When is enough, enough? Advocacy, evidence and criticism in the field of urban agriculture in South Africa

Abstract

In South Africa, the advocacy of urban agriculture as a means of improving the plight of the urban poor has been a major theme in the literature since the early 1990s. Ironically, the criticism of that advocacy has almost as long a history. To elucidate these two themes, this paper investigates the evidence on which the advocacy is based, outlines the criticism of the advocacy, and analyses the responses to the criticism. It suggests that continued advocacy in the face of disconfirming evidence is misplaced and calls for greater caution. It observes that responses to the criticism have been weak and concludes that continued advocacy and continued criticism, in parallel, hamper debate. Unless there is engagement between the advocates and the critics, not only will the field of urban agriculture suffer, but so will the urban poor.

Keywords:

• urban agriculture
• advocacy
• criticism
• stifled debate
• South Africa

1. Introduction

An important theme that emerges from recent urban studies in South Africa is the stark contrast between sections of the city that continue to fuel economic growth and those that are marginalised. The South African Cities Network (SACN, 2006:1–8 to 1–11) states that 77.31% of those living beneath the Minimum Living Level do so within 60 kilometres of the country's core urban concentrations. Further, Parnell (2004:1) asserts that inequality is increasing in South African cities. Poverty is strongly associated with cities here: it is a truism that millions of people will be eking out a living facing high levels of unemployment, crime, ill-health and inadequate service provision.

It is within the context of the poor and marginalised sectors of the city that urban agriculture in South Africa has enjoyed increasing attention over the last two decades, see Rogerson (1998b, 2003), Webb (1998b), Van Averbeke (2007). This is in line with the growing global interest in the practice (United Nations Development Programme [UNDP], 1996; Lynch et al., 2001; Mougeot, 1998, 2001; Van Veenhuizen, 2006) and points to the considerable potential that urban agriculture is said to have in ameliorating the plight of the urban poor. The potential is spelt out in terms of increased economic efficiency (Nugent, 2001; Danso & Moustier, 2006), higher levels of sustainable resource use (May & Rogerson, 1995; UNDP, 1996; Howorth et al., 2001), greater social cohesion at a community level (Rees, 1997; Slater, 2001; Van Averbeke, 2007) and a bundle of more specific benefits that could accrue to households that practise cultivation. Some of the benefits are improved nutrition and greater food security (Chiapa & King, 1998; Maxwell et al., 1998; Jacobi et al., 2001) and higher cash incomes as a result of selling produce (Binns & Lynch, 1998; Austen & Visser, 2002; Belete et al., 2005) or as a result of improved employment opportunity. Psychological well-being has also been mooted (Ngub'usim & Streiffeler, 1982; Slater, 2001; Van Averbeke, 2007).

To question the significance of urban agriculture in the light of the above claims might be considered churlish, yet the links between advocacy, evidence and criticism have yet to be examined adequately. This paper therefore considers such links in the South African situation and raises two related questions. The first is whether the point has perhaps been reached where advocates of urban agriculture will accept that there is sufficient counter-evidence to damp down their claims, and the second is whether it makes sense for critics of urban agriculture to continue to repeat their counter-arguments. To do this, I first scrutinise recent evidence that is being put forward to argue for the significance of urban agriculture. I then analyse the criticism that is presented and outline the responses that are being made to this. A consideration of these two questions should prompt deeper reflection on the nature of urban agriculture research, and this in turn should have positive implications for the cultivators and the urban poor in South Africa.

2. The emerging evidence

To place the recent evidence for urban agriculture in context, this section first outlines, chronologically, the types of findings used to advocate urban agriculture in South Africa in the 1990s. It then shifts the focus to a more detailed consideration of data that originate from early 2000 onwards, differentiating between studies based on detailed analysis and those based on weaker data.

The earliest evidence suggesting that urban agriculture might be a field demanding greater attention in South Africa came from two sources: case studies undertaken in other countries, and observations that maize production and the rearing of livestock had always been a part of urban life in certain South African cities (Rogerson, 1993b:25). Early writings such as those by Rogerson (for example 1993a:41–2) stressed that there were distinct lessons to be learned from Asian and other African countries. The most important were that urban agriculture should be encouraged, be seriously considered by urban policy makers, and be debated and researched. There was little other evidence to use in support of urban agriculture in South Africa at this time because, prior to 1992, specific investigations into urban agriculture did not exist in the academic literature. Development agencies such as World Vision, the Food Garden Foundation and Masibambane Christian Development Centre had gardening programmes with some based on experimentation (Hall, 1987:7). However, there was a dearth of information on urban agriculture at this time.

There were two problems with the early works that advocated urban agriculture. The first was that the supposedly successful case studies were based on statements that were already generalisations. The second was that these studies were repeated over and over again in the literature (Webb, 1996). The result was that the claims were substantially removed from any form of grounded data, but were nevertheless (by their constant repetition) gaining currency (Webb, 1998a). For example, over-generalisation often blurred important distinctions between rural and urban areas – see May & Rogerson (1995) for the blurring of boundaries between rural and urban issues in the case of Durban, and Wade (1981), Wayburn (1985) and Yeung (1987) for a similar blurring in the case of Shanghai. Unrealistic notions such as the benefits of planting fruit trees on street verges in India and squatter gardens in the middle of Jakarta (UNDP, 1996) also took on unwarranted importance.

From the mid-1990s onwards, quantitative data began to trickle in, necessitating the revision of early claims and advocacy. For example Schmidt & Vorster's (1995) study showed that the link between food gardens and nutrition was far from automatic. Furthermore, earlier notions that cultivation was mainly the work of women (Rakodi, 1985; Freeman, 1991) and that it benefited the most marginalised sections of the population most (Sanyal, 1987; Tricaud 1987; Jordaan, 1994) had to be moderated. In the former case, the claims in the above studies that urban agriculture was the domain of women were found to be exaggerated (Webb, 1996:150), and other studies began to emerge that demonstrated that, in some areas, urban agriculture was the preserve of men – see Binns & Lynch (1998) for conditions in Togo. In the latter case, it seemed logical that the ‘poorest of the poor’ would be the main beneficiaries of urban cultivation because of their extremely vulnerable position, especially as far as food security is concerned. Yet the data began to show a greater prevalence of urban agriculture among the less marginalised (May & Rogerson, 1995).

Thus, the data that began to cast doubt on earlier claims originated from both within and without South Africa. This led to a greater degree of caution and reappraisal. For example, my surprise at the incongruity between the very modest benefits obtained by cultivators in the Eastern Cape and the increasing advocacy in the literature prompted me to make a detailed study (Webb, 1996) of the urban agriculture literature in English. I was particularly interested in ascertaining whether the low levels of urban agriculture and the modest benefits associated with it were restricted to the Eastern Cape. The conclusion I reached was that the general literature at that time had failed to present a convincing case for urban agriculture and that its advocacy was ‘more congruent with development discourse than actual cultivation practice’ (Webb, 1996:xii). At this time, Rogerson (1998a,b) also moderated his stance – a point that I take up again in the next section.

A number of studies from further afield also began to take a more cautious stance on the matter of strong promotion of urban agriculture. For example, Binns & Lynch (1998:790) asserted that investigations had not always shown its clear advantages. Ellis & Sumberg (1998:221) stated more strongly that a ‘valid case for urban agriculture still has to be made’, and Smith (1999:15) felt that research had yet to show that urban agriculture was a source of food security and that very few attempts had been made to quantify yields.

These warnings in the late 1990s can be viewed as a positive development in that researchers began to take urban agriculture seriously. It seemed that shallow advocacy had given way to investigation of more substantive issues such as the theoretical frameworks underlying the practice of urban agriculture, methodological issues, and a recognition that far more empirical data were needed before urban agriculture could be recommended as a major development path. However, in the next spate of work undertaken in South Africa, the promotional stance emerges again in the works of some researchers. Ironically, it is based either on negative findings or flawed data. A short review of the later works is presented below. The first set of papers represents work grounded in fairly detailed research, while the second presents insufficient grounds for a generally positive view of urban agriculture.

From 2000 onwards, a number of detailed case studies were undertaken. My 2000 study (Webb, 2000) pointed to the lack of clear links between urban agriculture and nutrition in three instances. The first involved the inhabitants of Slough, a village in the former homeland of Bophutatswana. Here, no difference was found between cultivating and non-cultivating households in their dietary habits and nutritional status (Schmidt & Vorster, 1995). The second, based on my 1996 findings (Webb, 1996) demonstrated that links that were found between improved nutrition and cultivation on the basis of informant responses evaporated on further investigation. The third was a review of Chiapa & King's study (1998), showing that its claims could not be substantiated on the basis of its methodology.

Slater's study (2001) is illuminating in that it accepts that urban agriculture in Cape Town cannot be justified on traditional economic grounds. Instead, it sees the importance of such agriculture in personal and social terms. For the 14 women cultivators in this study, the practice was for solace in the face of abuse, to express new-found emotional and locational stability, to empower themselves, or to develop social networks. To prove this point, the study draws on intensive interviews based on the well-developed ‘life-history’ methodology. Questions can, however, be asked about the extent to which urban agriculture can be said to offer the depth of the benefits described. Such a characterisation of urban agriculture is close to viewing it as a panacea for an inordinate range of personal and social ills. Thus the response ‘In our culture there is no home without a garden’ (Slater, 2001:641) prompts the question as to why more women do not engage in this practice, given the benefits that have been enumerated.

Kasumba's study (2007) analysed urban agriculture from a ‘sustainable development’ perspective and considered four major aspects: food security, the contribution of urban agriculture to the cultivating household, the environmental benefits of the practice, and associated social and psychological benefits. Detailed findings revealed that such cultivation yielded produce for an average of only three months (Kasumba 2007:91), that the net value of the produce as a percentage of household income reached 6% for only 9.5% of the respondents (2007:96), and that the employment generated by such cultivation was minimal. Furthermore, environmental benefits were extremely modest, as were the social and psychological benefits as perceived by the informants.

Thornton & Nel's study (2007) on urban and peri-urban agriculture in Peddie found it to be a surprisingly low-key activity, given the levels of poverty and HIV/AIDS in this former homeland town. The following three examples illustrate this point. First, only 21% of households were found to practise such cultivation and, within this number, common plot sizes fell into the one to two square metres category. Second, for only 10 households did this cultivation have any importance – these households estimated that they saved R300 on food costs. Third, 95% of all the households felt that levels of urban and peri-urban agriculture were lower than a decade earlier. One of the contributing factors to this low incidence was that a large number of households received government grants, which provided some financial security, albeit at a low level. In the process of the investigation, Thornton was able to facilitate the sale of produce from a community garden to a local branch of a supermarket chain, and it is this development that caused the authors to suggest that urban and peri-urban agriculture carried some potential.

To collect quantitative data on the ‘material benefits’ of urban agriculture in five informal settlements in Atteridgeville, Pretoria, was the first aim of Van Averbeke's study (2007). The results were exceedingly disappointing for all classes of ‘urban farmers’ as can be seen in Table 1. The second aim was to elucidate the ‘meaning’ or significance of urban agriculture for the cultivators. In this switch from yields to meaning, analysis is downplayed. Three statements are made. The first is that cultivators view the activity as a response to poverty, unemployment and food security despite their low yields; the second that women said they were validated by cultivation, given that it was one of the roles that was customary for them to fulfil; and the third that urban agriculture provided a meaningful socialisation process for the cultivators. No details are given as to the number of informants voicing these views, the strength of their expression or their credibility.

Table 1: Selected characteristics of the three types of urban farming in the informal settlements of Atteridgeville (2001; n = 65)

Characteristic	Home gardening (n = 54)	Group-gardening (n = 3)	Crop production on open urban spaces (n = 8)
Gender of participants	89% female	100% female	60% female
Average age of participants (years)	42	57	53
Average size of individual unit (m^2/participant^−1)	26	196	351
Water supply	Irrigated	Irrigated	Rain fed
Choice of crops in order of importance	Swiss chard, maize, onions, other vegetables	Swiss chard, beetroot, tomatoes, carrots and lettuce	Maize, pumpkins, dry beans, melons
Average total monetary value of production (R/annum^−1)	41	211	213
Average monetary value of production per unit of land (Rm^−2/annum^−1)	1.58	1.07	0.61
Average contribution of farming to total household income in cash or kind (%)	0.2	0.7	1.5
Average allocation of gross value of production to home consumption (%)	100	78	100
Average allocation of gross value of production to sales (%)	0	22	0
Source: Van Averbeke (2007:340).

Yet, urban agriculture is advocated on these very grounds and the conclusion is reached that:

This study demonstrated the importance of urban farming for societies in transition, such as South Africa. (Van Averbeke, 2007:341)

Two collections of case studies on urban agriculture (Baudoin & Vink, 2001; Austen & Visser, 2002) also emerged in the early 2000s. In the first collection, only Karaan's paper (2001) had direct relevance to cultivators in South Africa, whereas the second collection contained short descriptions of eight case studies on this topic from around South Africa, with an additional fleeting reference to 36 other projects in Gauteng Province. Both collections seem to rely on inadequate data and are evaluated below.

Karaan's study (2001) relies on very weak evidence in its appraisal of urban agriculture in three poor areas of Cape Town. The data presented is vague, and the cultivators' perceptions and views are taken at face value without any analysis. The following are two examples of each. It is claimed that, by 1992, 1000 gardens were supported by Abalimi Bezekhaya, a Cape Town non-governmental organisation (NGO) that promotes home cultivation (Karaan, 2001:128). The history of this NGO suggests that support typically included training courses, garden demonstrations and the supply of seeds, seedlings and manure. However, no indication is given as to the level of support enjoyed by the cultivators at that time. There is little evidence to suggest that Abalimi Besekhaya was not fulfilling its aims in 1992, but there is just as little evidence to suggest that it was. A bald statement such as the one about the ‘1000 gardens’ means very little. The same can be said for the statements extolling urban agriculture. Karaan uses the conclusions of two newspaper reports to do this for him (2001:128), saying that the cultivation of vegetables allows the unemployed to be productive, and that cultivation on home plots is vital for household food security. These statements could very well be valid, but no evidence is given in support of them.

As far as the perceptions of the cultivators are concerned, Karaan (2001:132) correctly concludes that most costs in the production process are not apparent to the cultivators, as the following quote reveals:

I only buy manure for 60c a bag. I sold the vegetables to my neighbour and made about R349. From this I bought a sewing machine.

However, the following quotation is used as a basis for the advocacy of urban agriculture:

I find it cheaper to grow my own vegetables than to buy them because, for example, every time I need potatoes I have to go to the shop to buy, costing me money which I often don't have. So it is better if I can get them ready from my garden, whenever I want to cook food for soup. In this way I only have to buy when the garden does not grow. (Karaan, 2001:132)

This raises more questions than it answers. Is it really cheaper to produce one's own potatoes? This is often a blind spot on the part of the cultivators (Webb, 1996:245). Do the vegetables that are produced replace vegetables during the growing season or only substitute some at very low levels? For how much of the year does the garden produce crops? Without answers to these and other questions, the above quotations cannot be used as a basis for advocating urban agriculture.

Austen & Visser's report (2002) on urban agriculture in South Africa has three noteworthy features: their strong advocacy of urban agriculture, their view that debates as to the efficacy of urban agriculture in ameliorating the conditions of the urban poor are over and what is needed now is for local government to put clear and appropriate enabling policies in place, and their description of the eight projects that were used to prop up the above views.

As far as Austen & Visser's (2002) advocacy of urban agriculture is concerned, its justification is the widespread practice of cultivation in the cities of South Africa and beyond. Consider the following statements:

Urban agriculture is already a reality in South African cities, as this report will demonstrate. (Austen & Visser, 2002:1–1)The fact is that farming is taking place within urban areas all over the world. (Austen & Visser, 2002:1–4)

The above statements, together with the recommendation that urban agriculture deserves substantial institutional support, prompt many questions. For example, should an activity enjoy the full panoply of support of different tiers of government, NGOs and business simply because it can be observed? How far can the notion of ‘urban agriculture’ be stretched? Does the fact that people cultivate gardens – often measuring only one or two square metres (Thornton & Nel, 2007) – mean that they are farming? Is it reasonable for researchers to claim urban agriculture as a new discovery merely because it is found to be practised globally? Cultivation has always taken place within cities and the motivations for the activity are many and varied (Page, 2002). This particular assertion, that cultivation is an observable activity in cities and as such should be assisted so that the poor may benefit from its inherent advantages, has been the constant refrain of advocates of urban agriculture since the mid-1990s. It is based, however, on a flawed argument: the fallacy of asserting that something is right or good simply because it is has existed for a long time. And, as the previous section has shown, recent in-depth analyses seem to point in the opposite direction.

The second feature of Austen & Visser's report (2002), the claim that the debate about the efficacy of urban agriculture is over and that enabling policy is overdue for implementation, is misleading. Nevertheless such claims appear throughout the report:

Recently, however, discussion on urban agriculture seems to have moved away from this debate with the acceptance of urban agriculture as a reality in the cities of the developing world. (Austen & Visser, 2002:1–2)

The fact is that farming is taking place within urban areas all over the world. How to implement it in an effective and productive manner at a local level seems to be the next challenge. (Austen & Visser, 2002:1–4; emphasis in the original)

The statements are misleading because the debate is far from over. A number of researchers have called for further investigation into the issues around urban agriculture because there is so little rigorous data indicating that the practice could be sustainable, that it is an unequivocal source of household food security or that it provides the benefits that the cultivators are said to enjoy (Binns & Lynch, 1998; Smith, 1999). Almost every recent in-depth study in South Africa, as outlined above, is faced with findings that suggest urban agriculture is not as productive and important as might be expected. And in this report the authors themselves admit to problems. For example, in the paragraph that follows the above quotation, they say:

A distinct lack of data on urban food security, markets and urban food production is a constraint. So is a lack of environmental awareness, a lack of extension services, and difficulty experienced in accessing farm inputs such as technology, water and fertiliser. (Austen & Visser, 2002:1–4)

Thus, to promote urban agriculture on the simple grounds of its ubiquity seems not such a simple task after all.

The third notable feature of this report is the way its empirical basis is handled. Given that the data are presented to lend credence to statements about the ubiquity of urban agriculture in South African cities, it is surprising that the case studies offer little useful information. For example, in a section on urban agriculture in Gauteng, it is stated that 60 projects were being managed by the Gauteng Department of Agriculture, Conservation, Environment and Land Affairs (DACEL) and that 40 more were being considered (Austen & Visser, 2002:2–4). Brief descriptions of some of them are offered, such as:

Russell McGregor, the DACEL Deputy Director responsible for the sub-directorate, mentions another DACEL project with six cultivators selling their produce to the surrounding community. (Austen & Visser, 2002:2–4)

Or, describing a garden in Orange Farm (Johannesburg):

The garden is located on the premises of Radipabi Primary School and supplies fresh vegetables to 65 families. Some of the vegetables are sold, thus also providing a small income for the cultivators. (Austen & Visser, 2002:2–5)

The eight gardening projects visited by the authors are described in greater detail at the end of the report. An analysis of these descriptions shows that very little useful information is given about seven of them. The descriptions include generalisations that offer little help to an analyst trying to assess the significance of urban agriculture to the participants. The following are two examples of this inadequacy.

The authors conclude that the Feed the Child, Feed the Nation environmental project, an activity of the Banareng Primary School in Atteridgeville, Tshwane, is ‘a model project for urban agriculture and environmental awareness’ (Austen & Visser, 2002:A–21). The positive features of the project are the re-use and recycling of organic waste and grey water, the addition of vegetables to a daily meal enjoyed by the students, other educational benefits associated with making the students environmentally aware, and the sale of vegetables. That a school garden could provide such a range of benefits is laudable. However, no data are supplied to demonstrate that the stated benefits accrue to the recipients as claimed, and some of the claims are dubious. An example of the latter is the claim that a single bath (used as a container for grey water) can provide sufficient irrigation for vegetables for 670 students daily. This is simply not credible.

The Eastridge-Beacon Community Development Project, as described by the authors, is the only one of the eight case studies which does offer some quantitative data to establish some of the effectiveness of the benefits, but only in reference to the project's welfare function:

The project has assisted approximately 1500 recipients through the daily provision of a R100 food parcel for 10 people. (Austen & Visser, 2002:A–12)

The project provides daily and special weekend sales of vegetables to the community and a national supermarket chain also buys some vegetables. Yet the authors supply no specific data about these activities. A feature of the project that seems to undercut its sustainability is the extraordinarily high level of aid that it receives from outside bodies. Wages for the workers on the project are paid by funding from the Department of Social Development; and Telkom (a South African telecommunications company), government departments at various levels, various NGOs, and a faculty of a tertiary institution all play some role, the nature of which is not specified. It seems pertinent to point out here that one of the pillars of sustainability is that the dependence of livelihoods on external support should be minimal (Department of International Development [DFID], 1999:6).

The conclusions reached by the authors that the projects they reviewed were very successful, that they were models of urban agriculture, and that they benefit community members have not been substantiated. The projects might exhibit all the possible qualities listed above, but they have not been shown to do so. The findings must therefore be treated with caution, especially when more comprehensive analyses show opposite tendencies.

A more recent work suggesting considerable potential for urban agriculture is the study by Belete et al. (2005) on the Buffalo City Municipality. In these authors' view, the potential for urban agriculture in this area is a function of a suitable climate and the general benefits inherent in this type of agriculture. This positive assessment is of a most general nature. The benefits listed simply reiterate ideas from the early 1990s that have already come under criticism. The Municipality's potential is promoted in terms of existing market facilities, an adequate infrastructure and a regional climate that is considered conducive to good vegetable yields. No attempt is made to produce local-level findings or disaggregate them and the human dimension is absent from the analysis. Further undercutting the authors' argument is the admission that ‘there is insufficient available information to evaluate the extent to which urban and peri-urban food production can satisfy household needs’ (Belete et al., 2005:29). There seems thus little empirical basis for the potential they advocate.

A common finding of studies that have appeared since 2000 is that in-depth investigations show the benefits of urban agriculture to be disappointing. Yet studies based on insubstantial or unconvincing data continue to present a largely positive conception of the practice. It seems reasonable to suggest that this is a situation that calls for caution. The following section looks at criticism that has been made and the resultant responses.

3. The criticisms of advocacy and the resultant responses

Kasumba distinguishes three classes of scholars working on urban agriculture: those who advocate it, those who take a cautious stance towards it, and ‘those who are critical of common conceptions of it’ (2007:8). It is the third group that is of immediate interest here and this view is exemplified by the works of Page (2002) and Webb (1996, 1998a,c, 2000). Both employ the notion of Foucauldian discourse (a domain of knowledge that structures social practice) as the main analytical frame to their arguments and both draw on insights from Ferguson (1990).

It is important to note that neither Page nor I take a stand against urban cultivation per se. Our criticisms are levelled at way urban agriculture has been constituted by those promoting it. In Page's case (2002), urban agriculture is considered to operate as an ‘anti-politics machine’, in that

in the 1990s urban agriculture absorbed the dissatisfaction of citizens whose standard of living fell rapidly during the implementation of structural adjustment policies. (Page, 2002:42)

In my studies I view the common conception of urban agriculture as a construction of the development discourse designed to meet the needs of the development institutions. As such it ‘is a poor reflection of urban cultivation in practice’ (Webb, 1996:272). Thus Page and I both consider common conceptions of urban agriculture to be flawed. Because Page's study (2002) is based in Beau (Cameroon), it has only passing relevance to South African conditions. Therefore the rest of this discussion considers my criticisms.

A recurring theme in my writings is that insufficient grounds exist for promoting or advocating urban agriculture in South Africa. Initially, I levelled this argument at the common conceptualisation of urban agriculture itself, its growth as a field of investigation and its alleged benefits (Webb, 1996). Subsequently I focused on the inadequate way in which crops were treated in the literature (Webb, 1998b), the discrepancy between the ecological benefits attributed to urban agriculture and those arising from its practice (Webb, 1998c) and the link posited between urban agriculture and nutrition (Webb, 2000). In each case, I asserted that the claims that were made for urban agriculture had insufficient evidence to back them up.

Responses to my criticisms are instructive and appear in the works of Rogerson (1996, 1998a,b, 2003), Austen & Visser (2002), Kasumba (2007) and Thornton & Nel (2007). The following is a brief overview of these responses.

Rogerson (1998a:15–16) finds it ironic that at the time when local authorities were beginning to develop policies to promote urban agriculture, such agriculture began to be questioned. He acknowledges that some of my (Webb, 1996) findings confirm some of his own, especially as regards the economic status of most of the cultivators. These new findings moderate his earlier advocacy, his conclusion being that ‘the jury must, perhaps, still be out on the question of urban agriculture's role as a means to augment the asset base of South Africa's poor households’ (Rogerson, 1998a:17). The same conclusion is reached in a follow-up paper in the same year (Rogerson, 1998b:182). However, five years later his view seems to have changed. Rogerson (2003:146) considers local authorities to have made advances in policy to promote urban agriculture, with the city of Midrand (Gauteng) leading the way. In addition, he considers new research, notably by Slater (2001), to have demonstrated the social benefits of urban agriculture apart from any economic ones. These developments, together with six instances of urban agriculture that seemed to show promise in the Midrand area, have caused Rogerson (2003) to revert to an advocatory stance as follows:

These lessons of best practice offer starting points for the development of much needed guidelines to enable urban agriculture to become a pro-poor intervention in South African urban development planning. (Rogerson, 2003:151)

Austen & Visser's (2002) report tends to treat my (Webb, 1998c) objections to the advocacy of urban agriculture in two ways. First, the objections are acknowledged, but then by a twist of logic they are appropriated for their own brand of advocacy. Second, the criticisms are simply brushed aside in the wake of what they consider more relevant aspects. Here is one example. Consider the following statement:

The argument for a greater emphasis and analysis as an activity rather than a development path fits in with the new trend in this field focusing on policy and best practice for the implementation of urban agriculture, rather than promoting only its potential for development. (Austen & Visser, 2002:1–3; emphasis in the original)

This excerpt gives tacit recognition to my (Webb, 1998c) call for the practice (or activity) of urban agriculture to be subjected to deeper analysis. However, this analysis is glossed to mean preparation work needed for the implementation of urban agriculture. By invoking this type of analysis as a basis for implementation, the authors actually weaken their case – policy is purportedly being based on analysis of the practice of urban agriculture, but it is actually promoted on the basis of generalisations and very little analysis, as explained earlier. By also invoking ‘analysis’ as a springboard for ‘best practice’, they further weaken their case. The present paper has also demonstrated that examples of ‘best practice’ are far too crudely described to qualify as outstanding examples of urban agriculture and that the ‘best practice’ appellation has no objective basis.

Kasumba's (2007) study was set within the framework of sustainable development. As such, it sees urban agriculture as an activity that epitomises the sustainable development approach. Its response to my (Webb, 1996, 1998b,c, 2000) findings, is to seek to test these in a different setting in the Eastern Cape, given that my original study was based on three Eastern Cape centres. In this case the approach is sound – claims are tested in different contexts to either establish or refute their general applicability.

In Thornton & Nel's view, my studies reveal ‘built-in barriers’ that hinder urban and peri-urban agriculture (2007:14). They do not enlarge upon these in any way but use them as evidence that confirms their findings – that urban agriculture is not a significant activity for many of those who practise it. Their glossing over these barriers means that they are able to advocate urban agriculture particularly for marginalised households, as follows:

Finding ways to partially assist in the situation of the most marginalised is clearly of critical importance and within this context the role of urban and peri-urban agriculture should not be overlooked. (Thornton & Nel, 2007:19)

A summary of the responses to criticisms of urban agriculture in South Africa reveals the following. Rogerson's earlier acknowledgements of the criticisms (2003) pale in the light of a new impetus promoting and supporting urban agriculture, as well as new research emphasising non-economic benefits. The actual criticisms, however, are never addressed. Austen & Visser (2002) consider some criticisms but sweep them aside because, in their view, urban agriculture is already being practised in the urban areas of South Africa. Furthermore, they say it ‘should be promoted in order to fully harness its substantial social and economic benefits’, while policy guidelines will overcome problems and negative impacts (Austen & Visser, 2002:5–4 and 5–5). Thornton & Nel (2007) use the criticisms to support their own findings, while Kasumba (2007) seeks to test them in a new context. Overall, it seems that the criticisms have only had a very minor effect on urban agriculture research in South Africa.

4. Conclusion

This paper investigated the advocacy of urban agriculture in South Africa, the evidence on which the advocacy has been based and the criticisms levelled against such advocacy. It argued that the strong advocacy of the early 1990s moderated to some extent towards the end of the decade as a result of works that criticised this stance or called for caution. Since 2000, any caution seems to have been swept away in the face of higher levels of institutional support for urban agriculture and increased cultivation. The paper also argued that both the early and later advocacy was based on inadequate evidence, and that criticism of the advocacy and the data on which it was based has had only had a limited effect.

The fact that the advocacy of urban agriculture in South Africa first began in the early 1990s and that detailed criticism dates from the mid-1990s, prompts the question ‘When is enough, enough?’ When we look at the evidence that urban agriculture in South Africa does not provide the benefits so often attributed to it, then it seems there has been enough advocacy. There seems to be enough of such evidence for development institutions, local authorities and NGOs to at least adopt a cautious approach to the promotion of urban agriculture as a path out of poverty. The continued advocacy of such agriculture may be considered irresponsible at least until some valid evidence or grounds for the activity can be elicited from those practising it. It also seems there has been enough criticism, however valid it might have been. Little progress can be made in the field of urban agriculture if the advocates persist in promoting it, the critics continue to criticise common conceptions of it, and no real engagement between the two groups takes place.

A two-pronged course of action could be adopted in order to move the concerns about urban agriculture (as described in this paper) from abstract debate to serious consideration by local authorities. The first would be to have these concerns raised with two of the South African Local Government Association's (SALGA) working groups – Economic Development, Planning and Environment, and Social Development. If deemed urgent enough, these concerns could be raised at the National Members Assembly and the National Conference. Given the national nature of SALGA and its status, the potential outcome of such discussions could be substantial. The second course would be to make contact with the relevant directorates of local municipalities. Discussions could involve the relevant standing committees and the local councillors. A way forward would then be decided by vote. Raising such concerns at both local and national levels means they would be less likely to fall by the wayside.

It must, however, be stressed that, despite assertions to the contrary as expressed in Austen & Visser (2002), debates on the benefits of urban agriculture are far from over. In fact, sustained debate has yet to take place and, together with grounded analysis, is as urgent as ever. The urgency is not only to develop a viable and mature research field but also to have an impact on the livelihoods of the millions of urban poor themselves. Without clarity on the benefits and significance of urban agriculture, current and potential cultivators face high levels of uncertainty and risk.