (Lewis Reference Lewis2005: 57). Such ‘AIDS exceptionalism’ – that is, focusing on HIV/AIDS over and above issues of poverty, migration or food insecurity – may stem from a desire to simplify how practical, feasible (and often medical or technological) responses to human suffering are imagined. It could also be that framing research and interventions in narrow ways is sometimes done to appease institutional expectations; to comply with the narrow AIDS-specific mandates of funders, academic departments or non-governmental organisations (Marais Reference Marais2005). This too is overly simplistic.

AIDS is a profound stress but not an exceptional force. As discussed earlier, the pressures on the Warwick Junction grandmothers resulted from the cumulative effects of, among other things, AIDS, unemployment, poverty, displacement, stigma and chronic illnesses. They did not see AIDS as the main stress in their lives, but instead pointed to immediate poverty-related threats, like not being able to put food on the table, send their grandchildren to school, or afford transport home. It is true that the household maps showed that these pressures cannot be disentangled: AIDS might not be their most immediate threat, but it compounds pre-existing vulnerabilities (see Schatz and Ogunmefun Reference Schatz and Ogunmefun2005 for similar findings elsewhere in South Africa). More widely, ‘AIDS exceptionalism’ could hide the vulnerabilities of many grandmothers – vulnerabilities that probably exist whether or not they are directly dealing with AIDS-related illnesses in their families. If such complexities were embraced, this might lead to more meaningful and comprehensive responses. Certainly, responding to the multiple stresses in grandmothers' lives would make them better able to deal with the additional burdens posed by the epidemic.

Assumption 3: As a result of AIDS, South Africa is now speckled with ‘skipped generation’ households (of only grandparents and grandchildren, and no middle-generation members). Counter view: Skipped-generation households are the exception not the norm

The third common assumption is that in high-prevalence countries, a wholesale transformation of household composition is occurring, with many families struggling to survive with only elderly and young members (Desmond Reference Desmond2006). According to HelpAge International (2005: 2), ‘family structures are changing with “skipped generation households” comprising older people and children … emerging as a new category of household’. The evidence for this change is, however, scarce. The household maps from Warwick Junction did not represent ‘skipped generation’ living arrangements but rather multi-generational networks. Even where young children had lost their biological parents, middle-generation adults (aunts and uncles) were present in the structure of care. Similar evidence has been forthcoming elsewhere in South Africa (e.g. Madhaven and Schatz Reference Schatz2005).Footnote ⁷

In a recent presentation entitled ‘Emotional statistics’, Desmond (Reference Desmond2006) indicated that despite what is often reported in international campaigns to evoke emotional responses, ‘skipped generation’ and ‘child-headed’ households are not yet common in South Africa. His analysis of national data showed that the most common household configurations (defined narrowly) included two (non-elderly) adults or multi-generational members. The data showed that between 1995 and 2005 in South Africa, ‘skipped generation’ households increased from 1.69 to 2.29 per cent, but that over 83 per cent of households in 2005 were not ‘skipped’, ‘child headed’, ‘young-adult headed’ or ‘single-adult headed’. Desmond did not discount that even small percentages mean that too many children live in these conditions, but he clearly refuted the prevailing discourse that overstates the prevalence of these arrangements (see also Ford and Hosegood Reference Ford and Hosegood2005; Hosegood and Timaeus Reference Hosegood and Timæus2005a, Reference Hosegood, Timæus and van der Walle2005b). The stresses on grandmothers are growing with AIDS, but this does not mean that the South African countryside is speckled with ‘skipped-generation’ households; this assumption is not evidence based. The burdens on grandmothers may instead reflect something like we have seen in Warwick Junction: in the face of the growing demands for care, households are joining in grandmother-headed multi-generational social networks. Measuring the growth of such care networks is difficult with existing data.

Assumption 4: The grandmothers that care for AIDS orphans are old and frail. Counter view: The stereotype is false; not all grandmothers are old and frail

The fourth common assumption held by many researchers and advocates is that AIDS-affected African grandmothers are old and frail. ‘These old and unimaginably frail women often look after five, 10 or 15 kids, enduring every conceivable hardship for the sake of their grandchildren, alongside additional numbers of other abandoned waifs who wander the landscape of the continent’ (Lewis Reference Lewis2005: 50). One reason for such descriptions is probably the widespread conflations of ‘grandparenthood’ with ‘old age’, and of ‘grandmother’ with ‘elderly woman’. While it is not uncommon to find truly ‘elderly’ care-givers in South African communities, the images of old-age conjured by speaking about ‘grandmothers’ may not be entirely accurate. We have seen, for instance, that all the Warwick Junction grandmothers were fairly young, aged in their forties and fifties. But how representative were they? At what age does the ‘average’ South African woman become a grandmother?

One could argue that the Warwick Junction grandmothers are ‘demographic outliers’, but the data that suggest that this is not the case. Drawing on the 1998 Demographic and Health Survey of South Africa, Moultrie (Reference Moultrie2006) estimated the ages at which ‘African’ South African women first become grandmothers. The modal age of entering grandmotherhood was 39 years and the median age was 40 years. Fewer than one per cent of women first become grandmothers after 60 years-of-age. The fact is that there are many women aged from the late thirties to the fifties who are already grandmothers in South Africa, like the traders at Warwick Junction.

The association of grandmotherhood with old age appears to influence how research is carried out and which issues are given prominence. The scarce literature on the impacts of HIV/AIDS on grandmothers focuses almost exclusively on ‘the elderly’, often limiting research participants to women aged 60 or more years (e.g. Schatz Reference Schatz2005; Oburu and Palmerus Reference Oburu and Palmerus2005). The issues faced by younger grandmothers barely appear in this limited body of research. Moving beyond the stereotypes means understanding that not all South African grannies are old and frail – most are not.Footnote ⁸ It also means not constraining who is included as ‘grandmothers’. Younger grandmothers are struggling even though they are not ‘elderly’; they may be just as worthy of attention in that several generations depend on them. The exclusive focus on ‘elderly’ grandmothers in research may be obscuring the vulnerability of younger women in similar positions.

Assumption 5: In South Africa, grandmother-headed households pivot around old-age pensions. Counter view: The pivots are grandmothers not pensions.Footnote ⁹

Unlike in most African countries, older people in South Africa can access government old-age pensions – women at 60-years-of-age and men at 65. In this context, it is often assumed that large granny-headed households pivot around these social grants (May Reference May2003; Schatz Reference Schatz2005). While the implications of pension-centred research are critical, it follows from the previous argument that most women become grandmothers long before they are eligible for pensions. There is substantial evidence that pensions, valued at up to 740 Rand per month (approximately US $100), can ease the stresses on older people and their extended families. Amidst poverty and high unemployment, pensions are often the most stable and reliable form of household income (May Reference May2003). Indeed, in one part of Mpumalanga province, where one-in-10 poor older people cared for ill, young adults without basic services, and where older people were the main breadwinners in three-in-four households, the most important source of income was the government old-age pension (Makiwane, Schneider and Gopane Reference Makiwane, Schneider and Gopane2004).

In AIDS-affected families, older women often use their pensions as substitutes for unemployed children's incomes; they invest them in insurance and credit programmes to cushion AIDS-related shocks, and use them to cover costs associated with illness and funerals (Schatz and Ogumefun Reference Schatz and Ogunmefun2005). Channelling subsidies through older people (especially women) also has significant redistributive effects (Duflo Reference Duflo2003). Pensions once aimed at reducing poverty among older people have been transformed into a lifeline for younger household members (Marais Reference Marais2005; Legido-Quigley Reference Legido-Quigley2003). This evidence suggests the potential of state support for reducing vulnerabilities, and the effectiveness of distributing such support through older primary care-givers.Footnote ¹⁰ Thus, while it is unfounded to assume that granny-headed households necessarily pivot around pensions in South Africa, it is still the case that broader access to social security could mitigate some of the epidemic's impacts.

Assumption 6: Grandmother carers tend to be isolated and live in remote rural areas. Counter view: Grandmothers are extremely diverse. Many have extensive social networks, and some are urbanites who use the latest communications technologies

In part because mass rural-urban migration established a pattern of grandmothers being ‘left behind’ in the Bantustans to care for children, and in part because there is a dearth of research in urban areas (especially the ‘informal’ townships or shanty towns) (Posel and Casale Reference Posel and Casale2003), it is often assumed that to find the African grandmothers most affected by AIDS, one must travel to remote rural areas. Descriptions of ‘huts where women are dying in the presence of their children’ conjure images of an epidemic taking its toll in the unreachable depths of ‘Africa’ (Altman Reference Altman2006). The prevailing picture is that of the rural granny surrounded by squadrons of young children, while their dying mother suffers, isolated in a mud hut. The Warwick Junction evidence presented a very different image – while most of the grannies perceived themselves as ‘rural women’ and maintained close ties with their rural homes, they were also mobile labour migrants who lived in a large city and kept in touch with their families by cell phones. Many came together to save money, in prayer and to offer mutual support. They may have been stressed by their frequent displacements from home and by the silence around the epidemic, but they were neither geographically nor socially isolated. Again the question arises: are these grandmothers anomalies? The answer must be ‘not really’.

In South Africa, an increasing number of grandmothers are urban labour migrants. Since the end of apartheid, cities have been desegregated and laws around informal trading relaxed (Nesvag Reference Nesvag, Freund and Padayachee2002). More and more women, including older women, move to cities to work, and an increasing number work ‘informally’ and engage in circular urban-rural migration (Posel and Casale Reference Posel and Casale2003; Hosegood and Solarsh Reference Hosegood and Solarsh2004). The number of ‘informal’ traders in the Central Business District of Durban rose from 200 in 1984, to 800 in 1990 and to 19,800 in 1997; they are predominantly mobile women (Nesvag Reference Nesvag, Freund and Padayachee2002). In 2002, 58 per cent of the South African population lived in urban areas (World Bank 2005).

Furthermore, the phenomenon of grandmothers being at the centre of community responses is not unique to Warwick Junction: such involvement has been seen throughout the region (Mall Reference Mall2005; Chazan Reference Chazan2006). A clear example of social and political organisation among South African grandmothers is the Cape Town group, Grandmothers Against Poverty and AIDS (Broderick Reference Broderick2006). So not all grandmothers match the image of isolated, old women. Making this clear is not to deny that there are many AIDS-affected grandmothers in highly-impoverished and constrained conditions in rural South Africa, nor that it can be extremely lonely caring for those affected by HIV/AIDS, nor that city life can be alienating and can increase exposure to crime and environmental pollution (especially in trading hubs like Warwick Junction), nor that involvement in community projects offers social support but also means additional, strenuous and usually unpaid work. Rather, there is a need to recognise the diversity of situations in which grandmothers live and work, to document what is happening in rural and urban areas, and to analyse critically the effects of ‘community mobilisations’. Such critiques require moving beyond hegemonic images of grandmothers stranded in rural huts – these images only obscure the struggles of large and growing groups of women (Mohanty, Russo and Torres Reference Mohanty, Russo and Torres1991).

Assumption 7: The worst of the epidemic's impacts in South Africa are evident now, as young adults are dying and leaving behind orphans. Counter view: This is short-sighted. We must consider what happens when today's grandmothers die

At this point, it may seem that the consequences of AIDS are not as dramatic as frequently described. The prevalence of skipped-generation and child-headed households tends to be exaggerated; AIDS is not drastically transforming family dynamics; it is not the primary stress in the lives of many grandmothers; and not all AIDS-affected grandmothers are elderly, frail women living in rural isolation. The doomsday images of rogue, orphaned children, susceptible to lives of criminality, have not transpired (Marais Reference Marais2005). In South Africa, AIDS is not a ‘human apocalypse’ (Altman Reference Altman2006), or anyway not yet. The irony, of course, is that this ‘good news’ could be precisely because grandmothers are doing what they have always done, and doing it well. The Warwick Junction household maps suggest that grandmothers have mitigated many of the epidemic's negative consequences by caring for growing networks of dependants on shrinking incomes. They are also supporting other families in their communities. Cushioning the epidemic's impacts has come at a high personal cost, however, for many women are in survivalist positions, physically unwell, emotionally distraught, and struggling to make ends meet.

The question arises, ‘what do these growing struggles among grandmothers mean for the future?’ Or, as the chicken seller asked, ‘What will happen when we die? Who will look after our children and grandchildren?’ These questions have not been adequately considered by researchers, policy makers or the advocates of AIDS care. The final question addressed in this paper is whether the current assessment of the main negative impacts of AIDS/HIV is short sighted? Is the assumption that the worst of the epidemic's impacts are taking place now, as young adults die, blind to the intergenerational implications of AIDS?

The epidemic in South Africa is new: in 1990 national HIV prevalence was less than one per cent and it has been rising ever since. The intervals between when people become infected, fall ill and die indicate that the epidemic's impacts will unfold for decades. Some suggest that it will be at least another five years before prevalence peaks, and another 10 before mortality rates and orphan numbers stop rising (Whiteside Reference Whiteside2004). Figure 4 shows female mortality rates by age in South Africa from 1997 to 2004. The rising peak shows the remarkable increase in mortality among women aged 20–49 years; that is, more and more women have been dying in their twenties, thirties and forties. The trend is likely to continue.

Source: Statistics South Africa 2006. Mortality and Causes of Death in South Africa, 2003 and 2004. Statistics South Africa, Pretoria, South Africa.

Figure 4. Female mortality in South Africa, 1997–2004.

A prospective view of the impacts of AIDS suggests that grandmothers are shouldering today's burdens, while the ‘next generation’ of grandmothers is disappearing. This raises a series of important but rarely asked questions. With the next cohort of grandmothers decreasing in size, what will happen when today's grandmothers die? Will family structures and gender roles transform or adapt? Do the worst impacts lie ahead? There is much uncertainty, but it is clear that we should be looking to the future and where there are obvious vulnerabilities today, responding vigorously: providing financial assistance, counselling and health-care to keep ageing women healthy as long as possible. This requires keen foresight – a commitment to replace the current short-sightedness and to intervene in preventive rather than reactive ways.