How rural land reform policy translates into benefits

Abstract

Land reform policy in South Africa has been strongly criticised, especially its instrumentality. However, recent ethnographic studies indicate that it is a complex and deeply social process in which policy is understood differently by different actors. Rather than asking whether land reform works we should ask how it works. Using a case study of SLAG (Settlement Land Acquisition Grant) redistribution beneficiaries in a southern Cape village, this paper describes how these rural residents interpreted policy and used the resources put at their disposal by the state. These local actors' decisions and actions were based largely on their livelihood requirements and frequently determined by their historical experiences and social relationships. Although they behaved in ways that were not anticipated by officials, a number have gained tangible benefits. Beneficiary ‘success stories’ have given credence to the land reform policy, and state officials have responded by continuing to provide support to the project that was the subject of this study.

Keywords:

• actor-oriented approach
• agricultural development interventions
• land redistribution
• policy translation

1. Introduction

Since 1994 land reform has become synonymous with rural development policy in South Africa. Academics and consultants have made their contributions and scholars have voiced their criticisms, but generally the land reform process has been found wanting (e.g. Lahiff, 2007; Cousins & Scoones, 2009; Hall, 2009; Tregurtha et al., 2010).

Many of the criticisms have focused on the instrumentality of land reform, primarily considering the policies, models and plans for interventions, and the way they have been implemented. As Aliber et al. point out, some studies blame poor land reform on the usual suspects: ‘lack of money and equipment; lack of skills (both technical and managerial); lack of postsettlement support; lack of appropriate legal structures; and infighting’ (2010:292, emphasis in original).

Other scholars have adopted an ethnographic stance to unveil the complexities, controversies and contradictions that are integral to land reform. Robins & Van der Waal (2010) suggest that appearances are deceptive and that meanings are made and contested by the many actors involved. A study by Deborah James (2007) illustrates how potential and actual land reform beneficiaries either misunderstand policy or seek to reinterpret it in ways that would meet their own needs. The diverse aspirations of beneficiaries and the many interpretations by various actors have given rise to different views about the extent to which land reform projects have succeeded or failed. Land reform interventions, like all development projects, are political systems in which different perspectives contend for influence (Mosse, 1998).

If we take the ethnographic angle, which highlights the complexity of land reform, and also consider Mosse & Lewis's work (2006) on the role of actors and ‘translation’ in development, the question we will ask is not whether land reform works (a purely judgemental question), but rather how it works; not whether a development project has succeeded, but in what ways it has succeeded and in what ways it may have failed. Focusing on how development works opens up the ‘black box’ of implementation to reveal the complex social relationship between policy and practice.

According to Latour (1996), the success of policies, policy models and project plans cannot be taken for granted. The amount of effort invested since 1994 in getting South Africa's land reform right and the continual criticisms levelled at the policy, models and plans, despite summits and regular restructuring, gives credence to Latour's view. Furthermore, the control that policymakers, planners and implementers (state bureaucracies, development agencies or NGOs) have over development events and practices is far from hegemonic. Control is restricted because the components of the project system – policymakers, consultants, professionals, managers, planners, field staff, technicians and local residents – are often weakly linked. They function with a high degree of independence from one another, and their interactions involve negotiation and a diversity of interests (Mosse, 2005:10). Consequently, the influence of policymakers, planners and managers is limited by the extent to which their ideas and instructions are interpreted by other people to meet their own ambitions, intentions and goals.

These interpretations and their ensuing actions can then be reinterpreted. The way this happens is important to development agents because this is the basis for evaluating policy. The more interests involved and support secured, the more stable and dominant the policy models become (Fairhead & Leach, 1996). Support is thus given to models not because they are the best, not because they do what they claim to do, but because they can be interpreted in ways that suit the needs of both recipients and development agents. These interpretations can be seen as ‘translations’ of contingent outcomes that help to legitimise policy models (Mosse & Lewis, 2006).

To illustrate the social processes inherent in the workings of land reform interventions, this paper explores a SLAG (Settlement Land Acquisition Grant) redistribution project in the southern Cape using Norman Long's actor-oriented approach (2001) and Mosse & Lewis's ‘translation’ concept (2006). Local rural residents negotiated and ‘translated’ (i.e. interpreted) policy and interventions to suit their requirements and ambitions, and did so successfully enough to lend credence to the land reform policy. Decisions were often determined by collective and individual historical experiences and social relationships and based on the livelihood demands, strategies and priorities of local actors.

The actor-oriented approach and the function of translation are important features of this study. The next section discusses the theoretical orientations of actors and describes the research methodology. This is followed by an overview of land reform in a village in the southern Cape. The fourth section describes how the beneficiaries translated policy to suit their requirements and how their actions appear to give credence to policy, and the paper concludes by reemphasising the key question to be asked when assessing land reform policy.

2. Theoretical approaches and research methodology

The machine metaphor for development interventions is misleading. It suggests that the correct inputs will lead to the desired outputs and implies that history, i.e. the actors' experience and context, is unimportant. Also misleading is the idea that development is all-powerful and merely a new global form of cognitive control and social regulation (Escobar, 1995). How development works in practice has much to do with the different interpretations by the numerous actors involved. It is therefore much more dependent on social interactions and their consequences than on technical and material inputs. The focus on social interactions requires an actor-oriented approach, which Long describes as ‘an ethnographic understanding of the social life of development projects – from conception to realisation – as well as the responses and lived experiences of the variously located and affected actors’ (2001:14–15).

This description makes three important points. Firstly, it emphasises that a development intervention is social, rather than purely technical. Secondly, it notes that a project needs to be understood from conception to realisation. Since the entire project is socially determined, good or bad conception may not ultimately determine success or failure. Finally, it is important to understand the actions and experiences of the actors involved at different levels (e.g. officials and farmers), which demands studying upwards (focusing on those who have the power to exclude themselves, as pointed out by Cooper & Packard, 1997) as well as the usual practice of studying downwards.

In an actor-oriented approach, a degree of responsibility for the outcomes of development policy and subsequent interventions is attributed to the agency of the various actors at different points in the chain of developmental organisation. (This refers to the chain that links centrally located policymakers and managers to the outlying field sites.) Given the multitude of actors and different contexts, it is inevitable that various interpretations, political manoeuvrings and negotiations will take place in the course of policymaking, planning and implementation, all of which may well result in deviation from the planned intentions and objectives (see Crawford, 2003; Mallarangeng & Van Tuijl, 2004).

An actor-oriented approach reveals that local people and development agents are political actors often pursuing different agendas, while negotiating development outcomes in particular contexts. Beyond the local context, development interventions also occur within a broader context in that they ‘are always part of a chain or flow of events located within the broader framework of activities of the state and the actions of different interest groups’ (Long & Long, 1992:228). The relationship of policy and practice is not simply the instrumental translation of ideas into reality, but rather a process in which transactions are often uncontrollable and their outcomes uncertain, precisely because they are subject to the contexts in which they occur and the fact that all the actors translate them differently, depending on their experiences and requirements.

Different understandings become a question not just of perspective but essentially of translation, as it is the translation of events that emphasises some actions and objects while ignoring others (Mosse & Lewis, 2006). Multiple actors and their varied translations mean that ultimately we end up with ‘as many theories of action as there are actors’ (Latour, 1996:167). Actors are in turn able to translate these theories for their own purposes.

Translation allows for the ‘negotiation of common meanings and definitions and the mutual enrolment and co-optation into individual and collective objectives and activities’ (Mosse & Lewis, 2006:14). As a result, at different points in the development chain the actors use their interpretations to act collectively and individually, to work together and in opposition, to make use of the policy in ways that meet their requirements and to withdraw. The same applies at the level of the project field site. Essentially, the multiple actors are constantly involved in creating order out of disorder, through political acts of composition, in which ‘people, ideas, interests, events and objects … are tied together by translation of one kind or another into the material and conceptual order of a successful development project’ (Mosse, 2005:9, following Latour, 2000). Meaningful representations are achieved through translation and are never a matter of policy and project design (Mosse & Lewis, 2006). Of course, negotiations and translations do fail. In such instances, coherent representations are not achieved.

This paper explains how the land reform beneficiaries at the study site translated state policy to produce actions and events that enabled officials to reinterpret it and give it credence. The author's interpretation is confirmed by the fact that the state continues to support the beneficiaries. The focus is almost exclusively on the roles and actions of the farmers and recipients of land and how they managed to negotiate and interpret policy for their own ends. While both the actor-oriented approach and the concept of translation acknowledge the importance of broader contextual issues, such as the impact of political and economic factors occurring outside the village, these are beyond the scope of this discussion.

Research was conducted in the study village over several years from early 2000 until the end of 2009. Research methods included PRA (participatory rural appraisal), surveys and ethnographic fieldwork, using participant observation and informal interviews. Between 2005 and early 2009 the author lived in the village at various times for periods of up to eight weeks. Informal and scripted interviews and participant observation sessions were typically carried out during the course of interactions with farmers, land reform beneficiaries and other village residents. An actor-oriented theoretical approach to data collection and analysis was gradually adopted from 2001. The concept of translation is discussed in further detail in the sections below.

3. Land reform comes to a southern Cape village

3.1 Background and historical overview

The study site is a village situated at the foot of the Outeniqua Mountains, inland from the coast. The village was originally a farm in the early 1800s, becoming an NGK (Dutch Reformed Church) mission station in the mid-1800s. In the 1970s it became a designated ‘Coloured Rural Reserve’. Residential land was never owned by residents; they only had the right to occupation of their houses, often at the whim of the farm owner in the 1800s, the church after 1870 or the local town council after 1970. As of August 2008, virtually all the 168 households had access to electricity and potable water, but most of them still had no proof of ownership of their houses and residential land.

Most adult female residents work either on neighbouring commercial farms or as domestic workers or shop assistants in the surrounding coastal towns. Most of the male residents are artisans and work in the construction industry. A handful of residents work for local and provincial government structures in the village – the primary school, the health clinic, the day-care centre, the postal agency and the local municipality. There are four to five home-based (spaza) shops in the village, which provide limited self-employment and a small range of essential goods. A local general dealer provides other supplies such as gas, electricity and groceries. According to local residents, very few people in the village are extremely poor, although a few households fall into this category.

Until the 1960s most village-based agricultural production was conducted on residential plots for household consumption. Historically, these plots were one morgen in size (0.85 hectare) but as the population of the village increased so the pressure on land increased. This resulted in the ‘subdivision’ of residential plots and the allocation of some of the commonage for residential units. As a result of this loss of potential agricultural land, fewer people farmed and, more importantly, less land was available for agricultural purposes. The result was a reduction in the volume of agricultural produce obtained from home gardens and the commonage. In the 1970s a few male residents were engaged in dairy activities with neighbouring commercial farmers. Town planning after 1970 and the extension of the residential areas meant that by the mid-1980s only a handful of elderly male residents remained engaged in horticultural crop production on the commonage. Despite these agricultural activities, agriculture was not the main form of livelihood for most residents historically, unless they were employed on neighbouring farms. Farming on residential land and the commonage was often a supplementary livelihood, or practised in response to a lack of income from other sources. Many of the male land reform beneficiaries reported working outside the village for long periods, pursuing employment as artisans, clerks and truckers.

3.2 Negotiating land reform

In 1995 a local councillor heard about the state land redistribution policy. He discussed this with eight of his contemporaries (his son, his cousin, three pensioners and three of their sons) as a possible opportunity to secure land for farming. All nine expressed a desire to jointly farm some of the land and obtain access to land for future use. During a meeting with the SCDLA (southern Cape regional office of the Department of Land Affairs) the nine were told that to purchase any land in the area they would need to form a CPA (communal property association). Because the proposed grant of R15 000 per household for the nine households would not be sufficient to purchase any local farmland, the group approached other residents until 30 households, 28 of which were male-headed, had agreed to form a CPA. The core group of household heads was made up of family members or friends; many of the other household heads were not close friends or family and some joined the CPA in order to acquire land rather than to actively farm it. At the time of the submission of the application, eight of the household heads were pensioners and three were disability grant recipients, one was unemployed, and 16 were employed on a permanent, full-time basis. Three of the employed applicants worked at the local council offices in the village, while the remainder all worked outside the village, with some only coming home on weekends or less frequently. An income means test (combined household income of less than R1500) identified all but two of the applicant households as suitable beneficiaries. Two applicant households acknowledged having incomes above the mean.1

However, few of the proposed beneficiaries had been directly involved in agricultural activities or involved for any sustained period. At the time of application fewer than half of the CPA members had any extensive experience in agriculture. For many, their experience was confined to tending small vegetable gardens at home. Up to this point, most agricultural activities in the village had been on a micro scale. Residents said that residential vegetable gardens had been common until the 1980s.

In 1996 the Kagiso Trust provided a grant for a potato production project. Five hectares of land adjacent to the village was used to grow the potatoes. Local farmers, local agrochemical suppliers and the WCDA (Western Cape Department of Agriculture) supported the project in various ways. For example, the WCDA trained some CPA members in the use of mechanised implements, managed the funds and sourced inputs.

The potato project revealed very quickly that the CPA was not a homogeneous group. There was a lot of internal conflict as many members did not get actively involved, often because of their employment obligations. After harvest, when the CPA decided to pay a share of the proceeds only to those who had worked on the project, conflict arose between those who received a share and those who did not. In the end the CPA made very little money and much ill-feeling had arisen in the group.

Eventually, a neighbouring farm of 99 hectares was identified for acquisition and the WCDA conducted technical surveys and compiled production estimates. The land had not been farmed for a decade. A March 1998 WCDA technical report indicated that almost any crop could be grown on the land with the appropriate agricultural inputs. Tropical and sub-tropical crops, small grains and deciduous fruits were not recommended. The report recommended that only five hectares of irrigated vegetables, including potatoes, could be cultivated given the existing water constraints. The investigators also determined that commercial cattle production was unfeasible because of the very low carrying capacity of the natural rangeland.

In September 1999 the CPA formally took possession of the farm. Contrary to the original intention and in response to requests by senior CPA representatives arising from the conflict experienced during the 1996 potato project, the SCDLA made provision for the land to be subdivided. Each beneficiary household was allocated approximately two hectares of land, to be farmed on an individual basis. The remaining 39 hectares are held in trust by the CPA. Much of this land is overgrown and unsuitable for agriculture as it is mountainside and gullies.

Formal transfer of the subdivided land only took place in September 2007 as there were problems relating to water access and rights, compounded by the deaths of four CPA household heads. Only the smallholdings on the westerly side of the farm have access to water from the two dams fed by the local irrigation network. Those on the easterly side of the farm have virtually no access to irrigation water. The distances are great and the terrain so uneven that it is too expensive to channel water to the eastern side of the farm. Despite numerous visits by the WCDA and the Department of Water Affairs, this problem has not been resolved because of the exorbitant cost that the state would incur. An Agricultural Research Council (ARC) soil survey in 2002 indicated a high infestation of root-knot nematodes which primarily affect root crop production. At first glance one might be tempted to say that this was neither the most suitable piece of land nor an adequately cohesive group for a land reform project.

4. Effective translations: Meeting the objectives of land reform

The three primary objectives of land reform are: first, the return of land to those who were unfairly dispossessed; second, redressing the extreme racial imbalance in landholding, especially in agricultural land in rural areas; and third, the alleviation of poverty in rural areas (DLA, 1997). The first objective does not apply to land redistribution projects. What was important in the study was to consider how local actors (farmers and officials) translated their needs in respect of the second and third objectives.

4.1 Racial redress in a rural area

Through the redistribution project in the study area, a farm consisting of 99 hectares of agricultural land previously owned by a single white household was transferred through the ‘willing seller, willing buyer’ process to 30 black households, most of whom had never owned land. In 1999 the CPA (representing these 30 households) took transfer of the 99 hectares, but immediately a subdivision process was started (with SCDLA support). By late 2007 the 30 households took transfer and private property ownership of about two hectares each, making a total of 60 hectares. The CPA (made up of these 30 households) ended up holding the remaining 39 hectares in trust, access to and use of which was shared by all 30 households. Consequently, the area of land eventually owned by the CPA and by each of the individual households was much smaller than that owned by the former owner. Despite the fact that each black household only owned two hectares (and had shared access to 39 hectares), while previously one white household had owned 99 hectares, this redistribution process was considered an achievement by the SCDLA. Similarly, the beneficiaries were happy to own agricultural land for the first time, even if some had no intention of using it for agriculture and a large proportion of the land held in trust could not be farmed because it was mostly rocky mountainside and not arable land.

4.2 Poverty alleviation: Land use, livelihoods and assets

The intention of the initial proposals and the final project business plan was to address poverty at the individual and community level through the combination of household vegetable gardens, a collectively operated commercial farm, and individually operated commercial small farms. However, these good intentions were never realised in practice.

4.2.1 Land preparation and use

Immediately after obtaining the land, 13 beneficiaries prepared it for agricultural purposes. They erected fences, cleared the natural vegetation and ploughed their individual plots. Three pensioners, two disability grant recipients and two beneficiaries who worked in the village planted potatoes and vegetables. Two beneficiaries who worked outside the village returned irregularly on weekends and planted small patches of potatoes and maize which were tended by their spouses. Within a year their work commitments had obliged them to shift to livestock rearing. Three beneficiaries who worked outside the village but returned home every evening also planted potatoes but shifted to livestock rearing the following year. The only unemployed beneficiary planted a variety of vegetables on his land. Two other beneficiaries made their land available as a demonstration plot for an ARC honey bush (cyclopia spp.) project in the hope that other beneficiaries would help them produce the crop by providing the necessary labour. The demonstration plot component of the project was stopped after two years due to neglect.

Fewer than half the beneficiaries have ever actually used their land. Fluctuations in farming activity occur, with no more than 11 using their land for agricultural activity during any season. A visit in August 2008 revealed that five farmers had planted vegetable crops during that year, while in 2007 four had done so. The most active farmers between 2000 and 2008 tended to cultivate potatoes and vegetables. They also tended to be those who were present in the village most days (i.e. they worked in the village, or were pensioners or disability grant recipients). Those who worked outside the village tended to use their land as grazing for one or two cattle and occasionally planted fodder.

4.2.2 Fluctuating livelihoods and household circumstances

Fluctuations in the number of farmers and shifts from horticultural to livestock production have been due to death, illness and old age, or to changes in off-farm livelihood commitments and demands. Four of the beneficiaries have died and their heirs have been slow to take up farming. One pensioner is 90 years old and is too frail to work on his smallholding. Two initially active beneficiaries now work outside the village and do not have the time to grow vegetables. One of them, a veterinary assistant, took up cattle husbandry. Unlike many others, he has developed quite a substantial herd and makes use of the surrounding commonage for grazing. The other lends his land to an active beneficiary to plant vegetables. One household started producing cash-crop vegetables on its land, but stopped when the lack of water became a serious constraint and the husband obtained employment that took him away for long periods of time. The wife now manages a WCDA-funded chicken layer project on part of the trust land. Their smallholding lies fallow but they intend to rear livestock there. Another household has not worked their land for two years, following the husband's employment in a capacity which prevented him from returning home every weekend. Two beneficiaries have never worked their land but lease it to a non-beneficiary, who, with the support of his employer, produces vegetables for the formal market. Since 2003 some beneficiaries and at least two non-beneficiaries have leased uncultivated land from inactive beneficiaries. This is done in exchange for a small portion of the harvest, or in exchange for clearing alien acacia species from the land, or to strengthen intra-familial and peer relationships.

For those beneficiaries who do farm, farming is a secondary activity. Off-farm employment or access to a pension or disability grant is the primary source of income and farming is done to increase income or household food supply. Some farmers reported not farming for two or three seasons because they were too busy with off-farm employment.

4.2.3 Land as a tangible household asset

Obtaining land was seen as an asset in at least three ways. Firstly, beneficiaries could farm on the land and produce commercial crops which were sold locally or to hawkers. Secondly, for some it provided an asset which they could use to further increase household assets by rearing livestock. Thirdly, land was considered an investment and, after the transfer of the subdivided lands to the individual owners, they and their families could decide to sell the land, retain it as an asset or use it for collateral. Since transfer in 2007, three households have sold their land to people from outside the village, obtaining more than triple the initial purchase price. None of these households used their land after the transfer to the CPA in September 1999.2

4.3 Translations and sustained support

Most SCDLA and WCDA officials made it clear, either by word or deed, that they considered the project a success rather than a failure. Some points are worth discussing. First, the original 99 hectare farm had been allowed to stand fallow for over a decade and, while not all the beneficiaries actively farmed, a variety of farming activities could be seen every year following acquisition. Second, to some degree, a small farm model has been adopted and some of the land is being used productively and profitably. Third, beneficiaries occasionally hire labour from the village and surrounding areas. The implication is that previously unused land is being used productively, and this argument has prompted continued and regular support from the state.

To receive SCDLA and WCDA support initially, beneficiaries had to transform the CPA into a local farmers' association. This legally constituted association gives the appearance of being organised and continues to attract WCDA support. Since 2000, state support has continued to flow to the beneficiaries through a variety of projects and actions. The balance of the SLAG funds was used to purchase a tractor and some implements late in 1999. These implements are owned and managed by the CPA, which has since received other implements from the WCDA. The WCDA established a mechanisation centre on trust land in 2001. In 2003 the WCDA provided funds to build a structure for a chicken layer project on part of the trust land, which it continues to maintain and upgrade. This project is managed by a beneficiary who employs four village residents. In 2005 the WCDA established an office for an agricultural development officer in the village to assist farmers with funding and agricultural projects. A permanent store was built in 2005 for the mechanisation centre, using funds from the Comprehensive Agricultural Support Programme (CASP). The WCDA also refurbished a cottage to be used as an office by the farmers. Other projects continue to be explored.

5. Conclusion

Policy designers and implementers have very little control over the way a development process will unfold. The implementation of any development initiative is inherently complex. An actor-oriented analysis reveals that both the SLAG beneficiaries in the study village and the government officials have a sense of agency in this situation.

The land grant beneficiaries have been able to translate land reform policy in ways that have secured state interventions that satisfy their livelihood requirements. They have also been able to change their activities on the transferred land as their primary livelihood or other household circumstances change. They have not done this to satisfy the needs of state development agents and policymakers, nor have they conformed to existing models of how land reform is supposed to work.

Implementation agents have translated official policy to suit the beneficiaries' requirements, although, at times, their actions mimic policy models and intentions. Although the project beneficiaries have behaved in ways that were not anticipated by officials, a number have gained, and continue to gain, tangible benefits. Reinterpretation of these actions allows for beneficiary ‘success stories’ that give credence to the land reform policy, and state officials have responded by continuing to provide support to the project.

All the actors in this case are trying to create order out of a series of contingent outcomes. Both the development agents and the beneficiaries are generally satisfied with the outcomes. Against this background, the key question should not be whether land reform works, but how it works; not whether a development project has succeeded, but in what ways it has succeeded and in what ways it may have failed.

Acknowledgements

The author acknowledges the participation of the village residents and various development officials in the research process. Peter Jacobs and Stephen Heyns are thanked for their critical and supportive comments on earlier drafts of the paper. The views expressed are those of the author and do not necessarily reflect those of any other party.

Notes

¹MacDonald (1998:52) suggests that government officials are unlikely to challenge the identification of beneficiaries as this would challenge the representivity of the spokespersons with whom they engage.

²The original constitution stipulated that during the first five years the land could not be sold to non-beneficiaries.