Livelihoods in the wake of agricultural commercialisation in South Africa's poverty nodes: insights from small-scale irrigation schemes in Limpopo Province

Abstract

Small-scale irrigation farming is envisaged to play a progressively larger role in rural development and to help reduce some inequalities in South Africa's space economy. Since the late 1990s, the government has aimed to ‘revitalise’ government-owned small-scale irrigation schemes, many located in former homelands. Its macro-policy shifts seem to favour the creation of a black farming elite. Important questions are whether neoliberal policies will harm the poorest and most vulnerable in irrigation farming communities, and whether a new class of petty commodity producers can establish themselves in global commodity chains. This paper looks at vulnerability and marginalisation in selected small-scale irrigation schemes in Limpopo Province. The findings suggest that existing approaches to agricultural commercialisation may not reduce rural poverty and inequality. Although these approaches help to integrate resource-poor irrigation farmers into globalised commodity production sectors, they could undermine the livelihoods of the poorest and most vulnerable in these communities.

1. INTRODUCTION

This paper discusses the way the South African government-led drive towards agricultural commercialisation has affected the livelihoods of people living in small-scale irrigation schemes in Limpopo Province. It aims to determine whether prevailing approaches to agricultural commercialisation in this province can realistically be expected to achieve the intended outcomes of livelihood sustainability for irrigation farming communities in poverty nodes identified under the South African government's Integrated Sustainable Rural Development Programme (ISRDP), a programme launched in July 2000 as a renewed attempt to improve opportunities and well-being for the rural poor.

The main objective of the paper is to examine how some factors associated with agricultural commercialisation have enhanced or detracted from the livelihoods of the rural poor living in irrigation schemes; and in particular the way institutions have responded to the interests of people living in irrigation communities, on the one hand, and those of stakeholders outside these communities, on the other.

The paper begins by setting out the context of the research problem. This is followed by an overview of the location, historical background and characteristics of the three study sites. The research methodology is then described. The next section discusses the findings and provides a comparative account of the livelihood and coping strategies of individuals and households observed prior to, during and after initiatives towards agricultural commercialisation, followed by a discussion of key issues. Finally, conclusions are presented.

2. CONTEXT AND POLICY

Irrigation farming is widely seen as having a significant potential to enhance rural development, reduce poverty and increase small-scale farmers' productivity, employment and incomes (Norton, 2004). A prevailing view in South African government circles is that increased involvement in commercial agriculture by resource-poor black irrigation farmers, mostly located in the marginal former homeland areas, is a viable strategy for reducing rural poverty, inequality, food insecurity and unemployment. Agricultural commercialisation is also seen as contributing, through multiplier effects, to local economic growth and rural development.

According to the Provincial Growth and Development Strategy, solving the problems of poverty, unemployment, food insecurity and rural development in Limpopo Province will depend largely on investment and growth in the agricultural sector. Consequently, the Limpopo Provincial Department of Agriculture has set aside R224 million to fund the Revitalisation of Smallholder Irrigation Schemes (RESIS) programme, which focuses on the 126 existing irrigation schemes and aims at ‘re-building socially uplifting [and] profitable agribusiness’ through ‘a comprehensive programme to structure, train and capacitate smallholder farmers to run their scheme profitably and sustainably’ (De Lange, 2004).

Many of the irrigation schemes in Limpopo Province were developed after the publication in 1955 of the Tomlinson Commission report on ‘Socio-economic Development of the Bantu Areas within the Union of South Africa’. The report recommended, among other things, that irrigated holdings of between 1.3 and 1.7 ha could adequately ‘provide a family with a living that would satisfy them’ (Perret, 2001). The schemes became dysfunctional after state subsidies were suddenly withdrawn by the post-apartheid government after 1994. This appears to have deepened food and livelihood insecurity for rural communities in irrigation schemes. While the RESIS programme might therefore be construed as an attempt to reverse the adverse consequences of government decisions, it is also true that the programme is in response to an international drive to enhance water use efficiency and reduce the transaction costs of operating state-sponsored irrigation schemes, by transferring irrigation management to farmers and implementing agricultural commercialisation.

In addition to the RESIS programme, there are also moves by national, provincial and local government structures in Limpopo Province to revive commercial production in government-owned small-scale irrigation schemes that were historically used as white farmer settlement schemes and have since been allocated to black farmers under the land reform programme. This paper explores issues of livelihoods and agricultural commercialisation in both types of irrigation scheme, with specific focus on ‘poverty nodes’ identified under the ISRDP.

Moves to revive and revitalise commercial small-scale irrigation farming offer a promise of improved livelihoods and well-being for some among the rural poor in Limpopo Province. This paper is concerned, however, that such moves are concomitant with macro-economic policy shifts away from the focused anti-poverty strategies of the Reconstruction and Development Programme towards a hybrid emphasis on both poverty reduction and national economic goals espoused by the Growth Employment and Redistribution strategy. The moves are also concomitant with reforms in agriculture, water and other sectors, which are predicated on principles of neoliberal economics aimed at reducing transaction costs. Key questions therefore are whether farmers will be able to pay for irrigation services and establish a viable niche in highly competitive global commodity chains and whether neoliberal policies will have negative impacts on the livelihoods of the poorest and most vulnerable in small-scale irrigation farming communities. These questions spring from perceptions that although the South African government has demonstrated an intention to balance economic growth and development with social security, structural conditions underpinning reforms in agriculture, land and water sectors seem to favour the creation of a black farming elite.

The Agricultural Sector Strategy (Republic of South Africa, 2001) lists core objectives for supporting emerging farmers: more equitable access and participation in the agricultural sector, improved global competitiveness and profitability, sustainable resource management and food security. A key strategy for helping emerging irrigation farmers enter commodity production is the broad-based black economic empowerment framework for agriculture (Agri-BEE), within which joint ventures play a major role. These ventures, involving private investors and producers and supported by government agencies, are seen as a way to provide emerging black farmers with the requisite financial resources, technology, technical and managerial skills and access to markets. The problem is that the vision of a significant role for black petty-commodity producers involved in irrigation farming seems rather ambitious when viewed against the following factors.

Firstly, the majority of black arable farmers in South Africa practise commercial rainfed cropping rather than irrigation, and subsistence rather than commercial farming (Perret, 2001). Secondly, black farmers face significant constraints to effective participation in the highly competitive and globalised commodity production sectors owing to their lack of financial resources, technology, technical and managerial skills and access to markets, and therefore their requirement for significant investments in farmer support. Expectations that such support will be provided through joint ventures with the corporate sector must be underpinned by effective institutional mechanisms to ensure that the partnerships achieve the desired benefits for black farmers. Thirdly, the proportion of black farmers who can benefit from agricultural commercialisation is constrained by water scarcity. While the relatively dry and drought-prone climate of Limpopo Province is widely acknowledged to be responsible for environmental water scarcity, it is the structurally induced water scarcity that preoccupies ongoing debates in the South African water sector. White commercial farmers continue to command a significantly larger share of irrigation water and infrastructure in the ‘water-stressed’ Olifants River Basin (Tren & Schur, 2000; Ligthelm, 2001) where this paper's study sites are located. While the South African Government's Financial Assistance Policy to Resource Poor Irrigation Farmers is a welcome move to improve access to water by black small-scale irrigation farmers, the gradualist approach adopted by the government's Water Allocation Reform programme limits prospects for redistribution of water at a critical phase of entry by black farmers into commodity production.

Other constraints to effective participation by black farmers emanate from world market conditionalities. Trade liberalisation policies, such as the World Trade Organization's Agreement on Agriculture, compel developing countries to phase out subsidies, exchange controls and trade barriers without imposing the same conditions in countries in the north (Jacobs, 2001). Sanitary and phyto-sanitary measures, as well as the provisions of the European Union's Common Agricultural Policy, are used as non-trade barriers to protect producers in the north, thereby making it difficult for producers in developing countries to compete (Madonsela, 2001, in Jacobs, 2001).

Prospects for black irrigation farmers establishing and retaining a niche in globalised commodity production sectors therefore remain bleak. Without a radical transformation of the structural factors, South African government interventions will achieve limited gains for the rural poor in small-scale irrigation schemes.

3. STUDY SITES

Limpopo Province is situated in the northeastern part of South Africa and is among the three least developed provinces in the country. Its predominantly rural socio-economic landscape glaringly reflects the inequalities inherent over much of South Africa's space economy. This section provides a comparative overview of salient characteristics of the three selected case study sites.

3.1 Location

Two of the study sites are in Greater Sekhukhune District (Table 1), which is one of the initial 13 poverty nodes identified and gazetted by the ISRDP in 2000. The third is in Vhembe District (Table 1), one of the 17 nodes identified in September 2003 but still to be gazetted. All the schemes are in the Olifants River Basin.

Table 1:  Characteristics of study sites, 2004–2007

	Hereford Irrigation Scheme	Phetwane Irrigation Scheme	Makuleke Irrigation Scheme
Location
District	Greater Sekhukhune	Greater Sekhukhune	Vhembe
Local authority	Greater Groblersdal	Greater Marble Hall	Thulamela
Area (ha)	192	53	215
Number of plots	34	49	46 commercial plots
			273 food plots
			One greenhouse plot consolidated from four food plots
Average plot size	5 ha (plot sizes range from 0.7 to 14 ha)	1 ha (plot sizes range from 0.5 to 2 ha)	Commercial plots: 5 ha
			Food plots: 0.1 ha
			Greenhouse plot: 0.5 ha
Number of irrigation farmers	33	49	Commercial plots: 43
			Food plots: 273
			Greenhouse plot: four emergent commercial farmers in a consortium
Total number of irrigator and non-irrigator households	33	195	+ / − 2000
Distribution of irrigation plots by gender (%)	Men: 82	Men: 29	Commercial plots: men, 88; women, 12
	Women: 18	Women: 71	Food plots: men, 51; women, 49
			Greenhouse plot: men, 100; women, 0
Composition of the adult members of surveyed households by gender (%)	Men: 55	Men: 43	Commercial plots: men, 55; women, 45
	Women: 45	Women: 57	Food plots: men, 51; women, 49
Composition of adult members of surveyed households by first language spoken (%)	Pedi: 73	Pedi: 100	Tsonga: 100
	Zulu: 15
	Tswana: 6
	Swati: 3
	Ndebele: 3

Hereford Irrigation Scheme is situated on four portions of 53JS Loskop South Farm on the outskirts of the small town of Groblersdal in Greater Groblersdal Local Municipality. The scheme is located along the Olifants River and Hereford Canal, within the commercial farming area administered by Hereford Irrigation Board.

Phetwane Irrigation Scheme is on Hindustan farm in Ward Nine of Greater Marble Hall Local Municipality, in the former Lebowa homeland. The RESIS programme classifies Phetwane as part of the Upper Arabie Balemi Irrigation Scheme Trust, which comprises a group of four irrigation schemes in Cluster Eleven of the programme. This cluster consists of 19 irrigation schemes located along the eastern banks of the Olifants River immediately downstream from the Flag Boshielo Dam.

Makuleke Irrigation Scheme is in Ward Five of Thulamela Local Municipality in Vhembe District, in the Nthlaveni communal area along the western boundary of the Kruger National Park. Before 1994 this area fell under jurisdiction of the Gazankulu homeland government. Makuleke Irrigation Scheme is sited along the Mphongolo River, downstream of Makuleke Dam. It is classified by the RESIS programme as part of Cluster Four, which consists of eight schemes located close to or within the lower reaches of the Levubu River Basin.

3.2 Historical background

The Hereford commercial farming area was proclaimed in 1926 in terms of the Irrigation and Conservation of Water Act 12 of 1912 (Tren & Schur, 2000). After World War II, following the accession of the National Party government in 1948, Hereford was developed as a welfare settlement scheme for white workers who had been pensioned early from mines and other industrial sectors (Butler, 1994). The government developed the land comprising Hereford Irrigation Scheme specifically as a white settlement scheme for soldiers returning from the war. The war veterans abandoned the scheme in the 1980s for various reasons, including the failure of a tobacco joint venture and insecurity due to anti-apartheid activism. For 10 years the land lay fallow and infrastructure became dilapidated. On 1 February 1997, a group of 32 black farmers, members of the Tafelkop Farmers' Association, an 800-strong organisation of landless black farmers from across four of South Africa's nine provinces, occupied the unused land in Hereford Irrigation Scheme. Most of them had spent their childhoods working on Hereford smallholdings alongside their parents. Broad stakeholder consultations over the legitimacy of the occupation resulted in the Minister of Land Affairs and Agriculture approving the farmers' occupation of the scheme while cautioning against further similar land occupation.

Phetwane Irrigation Scheme was established in 1957 in a small village on Hindustan Farm in the Rakgwadi area of the Lebowa homeland. The scheme was developed for the BaKone-ba-Matlala people, who settled in the Rakgwadi area following their migration away from the Madibong area in Sekhukhuneland (Claassens, 2001). Lahiff (2000) observes that while plots allocated by the chief, Kgoshi Matlala, from 1969 onwards were based on ‘permission to occupy’ (PTO) certificates, it is not clear what form the original registration of plots took. What is clear, however, is that from 1957 onwards newly settled Phetwane farmers began producing maize and other grain crops using water from the Arabie Dam (now known as Flag Boshielo Dam). The apartheid government subsidised agricultural production until the 1970s, when Lebowa homeland authorities gained political and administrative ‘independence’ and assumed responsibility for the irrigation scheme. The withdrawal of state subsidies and subsequent degradation of the scheme after 1994 meant that, to revitalise crop production in Phetwane, the irrigation infrastructure had to be rehabilitated. Revitalisation began in 2003 during Phase 2 of the RESIS programme.

Makuleke Irrigation Scheme was established in 1991. The chief, Joas Phahlela Makuleke, allocated plots on the scheme, in consultation with the local agricultural extension officer. Crop productivity remained generally low because plots were allocated to part-time farmers who were gainfully employed elsewhere, and declined sharply in 1999 because government subsidies were withdrawn. In 2002 Makuleke Irrigation Scheme was included in the ‘Water Care’ Programme, a sub-programme of RESIS. The 215 ha of land comprising the scheme was redistributed in favour of full-time aspirant commercial farmers and indigent people in the community.

3.3 Characteristics

The characteristics of the study sites are summarised in Table 1. Phetwane Irrigation Scheme is the smallest of the three, but has a relatively higher concentration of plot holders than Hereford. Makuleke Irrigation Scheme is the largest, with 215 ha and a membership of 317 farmers. Whereas all the farmers in Hereford and Phetwane are involved in various commercial farming activities, the Makuleke farmers can be subdivided into three distinct groups: petty-commodity producers, subsistence food producers and a consortium of commercial vegetable growers. Farmers, rather than government officials, influenced the adoption of this land distribution approach, in line with the principles governing community development in Makuleke.

The allocation of plots to men and women in all three schemes reflects site-specific differences in gender dynamics. The predominance of male plot holders in Hereford resonates with commercial plot allocation to petty-commodity producers in Makuleke but contrasts sharply with the predominance of female plot holders, particularly elderly women, in Phetwane (Table 1). The gender allocation of food plots in Makuleke strongly reflects the gender composition of adult members of subsistence food producer households within the scheme. It is not clear, however, whether this is due to deliberate planning or not (Table 1).

4. METHODOLOGY

A case-study approach was adopted to examine empirically how various selected factors associated with agricultural commercialisation have enhanced or detracted from the livelihoods of the rural poor living in small-scale irrigation schemes. Two analytical frameworks underpinned the study: the Sustainable Livelihoods framework developed by scholars such Scoones (1998), and the Capability Approach and Entitlements Analysis Framework developed by Sen (1999). Three schemes were selected from among 126 similar schemes in Limpopo Province. The sites were selected to represent two key approaches to agricultural commercialisation and variations in socio-economic settings within poverty nodes in Limpopo Province. The data collection was staggered into two phases.

In the initial phase, baseline data on livelihood profiles were collected using questionnaire surveys. Since all Hereford households are principally oriented towards commercial farming, a 100 per cent survey was performed. In Phetwane, the aim was to survey all households in the community, both irrigation plot holders and non-irrigators. However, logistical constraints resulted in the study surveying 72 per cent of the Phetwane households and 98 per cent of irrigation plot holders' households. The author compared irrigation plot holders with non-irrigators who share the same community but have different relationships with the irrigation scheme. In Makuleke, 100 per cent of the 43 commercial plot holders and 16 per cent (44 out of 273) of the food plot holders were interviewed. The emphasis was on petty-commodity producers, who occupy the commercial plots, because the study was primarily concerned with issues of agricultural commercialisation. However, the author also considered that some comparison between petty-commodity producers and subsistence food producers would provide useful insights into the way agricultural commercialisation has affected the two main categories of irrigation farmers who share the same physical space and resources. The consortium of commercial vegetable producers, as a third distinct group, was not surveyed because it did not exist at the time of the study. However, 75 per cent of the consortium members are part of the group of 43 petty-commodity producers surveyed. While questionnaire data provided useful insights on individuals and households living in the three irrigation schemes, the scale at which data was collected meant that the livelihood profiles could only be generalised overviews. The second phase of data collection therefore sought to elicit greater detail and nuance on livelihoods. Data collection in the second phase also attempted to triangulate information gathered in the three irrigation schemes with data possessed by externally based institutions.

The methods used in the second phase of data collection included in-depth interviews, focus group discussions, direct observation, informal conversations and formal workshops. A key approach was to work with a few selected households in each research site and, over the course of approximately 2.5 years, to find out how the livelihoods of people in these households had been affected by or had responded to agricultural commercialisation initiatives. The emphasis was on strategies by which the rural poor in irrigation schemes gain access and safeguard entitlements to socio-economic and political rights in the context of government interventions in agricultural commercialisation.

The socio-economic rights data pertained to patterns, structures and issues of access to irrigated land, and the political rights data to issues of participation in decision-making processes for joint-venture arrangements. The data collection sought to elucidate whether agricultural commercialisation approaches and processes afforded the rural poor freedom of action and freedom to make decisions on issues directly affecting their livelihoods, and to determine whether they were able make optimum use of opportunities provided by agricultural commercialisation, given their personal and social circumstances. The study also looked at the role of institutional arrangements in interactions between agricultural commercialisation approaches and rural people's livelihoods. The following section synthesises the key findings.

5. FINDINGS AND DISCUSSION

5.1 Overview of livelihood and coping strategies in the study sites

Agricultural commercialisation in all three schemes has taken place in the context of generally low levels of formal education, employment, income and material asset ownership. Individuals and households in farmers' groups and communities rely on a range of strategies to cope with problems of poverty, unemployment, shocks and vulnerability. Social grants and social networks are particularly critical to many households' survival. Coping strategies include relying on relatives and neighbours for moral and material support, and many households depend on loans from community members, micro-lenders and burial societies. Others continually obtain credit from formal retailers and informal traders. A number of households have personal savings accounts and insurance, particularly through membership of burial societies and church organisations. Some forms of insurance are less tangible; for example, ownership of items such as catering utensils for use at funerals, weddings and similar events. These events are an important part of social life in communities and extended families. Contributions in cash, kind and labour towards such events strengthen social relations and, through reciprocity, ensure a broader array of future support for all involved.

In Phetwane and Makuleke, other coping strategies are symbiotic relationships involving shared access to arable land, including irrigated plots. Work and produce are shared through informal arrangements based on responsibility for relatives and indigent community members. For example, although irrigated plots in Phetwane are formally allocated to 49 farmers, in practice at least 56 households were identified by the study as having direct shared access to the land. In Makuleke, although shared access is more prevalent on food plots, this study found three examples of shared access on the commercial plots. The sharing arrangement, however, involves farmers using both their own plots and plots registered in the name of a relative, and not shared production on the land. Such arrangements are considered fraudulent and have caused landless people in the community to have misgivings. The fact that all official records document Makuleke Irrigation Scheme as having 46 commercial plots and 46 commercial plot holders indicates that the landless might have a valid case. Besides shared access, casual, seasonal and permanent employment in farm work provides a number of households in all three schemes with access to benefits deriving from the agricultural production on the land. The benefits from such access, however, are limited to the duration of employment.

Hereford coping strategies include ‘straddling’. Farmers' households retain the use of homestead sites in communal areas of Tafelkop or Motetema, some 10–15 km away, while occupying the smallholdings in the Hereford Irrigation Scheme. The dual occupation of homesteads is to optimise the available livelihood generation opportunities in Hereford while simultaneously gaining access to the social services in Tafelkop and Motetema that the irrigation scheme area lacks. The general trend is for many of the married women to divide their time between working on the scheme and caring for children and elderly relatives in Tafelkop and Motetema. The more consistent presence of women on the scheme is often associated with polygamous households (6 per cent), in which one wife – often the younger – lives and works in the scheme while the other – often the older – lives in Tafelkop or Motetema, caring for younger children and elderly relatives. It is also associated with elderly women with grown-up children (42 per cent) and women who either hold land rights in their own capacity or who are the main breadwinners in their households (18 per cent).

5.2 Socio-economic differentiation

Emerging black small-scale farmers such as those of Hereford, Phetwane and Makuleke are often described as ‘resource poor’. The study found that in all three schemes social grants were a critical factor for the survival of many individuals and households. Phetwane farmers relied most on these, with 83 per cent receiving old-age pensions from the South African government. For the broader Phetwane community, the majority (75 per cent) of the labour force was unemployed, and most (64 per cent) of the households had no source of income other than social grants, primarily child support grants and old-age pensions. Reliance on social grants by Makuleke and Hereford petty commodity producers was much lower.

Evidence shows that there is a degree of socio-economic differentiation within and among farmer households in all three irrigation schemes. Such differentiation is evident in the diversity of livelihood strategies, aspirations and capabilities, such as production and marketing skills and managing household financial resources. While individual and household incomes are a useful measure of socio-economic differentiation, many respondents were unwilling to disclose their incomes. For ethical reasons the author did not insist on income disclosure but instead used proxy indicators (such as household material asset ownership, monthly expenditure and household nutrition) to qualitatively and quantitatively determine socio-economic differentiation.

Some of the findings were that Hereford farmers have the highest proportion of vehicle and tractor ownership, while Phetwane farmers have the lowest (see Figure 1). Asset ownership also varies between households in each scheme and tends to be concentrated in particular households. In Hereford, while a significant proportion (57 per cent) of households do not own gas, electric, coal or primus stoves and rely on wood fuel for cooking, some (15 per cent) own two types of stove, sewing machines, tractors and bakkies (small trucks). In Phetwane, the differences between households with direct access to irrigated plots and those without is less visible, although plot-holding households own significantly more livestock, fridges and electric stoves. Non-plot-holding households own more mobile phones and the less expensive paraffin (primus or kerosene) stoves. Mobile phone ownership is not always an indicator of affluence but of the importance of keeping in contact with household members who work in more distant urban, industrial or mining areas. In Makuleke, the majority of subsistence food producers have far fewer assets than petty-commodity producers. However, a number of householders among the subsistence food producers were found to own more of the expensive material assets, such as cars and electrical appliances, than petty-commodity producer households. This suggests that not all Makuleke food growers are indigent.

Figure 1: Socio-economic differentiation: percentage frequency of ownership of material assets by households

In all of the study sites, expenditure on food is relatively low, the average monthly expenditure per household ranging from R300 to R400. In both Makuleke and Phetwane, a narrow range of food items is consumed per household per week, with the emphasis on staple foods such as mealie meal, and relatively cheap foods such as cooking oil, bread and vegetables. Most households only very occasionally consume protein-rich foods such as milk, meat, fish, chicken, trotters, peanut butter and margarine. Protein-rich food, particularly meat, is relatively expensive and may therefore be beyond the reach of many households in irrigation schemes. Although Hereford households spend roughly the same amount of money on food as those of Phetwane and Makuleke, a greater percentage of Hereford households consume a wider range of vegetables, milk and poultry more frequently. This is because these foods are produced by many plot holders in the scheme and are therefore more readily available and at relatively low prices. There does not seem to be any significant difference in food consumption between irrigation plot-holding and non-plot-holding households in Phetwane, or between petty-commodity producers and subsistence food producers in Makuleke.

5.3 Overview of joint ventures

Farmers in all three schemes have received financial, material, technical and managerial support from the government, private investors and non-governmental organisations. They have complemented this support with their own investments in the form of time, labour and other resources. The largest amount of financial support has accrued to Hereford farmers, who have received more than R10 million since their occupation of the irrigation scheme in 1997. Support from the various stakeholders has enabled farmers to engage in a series of joint ventures.

Hereford farmers have engaged in tobacco and vegetable production for local and international markets, while Makuleke and Phetwane farmers have largely been involved in cotton production. Hereford farmers have had more freedom than Makuleke and Phetwane farmers to make decisions about produce and the choice of joint-venture partners. Although both Makuleke and Phetwane farmers have had to acquiesce to decisions by government officials and private investors, Makuleke farmers have been more assertive in negotiating for the kind of support they require. Many of the joint ventures have, however, not yielded the planned outputs, particularly incomes for farmers.

Among the direct results of joint-venture failure in all three schemes are debts, food insecurity, job losses and mental stress, and the ripple effects of such failure include tensions and conflicts among farmers and between farmers and farm workers. In Makuleke and Phetwane, where irrigators live in local communities, these tensions and conflicts often spill over into the broader community, escalating antecedent tensions and social and political marginalisation, and generally impairing the quality of life. In all three cases, joint-venture failure and its attendant problems have led farmers to claim a more active role in production-related decision-making. However, several factors determine their freedoms to respond. One factor is land tenure.

Makuleke petty-commodity producers are constrained by tenure insecurity. The monthly rent that each farmer pays to the chief lacks the guarantees required to engage in commercial crop production. Farmers are therefore compelled to abide by collective decisions and actions for fear of losing access to land. However, despite limitations imposed by tenure insecurity, Makuleke farmers have become more proactive and successful in negotiating for the kind of support they require. The aftermath of joint-venture failure has seen them moving beyond arguments for RESIS projects to accommodate both commercial and food security concerns, to demands for farmers to determine the choice of crop and the contents of joint-venture contracts. This response is driven by individual farmers' entrepreneurial aspirations and a shared vision that considers the Makuleke Irrigation Scheme as one-half of a twin engine for community development, the other being the 21 887 ha of restituted land in the northern region of the Kruger National Park. The Makuleke farmers' increasingly proactive stance has garnered positive responses from stakeholders, such as the Limpopo Provincial Department of Agriculture, non-governmental organisations such as Geselschaft Technische Zusammernarbeit, local supermarkets and tourism operators in the restituted region.

In Phetwane, although PTO certificates give farmers a greater sense of tenure security, strong perceptions of state and corporate power limit their ability to voice strong dissent or negotiate as key stakeholders in the cotton joint venture of 2003–2004. Since then, Phetwane farmers have emerged from joint-venture failure with a keen awareness that, although the irrigation scheme is owned by the government, their PTO certificates and their livelihood and food security interests give them a strong stake in production-related decision-making. They therefore want to see an approach different from the prevailing rehashes of pre-1994 practices that limited their participation in agricultural production to the provision of cheap labour while privileging public institutions and ‘expatriates’ in production-related decisions and activities.

The Hereford farmers' individual and collective freedom to make decisions about production has not been limited by the fact that they have yet to acquire ownership rights to individual plots, or by their awareness of the substantial support contributed by various stakeholders. The main constraint has been their lack of resources for making informed choices. Hence, although they have generally had more freedom than Phetwane and Makuleke farmers to make production-related decisions and responses, they have failed to make optimum use of this opportunity because they lack technical and negotiation skills, managerial expertise and information, and technological and financial resources.

5.4 Risks associated with capital-intensive farming

Experiences with agricultural commercialisation in the three study sites reveal a pattern in which joint ventures fail to make optimum use of available resources in order to generate the planned outputs for farmers. The problem seems to relate to risks associated with capital-intensive farming.

Small-scale farmers' production accounts show that crops such as tobacco and cotton are capital intensive. Production costs for cotton in 2003 accounted for 360 per cent of Makuleke farmers' net receipts from sales. Direct production costs accounted for between 89 and 320 per cent of Hereford farmers' net receipts from tobacco sales in 2004. In 1999, tobacco production costs per Hereford farmer ranged from 39 per cent of the receipts from sales for the more successful farmers to 128 per cent for the less successful. Insurance costs were relatively high, accounting for between 30 and 34 per cent of the total production costs, implying that tobacco is a risky crop.

Studies by De Klerk (1996) of similar farming schemes in the Western Cape suggest that adequate financial support is central to any programme for establishing small-scale farmers. A critical factor determining the viability of farming activity is ‘gearing’. This is the degree to which farming activities are funded by a small-scale farmer's own funds relative to finance through creditors' funds. A ‘rule of thumb’ of the South African Agricultural Union is that farms with a debt of more than 30 per cent in their financial structure are unsound, while those with a debt above 50 per cent are unlikely to survive (De Klerk, 1996). In the cases of Hereford, Makuleke and Phetwane, production costs have often far exceeded 50 per cent of net earnings, with much of the cost financed through subsidies from private investors and government. In the event that such subsidies cease, farmers might resort to credit schemes to finance the production of capital-intensive crops. With the rising costs of credit, the high ratio between possible debts and net earnings will render small-scale tobacco producers particularly vulnerable to downturns in market prices. The high risks involved, the requirement for substantial capital outlay and the losses experienced in successive joint ventures in Hereford, Makuleke and Phetwane raise questions about whether small-scale farmers can sustain capital-intensive production without long-term financial assistance.

Another source of risk is deciding to produce capital-intensive industrial crops without first sufficiently analysing the opportunities in regional and global production and market trends. For example, the Hereford farmers' decision to grow tobacco, despite that crop's history of failure in Hereford, appears to have been influenced by the decline of tobacco production in Zimbabwe following the orchestrated ‘fast-track’ land reform. What farmers did not realise was that Brazil had increased its output of prime tobacco, thereby claiming a significant share of the market. Farmers were also not aware that BAT–PLC, a major role-player in the world tobacco trade, had recently nominated Brazil, the US and Zimbabwe as major sources of tobacco for the future. From the outset, therefore, entry by Hereford farmers into the established tobacco sector was not assured. With regard to Phetwane and Makuleke, decisions to grow cotton seem to have been informed more by a convergence of the interests of the private investor and the Limpopo Department of Agriculture than the existence of real opportunities in the world markets. The private investor's needs to increase supplies to a newly built ginnery and demonstrate corporate social responsibility resonated strongly with the RESIS objectives, and possible downturns in cotton prices were ignored.

5.5 Debt

Joint ventures in all three schemes experienced financial loss. Although subsidies have cushioned Hereford and Makuleke farmers from the effects of debt, farmers are increasingly reluctant to accept joint-venture propositions from stakeholders based on outside irrigation schemes, since they have seen households descend into livelihood insecurity and vulnerability as a result of debt. An example identified by the study is a Hereford plot holder who lost most of his productive assets when the Land Bank repossessed these in lieu of a debt that he owed, and has not been able to grow any crops since 2004 as he waits to recover from the loss. The vulnerability of this household might be of longer term than anticipated, given that the total household income comes from the plot holder's intermittent casual employment, social grants totalling R340 per month for two of his daughter's children, and his daughter's ‘employment’ by an informal trader in exchange for food and soap.

Subsidies have not sufficiently cushioned Phetwane farmers from the less visible debts that remained long after private investors had gone, the result of disparities between budgeted and actual cotton-picking labour costs. The Phetwane joint-venture budget allowed for cotton-picking costs at a rate of R0.40 per kilogram. However, many farmers were compelled to pay workers R20 per day. No mechanisms were put in place to ensure that workers' rates of pay did not exceed the budgeted labour costs, or to amend the joint-venture budget to reflect actual labour costs. Although the Phetwane farmers' management committee and the RESIS project implementation agent anticipated problems with labour arrangements that farmers were individually entering into, they were unable to intervene effectively. Many farmers were not fully aware of the risks involved, and even if they were they had no choice but to accept the price determined by workers or face the prospect of losing much of the cotton crop. Many therefore rejected advice and adopted various strategies to cover the costs of hired labour. Some attempted to manage risk by working together with their workers on the plots and paying them amounts lower than the average daily rate of R20. That way, they managed to reduce their labour costs. Others used their entire remuneration for picking cotton to pay hired workers the R20 per day, but still owed workers by the end of the cotton-picking season. These farmers, who were mostly pensioners, were then forced to use their pensions to pay off labour-related debts. Financial and food security for the elderly farmers became critically low and they incurred further debts by buying food on credit.

5.6 Farm worker wages

The above shows that the wage-setting mechanisms of joint ventures need to safeguard the interests of workers employed in irrigation schemes. Phetwane's pre-project consultations over wages evidently did not include farm workers' interests. Had Phetwane labourers accepted the R0.40 per kilogram offered by the joint venture, they would have earned an average of R489 per ha for the 3-month duration of the cotton-picking season, shared among two to three labourers. It is not surprising, therefore, that workers rejected the wages offered by the joint venture.

Joint-venture budgets for cotton-production labour costs in Makuleke were R300 per worker per month. Such wages were not attractive to unemployed Makuleke women, who have historically relied on the scheme for relatively low incomes (Tapela & Omara-Ojungu, 1999). Since the South African Government broadened access to social grants, many Makuleke women are able to access child grants – and this has reduced both their dependence on the scheme and their vulnerability to exploitative wages. Makuleke women farm workers have been replaced mostly by male and female refugees from neighbouring Zimbabwe and Mozambique. The illegal status and desperation of such workers makes them particularly vulnerable to exploitation and compels them to survive on the margins of society and the economy.

In Hereford, the tobacco joint-venture budget allowed for higher minimum wages of approximately R550 per worker per month. However, such employment was generally insecure as it tended to terminate with the end of each joint venture contract. Only one farmer was observed to have retained most of his workers over an entire 3-year period.

5.7 Social and political conflict and marginalisation

The joint-venture inception and failure in all three schemes had various effects on local social and political dynamics. In Makuleke, the failure of the cotton joint venture in 2004 led to tensions among farmers over whether to continue or discontinue the joint venture. These tensions spread into the broader community and cast people into two main political camps. Farmers who had voiced dissatisfaction were then marginalised from mainstream political life in the community. Among Hereford farmers, joint-venture issues have not erupted into open conflict but have remained as low-key but persistent tensions. While farmers are evidently kept in check by fears of their leader, the source of this leader's power is not clear.

In Phetwane, although plot-holding farmers firmly believe that their rights to irrigated land are secure, arguments by the landless youth for land reallocation remain an ongoing source of tension in the community. Unemployed youth, who constitute 57 per cent of the adult population and 59 per cent of the unemployed in Phetwane, invoke the Freedom Charter of the African National Congress to argue that ‘the land shall be shared among those who work it’ and call for a reallocation of irrigated land away from elderly plot holders, who constitute 83 per cent of Phetwane irrigators. The Communal Land Rights Act indeed provides for PTO certificates (or ‘old order rights’) to be upgraded into ‘new order rights’. However, it is still not clear what tenure arrangements will apply to land in the government-owned scheme. Arguments for land redistribution also need to be balanced against obligations to avoid undermining the livelihood security of households presently holding PTO certificates as well as those with informal shared access to land on the scheme.

There has emerged a peculiar form of social marginalisation in Phetwane, which is markedly absent from Hereford and Makuleke. This takes the form of allegations of witchcraft, which became rife after the failure of the cotton joint venture in 2004. Vulnerability to witchcraft is best understood in the context of Sekhukhuneland's historical reputation as the 1980s epicentre of witch burnings in South Africa, documented by Peter Delius (1996). While witch burnings in Sekhukhuneland have historically been linked to rural resistance to colonial and apartheid repression, the current wave of allegations seems to be driven by a widespread failure of coping strategies following the shocks to livelihood and food security. The fact that current witchcraft allegations tend to single out those elderly women who manage to cope against odds that many cannot overcome is in itself indicative of problems of gender power dynamics in irrigation schemes (Bastidas, 1999; Van Koppen, 2001).

The challenge for livelihood sustainability in small-scale irrigation schemes is not about the presence or absence of political and social dynamics within communities, but about the implications of unequal political power relations for the interests of the less powerful people in these communities. Agricultural commercialisation initiatives should therefore guard against reinforcing the ‘voicelessness’ of such people, and yet be powerful enough to ensure that all local views are foregrounded in crafting the desired interventions in irrigation schemes.

6. CONCLUSIONS

Research findings show that agricultural commercialisation of small-scale irrigation schemes in Limpopo Province has reinforced various forms of marginalisation. Attempts to integrate black farmers into globalised commodity sectors highlight the magnitude of the challenges these farmers face. They encounter stiff competition from established producers both in South Africa and internationally, and have to contend with stringent requirements imposed by phyto-sanitary regulations governing supermarket chains and international markets. At the same time they have to overcome their lack of financial, technical, technological and managerial resources. Assumptions that joint ventures will enable black small-scale irrigation farmers to participate effectively in the highly competitive and globalised commodity sectors should be questioned.

The recurrent failure of joint ventures points to a need to scrutinise the design principles of such projects. With the exception of Makuleke, interventions to commercialise agricultural production do not seem to have taken into account the farmers' socio-economic differences, but have tended to replicate many features of the pre-1994 models. A ‘one-size-fits-all’ approach has been taken to involving farmers in the projects. The decision-making processes need to be examined, since concerns voiced by the less influential stakeholders in agricultural commercialisation, such as the elderly and illiterate farmers of Phetwane, have not been readily accommodated.

It is particularly important to clarify why joint ventures should continue to produce capital-intensive crops when past experience demonstrates that market trends, particularly declines in commodity prices, could exacerbate farmers' marginalisation by pushing many into debt from which they might not recover without the cushioning effect of subsidy grants. What is evident from the study is that risk is often downplayed, as government officials and private investors select crops that ‘justify’ the high levels of investment. Such decisions converge with farmers' anticipation of high incomes, and any voices of caution are silenced. Gaps in institutional arrangements for project monitoring and accountability mean that problems are often detected too late, and mechanisms for averting similar pitfalls in future projects are either ineffective or nonexistent. A robust framework for monitoring decision-making processes is clearly needed, both in site-specific joint venture projects and wider-ranging agricultural commercialisation programmes.

Stephen Friedman's observation that attempts to deal with poverty are ineffective because they do not reflect what the poor want (Friedman, 2005) suggests a need for greater and more effective participation by the poor in shaping decisions that affect their lives. However, there are no guarantees that such measures will result in farmers taking decisions most appropriate to their interest, because, although farmers have limited powers to decide how to commercialise, they nevertheless have some freedom to make choices. Such freedom can unwittingly be used to make suboptimal decisions, to the farmers' own detriment. A case in point is the farmers' hunger for incomes, demonstrated in all three irrigation schemes, which has resonated with arguments by government officials and private investors for producing capital-intensive high-value crops, thus exposing the farmers to the risks associated with this kind of farming.

The dominance of global discourses by neoliberal economist thought leaves little scope for small-scale black farmers to establish a niche in global commodity production sectors, let alone sustain their livelihoods. Some of these farmers have resorted to orienting their activities towards local informal markets, but their survival on the margins of the mainstream economy is not assured. Economies of scale are in favour of supermarket chains, which are rapidly making inroads into places and markets that have historically been the domain of local petty-commodity producers, informal traders and formal retailers.

In the face of constraints that militate against black small-scale farmers' entry into the highly competitive and globalised commodity sectors, it is perhaps worth rethinking the prevailing preoccupation with agricultural commercialisation. The negative effects of joint venture failure on the livelihoods of people in small-scale irrigation schemes point to a need for alternative approaches to addressing South Africa's challenges of rural poverty and inequality. Cousins (2003) observes that, although it is clear that the deep poverty in rural areas requires radical measures, not least a redistribution of resources including land, a sustainable livelihoods approach that builds on rural people's current land-based livelihoods and enhances their economic value might be more appropriate than attempting to replace these livelihoods with fully market-oriented or commercialised approaches. A possible way forward could therefore be for various stakeholders, particularly locally based civil society organisations, to engage with farmers in defining the most appropriate ways to enhance livelihoods and well-being.

The revitalisation of small-scale irrigation schemes in Limpopo Province reflects a welcome shift towards aligning the South African Government's interventions to local realities in poverty nodes. It is laudable that there have emerged initiatives such as the RESIS programme that clearly grasp the need to respond positively to commercial farming and food security concerns and to promote connections between on-farm and off-farm livelihood strategies. This shift, however, has yet to overcome the preponderance of neoliberal economics criteria for gauging ‘viability’. Evidence from the study shows that definitions of viability urgently need to take into account a broader range of criteria than the economic; for example, finding creative ways to include criteria such as livelihood sustainability.

With improved project and programme instruments, investments in irrigation schemes could significantly enhance the livelihoods of people living in and around these schemes. However, the other reality is that such investments benefit a very small fraction of the rural poor. There are plans to broaden the scope of interventions beyond irrigation schemes; for example, through rainwater harvesting in the drier rural areas. However, environmental limits to water availability in Limpopo Province mean that a wide-ranging programme to enhance livelihoods and food security for the rural poor is not tenable without a similarly wide-ranging programme to redistribute water and land.