Mobility, marriage decline, and the ceremonial economy: socio-cultural factors influencing farming in South Africa and implications for land reform

ABSTRACT

This article reviews the literature on the social dynamics influencing small-scale agriculture in South Africa. These include three primary factors: the trans-local character of livelihoods; the role of social hierarchies of gender, age and marital status in allocating rights and responsibilities at home; and the ceremonial economy. South African land reform policies must recognise these local practices of distribution and social reproduction as integral to people’s livelihood strategies. By doing so, land reform can move beyond the narrow emphasis on productivity and ‘self-reliance,’ instead focusing on aligning policies with the strategies of the poor.

KEYWORDS:

• Small-scale agriculture
• land reform
• migration
• distribution
• livelihoods
• South Africa

Introduction

There is widespread agreement that government interventions to support small farmers in South Africa have often been poorly constructed and have failed to produce positive impacts (Aliber and Cousins 2013; Okunlola et al. 2016).¹ Political factors play a key role, such as the ‘elite capture’ of land reform processes and resources by better-established farmers, limiting the support available for small farmers (Hall and Kepe 2017). This power struggle is aided by a ‘modernist’ bias within land reform policy whereby ‘commercial viability’ underpins entitlement to redistributed land (Cousins and Scoones 2010). This is a narrow view of land reform that prioritises production over the potentially multiple uses of land at different scales, and fails to recognise the socio-cultural contexts in which diverse farming practices take place.

Explanations for the limits to small farming have focused on the high replacement costs of production. These include the cost of inputs, the availability of fencing, as well as costs required to mitigate risks, for instance, crops being damaged or eaten by animals, or ecological risks such as droughts, flooding and declining soil fertility (Andrew, Ainslie, and Shackleton 2003; Aliber et al. 2009). However, the role of social and cultural factors is poorly understood, in part because of a tendency in land reform debates to treat ‘culture’ as separate from, and incidental to, economic behaviour. For example, this tendency is implicit in the assumption that farming is the only legitimate use of land, and moreover that if people are given access to land, inevitably they will take up farming (Hornby and Cousins 2019, 202). The ‘non-productive’ use of available land in many rural areas calls into question this assumption and suggests a need for a deeper understanding of the diverse issues affecting farming decisions (Hey and Beinart 2017; Hull 2014). Our aim here, drawing on and bringing together a range of examples, is to highlight the social dynamics of agriculture that impact on production and livelihoods.²

In characterising these as ‘social’, we do not regard them as distinct from economic factors; instead, we conceptualise them as aspects of economic life that are socially embedded and therefore cannot be understood outside of the social context. Reviewing the literature on this theme, we consider three main factors: migration and resource exchange between urban and rural areas; uneven responsibilities between family members in the context of changing marriage practices and household demographics; and the circulation of resources in the ceremonial economy. All three involve articulation between domestic lives and the broader economy characterised by migrancy and unemployment. In Marxist political economy this relationship has been theorised in terms of the interconnected dynamics of production and reproduction. In this framing, social reproduction includes ‘the reproduction of distinctive forms of marriage, systems of kinship and community membership, as well as property relations that are not characterised by private ownership’ (Cousins et al. 2018, 1060). Central to this discussion is the observation that unemployment and de-agrarianization have placed mounting strain on social reproduction which consequently has become contested and contradictory. Anthropologists have also contributed to this literature, seeking to overcome the dualism between production and reproduction. The anthropologist James Ferguson recently challenged scholars writing on Southern African economies to think beyond the analytical centrality given to production in both neoclassical and Marxist economic approaches (Ferguson 2015). In a context of unemployment and low agricultural productivity, he maintains, economic survival relies primarily on distributive rather than productive practices. These include the state mechanism of cash transfers on which there is widespread reliance to meet basic food needs, as well as urban-rural remittances and other social networks through which ‘distributional demands’ are made locally (Ferguson 2015). In farming households, as well as to meet consumption needs, mitigate risks and re-invest in production, income is often used to manage social claims from others (Neves and du Toit 2013).

The emphasis on distribution as a key site of value reveals the interconnections between production and reproduction, highlighting the significance of ceremonial and other social obligations that impact on farming prospects. It calls into question the assumption in mainstream agricultural development policies that surplus gains from agriculture will necessarily be reinvested in production. Instead, distributional obligations often take priority over production and accumulation. This also highlights concerns about the goal of land reform policy for small farmers to achieve ‘self-reliance’, raising the issue of whether this is either realistic or politically desirable in a context in which distribution and interdependence are constitutive of economic activity. Some small farmers do of course establish themselves independently. Rural communities are fragmented by class differences that produce differentiated outcomes and prospects for farming. Our point is to emphasise the range of practices and constraints that affect how land is used, thus questioning the normative prioritisation of self-reliance in land reform discourses. Therefore, unlike Ferguson who views the two as ‘starkly disconnected’ (Ferguson 2015, 192), we consider it essential to understand the ways that distributive practices are linked diversely to those of production.

In this article, we bring together evidence and insights from both Marxist political economy and anthropological literature to demonstrate how through distributional activities, social relations impact on farming prospects. We refer to examples from both the former ‘homelands’ areas and land reform contexts, identifying similarities across the two.³ After considering three major features of the literature – trans-locality; household members’ roles and responsibilities; and the ceremonial economy – we reflect briefly on the implications for land reform policy. The discussion highlights the inadequacy in mainstream policy of treating farmers as simply responding to economic constraints or incentives since the dynamics of social reproduction – including distributional networks – impact on how small-scale production is organised and its viability.

Households as trans-local, distributional networks

In South Africa in the late nineteenth and twentieth centuries, successive colonial and apartheid governments created and institutionalised a class of cheap, migratory wage labour through land expropriation and political disenfranchisement. Large numbers of African men left rural households to earn cash income imposed by the imperatives to meet tax obligations and to supplement declining agricultural resources. As agricultural production fell, the state-maintained food supply by subsidising and supporting large-scale white-owned farms. This created a highly unequal structure in which white farmers produced the bulk of food and export goods. This marks South Africa as distinct in comparison to surrounding countries where peasant agriculture was not so aggressively dismantled. Nonetheless, there are also similarities as far as the social dynamics of agriculture are concerned, not least because these countries experienced considerable migration to South Africa, shaping family life and economic practices in home communities.

Despite the aspirations to redress this dualistic structure by the democratically elected African National Congress (ANC) in 1994 through an ambitious programme of land reform, inequalities in the sector have been further entrenched. A much smaller portion of land has been redistributed to black farmers than was originally intended, and insufficient post-transfer support has hindered the success of land reform projects. Moreover, the removal of subsidies has ensured the concentration of market power in still fewer hands (Bernstein 2013). These agrarian dynamics have combined with deindustrialisation to place immense pressure on livelihoods, reconfiguring social relations in various ways and producing a perceived crisis of social reproduction.

The patterns of circular migration underpinning this agrarian structure have created dual urban-rural homesteads as enduring forms of social organisation (Amoateng and Richter 2007, 3). Harold Wolpe’s (1972) seminal article described this spatial dynamic in terms of the production-reproduction model. While men moved to the mines and other industrial centres, he argued, women remained in the rural base reproducing the household. This discussion initiated rich debates about the trans-local character of livelihoods, albeit leaving little room for understanding women’s experiences of paid work. Women began migrating to the cities much sooner than was credited by the longstanding ‘male-migrant’ model (Phillips and James 2014). In this section, we describe how the pressures of unemployment and limited farming prospects mean that livelihoods frequently rely on mobility and family networks – that is, on distributional practices as well as productive ones. These practices, in turn, affect farming prospects because there are many claims on urban incomes aside from those that support agricultural production, and because mobile lives and shifting obligations within the family limit the availability of labour or resources for farming.

In the 1990s, liberal democracy was established alongside a rapid increase in urbanisation and permanent ruptures occurring in some urban-rural household relations (Smit 1998). Sustained out-migration of young adults in the early 2000s in search of better prospects has profoundly affected rural households leaving many without a middle generation. Those who stayed behind described rural locations as ‘emptied outs’ after 2000 (Bank 2015). As Leslie Bank notes, ‘By 2006, figures showed that young men and women from traditionalist communities like Shixini were leaving in droves without waiting for promises of jobs before departing’ (Bank 2015, 1073).

However, limited urban work opportunities created a counter-pressure. In June 2020, Trading Economics reported on its website that South Africa’s youth unemployment rate was 52,3%, third highest in the world after Angola (56,4%) and Kosovo (54,1%). While the precise extent of unemployment in South Africa is open to debate, it is clear that it is severe, growing, and endemic largely as a result of technical changes in, and the closure of marginal mines on the reef, alongside the absence of alternative, job-creating industries. The loss of work was initially partly offset by wage rises in the post-apartheid period, but this had the effect of sharpening rural differentiation, while shifting the burden of escalating unemployment onto young, mainly black adults (O’Laughlin 1998).

In land reform debates, an assumption underpinning the focus on land transfer is that in a country with high rates of unemployment, land rather than labour is the key constraint. Hence, if people have access to land, they will automatically take up farming. The widespread practice of migration introduces an obvious complexity to this issue that has been debated for several decades. On the one hand, migration has contributed to declining agriculture by reducing labour on the land (Bundy 1988). On the other hand, this could be mitigated by the gendered division of labour in which women became responsible for farming, and by the investment of remittances in farming inputs. For instance in Mpondoland, migration and smallholder production expanded simultaneously between the 1890s and the 1960s (Hey and Beinart 2017, 754). A personal account by a farmer near to Johannesburg suggested that young people with salaried income from formal employment have the greatest likelihood of success in farming because among other things, they have the advantage of being able to access loans (Ngubane 2018). Unemployment may preclude household farming because a family will have insufficient funds to cover the costs of production or to absorb risk.

The shortage of wage work in the urban labour market has meant that the rapid outmigration of youth from rural areas has not necessarily been accompanied by a growing commitment to urban permanence (Bank 2015). Although many households identify themselves as urban, many believe the rural link is important both as a secure base and as a means to a more fulfilling life. Due to clustering around rural-based pensions and the need for a safety net should the household be unable to survive in the city, many urban dwellers continue to maintain links to a rural home. The rural home is less a place of investment than it was during apartheid and more a form of insurance against unemployment and misfortunes, in part because young, unemployed adults have not been able to marry and create their own homes (Scully and Britwum 2019; Hull 2014). In northern KwaZulu-Natal, mobility may be high even without employment, since some young people move frequently between country and city in search of work (Hull 2014, 454). In other cases, young adults are aware of the scale of urban unemployment and are not prepared to take the risk of migrating. Instead, they are left feeling trapped indefinitely in a rural setting devoid of work (Hull 2014, 454). As we elaborate in the next section, this has unpredictable outcomes in terms of their availability to contribute to family farming due to changing expectations within the home.

Remittances have changed over time due to changing work opportunities. Where (mainly) men employed in urban areas used to regularly remit part of their wages to rural homesteads, they are now partially employed and tend instead to make occasional lumpy investments:

Today men say to their wives ‘you get your own income from child grants. Use that for food.’ So they’ve stopped remitting on a regular basis. But this doesn’t mean they don’t bring money home. They do, in the form of big investments: building a new house, investing in or expanding livestock (goats or cattle); household purchases, like TVs, generators, furniture. (Rauri Alcock, personal communication, 23 July, 2019)

In addition, while urban areas send remittances to the rural home, rural areas transfer food to the city, making it possible for urban households to survive (Frayne 2005). These exchange circuits underpin the sustainability of productive activities both on and off farm. Rural households with an urban pole or urban households with a rural pole are less vulnerable than either urban or rural households without a rural or urban pole (Neves and du Toit 2013; Neves et al. 2009, 110).

The interdependence of rural and urban areas is also due to the differing advantages they offer, unrelated to resource transfers. Urban sites offer access to employment, better amenities and social services (schools, healthcare, electricity, potable water and transport). Rural areas, by contrast, offer retreat from urban labour markets, are places of retirement and recuperation, safety nets for livelihood shocks (death, retrenchment, illness) and incur lower costs of living as a result of relatively non-commodified land, housing, water and fuel costs. This also makes them economical spaces in which to raise children, and care for the elderly and sick (Neves and du Toit 2013, 101). People attach great value to rural life and many wish to retire there. This relates to how the former ‘homelands’ are the ‘socio-geographical heartlands of distinct and dynamic identities, cultures and languages’, a point frequently overlooked in discussions about marginalisation and livelihood struggles (Beinart 2012, 19).

The patterns of mobility described here reflect three key dimensions of livelihoods. First, households are spatially ‘stretched’ with fluid membership (Spiegel, Watson, and Wilkinson 1996), as livelihoods straddle urban and rural spaces. Second, livelihoods are diversified across a range of activities in order to spread economic risk (Francis 2000). Even if people farm, they are likely to spread their time between other activities as well. Third, households rely on ‘clustering’, which involves settling around one regular income, often a government old age pension grant (Francis 2000). Fundamental in each of these strategies is interdependence between family members as well as within wider social networks, signalling the importance of distributional practices as a core feature of rural economies.

One implication is that households are held together through networks of distributional obligations, rather than because they are a cohesive unit of production. For this reason, mainstream understandings of the ‘household’ do not map easily onto this setting. The South African census includes a person as a household member if they were present on the night of the enumeration, whereas other definitions define a household as a ‘group living on the same property who eat from the same pot of food’ (Agincourt Health and Demographic Surveillance System, cited in Atkinson 2014). Instead, households are better characterised in terms of patterned, regular transfers of resources than in terms of units of co-production, co-residence or co-consumption.

Processes of de-industrialization have resulted in two important trends: an increased reliance on income generated by insecure and temporary employment, often secured by women in the informal sector (Neves et al. 2009, 33; Fakier and Cock 2009); and an increasing number of households dependent on social grants, particularly the old age pension. This wider economy of unemployment together with the distributive practices that emerge in the absence of work entail shifting roles and responsibilities in rural homes. These have implications for farming, issues to which we now turn.

Gendered and generational roles and expectations at home

The uneven allocation of responsibilities and rights within households is partly determined by socially inscribed hierarchies through which inequalities are normalised. A key fracture is gender, given that the labour burden of social reproduction falls disproportionately on women. Age and marital status are also significant factors in the attribution of responsibilities and rights.⁴ They influence the viability of family farming which relies on people’s social entitlement to the use of land as well as the time, capability, and willingness of household members to provide unpaid labour. Far from being fixed or predetermined, these categories and the roles attached to them are negotiated through ‘domestic struggles’ that also interact with, and are contingent upon, the dynamics of the capitalist economy, as Belinda Bozzoli articulated in her well-known article on migration and patriarchies (Bozzoli 1983). An economy marked by chronic unemployment has caused lasting changes that affect gendered and generational power relations, which in turn, affect the organisation of farming within the household.

One important outcome of de-agrarianisation and unemployment is the unaffordability of ilobolo (bridewealth), precluding the option of marriage for many men. Marriage rates have declined drastically since the 1960s. The proportion of African adults who were married in South Africa fell from 57% in 1960 to only 30% in 2011 (Budlender et al. 2011).⁵ An important clarification, however, is that the binary view of conjugal relationships as either married or unmarried is inaccurate in African contexts in which marriage occurs gradually over a number of extended transactions and inter-family engagements. Historically, these included ilobolo (bridewealth) negotiations and cattle transfers, umembeso (the transfer of gifts from a man’s family to his future wife’s family) and umabo (the transfer of gifts from the wife’s family to her husband’s family). The high costs of completing these transactions means that many couples adopt a truncated marriage process that involves the transfer of three to five cattle (including a ‘fine’ for pregnancy, which allows the child to be recognised by the paternal lineage), the relocation of the woman into the man’s household as makoti (young wife), and official registration of the marriage. Given the economic constraints, this stage of marriage – uganile or uganiwe – may be displacing older forms of marriage (Cousins 2017). Moreover, anthropologists have shown that bridewealth among the Tswana is seldom paid in full and that this is not new (Comaroff and Roberts 1981, 136). In this respect, marriage is ‘a process and not an event’ (Peters 1983, 110). One implication for farming is that the bureaucratic assumption of a couple as either ‘married’ or not in some cases leads to tenuous land rights for women who cannot demonstrate legally binding marriage (Weeks 2013).

Here, we focus on three key aspects, all linked to unemployment and declining marriage rates, with important implications for farming: customary law and land access; cash transfers; and changing household demographics.

Women’s access to land in communal areas

Today, land in South Africa’s communal areas, which are home to about 16.5 million people, continues to be governed by localised rules and practices that find legitimacy under customary law and which can often be clearly articulated by the residents in these areas.⁶ The administrative authorities operate at different socio-spatial layers in systems that are socially-embedded (Cousins 2017). Chiefs control land while also granting citizenship rights. As part of the bundle of rights associated with membership is access to land for residence, cropping, grazing and natural resource harvesting. The actual authority to allocate, however, lies in the first place with the household heads in the immediate neighbourhood (‘ibandla’) who can refuse an allocation. Furthermore, even though residential and cropping land is held under the authority of a particular household head, there is a common understanding that it is an allocation to the whole household to support their livelihoods.

Historically, ‘married couples’ and their children tended to live within their extended family homesteads before they established their own home, resulting in large homesteads containing multiple households. According to stereotypical customary law, new households would be formed when a man with his wife applied to the chief for land to establish his own homestead. Each household would be allocated its own plots within the homestead to build their home and fields to grow food. Married women would have access to land through their husbands, making marriage a key mediating institution in land access (Budlender et al. 2011). Unmarried men and women would be expected to stay in their natal family homesteads until they married and became entitled to land.

Although these household organisations are still evident across the South African countryside, variable marriage practices generate a diverse set of household arrangements, including co-residence of partially married couples (Hosegood, McGrath, and Moultrie 2009, 294) and more households formed by unmarried women with children. These realities have begun to put pressure on formal customary law. Notwithstanding a great deal of regional variability, unmarried women frequently are being allocated land to build homes and farm (Claassens and Smythe 2013; Cousins 2017). The changes mean that women with children can now ‘hold’ land in their own right and establish their own households.

Although the changes are heavily contested within households, they have been prompted by the increasing numbers of female-headed households on the one hand, and the contributions made by unmarried daughters in developing the home and caring for their parents on the other hand. In the Eastern Cape, evidence shows that women are increasingly taking control of inherited property, or at least sharing the responsibility with male kin (Kingwill 2016). Because women are judged as a more reliable ‘insurance’ against the sale of titled family property than men, an increasing number of females are appointed as custodians. Rosalie Kingwill concludes that these trends complicate the notion of patrilineal succession and confirm a growing recognition that women can be de facto successors to family property.⁷

It remains unclear how widespread these locally negotiated changes are. In the context of wider struggles over traditional leadership, there have also been attempts to reassert static notions of customary law, in order to reassert authoritarian powers of chiefs and remove the gains made by women (Claassens 2013). Ambiguity also arises in the mismatch between legal binary categories of marriage and social realities. Women who lose residential land rights at the hands of family members are less likely to receive legal redress in part due to ambiguous marital status (Weeks 2013). Women’s tenure security or otherwise therefore remains embedded in social relationships. State support is needed for processes of securing land rights for women. Instead, the state has frequently ‘opted to bolster the power of traditional leaders to unilaterally declare the content of custom’ and therefore undermined the potential for grass-roots change (Claassens and Smythe 2013, 26).

Moreover, despite evidence of a shift in the allocation of land to unmarried women with children, this has not resulted in an automatic increase in investment into agriculture, which remains sporadic and conditional (Shackleton 2019), an issue taken up in the next sections.

Cash transfers

South Africa has a well-developed social protection programme, including old-age pension, child support, and disability grants among others. 44% of households receive at least one grant, up from 30% in 2003 (Stats 2018). Cash transfers expanded further in the wake of the food and livelihood crisis precipitated by the Covid-19 lockdown measures, albeit hindered by implementation failures (Devereux 2021).

State social grants, with mostly the elderly as recipients, have become the basis of survival for many rural households with farming providing mainly a supplementary livelihood for some. The pension is the largest, most regular transfer of cash into rural households and is thus ‘the material lynchpin for many rural households, particularly the three-generation or skipped-generation’ household (Neves and du Toit 2013, 106). Since women have higher life expectancy than men, recipients of pensions are often elderly women, around whom the rural household forms. This has weakened the link between old age and extreme poverty and reversed the direction of economic dependency (O’Laughlin 1998). The consequence is that the poorest people are prime-age adults without a regular wage or remittance income, living in households in which nobody is receiving a pension. They may lack marketable skills, have too many dependants to be able to work, or are too poor to look for work, because of the costs of searching for work (Francis 2002). Young kin in poorer households may be less likely to access urban incomes because they keep in contact less frequently with relatives.

Social grants further influence these shifting power relations, contributing to changing gender and generational relations. One key aspect of this distributional economy is the reversed dependencies whereby men rely on women’s income via the pension. As well as the pension, the child support grant is also widely controlled by women, sometimes leading to conflict within homes about how the money should be spent. While some show that women’s autonomy has increased (Granlund and Hochfeld 2020, 1233), Elizabeth Hull’s research in KwaZulu-Natal suggests that some families experience ambiguity or conflict over how grants should be spent. Both grants and home gardens can also expand interdependency between women occupying different households because of their ability to share gifts of food and to participate in informal savings clubs known as stokvels (Granlund and Hochfeld 2020, 1234; Hull 2016).

While pension grant income is sometimes used as a source of agricultural investment (Olofsson 2020), David Neves and co-authors (2009) demonstrate that grant income is more often used to support unpaid socially reproductive care work. This frees up working-age adults to engage in employment. Social grants thus make possible ‘cycles of reciprocity’ that ameliorate vulnerability and are not simply a means to boost consumption or facilitate fiscal investment. The investments in inter-generational relations thus do not necessarily translate into agricultural investments.

There is some indication that women with child support grants are less likely to engage in farming (Fay 2013; Hey and Beinart 2017; Granlund and Hochfeld 2020). While elderly people sometimes attribute ‘laziness’ of younger relatives to the availability of grants, young people’s limited participation in farming occurred prior to the introduction of the child support grant, pointing to a complex array of factors (Hey and Beinart 2017). In Hobeni in the Eastern Cape, residents explained that school attendance had risen with access to child support grants, meaning children were spending less time tending fields (Fay 2013). While residents have given up field cultivation, they have continued to cultivate in small homestead gardens, using the child grant to buy the staple maize while planting vegetables and tree crops. A similar observation has been made elsewhere in the Eastern Cape (Granlund and Hochfeld 2020).

Therefore, to the extent that they do invest in agriculture, women-headed households may be more likely to allocate resources to gardens for consumption, that can be protected from livestock damage and that have low input requirements in terms of cash and labour (Shackleton 2019). However, this appears to be highly variable across the country. In former homeland areas, where villages were planned through the notorious Betterment schemes, fields were forcibly separated from homesteads creating a clear-cut distinction between fields and gardens that does not necessarily pertain to other rural areas (Hey and Beinart 2017, 770).

Livestock preferences are frequently chickens and goats, which are traditionally owned by women and children, as they are smaller and are therefore easier to manage and use for immediate consumption (Rauri Alcock, personal communication, 23 July 2019). The gendered nature of cattle ownership may, however, be changing as a result of women’s access to social grants together with rising male unemployment resulting in women purchasing cattle stock in their own right (Hornby 2015). Decisions about how to use grants, how and who should farm, and the entitlements that family members have to these resources, is also shaped by changing demographics at the level of the household.

Changing household demographics

The discussion so far highlights that even when land is available and people are at home without work, they may not always contribute to family farming, a tendency particularly noted among young adults. This has important implications for land reform processes, which have frequently assumed that labour will automatically be available for farming. In the wake of declining marriage rates leading to the inability of young adults to form their own homestead, multiple household arrangements have emerged that are changing people’s roles and responsibilities at home. As well as ‘skip-generation’ households where an adult middle generation is missing (Posel 2001), there are also households containing multiple generations, sometimes as many as four, often under the head of a pensioned grandparent and with resident unemployed, unmarried adult children (Hull 2014). Another study finds that rural households vary in terms of size (4–29 members), number of generations (50% consisting of three generations or more living together) and membership, including mothers, daughters-in-law and grandchildren of the head, the household head’s brothers and sisters, nieces and nephews, and widows of deceased siblings along with live-in domestic helpers and adopted orphans, in various combinations (Hazell 2010).

Due to these differing arrangements frequently not based around a marital partnership, adults may have fewer responsibilities than they would have had as part of a ‘conjugal contract’ (Hull 2014). In a context in which people make distributional claims on one another, this also entails efforts to ring-fence resources at an individual level and to pursue individualised livelihoods, making the prospect of communal household production less likely. Cash income secured by members of households is often not pooled because of a conceptual boundary between money and subsistence, where an individual’s cash earnings are privately owned but subsistence is a collective right (Hull 2014, 455). The effects of the non-pooling of cash income are a burden that falls mainly on older female pension recipients because of their responsibility for food provisioning and their dependence on cash income for food purchases, which reduces the available capital for investment in agriculture. It also becomes more difficult to mobilise the unpaid labour of young adults in farming, suggesting changes to generational authority potentially linked to declining marriage. The reluctance of young people to engage in farming is also partly due to a collective memory of the rural household being forged on the basis of a salary-earning male, creating aspirations that are often focused on achieving paid employment rather than on family farming. Ideas of masculinity and citizenship are linked to the status of formal employment (Hull 2014, 455).

In fruit and nut orchards in the Vhembe District in Limpopo adult children associate farming with failure, hardship and lack of profitability (Olofsson 2020). This makes them reluctant to pursue it as a livelihood while new opportunities for education and professional careers make farming undesirable by comparison. In order to motivate their children to take up farming, the older generation has adopted particular strategies, such as ceding the harvest to the child in return for his taking responsibility for it. None of these elderly farmers considered selling their farms, and in only a few cases did the subsequent generations view farming as a potential career. Instead, they viewed the orchard as an asset that required few inputs and labour but provided an additional income source.

The organisation of the household, its capacity to marshal resources and co-ordinate the actions of its members, and either to secure the labour of younger generations and women or to purchase the surplus labour power of other households, are all key to the class dynamics of unequal resource access in conditions of inequality. Central to this is effective decision-making associated with generational authority (Neves et al. 2009; Hull 2014) . The erosion of authority as a result of the changing composition of households has generated contradictory and complex negotiations in the rural household, anchored on the separation of production and subsistence: on the one hand, younger kin are entitled to subsistence from members with access to a state pension. On the other hand, elderly kin may struggle to claim the labour or income of young adults. This varies across class differences however, since for older farmers who have sufficient resources to pay for piecemeal labour, they may hire young people on a casual and piece-meal basis for weeding and harvesting of fields: ‘the cheap labour of young people who make up a reserve pool of casual waged-labour is appropriated by better established, older residents with access to resources such as land and credit who can afford to buy labour’ (Hull 2014, 456–457).

There are also regional differences and variations in land access arrangements, as Hull’s study was conducted in an area where these older residents had access to plots inside an irrigation scheme. Younger women, in truncated marriages with the (grand) sons of the older women heads of households in which they resided, undertook domestic work, caring for their own and other children in the household, which released the older women to tend to their cash crops (Hull 2014). The possibility of accumulation from small, irrigated fields is thus conditional upon labour exploitation for production and social reproduction along the two axes of generation and gender, and on declining rates of marriage, which make young women’s socially precarious positions in their partners’ homes contingent upon their ongoing participation in domestic work.

Similarly, Cousins (2013) and Neves and Du Toit (2013, 103) document cases of older women farmers who hire in agricultural labour, sometimes in response to neighbours’ requests for material assistance. While the remuneration is low and the jobs ad hoc, the response is an effective way of enabling women farmers to develop relationships of mutuality and petty patronage that cement their positions as minor benefactors. As the authors observe, ‘[t]his is a noteworthy consideration in a patriarchal context in which an elderly, mobility-impaired, female (de facto) household head might otherwise be more easily marginalised’ (Neves and du Toit 2013, 103).⁸

In summary, the political economy of unemployment may be reconfiguring a new gender-generational organisation of productive and care labour. Intergenerational relations are changing in contradictory ways. On the one hand, family farming may be limited by the difficulty of the elderly to mobilise the labour of their younger kin, who aspire to work in salaried jobs in towns and cities. On the other hand, in the context of deepening unemployment, class formation occurs along generational lines when older people in more established positions and can utilise the labour of a younger class of unemployed people. In this process of differentiation, class, gender, and generational roles and expectations have contradictory effects that influence the uneven use of land in rural areas.

The ‘ceremonial economy’

Integral to rural life is a range of customary practices centred around ceremonies to mark major life-cycle events and to maintain connection to ancestors (amadlozi). Animal slaughter is an essential part of these festivities, and its frequency may if anything have increased in recent years (Beinart 2012, 20). Ceremonies create collective life and symbolise its reproduction through time. They require households to put aside a ceremonial fund for ongoing ceremonial use. In small farming households, the ceremonial fund can be a drain on the reproduction of labour and capital since it is a demand on the total funds necessary for securing the conditions for future production (Bernstein 2010). Therefore they are a key aspect of the distributional demands made on households and may often be prioritised above productive activities.

So powerful are the social obligations inscribed in the body of cattle and goats that strict distinctions are maintained between ceremonial meat and meat for monetary exchange, making it difficult to convert one into the other. As a rural NGO worker noticed: ‘You can’t slaughter a cow and cut it up and sell it as pieces of meat. It can’t be done. But if it’s processed – say into sausages or burgers and packaged – then it shows you’re producing cattle as a business. The meat is for selling. It’s not dlozi (ancestral) meat anymore’ (Gugu Mbatha, personal communication, 23 July 2019). While the ceremonial uses of cattle may be viewed as economically ungainful from the perspective of commercial production, they are also part of a web of social entanglements (Ainslie 2013) that underpin household formation processes.

A distributive principle is central to the ceremonies as cattle stock, meat and food are shared by the whole community, which is crucial for cementing kinship and neighbourly relations. This principle also extends well beyond the ceremony itself. Household cattle herds represent a dense network of social debts and credits. A man whose now deceased father accepted ilobolo for his sister would owe support to that sister and her children should her marriage end. Although ilobolo is often viewed in the literature as a pre-marriage transfer of ten or eleven cattle, in practice, the number of cattle (or cash) transferred during ilobolo is highly negotiable, and the transfer takes place in small amounts over a lifetime and beyond, remaining a debt to be extracted from a women’s conjugal partner and his lineage by her natal family.⁹

The ceremonial fund can be a burden for many families. The cattle used in ceremonies can constitute a significant reduction in herd size. The smaller the herd to begin with, the greater is the withdrawal from the breeding stock. Not everyone has the four or five cattle and the goats to carry out mourning, celebration and marriage feasts or the cash involved for the purchases of food, goods and services entailed in these ceremonies. The integration of ceremonials or rituals into commodity circuits makes visible both the growing inequalities and who is able to meet the social imperatives of maintaining relationships between the living and the dead, which are key to the establishment and maintenance of kin relations. Those with the cattle and cash required for the full ceremonial feasts are not just rich and capable of largesse: their actions are also deemed morally good in that they reflect good standing with one’s ancestors and with one’s living kin. Many widows go to great lengths to complete burial rituals and their marriage to their deceased husbands using the cattle stock inherited from their husbands and loans from mashonisa (informal credit lenders) or stokvels, in order to maintain their relationships with their husband’s kin. The socio-moral importance of ceremony together with the costs of performing them means it is not entirely surprising that there is an overlap between wealth, ceremonial performance, position amongst kin and social status on the one hand, and the election into positions of local power, such as Communal Property Association committees that oversee land reform farms, on the other (Hornby 2015).

The logic of the ceremonial animal market is different from that of an animal raised as beef for a market geared at urban meat consumption. The criteria used to select dlozi animals are integral to the specific ritual and its purpose, including negotiation between families and kin, whereas beef is about class-based consumer preferences. Formal markets, which are geared primarily towards feedlots, abattoirs and butchers, use selection criteria geared at beef consumers, such as age and meat-to-bone ratios whereas the dlozi and ilobolo animal selection criteria include sex, size and colour. These market criteria in turn shape production practices and objectives in that beef farmers aim to achieve fast weight gain by rotating grazing and providing supplementary feed, while dlozi animal producers aim to produce the maximum number of animals in low-input systems. These very different production objectives put small-scale farmers at a price disadvantage in formal markets that focus on cattle-for-beef transactions.

Nevertheless, the production of ceremonial cattle and goats is an integral component of African farming systems, which constitutes a significant, albeit under-acknowledged, market. The Department of Agriculture, Forestry and Fisheries (DAFF) reported on their website that in 2002, South Africa’s total cattle stock (dairy and beef in commercial and communal areas) was estimated to be 13.9 million animals. In 2020, approximately 9.1% (5,4 million) of the total human population was aged 60 years and older (Stats 2020, 10), of which about 4,4 million (80.8%) were African. If we assume that most of them will die in the next 20 years, and that the burial processes of each will entail the slaughter of one animal (a conservative estimate), then the ceremonial market will be nearly 33% the total cattle stock of the country in any one year and worth approximately R22 billion over this period (calculated at a low R5000 per head). This estimate excludes the use of cattle in weddings, coming of age events and other celebrations.

The contrasting production objectives of cattle for commercial and ceremonial use illustrate the ways in which farming practices – in this case, livestock farming – can be geared not only towards private accumulation but towards meeting the demands of local distributional economies centred around ceremonial practices. Moreover, these objectives can compete with one another, calling into question assumptions that farming decisions are made purely to increase self-reliance.

An example of land reform: Communal Property Associations (CPAs)

A final example is instructive in revealing how underlying social inequalities and the redistributive demands of ceremonial economies play out in ways that undermine farming prospects following land transfer, unless risks are carefully managed and mitigated. CPAs are a new post-apartheid form of land-holding organisation created in law¹⁰ specifically to enable groups of people to hold land and other property communally that they have received through land reform. Together with land trusts, the 1,490 CPAs that exist owned most of the 8000 hectares of land transferred through redistribution and restitution by 2015. They have been widely criticised in government reports as ineffective and conflict-ridden.¹¹ This suggests another instance of a ‘productivist’ approach which assumes that because land has been transferred and because rural communities have engaged successfully in communal farming practices historically, cohesive arrangements for production will necessarily arise. While the forms of organisation, production and property holding resonate with those in communal areas, they are also distinct and entail specific challenges.

A case study of six farms redistributed to former labour tenants in the KwaZulu-Natal beef and maize producing area of Besters suggests a mixed picture (Hornby 2015). On the one hand, the CPAs have enabled women-headed households to engage in cattle farming, previously a strictly male activity. Where CPAs maintain commercial beef herds and distribute dividends to members in the form of cattle, cash and farming inputs, women-headed households have been able to invest in cattle, and sustain and grow their herds, providing an important layer of diversification and resilience to grant-dependent livelihoods. An important additional component at Besters is that women do not have to manage their cattle on their own because the CPAs employ men skilled in cattle farming to manage the household and collective herds.

On the other hand, wealth differentiation has emerged quickly, leading to conflict. Households that had wage and pension incomes and had inherited cattle stock had household herds that were growing, often rapidly. Other households experienced sharp production declines as a result of high rates of herd off-take through sales and the ceremonial use of cattle, often at times of ‘shock’ such as the death of a pensioner or retrenchment of a wage earner. In this respect, expanding agricultural production can clash with household social reproduction, since deaths and incidents of misfortune both require ceremonies involving animal slaughter. Where off-takes exceeded rates of replacement in raising calves, purchases, bridewealth and other cattle transfers, the farmers faced the prospect of dropping out of farming altogether. However, where CPAs disbursed cattle (heifers) and cash to household members and subsidised household farming to a significant degree, households were able to re-invest in cattle production and reverse the decline in herd sizes, or maintain their cattle herds at more or less constant numbers. Where households received no cattle or subsidies from the CPA, the drop-out rates were striking. In these situations, the farmers most likely to expand cattle production were those with larger herds (accumulated prior to land reform) and were thus able to appropriate a larger share of the common grazing land. In some cases, the differences led to irresolvable conflicts in the CPA.

The example demonstrates that farmers are struggling not only with the demands of production, but also the demands of social reproduction through the distributive practices of ceremonies that require considerable resources. The profitability of CPA cattle enterprises and the distribution of surplus income are at the centre of a range of tensions and conflicts on land reform farms. Abandoning CPAs as a form of land and property holding will not resolve these tensions. Alternatively, proper support would include assisting CPAs to make decisions that enable their members to survive the distributive demands and social reproduction pressures on their farm production.

Conclusion: implications for land reform policy

A legacy of land dispossession combined with chronic unemployment has produced an economy characterised by diverse livelihood strategies spanning country and city, and underpinned by the distributional mechanisms of cash transfers and the dense social and relational networks along which resources circulate. The relationships through which people access resources are mediated by gendered and generational roles and expectations that are both strained and in flux. As marriage rates decline and household composition changes, people’s roles and responsibilities have become more ambiguous, destabilising expectations about who is responsible for the work and risk of farming, and the replacement, consumption and ceremonial costs that households must cover in order for farming to be viable at all. Moreover, those who produce surplus may struggle to ringfence their resources given the social claims made on them. While scholarship increasingly recognises the contradictory dynamics of production and reproduction in shaping the prospects of small-scale farming (Cousins et al. 2018), distributive logics are also central in determining how resources are used. Distribution operates at different levels and includes: the state system of cash transfers and other claims on the state such as through land reform; customary rituals and ceremonies; and a diverse range of informal and piecemeal social claims on the resources of kin and neighbours, including support to urban-based kin for whom the rural home is insurance against the unpredictability of urban life. How these practices will alter in response to the shock of Covid-19 remains to be seen. Emerging evidence suggests that subsistence farming may be increasing as hunger becomes more widespread and job losses lead migrants to return to rural homes, though this may also be destabilising women’s land rights (ten Brink 2022).

Ferguson argues that these distribution-based livelihoods are ‘often starkly disconnected from both agriculture and wage labour’ (Ferguson 2015, 192, italics in the original). However, the evidence we have presented suggests the strongly interconnected nature of productive and distributive activities. For example, the expense of managing social claims may prevent certain productive strategies in agriculture because they are prioritised over replacement costs. Moreover, families containing a wage earner frequently have more success in farming because wages pay for input costs needed by rural-based relatives and mitigate risk. Therefore, it is the interrelationships and trade-offs between the productive, reproductive and social demands on resources that is key to this discussion.

Land reform policy needs to be tailored more closely to these local realities. Yet growing budgetary allocations (in a declining overall budget) to smaller numbers of beneficiaries accessing ever larger farms indicates a different approach. It implies a commercial farmer discourse that to survive market pressures requires ever larger farms, that only full-time, commercial farmers are ‘real’ farmers, and that the implementation mechanisms for targeting land and agricultural services for different categories of beneficiaries either do not exist or have lost political favour. The effect is to exclude poorer and part-time farmers and those at risk of dropping out, to prioritise women and youth only at the level of discourse without practical instruments for such targeting, and to create an environment for elite capture and political patronage. Limited state resources are thus spent on a small number of relatively better off individuals, who are mostly older men.

The ‘Final Report of the Presidential Advisory Panel on Land Reform and Agriculture’ (2019), which is a recent high-level intervention in land reform debates, reasserts the importance of targeting the poor (social grant beneficiaries and those who are food insecure), land constrained households (landless cattle owners, farm dwellers including labour tenants, commonage farmers, smallholders etc), and identifies women as a primary neglected category. However, it retains the notion that the purpose is to ‘graduate’ beneficiaries from ‘social protection to self-reliance’ (18). Thus:

Land Reform products should consider that there is a need to balance the need for large scale farming to maintain output however there is still a need for low input farming which will increase employment opportunities and be a platform to leapfrog small holders into commercial farmers. (54)

The social dynamics described in this article, however, show that successful small-scale farming depends realistically not on independent commercial viability, but on diversified income sources and secure distributional networks. In contrast, ‘self-reliance’ (the absence of pensions in particular) is associated with small farmers dropping out of agricultural production often into extreme poverty. Moreover, the idea of ‘self-reliance’ is antithetical to the distributive logics of local economies in which, as James Ferguson (2015) has argued, relationships of interdependency are integral. Meeting the wider obligations necessary for social reproduction may take precedence over reinvesting in agriculture, calling into question the assumption that expansion – leading to progression through stages of greater self-reliance – is either possible or desirable. Given that the evidence strongly shows that small-scale farms are successful only when combined with income through grants or most especially through waged employment, land reform policy should abandon productivist aspirations toward self-reliant small producers, an idea that has led to ill-conceived and ultimately failed projects.

More usefully, the Presidential Advisory Panel Report argues that ‘the starting point for redistribution must be serious engagement with the nature of demand – who wants what land, where, for what purposes’ (2019, 79). It proposes an area-based, proactive targeted land acquisition programme, which includes public and privately owned land that is well located for its intended purposes and beneficiary target groups. As part of this approach, it is necessary to recognise that not everyone seeking resettlement through land reform has or aspires to have agriculturally-tied livelihoods. Settlement options should be accordingly disaggregated, with agriculturally-tied wage workers having the option to settle in rural development nodes, close to places of employment, where services (housing, water, electrification, cell phone networks, pension pay points) and infrastructure (roads, hospitals, schools) are prioritised (AFRA 2019). These settlements could include sufficient land for large gardens and commonage. Agriculturally-tied small farmers could be prioritised for land redistribution, with proactive land acquisition identifying land close to input and output markets. Farm dwellers with non-agriculturally-tied livelihoods would be disaggregated into those who, for socio-cultural reasons, wish to live out their lives on the farm and those who would prefer to live closer to job opportunities and government services. The former would be given life-long usufruct rights while the latter would be prioritised for housing settlements in towns with growing economies (AFRA 2019). Job growth could focus on commercial farms, alongside prioritising improved pay and conditions of farmworkers, reflecting the reality that food security in South Africa largely depends on commercial farm outputs.

This approach recognises the diversity of livelihoods and uses of land, and the social dynamics that shape farming prospects in rural communities. Without recognising and aligning with these factors, the prospect of a land reform process that genuinely meets the social and economic needs of marginalised communities remains remote.

Acknowledgements

The authors would like to thank Emeritus Professor Ben Cousins for valuable comments on an earlier draft.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by EU/Capacity Building Programme for Employment Promotion (CBPEP).

Notes on contributors

Donna Hornby

Donna Hornby is a senior researcher at the Association for Rural Advancement (AFRA), KwaZulu-Natal, South Africa.

Elizabeth Hull

Elizabeth Hull is a senior lecturer in anthropology at SOAS University of London and Chair of the SOAS Food Studies Centre.

Notes

1 We use the term ‘small farmer’ instead of ‘small-scale farmer’ or ‘smallholder’ to avoid the ambiguities of the latter two terms. For instance, some production systems may be small but intensive, while ‘smallholder’ tends to conflate those farming for subsistence, commercial activity or wage, and hence obscuring class-based differences (Cousins 2010).

2 This article builds on an earlier report by Donna Hornby entitled ‘Thematic study of the social and cultural aspects of small-scale agricultural production in South Africa and the implications for employment-intensive land reform’ (Hornby 2020; available at https://www.cbpep.org/landreform-resources).

3 The ‘homelands’, or Bantustans, were territories set aside by the apartheid government for the purpose of creating ethnically homogenous, independent states for South Africa’s black ethnic groups. Ten were created altogether and were subsequently dismantled following the democratic elections in 1994.

4 These aspects are widely documented in South Africa and beyond (e.g. Guyer and Peters 1984; Berry 1993).

5 The reduction in the number of married adults of reproductive ages is the result of non-marriage rather than widowhood or divorce (Hosegood, McGrath, and Moultrie 2009; Budlender et al. 2011), although there are locational variations.

6 South Africa’s countryside is characterized by complex but distinct tenure contexts, including, state-owned ‘communal’ areas in which traditional authorities and leaders play important roles in land administration; privately owned commercial farms occupied by owners, managers, farm dwellers, labour tenants and farm workers; land reform communal property institutions (CPAs and trusts) owned and sometimes occupied by groups of households or individuals but also sometimes leased out to third parties; and state owned commercial farms leased to land reform beneficiaries for the purposes of undertaking commercial farming.

7 Similar trends are reported in neighbouring Botswana (Kalabamu 2009).

8 As a result of rural labour constraints, small-scale producers in KwaZulu-Natal have urged Government to consider subsidising minimum wages or providing a stipend to assist with labour costs. Current government funding does not assist with labour costs but with the purchase of equipment which can reduce the need for labour, but this support is uneven and erratic, and does not contribute significantly to unemployment reduction (KZNDARD 2018, 43).

9 The combined value of ilobolo transfers and outstanding ilobolo debts on the land reform farms in Hornby’s case study (2015) was in the region of between 550 and 1400 cattle (based on lobolo ranging between six and 15 cattle, the outer ranges found in the case study), with a total estimated value of between R2,7 and R7 million spread across 84 households or R32,000 to R83,000 per household. Bridewealth or ilobolo as a component of the ceremonial fund is thus a fund of considerable value.

10 The Communal Property Associations Act, 1996 is the enabling legislation. The Act was amended in 2017 by the Communal Property Associations Amendment Bill, which is still awaiting the President's signature, and which is on hold following the Presidential Panel Report and the High Level Panel Report recommendations on how to improve the tenure security of unregistered land rights holders.

11 See ‘Final Report of the Presidential Advisory Panel on Land Reform and Agriculture’ (2019) and ‘Report on the High Level Panel on the Assessment of Key Legislation and the Acceleration of Fundamental Change’ (2017).