Reframing Vulnerability: Mozambican Refugees’ Access to State-Funded Pensions in Rural South Africa

Abstract

Researchers at the South African Medical Research Council/University of the Witwatersrand Rural Public Health and Health Transitions Research Unit (Agincourt) fieldsite in rural South Africa consider Mozambican residents more vulnerable than others in the local population. These self-settled refugees, many of whom are still not South African citizens, primarily came to South Africa in the 1980s during the Mozambican Civil War. This perceived economic vulnerability is rooted in their difficulties in accessing social grants, until recently legally available only to those with South African citizenship documentation. This paper focuses on semi-structured interviews with 30 ‘older’ women of Mozambican-descent living in the Agincourt area. These interviews highlight three important aspects of vulnerability; the respondents: (1) perceive a risk of deportation despite their having lived in the country for 20 years, (2) are unable to easily access social grants, namely the state-funded old-age pension, and (3) struggle to make ends meet when faced with daily needs and crisis situations. All three of these vulnerabilities were mediated to some extent by these women’s resourcefulness. They generated ties to South Africa through obtaining identification-documents, used these documents to access pensions, and used the pensions to help them sustain their multigenerational households.

Introduction

In the AIDS era, the word vulnerability is used freely and frequently uncritically, often not explicitly defined or contextualized to explain what makes a particular group vulnerable per se, or vulnerable to what in particular. The term has proliferated to such an extent that “orphans and vulnerable children” (OVC) is now a widely used and commonly abbreviated term in much policy and research literature on AIDS. Many articles highlight the plight of the vulnerable—primarily women, orphans, infants or children—while very few articles problematize the use of the term, fewer still question how the term’s use might hide or even hamper the vulnerable individuals’ agency and empowerment (exceptions include DeGuzman 2001; Delore and Hubert 2000; Gilbert and Walker 2002; Kalipeni 2000; Mayer 2005; Skinner et al. 2006). While the generic term vulnerable has some utility in developing, implementing and disseminating anti-poverty policies, evaluating these policies and the use of this terminology from the perspective of those defined as vulnerable is crucial.

This paper turns a critical lens on the notion of vulnerability, and examines the actions and agency of a group of Mozambican-born older women living in rural South Africa, in the South African Medical Research Council/University of the Witwatersrand Rural Public Health and Health Transitions Research Unit (Agincourt) research site. Rather than defining vulnerability for these women, the purpose is to outline the ways in which these older women define their own vulnerabilities and the ways in which they contest the notions of vulnerability placed upon them by outsiders. Even among residents in this rural South African community, a community within which many non-Mozambican residents could also be identified as vulnerable for many reasons—poverty, HIV/AIDS, poor access to health care, increasing numbers of maternal and double orphans—Agincourt researchers and the local community often have labeled this particular group as vulnerable due to differential access to economic opportunities and social services.

Until recently, Mozambican permanent residents living in South Africa were not legally eligible to access government sponsored social grants (e.g. old-age pension, disability, and child grants). A South African Constitutional Court ruling in March 2004 made it possible to access social grants with a Permanent Resident identification-document, as well as a South African citizen identification-document (Khosa and Others v Minister of Social Development, March 2004).¹ Social grants are a fairly substantial and very reliable source of income for many poor families in rural South Africa, with old-age pensions being among the most desirable due to fewer required supporting documents and comparatively high cash allotment (Booysen and Van der Berg 2005). Obtaining access to social grants may itself be considered a survival strategy and a way to reduce certain types of vulnerability.

The women interviewed for this paper belong to a specific category of Mozambican permanent residents, those who are age-eligible for the South African state-funded non-contributory pension program.² The women, who are the focus of this paper, self-identified as ‘Mozambican’ in the Agincourt census, which began in 1992. Our in-depth interviews from the Gogo (Grandmother) Project³ with 30 Mozambican women aged 60–75 show that in September-December 2004 (6 months after the ruling) our respondents did not know about the law permitting permanent residents to receive social grants, yet the majority had already devised ways to access the pension program. Our respondents’ narratives point to three important issues related to the notion of vulnerability; our respondents: (1) continued to worry about the possibility of being deported despite their having lived in the country for 20 years, (2) were concerned about the difficulties and legality of accessing social grants, namely the old-age pension, and (3) struggled to make ends meet when faced with daily subsistence and crisis situations. The interviews show, however, that all three of these vulnerabilities were on some level mediated by these women’s resourcefulness and ability to manipulate the social grant system, even before knowing that the law had changed in their favor. Although often labeled vulnerable, their agency worked to their benefit by creating stronger connections to South Africa through identification-documents, using these documents to access pensions, and using the pensions to help them maintain and sustain their multigenerational households. In short, these older female self-settled refugees showed incredible resilience in their attempts to improve their own lives and those of their families.

Background & Significance

Vulnerability and AIDS

The term vulnerability is common in a number of literatures: the disaster literature (e.g. Fordham 1999; Levine 2004), literature on conflict and refugees (e.g. Haour-Knipe 1997; Kalipeni and Oppong 1998), literature on poverty and health (e.g. Galea et al. 2005; Kalipeni 2000; Leatherman 2005;), and more recently, literature relating to the AIDS crisis (e.g. DeGuzman 2001; Gilbert and Walker 2002; Mayer 2005; Skinner et al. 2006). Social vulnerability as a concept usually focuses on a certain social group’s insecurity and risks to an impending danger—whether natural disaster, disease, or violence and conflict (Delore and Hurbert 2000)—describing differential access to resources among groups or individuals (Kalipeni 2000), and generally juxtaposing vulnerability, empowerment and agency (DeGuzman 2001). Further, the AIDS literature describes those children who are orphaned by or otherwise affected by AIDS as vulnerable (Skinner et al. 2006).

Many authors use the term without fully defining the meaning of vulnerability in the context analyzed; even those authors that define the term, often do not consider how the identifier may strip agency and power from the individuals labeled as vulnerable. Feminist scholars have critiqued Western researchers for homogenizing third world women as powerless, vulnerable and in need of saving (Mohanty et al. 1991). A similar critique could be made of the use of the term vulnerable to group individuals affected by a common factor. Although differential access to resources is the basis of vulnerability, the tendency of researchers and policy makers is to focus on group rather than individual access (Kalipeni 2000; Oppong 1998). Consequently, the term vulnerability often pigeonholes individuals into a single social class despite possible differences among members of the group (Delore and Hubert 2000; Levine 2004). Such homogenization might cause researchers to overlook the subtle ways in which individuals are empowered and how they use their agency. In fact, DeGuzman (2001) and Levine (2004) both discuss ways in which vulnerability is the opposite of empowerment. Thus, to speak of the Mozambicans living in the Agincourt area as a homogenous vulnerable group ignores within group differences, and hides how individuals within this group avoid vulnerability by negotiating and managing their social environments.

Mozambicans in South Africa

During the 1980s, 250,000 to 350,000 Mozambicans crossed the border from Mozambique into South Africa fleeing the civil war ravaging their country.⁴ The Mozambicans who fled were mostly peasants, and left because their home districts were disrupted by severe bouts of violence. Because they arrived in South Africa while it was still under apartheid, they were never given refugee status by the South African government, although they are often referred to as “self-settled refugees” (Polzer 2007; Steinberg 2005). Instead, tribal leaders in the Bantustans gave many Mozambicans refuge, including the former apartheid Bantustan of Gazankulu⁵; this area is now primarily in South Africa’s Mpumalanga Province⁶ where the current study was conducted. Although Mozambicans were welcomed to this area for reasons relating to ethnic solidarity—the majority of both the local residents and the refugees shared a language and culture as part of the Shangaan ethnic group—neither the South Africans in the region nor their new Mozambican neighbors had much wealth or access to resources (Steinberg 2005).

Mozambicans who were given land by the local tribal head and settled in this area came in shifts from Mozambique. Once a family was settled, they would often send for members who were left behind in the war zone (de Jongh 1994; Dolan 1997). The Mozambican settlement areas were generally less fertile areas, and were enclaves within villages, such that locals can still clearly demarcate which lands are “Mozambican areas” and which are “South African areas” (Dolan, Tollman and Nkuna 1997). Dolan et al. (1997) also found that Mozambicans living in the Agincourt area were disadvantaged in terms of access to water, sanitation, fuel and housing. A recent study shows that the Mozambicans living in this area are still falling behind their neighbors on certain social and demographic indicators. For example, Mozambicans living in the Agincourt area have a higher rate of child mortality than their South African neighbors (Hargreaves et al. 2004). Perhaps the truest sign of non-integration is the fear expressed by many Mozambicans in the area that they could still be deported (Polzer 2007).

In the post apartheid era there have been a number of campaigns that have either attempted to repatriate Mozambicans back to Mozambique⁷, or have opened processes for Mozambicans to obtain South African residency, if not citizenship (Handmaker 2002; Handmaker and Schneider 2002; Polzer 2007; SAMP 2001). The amnesty conceived in the mid-1990s, allowing for Mozambican refugees to become permanent residents, was implemented in 1999–2000 (SAMP 2001). Despite these campaigns, the majority of Mozambicans living in South Africa have had difficulty obtaining citizenship through legal routes (SAMP 2001). Although most Mozambicans in the Agincourt area have sustained residency in South Africa for nearly two decades, many did not have documents to prove when they came to South Africa, or that they are in fact from Mozambique. The absence of such documents was an obstacle to obtaining South African identification-documents as permanent residents or citizens (SAMP 2001).

Post apartheid, without documentation, Mozambicans and other foreigners face police tactics that do not differ drastically from those used under apartheid (Klaaren and Ramji 2001). Klaaren and Ramji argue that arrest and detention of undocumented migrants, as well as corruption and bribery, continue to plague the migration policing policies in democratic South Africa. Kloppers (2004) contends that it is the illegal immigrants, who have crossed from Mozambique to South Africa in recent years for economic reasons, who are constantly searched for, arrested and deported. Given these policies, however, it is not surprising that ‘Mozambican inhabitants of the borderland’ (Kloppers 2004), who have established roots and families in South Africa, worry about their own vulnerabilities to deportation.

Since unemployment rates are extraordinarily high in rural areas of Mpumalanga Province (as well as much of South Africa), social grants are often the most stable and reliable income sources that families have (Barrientos et al. 2003; May 2003). Social grants, including the old-age pension, disability, and child grants are all means-tested government sponsored social assistance. In 2004, pensioners and disability grantees received ZAR740, the equivalent of approximately USD100, per month; the child grants are substantially smaller. The non-contributory old-age pension is the most accessible of the social grants as its only requirement beyond the means-test is having a 13 digit bar coded identity document. Other social grants have additional requirements or are only granted for special circumstances.⁸ In the case of child grants, the requirement to produce a child’s birth certificate is a substantial obstacle to application and receipt (Twine et al. 2007). Still, when accessed, these grants can provide important sources of household income (Booysen and Van der Berg 2005). Social grants play an important role in alleviating poverty—reducing its incidence, depth and severity (Booysen and Van der Berg 2005). Much work has shown the importance of the old-age grant in particular in increasing household expenditures on food and improving the well-being of all household members, not just the well-being of the pensioner (Barrientos et al. 2003; Booysen and Van der Berg 2005; Case and Deaton 1998; Duflo 2003; Maitra and Ray 2003, May 2003).

Mozambicans who had managed to obtain a South African permanent resident identification-document did not have access to these social grants prior to the Constitutional Court judgment in March 2004. In order to access South African social grants, one had to hold a South African citizen’s identification-document. Mozambicans living in South Africa without such a document did not have equal access to these social grants. The inability to access grants has been one of the main reasons that researchers in the Agincourt area have regarded Mozambicans inhabitants as economically more vulnerable than the general population.

Before access to social grants was available to permanent residents, it was worthwhile for Mozambicans, who in general have little intention of returning to Mozambique (Polzer 2004; Golooba-Mutebi 2005), to try to obtain South African citizen ID documents through legal and extra-legal means. Although most of the campaigns mentioned above gave Mozambicans access to permanent resident documents, there were several campaigns, including voter registration for the 1994 election, during which those of Mozambican descent might have been able to obtain South African citizen-identification-documents (Dolan 1997). In addition, because many Mozambicans have South African relatives, they sometimes use relatives’ citizenship, or a “South African name” to convince local Home Affairs officials to give them citizen rather than permanent resident documents (Dolan 1997; Polzer 2007). Local tribal leaders also sometimes vouch for individuals in order to convince government officials that a person of Mozambican-descent is actually South African (Polzer 2004). The qualitative data from the Gogo Project uncovers a story of resilience and resourcefulness in manipulating the system prior to the 2004 ruling in order to access state-funded old-age pensions.

Methods

The data for this paper come from the Gogo Project, a qualitative study conducted in the Agincourt fieldsite in the northeastern corner of South Africa. About a third of the Agincourt population is of Mozambican-descent (Tollman et al. 1995). The only land separating the study area from Mozambique is the Kruger National Park, one of the largest game parks in Africa. Low rainfall and high population density make this area inadequate for subsistence farming; it is more suitable for cattle or game rearing. The population has low levels of education and high rates of unemployment.

Agincourt runs the longitudinal Agincourt Health and Demographic Surveillance System (AHDSS), which has been collecting data annually since 1992. The AHDSS data include: a household census, updates of vital events, verbal autopsies, and occasional modules on labor migration, household assets, temporary migration, and education.⁹ In 2003, preceding the Gogo Project, the study site was home to about 70,000 people, over 11,600 households in 21 villages. Upon first entry into the census database, individuals self-classify as Mozambican or South African.

The 2003 AHDSS census provided a sampling frame for the qualitative study. The qualitative project was conducted in two phases. During the first phase (July–September 2004), we randomly selected 30 households with a ‘South African’ woman between the ages of 60–75.¹⁰ For the second phase of the project (October–December 2004), we randomly selected 30 households with ‘Mozambican’ women aged 60–75.¹¹ We interviewed Mozambicans to compare the coping strategies of older women without pension access with our South African sample.¹² Over the three in-depth semi-structured interviews with each respondent, we asked a battery of questions relating to beliefs about HIV/AIDS, sources of income and expenditures, pension usage, and caregiving of orphans, of fostered children, and of the sick, who are mainly adult children suffering from HIV/AIDS.

The interviews were conducted by three local women over the age of 40, who were trained by the author in qualitative interviewing.¹³ Each interviewer was responsible for one-third of the respondents with whom she conducted and digitally recorded all interviews; in addition, each interviewer translated and transcribed her own interviews. Since the sampled population is largely illiterate, respondents provided verbal consent to participate in the study. While the project was in the field, the author read each interview, reviewed queries with the interviewers, and wrote a unique interview guide for each respondent for the second and third interviews. These guides allowed the author to fill gaps, follow up on interesting issues and explore new questions, despite not conducting the interviews herself.

The impetus for the first phase of the study was to understand the ways in which older women are experiencing the HIV/AIDS epidemic.¹⁴ For the majority of older women, their experience is not one of being infected, but rather being affected. They often become the caregivers for HIV-positive adult children, and for fostered and orphaned grandchildren (Schatz 2007). It became clear during the interviews that pensions were playing a crucial role both in day-to-day subsistence and when crises occur, e.g. paying for healthcare and funeral costs related to AIDS morbidity and mortality (Ogunmefun and Schatz 2009). All but one of our original South African respondents were receiving and using pension money to help sustain and maintain the multi-generational households in which they lived (Schatz and Ogunmefun 2007).

Since we believed the Mozambican respondents would be without the benefit of pension money in their households, expanding the study to women of Mozambican-descent was meant to reveal their coping strategies in contrast to the South African female pensioners. To our surprise, we found that two-thirds of the Mozambican respondents were receiving pensions. The remainder of this paper focuses on the 30 Mozambican respondents’ narratives, particularly how and why they spent much effort to obtain a South African identification book, and thus access the old-age pension.¹⁵

The analysis for this paper was loosely based on grounded theory (Strauss and Corbin 1990). The author read each of the interviews for emergent themes, beginning with an open-coding system, then created a coding-tree from that coding list, and finally closed-coded the remaining interviews. Once all 30 interviews were coded, the author used Nvivo software to extract sections of the interviews on the core themes outlined in the paper, and assessed the themes for differences and similarities among respondents.

Results

The original impetus for stratifying the sample by nativity was driven by the idea, which we acquired through the literature and talking with local researchers, that Mozambicans were more vulnerable members of the Agincourt population in part due to their inability to access social grants (Dolan et al. 1997; Hargreaves et al. 2004, personal communications with Mark Collinson, Kathleen Kahn and Tara Polzer). Once we began interviewing the Mozambicans, however, this basis for economic vulnerability became problematic: the majority of our respondents had managed over the years, some through legal and some through extra-legal means, to obtain South African identification-documents; yet the interviews exposed other vulnerabilities. In general, our respondents were very wary of being interviewed, worried that we might share their information with the authorities, leading to their deportation. Although respondents with access to the old-age pension were able to assist their households, those without access—the one third of our respondents not receiving a pension—struggled to survive. Our respondents expressed insecurities in their claim to permanence in South Africa and in providing for their families. Below we explore these vulnerabilities as well as the agency portrayed by these women in attempting to and often succeeding in accessing documents and pensions, and providing for their families.

Wariness

The AHDSS data identified how each of our respondents self-reported her nativity or nationality in 1992, when the AHDSS began collecting information, or in the year in which she moved into the study area. Despite annual AHDSS visits and a very high response rate to the census, our respondents were wary of sitting down for interviews with our team. Several women told our interviewers that they were not the person whose name was listed or that they were not in fact Mozambican. The following note prefaced one respondent’s interview:

On the day, I [first met the respondent and] set up an appointment [for the first interview], the gogo [grandmother] told me that she is a South African citizen, but on the day of the interview, she changed and said that she is a Mozambican citizen. I asked her, “Why did you tell me yesterday that you are South African?” She laughed and said, “I was afraid. I thought maybe you want to take me back to Mozambique. Ha, ha, ha.” The gogo laughed, but we [went on to have] a nice interview (interviewer’s note about Tinyiko, Pension Recipient).

Only with the assistance of several community members who have worked for Agincourt for close to a decade, were we able to assure these individuals that we meant no harm, and had no connection to the government. Although many of the respondents were reluctant at first, in the end we did not have any respondents refuse to be interviewed.

Despite many assurances that we had no ill-intentions, several respondents used the ‘life history’ portion of the interview to legitimize their presence in South Africa. Six of our 30 Mozambican respondents claimed to have been born in villages on the South African side of the Mozambican border; yet two of these six respondents also spoke about having lived in Mozambique and fleeing during the war. An additional seven of our respondents claimed to have come to South Africa before the war; they mainly claimed either to have come to South Africa when they married a man who was working in Johannesburg, or to reunite with family who had been moving back and forth across the border for a century. Still others who came during the war claimed similar connections to South Africa that pre-dated the war—relatives who were South African, or husbands and brothers who worked in South Africa and had South African documents.¹⁶

One respondent, who elaborated on her sons who worked in the South African mines and a sister who had lived in her village before the war, explained about why she was not open to being interviewed at first, “I was scared, thinking that maybe you [the interviewer] was an investigator coming to investigate us so that we must be sent back to our place in Mozambique” (Aletha, No Pension). Another respondent, Nelly, a widow who lives with her six grandchildren and a widowed daughter-in-law, explained how her South African connections both provided advantages and disadvantages when she fled the war in Mozambique:

My husband was in Johannesburg [when the war began], but he heard from people that we left Mozambique because of the war. He tried to get to us. Then he came to [our new village] and stayed with us. But my husband had a South African ID. When we were staying with the Mozambican people, the headman said that we must move from the place we were given and go to the headman in [our new village]. He was going to give us another place to build our houses because [we were] not allowed to get free food and clothes, while my husband was working and had an ID. My husband went to the headman of [our new village] and explained why we came to him. He gave my husband a compound and we started staying there (Nelly, Pensioner).

The interesting element is not whether or not this and the other stories about ties to South Africa are true, but rather how their beliefs about their vulnerability to deportation led to an emphasis on their long standing connections to South Africa.

Given the harrowing stories of family members and neighbors killed by RENAMO¹⁷ soldiers in front of their eyes, and the arduous journeys that included bribing people for assistance, walking long distances, fear of being eaten by lions, and having to leave dead family members along the way without a proper burial, it is not surprising that respondents feel vulnerable to deportation. Jane’s concise story underscores elements reported by many of our respondents:

It was during Autumn when we heard that there was a war between FRELIMO soldiers and RENAMO soldiers. It started in other villages, far away from ours, but as time went on, it reached our village. Many people had been killed by using a mortar and pestle, and also sharp knives. When they arrived at your house, they took everything you had e.g. cows, goats, chicken and also money. After that, they would choose whom they wanted to kill. We ran during the night. We went by foot, not transport, to come here. We used paths in the forest. During the night [wild] animals were there [in the Kruger National Park]. Some of us didn’t arrive here; they were killed by animals e.g. lions, but me and others arrived here safely (Jane, Pension Recipient).¹⁸

Even though life in South Africa is hard, none of our respondents mentioned a desire to return to Mozambique.¹⁹ Several had older relatives or siblings, or even children, who still live in Mozambique, but our respondents knew little about these relatives and claimed that they did not see them because of the distance between Mozambique and their villages. Despite having lived in South Africa for 20 or more years, the main concern our respondents raised was that we were collecting the stories of their journeys and current situation in order provide this information to local authorities who would then deport them to Mozambique. Fear of deportation was not what led researchers in the area to label Mozambicans as vulnerable, yet in many ways this was the most real and visceral vulnerability that emerged from the interviews.

Getting documentation

Documentation is one way to establish roots and connections to South Africa. While South African documents are clearly desired by our Mozambican respondents, many believe that it is difficult, or even illegal, for them to obtain South African documentation. Even among the two-thirds of our respondents who had South African IDs, several told of difficulties obtaining an ID. It is unclear from the narratives whether the attempts, or success, in obtaining South African citizen documentation occurred during amnesty periods when it would have been legal to receive permanent resident documentation (e.g. 1999–2000), during the 1994 election voter registration when some Mozambicans were able to obtain South African citizen documents, or at some point in time more recently through extra-legal means. Given the connections that respondents made between needing a South African ID and getting access to old-age pensions in particular, there is some indication that at least some of our respondents were referring to attempts made more recently than 1994. For the recent attempts, it seemed that luck was a key element to obtaining documentation.

Respondents gave a range of answers when asked about their attempts to get South African documents. As one respondent said, “[It is difficult to get South African-ID books] because we are not legally allowed to have a South African ID document” (Jane, No Pension). When asked why she had not applied for a South African ID, another respondent said, “I don’t have money to apply for a [South African] ID, but it is difficult to find an ID if you are from Mozambique. I did try to [get one, I] appl[ied], but I failed” (Refilwe, No Pension).

The process of applying for a South African ID book is a primary obstacle. Several women, both those who were successful in getting an ID and those who were still without, said that they had to apply for an ID at Home Affairs more than once, and in different venues. “I applied for an ID first at Nelspruit. I didn’t get it. I applied for a second ID at Hazyview. I got it. They never asked me where I came from” (Loveness, Pension Recipient). Another respondent is yet to be successful, “I applied for the first ID at Home Affairs at Thulamahash, but I never got an ID. The second ID, I applied at Mkhuhlu. I never got it” (Abigail, No Pension).²⁰

Another barrier to obtaining an ID book can be financial, as Refilwe mentioned above. Each of the 20 respondents who had IDs gave different stories of how much they paid for the document, if they paid at all—this may be an artifact of timing of application and receipt, but is likely to be connected to varying bribes requested by officials at Home Affairs. Some paid just ZAR20 for the photograph to put in the ID book, others paid ZAR200-300 to someone at Home Affairs. It was unclear from the information given by respondents whether they believed they paid an official fee or a bribe. One respondent found out it was a bribe only after she lost her money:

Myself, I went to Nelspruit for the first time to apply for the ID, but I never got the ID. I gave them R200. [I gave it to] the people who were working in Home Affairs, two of them… When I gave them the R200, I never got the South African ID. Those people were arrested because of robbery of ID books. … The second time I went to Thulamahash Home Affairs where I gave them R300 and I got the South African ID within 3 months (Linda, Pension Recipient).

Despite beliefs that obtaining an ID might be illegal, difficult, take a few tries, or cost money, all of the Mozambican respondents desired a South African ID. Those who did not have one at the time of the interview said they were in the process of applying, were planning to apply, or asked our interviewers for assistance in applying. The main reason all respondents mentioned for wanting an ID was to gain access to the pension. “I wanted an [South African]-ID in order to get a pension with it.” (Nelly, Pension Recipient). It is likely that most of the pension non-recipients were in possession of a Permanent Resident ID-book (we did not directly ask if they had this type of ID book, or ask them to show us any ID documentation in their possession).

Despite the change in law in 2004, our respondents still believed that without a South African ID Mozambicans were not officially allowed to access social grants. Our interviews took place 6 months after the Constitutional Court judgment, yet local knowledge and beliefs about the need to have a South African citizen’s ID book in order to receive a social grant had not yet changed. Not a single respondent mentioned new access to social grants using a permanent resident ID, which many of Mozambican residents in the area held.²¹ Because of the importance of pension to cushion economic household shocks, the ten respondents not receiving the pension would in all likelihood have applied for the pension if they had been aware of this legal change.

Economic and emotional shocks

Obtaining access to resources to reduce the impact of economic blows is one strategy to reduce (economic) vulnerability. Due to our respondents’ living arrangements and cultural norms, their personal welfare depended on kin, but their kin depended on them as well. Two of the 30 Mozambican respondents lived alone, one other lived with just one other family member; the remaining respondents lived in extensive multigenerational households (membership ranging from 3–25). The women with whom we spoke had suffered various economic and emotional shocks, including illnesses and deaths of spouses, children and grandchildren, the loss of income from children’s jobs as they were retrenched, unable to find work, or stopped working due to illness or death, and the addition of household members for whom they took over care both physically and financially when possible.

It is clear why pensions are desirable, and the necessity of persistence in accessing them. As with the South African grandmothers we interviewed who reported that their pension is meant to support themselves and their grandchildren (Schatz and Ogunmefun 2007), Mozambican pensioners also primarily used their pensions to maintain and sustain their multigenerational households. Nelly describes her life prior to receiving her pension. Her pension helps to defray her family’s costs, but still is not sufficient to make up for her son’s lost income and her husband’s missing support:

Before I got pension, life was easy because my husband was working in Johannesburg and he was taking care of me. When he was old, he was getting pension and my son was working, and he was supporting us. But when I started receiving pension, life became worse because my husband and son [had just] died. Now I must take care of my grandchildren and other children with my pension money of ZAR740 and things are very expensive (Nelly, Pension Recipient).

Sharing pension funds often meant that they disappeared quickly. Dudu, who lives in a very large household with 25 members, alluded to using up her money the day she receives it. “[After receiving my pension], I [first] buy 80 kg of mealie-meal, peanuts, beans, cooking oil, tea and 12.5 kg sugar. [After buying all these, there are] no leftovers. As I am speaking I don’t have anything in my pocket. Maybe if I was receiving help from others, there will be leftovers” (Dudu, Pension Recipient). Getting and giving help are mainstays of this community, and pensioners are an important source of income for their households, even if older women feel that the money they receive does not go far enough. Having no leftovers was another commonality between South African and Mozambican pensioners (Schatz and Ogunmefun 2007). This monthly running out of funds to support one’s family was a source of emotional distress and felt vulnerability that many of our respondents expressed.

Despite the insecurity among pensioners of making ends meet, pensions make a crucial contribution to poverty alleviation in our respondents’ households, getting families closer to the national poverty line. Although it is clear that pensions are not eradicating poverty in these households, the importance of the monthly pension to households can be seen clearly when examining the livelihoods of those not receiving the pension.

One-third of the Mozambicans respondents were not receiving the pension at the time of the interviews. The majority of these respondents were not getting the pension because of their difficulties in obtaining a South African ID; a few had South African IDs, but the age on the ID was “cut” (appeared as younger than the age they claimed they were) making them age-ineligible to receive a pension. Whatever the reason for not receiving a pension, these women largely felt that it was inequitable that their peers were getting money from the government to help their families, and they were not. Jane, a respondent whose ID showed her age as younger than she believed she was, claimed that it is unfair that some people get pensions and others do not, “because they are getting money to buy enough food, but myself I am suffering. I have no food to eat.” She went on to explain how hard it is to take care of the two foster children for whom she is responsible. “It is too difficult because I don’t have money to buy food for them or even to pay school fees” (Jane, No Pension). In households with no pensioner, elderly women must sustain their households through alternative coping strategies.

In households where there is no pensioner, women were more likely to be doing temporary labor and looking to family and neighbors for support. Linah stays with her sick co-wife, her co-wife’s son, two of her own daughters, and two grandchildren. She described how her very poor household finds money, “Some other people call to give us temporary jobs and [they pay us with] mealie-meal…. I’m not getting support from family members or relatives because they are also poor. They are not working” (Linah, No Pension).

Lilly, age 73, lives with 19 kin, but is only supporting her last born child. She survives with the assistance of her sons and a monthly payment from her former place of work, compensation from a tractor accident:

So my hand was broken. Then my boss took me to the hospital. And the doctor said that I must stay at home a month in order to recover. So when I went back to my work, my boss said that there are no more [jobs]. So I went back to my home until now. … [After it happened,] I went to a social worker to explain what happened. So the social worker wrote a letter to my boss, demanding him to pay me for the accident. I came across, so now he is paying me ZAR260 per month.

My sons and my ZAR260 are what support me. … I buy mealie-meal, soap, cooking oil, and other groceries for the house. As I told you that we do share with my son. … [Last month,] I took ZAR100 to buy a goat and the rest I used to buy other needs for the house. …I do not [have leftovers] because I do receive a small amount of money (Lilly, No Pension).

Although Lilly gets assistance from her children and from her former place of work, she lives on very little money each month. If she had not broken her hand, it is likely that she would still be working.

Living alone after the death of her son, and her daughter-in-law’s leaving for her own compound, Qeliwe’s situation is equally if not more desperate:

I don’t have money to buy food. I even don’t have food to eat. Some other days, I sleep without eating anything. [My granddaughter gets food that she gives me] because their mother [my daughter] is receiving a social grant… the disability grant. After receiving the grant, she gives me money to buy snuff. Also she gives me meat, that is, they give me raw meat, so that I may cook it (Qeliwe, No Pension).

Despite not having other people to support, Qeliwe often goes hungry—the small amount of money she uses to buy snuff is not likely to make much difference in her ability to feed herself. Clearly older women who are unable to access the pension, especially those without working children, are particularly vulnerable. Qeliwe’s situation was made worse by the hardships she experienced while caring for her ill son:

When it started he didn’t have an appetite, he wanted to drink water only. He didn’t have to eat food. He became weak. When you asked me about it, my heart becomes painful, my daughter. He died on the way to a sangoma [traditional healer]. He was on my back because I did not have [money to pay for] transport to take him to the sangoma. On the way, I fell down with him. [Qeliwe started to cry, the interviewer comforted her.] When you comfort me, I feel as if I can see my son. Because he left me poor, as I am, I am suffering (Qeliwe, No Pension).

As Qeliwe tells her story, it is clear that she suffered not just because she herself is poor and hungry, but also because she could not provide the assistance and nourishment her son needed when he was ill. Most of the women who were not getting a pension claimed that they wanted to receive a pension to use it for many of the same household essentials on which pensioners spent their money. Some said that they wanted to be like other grandmothers and help their children and grandchildren. Abigail said, “If I was getting a pension, I would buy clothes and a bed to sleep in. But now money becomes a problem for me to get it. Even my grandchildren, I want to give to them like what other gogos [grandmothers] are giving to their grandchildren” (Abigail, No Pension).

Our interviewers confirmed that the women in non-pension homes were more troubled financially than those with a pension. In classifying households as poor, average or wealthy (an internal comparison of the 30 respondents’ homes) two-thirds of the non-pension homes were classified as being poor, as compared to only a quarter of the homes with a pension recipient.²²

Not receiving a pension has economic, physical and emotional consequences. Some women are unable to feed themselves and their families; for others, health is jeopardized by needing to work when they are aging and frail; still others suffer because they cannot provide the support for sick family members that they feel they must. These consequences are magnified when an elderly woman does not have financially secure family members to support her. These stories of economic and emotional hardship do in fact point to Mozambicans being economically vulnerable in much the way that researchers in the area assumed them to be. Whether they are more vulnerable than the many South African households in the area without access to employment or pensions is questionable, however.

Discussion

Although all of the women in our study were legally eligible to receive the pension after the South African Constitutional Court judgment in March 2004, barriers remained to our respondents’ realizing their right. The primary barrier was knowledge of this right. While the focus of local researchers was on the economic vulnerability of Mozambicans due to limited access to social grants, on the ground it was the struggle to obtain South African documents that made our respondents vulnerable in three ways. First, it reduced their perceived right to stay in South Africa, thus making them wary of talking to us about their nativity. Second, the energy and finances spent to obtain South African IDs would have better served their families in other ways. Third, for those unable to secure a South African ID, their lack of access to the old-age pension diminished their ability to provide for their families. In addition, worries about their insecure status as non-citizens were most likely made worse when attempts to secure South African identification-documents failed.

Our respondents’ desire for South African IDs was not simply about securing status. These older women were also aiming to access the old-age pension. In most cases, those who had found means of obtaining a South African ID had done so before the March 2004 ruling, seemingly with the explicit purpose of gaining access to the old-age pension. Respondents in our study who were receiving a pension were much better off financially than those who were not, and were more able to assist other members of their households. The narratives above give further evidence of the dire straights in which many non-pensioners find themselves and their families, and the emotional strain they feel by not being able to help ailing or unemployed children and young fostered or orphaned grandchildren.

The fact that two-thirds of our respondents had managed to secure these documents and were accessing pensions calls into question the primary reason local researchers and community members had labeled them as particularly vulnerable. The narratives highlighting economic vulnerability bring into sharper focus the activism and agency of the majority of our respondents in creating opportunities to access pensions. Our respondents outlined concrete obstacles that they face. Their attempts to make use of the resources and support within their reach, however, highlight the distinction between generalizing the group as vulnerable and allowing these women an opportunity to name and therefore take ownership of their own vulnerabilities. These women did not define themselves as totally impervious. However, the insecurities they articulated did not necessarily match previously conceived notions of the population. By looking beyond the label to vulnerabilities actually experienced by this population, we begin to see the areas of struggle and action. We see the agency and empowerment that our respondents embody as actors in their households and communities, something which the label of vulnerable often hides.

Conclusion

By the time of writing of this paper, we hope that all of our respondents, all of whom were eligible to possess a Permanent Resident ID book, are now receiving pensions. One of the notable issues this paper raises is the difference between the legal decisions and the reality on the ground. In September to December 2004, all 30 of our respondents had a desire to receive a pension and were eligible to do so. Yet, 6 months after the Constitutional Court judgment, a third of them were not in receipt of a pension. The other two-thirds were not receiving the pension as a result of the judgment, but because they had been resourceful (and fortunate) in obtaining a South African citizen identity document at some earlier point in time. In response to this gap in knowledge, our research team shared this information with the local community liaison team, which set up sessions in a number of central villages to help Mozambicans, and other under-documented residents of the area, complete the required forms to obtain necessary documents and access social grants.

Because many Mozambicans found ways to obtain South African ID documents before it was legal to do so, it is difficult to estimate the number of Mozambicans who were already accessing social grants in 2004. The most recent rounds of the Agincourt census have begun collecting identification-document information as well as records of social grant receipt. Together these data provide future researchers the opportunity to explore Mozambican vulnerability with respect to social grants in a more quantitative and unequivocal way, which could help local policy makers create programs to adjust for any continuing inequality.

While the generic notion of vulnerability suffers greatly from limitations, there is utility in having blanket descriptions of vulnerable populations in order to forward anti-poverty policies. And, often outsiders have knowledge that can enhance the lives of so-called vulnerable populations. Yet interventions and programs will only succeed if there is a resonance between the policy makers’ views and how the population perceives their own risks and assets. Thus, critically evaluating programs with attention to the locally articulated vulnerabilities, in other words bringing in the voices of the vulnerable, is essential. Particularly given the need to assist populations rendered vulnerable by AIDS, defining vulnerability through their eyes and experiences, should be a key aspect of AIDS policy. Although perhaps a daunting task, understanding how and why an identified population feels vulnerable in a certain situation will allow policy makers to address those specific issues, which likely will assist these individuals more than blanket remedies. In particular, results will improve through the creation of strategies that aid individuals in areas where they have already demonstrated desire and agency, rather than simply addressing vulnerabilities as they appear to those outside the community.