Introduction

1Women’s land rights feature prominently in contemporary policy debates on agrarian change and gender transformation in Africa, although often accompanied by a certain weary scepticism about what this prominence signifies. On the one hand, policy-makers invoke women as an important category for attention, while on the other hand, analysts and activists regularly denounce the gap between high-level policy commitments and implementation. This exchange has become predictable, conventionalised.

2The 2009 endorsement by African Union (AU) Heads of State of the Framework and Guidelines on Land Policy in Africa offers a recent example of formal recognition at the highest level of the importance of women’s land rights, as both a human rights issue and a developmental imperative for the continent. Clause 2.5.2 of the Framework notes:

[T]he system of patriarchy which dominates social organisation has tended to discriminate against women when it comes to ownership and control of land resources. […] If law and policy are to redress gender imbalances in land holding and use, it is necessary to deconstruct, reconstruct and reconceptualise existing rules of property in land under both customary and statutory law in ways that strengthen women’s access and control while respecting family and other social networks… This is all the more important as women remain the primary users of agricultural land in most African countries.

3The document goes on to state (clause 3.1.4):

Throughout Africa, agricultural production and preservation of land resources is primarily the responsibility of women and children. It is still generally the case, however, that gender discrimination in access to land resources is a serious problem particularly in rural Africa. This is both undemocratic and a constraint on economic development1.

4This Framework and Guidelines document reflects several decades of African research and activism that have nudged the policy debates along. There is now a substantial body of work that confirms not simply the centrality of women’s land rights for land policy reform, but also the diversity and dynamism surrounding women’s land rights in practice, as well as the importance of grounding the analysis of these issues in specific contexts, both national and local. Notwithstanding the socio-economic commonalities and overarching political realities that justify continent-wide statements of principle, the promotion of women’s land rights is poorly served in practice by unproblematised constructions of “Africa” or “sub-Saharan Africa (SSA)” or, indeed, of “women” as unitary targets of attention2. At the same time, one of the more telling criticisms of the academic literature is that it is heavy on critique but light on solutions – it remains easier to expose the gap between policy and implementation than to bridge it. In terms of the contribution of research, it is time for researchers such as myself to bring something more to the familiar exchanges with policy-makers.

5With this as my starting point, I set out key elements of a conceptual framework that can inform the development of more grounded policy interventions around women’s access and rights to land in South Africa specifically. My main point is that effective policy development in this area requires an understanding of both the trajectory of agrarian change and the profound changes and contestations that are reshaping what can be termed the domestic sphere in this country. Ultimately, effective policy development also requires an analysis of the state and its power to implement policy and manage or drive social change, but that is not something I tackle directly here. “Women”, as a general social category, are discriminated against because of their gender. However, acknowledging women’s varied interests in and claims to land, in terms of both agrarian and domestic change, throws up a number of policy challenges. The central challenge is thus to disaggregate “women” and to design policies that not only recognise differences among them, hence the diversity of their interests in land, but also appreciate the relationship between land and non-land issues in advancing women’s well-being and promoting gender equality. A further challenge is to integrate the communal areas (South Africa’s former “bantustans” or “native reserves”) into the mainstream of land reform and agrarian policy, and to expand the concern with women’s land rights in these areas beyond the ghetto of “custom” and “tradition”.

6These challenges have to be understood in the very particular context of South Africa, which differs significantly from other countries in sub-Saharan Africa. Issues to consider here include the country’s particular history of colonisation and settlement which led, in the second half of the 20th century, to the abomination of apartheid, as well as the extent and form of contemporary urbanisation and the relatively small part played by agriculture as a sector in the national economy. These features of contemporary society are profoundly gendered, not simply in their differentiated impacts on men and women but in the manner in which they have been shaped historically by the social organisation of gender relations in the region. Here I am understanding “gender” in the way that R.W. Connell has defined it, as deriving from, but not confined to, the ordering of social practice “in relation to a reproductive arena… [that] includes sexual arousal and intercourse, childbirth and infant care, bodily sex difference and similarity” (Connell 1995, 34).

7Rethinking the framing of women’s land rights in South Africa is particularly timely, given the stated intention of the African National Congress (ANC) government to overhaul its 1997 policy framework, the White paper on South African land policy (DLA 1997). There is widespread concern across the political spectrum at the failure of the post-1994 land reform programme to deliver on its promises. This is compounded by scepticism at the capacity of the recently restructured Department of Rural Development and Land Reform (DRDLR) in providing new policy ideas and implementation capacity3. This extends to its capacity to design and manage a seriously gender-sensitive programme. In this regard, it is worrying that an unofficial copy of the DRDLR’s draft Green paper (in circulation since 2010) is depressingly silent on women’s land rights and gender equality, apart from a number of textual flourishes that academics such as myself love to criticise; these relegate women to an undifferentiated subset of an amorphous rag-bag of “vulnerable groups” comprising, at its most extensive, “rural people, women, youth, the unemployed, people living with HIV/AIDS, people living with disabilities, child-headed households and older persons” (DRDLR 2010, 47). In short: the majority of the population!

8While my focus is South Africa, the debates on women’s land rights in this country do have wider relevance for policy development in Africa, if only because they underscore the importance of locating these issues in a broader analysis of social and economic change and not assuming a one-size-fits-all approach. My attempt to de-centre the relative importance of land issues in the South African debate on gender equality in no way implies that this is true for the rest of the continent – on the contrary. Dzodzi Tsikata has recently drawn attention to “the importance of situating analysis [of gendered livelihoods] within the wider political economy” (Tsikata 2009, 12) and this is certainly a critical requirement. But equally critical – although less well theorised, certainly in the South African debates – are the changes and contestations taking place in what I am calling the domestic sphere: that highly charged sphere of affective relationships, biological and social reproduction, care, consumption and production that is organised through families and households. Inasmuch as this sphere is pre-eminently concerned with the ordering of the “reproductive arena” identified by Connell, it can be seen as the crucible of gender relations and gendered identities within society. It cannot, of course, be understood in isolation from the wider political economy, but equally, its dynamics cannot be simply subsumed – assumed – within that.

9I begin my discussion with a brief background overview of the current land reform programme in South Africa. Thereafter I address the question of agrarian change and the significance of the domestic sphere, and then conclude with a brief discussion of the policy implications.

Land reform in South Africa

10The basic elements of the post-apartheid land reform programme were thrashed out in the constitutional negotiations between the apartheid government and the ANC in the early 1990s, which made possible the transition to formal democracy in 1994. This resulted in a core compromise between, on the one hand, constitutional protection for existing property rights (at the time, essentially those of the white minority) and, on the other, a constitutional commitment to land reform “to bring about equitable access to all South Africa’s natural resources”. While the principle of gender equality is implicit in the provisions of the resulting “property clause” of the South African Constitution4, the primary concern of this section of the Constitution has always been to redress the deeply racialised inequalities in access, use and control over land that is rooted in South Africa’s history. Clearly, past racial discrimination has affected both black women and black men, but the specification of “race” in the property clause obscures the significance of gender inequality as a general, rather than an ancillary, concern with regard to land rights. It also obscures the significance of the intersection of race and gender discrimination for black women’s social, legal and economic status historically.

11Within the framework provided by first the 1993 and then the 1996 Constitutions, post-apartheid land reform was designed as a three-part programme:

• Land restitution, designed to compensate the victims of racially motivated land dispossession since 19135, either through the restoration of their land or through financial compensation or other forms of redress.

• Land redistribution, targeting commercial agricultural areas, to deliver land to the landless and land-hungry as well as to black aspirant farmers; this sub-programme has aimed to transfer 30% of commercial agricultural land – approximately 25 000 hectares – into black ownership, at first within five years of democracy, i.e. 1999, but later rescheduled to 2014.

• Tenure reform, aimed chiefly at farm dwellers and people living in the former Bantustans or communal areas and intended to “improve the tenure security of all South Africans and to accommodate diverse forms of land tenure, including […] communal tenure” (DLA 1997, v).

12Because of the political authority accorded to the history of black land dispossession in South Africa (described below), the primary focus since 1994 has been on land restitution and land redistribution in the commercial agriculture sector, at the expense of land reform in the communal areas. In part, this is because in these areas, land reform – here collapsed into a difficult and unresolved programme of tenure reform – fails to reduce the aggregate inequalities in white and black land ownership nationally. The struggle by “traditional leadership institutions” (whose reconstitution is based on the idea, if not necessarily the practice, of hereditary chieftainships and “tribal” councils) to entrench their authority over communal land in the name of an unchanging and deeply patriarchal “African culture” is also a factor.

13By 2010, after 15 years of effort, the state’s official land reform programme had fallen far short of its initial redistribution targets in terms of the number of hectares transferred. By then a little over a quarter of the redistribution target had been achieved, with some 7 million hectares6 reportedly transferred into black ownership through all components of land reform (DRDLR 2010, 20). The goals of “development” and “poverty reduction” had also proved elusive. Serious problems with both governance and production on many (but not all) redistributed projects have led to several policy reformulations for the redistribution sub-programme since the late 1990s, broadly involving more emphasis on deracialising rather than fundamentally transforming commercial agriculture, and directing more support towards “emerging” (i.e. black) farmers and production-oriented business plans for projects. However, although it is possible to find individual success stories, the direct economic benefits to beneficiaries have continued to be very disappointing overall, as acknowledged by the new Minister of Rural Development and Land Affairs in early 20107.

14In terms of restitution, by early 2011 around 4,000 “complex” rural land claims were still waiting to be settled. While this means that the great majority of the approximately 80,000 land claims that were lodged by the cut-off date (end of 1998) have been formally settled, the livelihood impacts of this sub-programme are difficult to measure. The contribution of the restitution programme to the national goals of social justice, redress and reconciliation are also very difficult to quantify, but can at best be summarised as “mixed”8. Most of the claims that have been settled to date are urban and have involved the payment of financial compensation – not necessarily a negative outcome for the sub-programme, as I have argued elsewhere, but not making as significant a dent on racially-based land ownership and spatial patterns as the raw numbers might suggest. In the land-based restitution settlements, furthermore, many of the problems of leadership conflicts and development outcomes that are faced by redistribution projects are also visible. And in the communal areas, the state’s attempt to turn insecure “old-order” land rights into more secure “new-order” rights, under the authority of purportedly reformed traditional councils, floundered to a standstill in mid 2010, with the striking down by the Constitutional Court of the Communal land rights act of 2004 as unconstitutional, in part because it was found to be discriminatory against women.

15The DRDLR’s draft 2010 policy formulations was flagged as an attempt to revitalise the land reform programme but it does not appear to shift the conceptual boundaries significantly; its status (as of early 2011) is also unclear as it has not been officially released for comment and there are suggestions it is still being subjected to internal revision. The draft document proposes a hierarchy of land-based livelihoods that land reform should serve. The hierarchy starts with a “homestead” category, comprising an estimated one million households who “are not regarded as farmers but require land for shelter, ploughing and grazing to meet their basic household needs”. The majority of this category are presumably in the communal areas, where, it is proposed in another section of the document, the state should retain underlying ownership of the land. The hierarchy then moves up through small-scale and medium-scale commercial farming to large-scale commercial farmers, the latter described as consisting of “relatively established black farmers who are already entrepreneurs”. This group, together with medium-scale farmers, is described as “stand[ing] a fighting chance to compliment [sic] the country’s ageing cadre of established commercial farmers in providing food security for the country”(DRDLR 2010, 114-115). In so doing the document makes explicit the importance the state attaches to national food security (embracing, then, the provisioning of the economically and politically dominant urban sector) as a major objective that “pro-poor” land reform not only cannot secure, but should not disrupt.

16With regard to women’s rights, it is hardly surprising, given the problems described above, that South Africa’s post-apartheid land reform programme has not been an effective instrument for the delivery, at scale, of either secure land rights or improved livelihoods to the women who need it most. Compounding the general problems is the way in which first the DLA, and now the DRDLR, have regarded “women” as a discrete special-interest group, rather than full (and diverse) members of its core client base. In the first phase of land reform (1994-1999), the DLA’s main strategies for including women were to set quotas for women’s representation on community land reform committees (generally 30%) and to target female-headed households for inclusion in beneficiary lists9. After 2000, the DLA continued with the strategy of quotas for women’s participation, but directed this more clearly at the class of emerging black farmers that it was by then championing. Here it had more success, but on a very limited scale in terms of the numbers reached and the involvement of women from the poorest sectors of rural society. While reliable numbers are hard to come by, I have estimated that by 2005 “several hundred thousand women would have probably benefited from the official programme [of land reform] as recipients of land or cash, both in their own right and, more commonly, as members of beneficiary households and communities receiving land” (Walker 2009, 488).

17In summary, then, the land reform programme that has been put in place has not served women well – but nor has it served the majority of men particularly well either. There are several dimensions to this, but a critical weakness, I argue, has been the conflation of land reform with aggregate black ownership of the formerly white countryside. The primary goal of the post-apartheid programme has been to deracialise land ownership in the commercial farming areas, without a sufficient appreciation of both class and gender differentials among black South Africans, and without a sufficiently critical engagement with the challenges of land-based livelihoods for poor rural people, including, more particularly, poor women. These issues are developed further in the following sections.

The trajectory of agrarian change

18Most discussions of the land reform programme in South Africa begin with a highly formulaic history of land ownership over the past 350 years, a narrative that is seen as offering an almost self-explanatory rationale for the programme. Central to this history are the radical land dispossessions of indigenous people as a result of European colonisation from the mid 17th century onwards and, much later, the development of a migrant labour system in the wake of the mineral discoveries of the late 19th century, the rapid expansion of capitalist relations of production across the region from this time, and the consolidation of white political and economic power in the form of “apartheid” after 1948. The apartheid regime saw the division of the country into a notionally “white” area comprising 87% of the land, all the major urban centres and most of the country’s wealth, and a black periphery that was divided into ten ethnic homelands or “bantustans”. In their final form these enclaves were planned to comprise just 13% of the land. The attempt to enforce this spatial dispensation involved the mass uprooting and relocation of several million black people, alongside major controls="true" on black urbanisation to the “white” core; this “displaced urbanisation” meant that the proportion of black people resident in the supposedly “rural” bantustans was in the region of a third of the population by the mid 1990s, many living not in farming villages but in dense though under-serviced “closer settlements” (for an overview, see Platzky and Walker 1985). Maintaining this system also involved increased levels of state repression as political resistance became more militant, both internally and externally, from the 1960s.

19All of this has left very deep scars on both the economic structure and the social fabric of post-apartheid South Africa. But, as I have argued at greater length elsewhere, while these developments have been profoundly important in shaping the country’s history and its contemporary discontents, the essentially political “master narrative of loss and restoration” that has been distilled from them fails to account for the full complexity of contemporary South Africa, and to come to terms with the challenges facing a state-sponsored land reform programme in the 21st century (Walker 2008). The social forces shaping post-apartheid South Africa are far more dynamic than this “headline history” allows. South Africa’s colonial history involves not simply loss and dispossession for black people, but also the emergence of new classes, new identities and new ideas about land, property and livelihoods among black (and white) people. It is also no longer appropriate, if ever it was, to think about land reform in terms of the conceptual silos laid down by the spatial and administrative categories of apartheid: on the one hand, a “white” commercial sector in which the core work of land reform, i.e. land restitution and land redistribution, takes place, and on the other hand, a residual “black” communal sector in the former homelands, in which the less urgent business of tenure reform for “traditional” land systems takes place.

20Furthermore, alongside the serious and all-too-familiar problems of inequality, poverty, dispossession, discrimination, anomie (and the list of negatives could roll on), as well as the equally serious but generally less well recognised ecological challenges facing the country, new social dynamics in the post-apartheid era – new opportunities, if you will – urge new ways of thinking. A more robust land reform programme than what is currently in place needs to understand the new social forms and economic relationships that constantly diminish the analytical usefulness of the old apartheid dichotomies. Such dichotomies do not account for the new constellations of social identity, sociality, popular culture, political mobilisation, consumerism, information networks and economic dependencies that are shaping new expectations around livelihoods and relationships, including between men and women.

21A key issue here is that South Africa is not, and has not been for some time, a predominantly agrarian society, neither economically nor, equally important, socially in terms of the way that the majority of people think about themselves and their aspirations for their children. Unlike in the majority of African countries, agriculture, both commercial and house-hold-focused, is not the dominant sector in the national economy. At 2.3% in 2009, the contribution of agriculture to the country’s gross domestic product (GDP) is not insignificant but is relatively small compared to that of financial services, manufacturing, commerce, transport and mining (Statistics South Africa 2009). This contrasts markedly with the situation in most of sub-Saharan Africa, where the contribution of agriculture to national GDP is 25% or more (AU, ADB and ECA 2009). Linked to this, today some 60% of South Africa’s population (or more, depending on the definition) is classified as urban (Statistics South Africa 2003), half of which is concentrated in the ten metropolitan areas of the country (Urban Sector Network 2003). Here the major battles over land are not for a plot on which to grow food, raise cattle and produce crops for the market, but for a place on which to build a shelter and secure access to basic services and job opportunities, whether these can be found in the formal sector (the first prize) or the informal sector, the default reality for many. While the factors driving urbanisation and its regional patterns are complex, there is no indication that the overall trend towards increased levels of urbanisation is in decline.

22Also significant are the ecological challenges facing the country as a whole and the agricultural sector in particular. Much of South Africa’s commercial farm land is not arable and the rural population is heavily concentrated in the better-endowed regions of the eastern half of the country which encompasses both commercial farming and communal areas. Over a third (36%) of historically white commercial farmland, the current focus of land redistribution, is located in the dry Northern Cape where only 2% of the population lives, while 21% of the population (just over half of which was classified as “rural” in 2003) is located in KwaZulu-Natal, one of the smallest provinces in terms of area (just under 7% of the national total) but one which is relatively well-endowed in terms of climate and agricultural land10. Biophysical externalities continue to impact on land-based livelihoods in ways that cannot be reduced to the legacy of apartheid. Climate change is expected to increase volatility around rainfall patterns, with one scenario forecasting less rainfall in the west of the country and more unpredictable patterns in the east, with major implications for future land-based livelihoods and settlement patterns (Jaarsveld and Chown 2001, 13).

23Of course the relative weight of the non-agrarian economy does not mean that agriculture is unimportant in South Africa, or that land reform should only concentrate on urban land issues and needs. There are significant linkages between agriculture and other sectors of the economy, including manufacturing, energy and transport. The commercial farming sector is a major source of employment, although there has been a substantial decline in the agricultural labour force since the 1950s and levels of remuneration in this sector are very low. Agriculture continues to make a significant, albeit declining, contribution to household livelihood strategies in the communal areas. According to one recent study, between 2000 and 2006, “the number of black people practising agriculture at some scale increased from 3.5 million to 4.5 million, generally keeping pace with rural population growth”, with most of this increase assumed to be located in the communal areas; the majority of those involved in farming were women, with “approximately 50% more black women engaged in agriculture than men” (Aliber, Baiphethi and Jacobs 2009, 138). However, this study also notes how, during this time, “there was a conspicuous move away from producing as a main source of food towards producing as an extra source of food”, which the authors attributed primarily to the “evergreater penetration of social grants” in these areas (136). Furthermore:

On average, households that engage in agriculture as a main source of food are poorer than those that practice agriculture for some other reason. In addition, the shift from agriculture as a main source of food to some other reason is typically associated with an increase in income. (Aliber, Baiphethi and Jacobs 2009, 137)

24While the urban sector strongly dominates popular culture, it does not mean that land-based livelihoods and forms of sociability are not still significant in many parts of the country. What we have is, as I have described it elsewhere,“urbanisation with a strong rural alloy”:

Many people still live in nominally rural areas and many people living in town retain strong links with rural localities, through extended family networks and histories of migration and dispossession. Today, […] these linkages are acquiring new dimensions as a result of land reform itself. (Walker 2010, 2)

25At the same time, land is infused with social meanings that cannot be reduced to economic utility alone. In the words of Kepe, Hall and Cousins,

land carries a powerful symbolic charge for many black South Africans not only because of their recent memories of racialised dispossession of their land, but also because inequalities in land ownership “stand for” and evoke the broader inequalities that post-apartheid policies have yet to undo. (Kepe, Hall and Cousins 2008, 145)

26Nevertheless, despite the strong political resonance of land reform’s “master narrative”, the extent of actual land demand among poor black rural South Africans appears to be quite modest overall, with land widely regarded as but one element in a larger process of income generation that straddles farm and off-farm activities. A 2005 study of land demand in three provinces found that most respondents were interested in relatively small amounts of land (five hectares or less), while those living in the communal areas were far less likely to indicate a (hypothetical) willingness to relocate in order to obtain land than farm dwellers on commercial farms (Aliber, Roefs and Reitzes 2006, 15). Clearly, these dynamics reflect a complex interplay of factors, including the absence of resources to invest in land, along with respondents’ expectations of what is possible. However, they also confirm the relative rather than the absolute contribution of land in people’s considerations around household and individual wellbeing, including in the rural areas. Furthermore, my own work on land restitution suggests that the authority of the national symbolism of land in shaping actual choices at individual or household level fluctuates in relation to numerous other considerations, including people’s gender, age, locality and employment prospects.

Gender relations and the domestic sphere

27Designing a land reform programme that engages effectively with all these considerations is clearly a challenge. A land reform programme that is serious about empowering women must also be responsive to the significant but uneven processes of social change that are reshaping the domestic sphere. Gender relations in contemporary South Africa, it is well documented, are highly volatile. Change and instability in this domain are not new but they are certainly intense in the current era, with established norms about male and female identities, roles and responsibilities under considerable pressure as a consequence. The conventional history of dispossession that dominates the land reform debate does not begin to address the unstable intersections of continuity, contestation and change that are redefining social relationships between men and women, across generations, within families, households and communities; these developments have major implications for domestic struggles over resources such as land and housing. Of course South Africa’s history of colonisation, migrant labour and apartheid has had a profound effect on the “ordering of social practice” in the reproductive arena, including the regulation of sexual relations and fertility and the allocation of responsibility for childcare in society. But other, more contemporary dynamics also need to be factored into the analysis, including new technologies of contraception, the impact of mass schooling and the mass media, the influence of new discourses of equality and individualism, especially post 1994, and the spread of HIV/AIDS.

28It is not possible in such a short space to do much more than signal some of the dimensions of change. Between 1960 and 2008, the fertility rate in South Africa decreased from over 6 to about 2.5, one of the lowest, if not the lowest, in Africa11. While there are important regional differences which need to be understood and taken into account in designing policy interventions at the local level, nationally marriage rates have been in decline since about 1960 (Budlender and Mhongo 2010). At the same time the average size of households is decreasing, down from an average of 4.5 persons in 1998 to 4 in 2003 at the national level (Walker, Aliber and Nkosi 2008, 28; 46), but still larger in rural areas than in urban settings. These developments are reflected in an increase in the absolute number of households in South Africa that is disproportionately greater than the actual rate of population growth, and this process of “nucleation” is fuelling the demand for housing and services. At the same time, the number of house-holds classified as “female-headed” is growing, signalling a complex set of changes in family, household and gender relations. By 2004, 37% of all households in South Africa were reported to be headed by women, with the proportion in areas classified as rural even higher, at 44% (Walker, Aliber and Nkosi 2008, 28; 20). In addition to what might be seen as positive indicators of change in terms of women’s autonomy and independence, there are also negative indicators of struggle and change in the domestic sphere. South Africa has extremely high levels of gender-based violence. The HIV/AIDS pandemic is putting heavy pressure on household stability and composition, as well as impacting negatively on the tenure security and livelihoods of affected households (for a general discussion on these issues, see Swaminathan, Kes and Ashburn 2008). Currently the prevalence rate appears to have stabilised at about 10.6% of the total population, which is still extraordinarily high and is disproportionately affecting women12.

29These issues are leading to major shifts in how households are organised and children raised. A study I was involved in, in the early 2000s, on women’s property rights, domestic violence and HIV/AIDS in the province of KwaZulu-Natal found high levels of ambivalence among women towards marriage and intimate relationships with men (Walker, Aliber and Nkosi 2008, 28; 47). In contrast to patrilineal social norms, the primary domestic network for the majority of the women we interviewed was natal rather than marital, with the responsibilities for household reproduction and childcare falling increasingly on women living in matrifocal households, often in unstable or non-existent relationships with the fathers of their children (Walker, Aliber and Nkosi 2008, 28; 54)13. House-holds were also dynamic in their composition, with considerable movement in and out of the place of residence by members, and complex links stretching over more than one place of residence. Similar developments in other parts of the country have seen an increase in demand for independent rights in land in the communal areas by rural women, which is being accommodated to varying degrees at community level (Claassens and Mnisi 2009). In the study on land demand, cited above, revealing gender differences emerged – “demand in rural areas was higher among males than among females, while the reverse was found in urban areas” (Aliber, Roefs and Reitzes 2006, 8). Concerns around tenure security and interest in land for household food production were expressed particularly strongly by the women respondents.

Policy implications

30What, then, are the implications of agrarian change and gender transformation for gender-sensitive land reform and policy development in South Africa? In my introduction I have emphasised the importance of disaggregating women and designing policies that recognise class and other differences among them, and cater for the diversity of women’s interests in land. I have also pointed to the importance of non-land issues in promoting women’s well-being and advancing gender equality. Linked to this is the need to address land reform in the communal areas as a central, not marginal, concern, in which the concern with strengthening women’s land rights is extended beyond a narrow focus on tenure reform and its current ghettoisation in the debate on authenticity and plurality in culture and custom.

31In my concluding comments I pick up on the first two issues; the third, as important as it is, is a subject that requires a paper in its own right.

Disaggregating “women”

32In the policy literature on gender and land reform, “women” are generally presented as a unitary category, with “gender” applied as an unquestioned proxy for “women”. This is very evident in the DRDLR’s draft Green paper described above, where “women” are discussed as an undifferentiated subset of an amorphous category of “the vulnerable”. In the process, important differences among women disappear; at the same time, the adversarial relationships between men and women tend to be emphasised, over and above the rather more ambiguously patterned relationships of dependence, reciprocity, care and solidarity that also characterise relationships between men and women within households and within communities. In South Africa, in the current context of land reform, racial solidarity among men and women is particularly evident at the macro level of national political debate, where black and white South Africans respond to the weighty symbolism of the land reform programme in terms of their very different collective histories and expectations of race-based redistribution. However, as one moves to the micro level of social interaction, within communities and especially within households, the issue of “race” is likely to be externalised beyond the primary group and gender dynamics are more likely to move to the fore, although always in context-specific articulation with other considerations such as class, status and age.

33The elision of differences among women in policy formulation is taking place notwithstanding a vigorous history of critique in South Africa of the assumption that all women necessarily share common interests because of their gender. Today academic analyses of women and gender emphasise difference, stratification and cultural variation as a matter of course. Yet despite this, and despite the regular acknowledgement that men’s interests in land (and other resources) are also gendered, the slide from “gender” to “women” and back again persists in much of the post-1994 policy literature, including that on land14. In part this is because the divisions of race in relation to land are taken so much for granted in local debates, that there is an unstated assumption that when one is discussing women in relation to land one can obviously only be talking about black women.

34Yet women are stratified not only by “race” and ethnicity, but also by class and, within the domestic sphere, by other markers of social difference, including age, marital status, sexuality, position in the family, and the nature of the household as a domestic unit. The women who are able to put themselves forward to the DLA and DRDLR as “emerging farmers” are likely to be better educated and to come from better-resourced backgrounds than the average rural woman. Within patrilineal house-holds, older women who are the mothers of sons are likely to stand in a very different relationship to household assets, including land, from the women that these sons bring into their households as wives, or with whom they have children (women who may themselves be living in matrifocal households and pressing the sons – the fathers of their children – for maintenance)15. The location of women along the rural-urban spectrum is also very significant, not only in terms of their access to land but also for their expectations of what they can and should do with that land.

The importance of other social policy interventions for successful land reform

35Internationally a broad case for stronger land rights for women is often framed in terms of the fourfold justification that Bina Agarwal has developed: namely, considerations of welfare, efficiency, equality and empowerment (Agarwal 1994; 2003). However, as I have argued more fully elsewhere, the causal relationship between stronger land rights and enhanced wellbeing is not a simple and unmediated one. In the South African context, land on its own is certainly not a guarantee of economic benefits nor enhanced livelihoods, as my discussion above has illustrated. Thus, while a compelling case can be made for gender equality in land rights in the abstract, giving content to the general principle requires a more careful and context-specific analysis. As Cecile Jackson has argued, while “gendering land rights is critical”, this cannot be “reducible to an argument for land rights for women” (Jackson 2003, 453).

36More work is clearly needed to theorise women’s diverse interests in land and to tease out the policy implications of this. In this theoretical work, the term “gender” should be seen as an important but insufficient conceptual tool for understanding the range of issues that need to be addressed. Understanding the trajectory of agrarian change as well as the significance of the social changes impacting on the domestic sphere is also critical. What these considerations suggest is that the struggle to engender the constitutional imperative to “bring about equitable access”16 to land for all South Africans needs to be grounded in a robust understanding of the particular environments in which women’s rights in land are exercised in practice, which are not everywhere the same. The content of substantive gender equality in land rights in practice is dynamic and context-specific, and cannot simply be read off or deduced from general legal principles and high-level policy commitments. It certainly cannot be reduced to a responsibility of the current land reform programme alone.

37Substantiating land rights for women thus turns out to have a number of dimensions. My analysis confirms the fundamental importance of harnessing the constitutional guarantee of the principle of gender equality to the interpretation of the reach and scope of the property clause – to engender the commitment to land reform for the landless and tenure-insecure, as well as for those unjustly dispossessed under apartheid. It also points to the need to develop a policy framework that is responsive to the variety of local conditions in which land rights and land reform are to operate – that supports a suite of tenure and land use options that can address the full range of women’s differently configured, gender-inflected issues and needs. This includes the search for more secure settlement and housing opportunities for women in both urban and rural areas, operating at different scales of affordability and flexibility, in order to incorporate greater responsiveness to changing household membership and financial circumstances as these unfold. Possibilities here include more attention to affordable rental housing as a policy option and more imaginative and flexible public housing design. Women’s role in smallholder and other farming enterprises in more rural and peri-urban contexts has also to be fore-grounded, along with making explicit women’s rights in common property resources where these are at stake. In all these contexts provision should be made for securing women’s rights in different capacities, so that women have options in terms of securing their claims through individual rights, or through joint family and group rights. Here too, blanket prescriptions appear incapable of responding satisfactorily to the plurality of circumstances and preferences at hand.

38At the same time, what I have also tried to show is that the issue of substantive gender equality in relation to land cannot be separated from women’s struggles for other socio-economic rights. Access to land, I have argued, is not an absolute and unmediated pre-requisite for well-being. Nor can land rights be exercised in a vacuum. In contemporary South Africa land reform that puts women at the centre has to go hand in hand with interventions in other areas of social policy, such as healthcare and education. My analysis points to the importance of linking the roll-out of women’s land rights to complementary policy interventions that go beyond the domain of land per se – for instance, ensuring that HIV-positive women are able to access antiretrovirals and tackling the scourge of gender-based violence. Investment in education to open up more choices for women (and men) beyond land-based livelihoods is also a critical issue. Such complementary interventions are necessary to ensure not only that women are in a stronger position to claim new land rights but that they can make the most of the rights in land (and property more broadly) that they already hold or come to acquire.

39In certain contexts such interventions may well require prioritisation ahead of sector-specific interventions around land.