Donor-driven local economic development in peripheral areas of KwaZulu-Natal: The Gijima Programme

Abstract

The Gijima Programme in one of KwaZulu-Natal's peripheral economic areas represents a fourth phase of local economic development (LED) approaches in South Africa. This paper assesses the programme by comparing it with international trends and recent developments in South Africa. Despite some noteworthy innovations, the overall programme reflects a supply-driven approach that seldom considers projects funded from the market side. This is evident from the overall approach and the quality of the plans, both of which tend to focus on ensuring legal compliance rather than on the quality of the end product. The sectoral distribution of projects seldom takes account of the latest knowledge economy requirements or the need for differentiated products and services that are appropriate for the poor. Partnership formation remains a major stumbling block. Consequently, projects are dominated by supply-driven approaches and their pro-poor nature is not well defined.

Keywords:

• local economic development
• local economic development partnerships
• pro-poor local economic development
• Gijima
• KwaZulu-Natal

1. Introduction

Since the early 1990s, local economic development (LED) practice in South Africa has had mixed outcomes (Nel & Rogerson, 2007). Significant outcomes have mostly been achieved in the larger urban areas. The smaller urban settlements and the rural areas have generally achieved few remarkable outcomes (Nel & John, 2006). Nel (2005:253) argues that, despite there being a bias towards LED research in larger urban areas, ‘some of the most profound development challenges are being faced and fought in the innumerable smaller centres’ (see also Nel & Rogerson, 2007). There is thus an urgent need to expand the successes of LED to include smaller urban settlements and rural areas. In KwaZulu-Natal, the eThekwini metropolitan area (mainly Durban) dominates the provincial economy, since it generates one-third of the GDP. Against this background, the Department of Economic Development and Tourism (DEDT) in the province, supported by European Union funding, initiated the Gijima Programme so as to support LED outside the metropolitan area. (The programme excluded eThekwini.)

This paper evaluates the projects initiated under the Business Enabling Fund (BEF) and the Local Competitive Fund – Competitive Action Plan (LCF CAP). It argues that, because LED is mainly based on donor-driven and government-driven approaches, the overall approach to LED funding in this programme is supply driven, with the result that it prioritises inputs (infrastructure, donor finance systems, technical expertise) and accords scant consideration either to markets or to ways to open markets for the poor. The paper recommends a more balanced approach, one which also takes account of the demand side (i.e. the market).

The paper begins by reviewing the phases of LED internationally and nationally and then evaluates the Gijima Programme in terms of the international and the national frameworks. Methodologically, the paper is based on an evaluation conducted during 2008, which focused on 17 projects (14 LCF CAP and three BEF). The main methods were in-depth interviews with all the partners in the 17 projects and with key officials of the DEDT. Where possible, focus group discussions were held with project beneficiaries. At the same time, specific information about the projects was also taken into account.

2. Evolution of LED and LED funding mechanisms

There seems to be no consensus about what caused the rise of LED. Nel (1999) argues that two contributing factors were increasing levels of globalisation and an overall tendency to roll back the state in economic development (and also in other spheres of public involvement). Added to this were increasing levels of decentralisation that necessitated local responses to those two factors. Some scholars believe that the limited success of state controlled regional planning also played a role (Tomlinson, 1993).

The World Bank (2005) distinguishes four waves in the development of LED thinking. The First Wave (1960s–1980s) was characterised by investments in infrastructure and manufacturing, the former usually related to the latter. The Second Wave (1980s–1990) saw a continuation of the focus on manufacturing, but new aspects came to the fore, such as a focus on inward investment, sectoral strategies, business extension and expansion, and city regeneration strategies. By the 1990s, the Third Wave had brought a shift in emphasis to a range of far more complex aspects, such as improving business environments; working in functional economic space instead of just political territories; an additional focus on soft infrastructure development, such as human resource development; partnerships; multi-sectoral approaches; institutional approaches; targeted strategies; and the coordination of government strategies. The increasing prominence of network development and network exchange meant that this wave also saw an initial focus on knowledge-based issues. The Fourth Wave was further influenced by these issues because human resource development and information technology development and innovation had become essential. Other prominent aspects were strategic (as opposed to general) investments and the link between regional and local development. Although some of the defining characteristics of a particular phase may be discerned to some degree in one or more of the other phases, the overall picture is a reasonably accurate reflection of the change in the world economy from a production-dominated one to one where knowledge plays an essential role. It is therefore not surprising that Saxenian (1989) identifies the following factors that had played a role in successful regions and localities: a distinguished research centre, access to venture capital, a skilled labour force, a nearby international airport, cultural and natural amenities, and a high quality of life. Anderson and Tromguist (1980, cited in OECD 2007) identify five crucial factors that they refer to as ‘the five Cs’: competence, capital, communication, creativity and culture.

South Africa's transition from apartheid to democracy opened the South African economy to the influences of globalisation (albeit a decade or two later than the rest of the world), and an extensively decentralised system of governance was implemented at both the local and the provincial levels (Rogerson & Nel, 2005). Decentralisation meant that localities were required to plan for their own development by means of a range of methods. The most prominent type of local response entailed integrated development plans, which also included planning for LED. In contrast to the international context, in South Africa the factors that led to the rise of LED were the high levels of political decentralisation and the realities of globalisation since 1994.

Nel (2007) has provided a fairly comprehensive overview of the changes in LED in South Africa. The description below also builds on studies conducted by Meyer-Stamer (2002, 2003), Nel & John (2006) and Rogerson (2008). Economic development during the apartheid period (before 1990) was heavily influenced by apartheid planning and economic decentralisation policies aimed at supporting the apartheid ideology. Overall, LED practice was limited. In cases where it did feature (mostly in the larger urban areas), it was focused on economic boosterism. Some of the key attributes and activities of LED during this phase were the attraction of industries, place marketing, infrastructure development, land development, tourism development and public works programmes (mainly directed at poor whites).

During the second phase (1990–1998), two opposing approaches developed, community based and city based. The Stutterheim Project is one of the better examples of the former (Nel, 1999). However, the most impressive results obtained during this phase were probably due to the approaches followed by larger urban areas (Nel & Rogerson, 2007). Not only did these cities spend a great deal of time developing policies, but urban renovation (in line with the international trends) was also high on the agenda (Nel, 2005). By the mid-1990s, many of the community-based initiatives were beginning to come under pressure because donor funds were more likely to be channelled through government structures. At the same time, the need to put a pro-poor response in place (as the community-based initiatives started to diminish) became prominent (Khanya-aicdd, 2006).

Against this background, the government introduced the LED Fund, which heralded the third phase (1998–2002). This Fund financed a range of pro-poor projects, with a poverty relief focus. However, the success of these projects beyond the funding cycle of the Fund was limited (see also Marais & Botes, 2007). They were typically hampered by poor design, the exclusion of local business expertise, domination by the public sector, and inappropriate technical advice (mostly provided by private sector consultants) (Nel, 2005; Marais & Botes, 2007). Consequently, the Fund was discontinued in 2002. Meanwhile, parallel to the programmes financed by the Fund, larger urban areas were implementing inner city renewal programmes, establishing convention centres (places for sharing knowledge) and generating active partnerships between business and the public sector (Nel & Rogerson, 2007). With the demise of the Fund, the focus shifted to the national policy development process. Although the outcomes of this process reflected an adequate policy approach, this meant, in practice, that LED would be actively linked to infrastructure development. At the same time, during the mid-2000s, the Industrial Development Corporation began establishing LED agencies across the country.

The question to ask at this point is how far and how precisely South Africa has followed international trends in this regard. Nel succinctly summarises the situation:

With reference to 3rd/4th Wave strategies, it is apparent that South Africa is pursuing some of them, e.g. coordinating government strategies, having quality of life goals, and targeted strategies. However, action in other areas is limited or lacking, including: working in functional economic spaces [and] the widespread use of partnerships, and there is only limited knowledge exchange, and HR and IT development. (2007:32)

Other problems identified in the literature are: an extensive and apparently ongoing debate about the role of the state, dominance by the public sector, narrow sectoral focuses, limited capacity to drive the process, rapid turnover of staff, and an excessive focus on narrow projects with only limited economic viability. Moreover, evidence of well-implemented pro-poor LED has been limited (Meyer-Stamer, 2002; Marais & Botes, 2007; Nel, 2007; Rogerson, 2008).

I would add that the donor nature of the Gijima Programme, and its requirements and systems, played a specific role in creating a supply-driven approach to LED – an aspect often neglected in research (Wallis, 1999). Other research that has noted the impact of the donor community in South Africa is Harrison (2006), which clearly indicates the role of the GTZ (German Technical Cooperation) in creating development planning guidelines. My argument in this article is that it is the systems requirements in respect of donor funds – not only the conceptual influences in respect of projects – that have significantly shaped the outcomes of the programme.

3. The Gijima Programme

It was against the backdrop of limited LED successes in smaller and specifically rural settlements and the practical problems outlined above that the Gijima Programme was designed. Also known as the KZN LED Support Programme, this is a 6-year programme (designed in 2003) aimed at supporting the provincial DEDT and a broad range of interested parties, and ensuring that LED is implemented more effectively in order to achieve equitable economic growth in KZN. Its three main objectives are to promote pro-poor LED, build local government capacity, and increase local competitiveness by building partnerships (DEDT, 2007).

There are three operational sub-programmes, the BEF, the LCF CAP, and the Local Competitive Fund Implementation (LCF IMP) – a fund that provided finance for project implementation (DEDT, 2007).This paper discusses the first two, but their relationship with the LCF IMP should not be ignored. The BEF sub-programme focuses on role clarification and regulatory issues in relation to LED, LED strategic planning, government coordination and the link between critical infrastructure and LED. At the time of writing, 87 BEF projects were being funded to the amount of R23 million. The three BEF projects evaluated in this study were a manufacturing development plan, a plan for an economic development agency for the Amajuba District, and a tourism development plan for the Kwa Sani Municipality.

In turn, LCF CAP funds were made available for developing business plans, financial plans, feasibility studies, operational plans, sectoral plans, LED plans and marketing plans (DEDT, 2007). It should be noted that the planning processes and the implementation processes were separated. This was in direct contrast to the practice followed in the LED Fund process, where there was a much closer relationship between the planning and the implementation processes. In the LCF CAP process, an attempt was made to minimise the risks of poor planning, and it was argued that implementation should take place through a range of funding mechanisms, that Gijima should not be the only financier, and that private sector financing needed to be procured. By June 2008, 74 LCF CAP projects were being funded. The monetary value of these projects was R27 million.

This study evaluated 14 LCF CAP projects: four business support or skills development projects, two manufacturing projects, five agricultural projects and three tourism projects. The 17 projects considered in this study are only a small sample of the more than 150 projects eventually funded. However, at the time of the study these were the only ones to have been completed.

4. Evaluating the Gijima Programme

Although a much broader evaluation is available (Marais, 2008), this section evaluates only four aspects of these 17 projects that were crucial to the design of Gijima: the funding process and support for and quality of plans; sectoral choices; partnerships; and supply-driven versus demand-driven approaches to the pro-poor nature of projects. The concept of partnership was important as it was one way Gijima distinguished itself from other similar programmes; its pro-poor nature was important because it was fundamental in the development of the LED Support Programme; and sectoral choices were important because several sectoral changes have occurred in the national and provincial economy over the past 30 years, while Gijima also had some objectives in this respect. (See Table 1 for a list of the projects.)

Table 1: List of Gijima projects evaluated for this paper

Local Competitive Fund – Competitive Action Plan	Types of LCF project
Local Business Support Centre – Umgeni	Business support or skills development
LED financial services model for excluded areas
Edendale Construction and Training Centre
Project Gateway
Nkandla Oils – Gelfuel Manufacturing	Manufacturing
KZN furniture manufacturing cluster
Illovo Sugar Ltd	Agricultural
Ugu Small Growers – organic farming for small-scale farmers
Ikususa Forestry Cooperative
Banana Outgrowers Business Plan Development
Kwanalu (Kwazulu-Natal Agricultural Union) BEE desk
Midlands Rail and Tourism	Tourism
Southern Midlands Railway Tourism
Kwasonto Tourism Development
Business Enabling Fund
Tourism development plan
Amajuba Regional Development Agency
Amajuba Manufacturing Development Plan

4.1 The funding process, support and quality of plans

The European Union logical planning framework format was used in the compilation of application forms, and applications had moreover to include a budget and a partnership agreement. Some interviewees mentioned how cumbersome the application forms were, which in many cases meant the partnership group had to source the skills of a consultant to complete them. Area managers appointed by the Gijima Programme and responsible for longer-term quality insurance also provided technical advice to help with the application processes.

The evaluation of plans was entrusted to technical experts, who judged the applications independently and impartially, without any political bias. However, the logical planning framework, as a template against which business plans (LCF CAP) can be assessed, is questionable. In essence, it fails to provide preliminary evidence of business viability, and is primarily a project management rather than a business management framework – although there certainly are similarities between the two types.

Overall, an efficient process, one ensuring that the most promising projects were funded, was in place at the start. However, this level of efficiency was not sustained in the remainder of the project so as to ensure quality. Planning was poor, for several reasons. One problem was that the emphasis on LED was based on the assumption that local actors would ensure good quality. This assumption seems to have been unfounded: there was little evidence that councillors, local government officials and other role players were able to engage with the work of consultants. For example, in some of the BEF projects there was evidence that material had been cut and pasted from other reports, as some irrelevant statistics relating to the Free State Province were included. The obvious reaction in such cases is to blame the consultant. Yet neither the officials nor the council who had endorsed the plan had detected this problem.

Another problem was inability to assess the available data and draw appropriate conclusions. For example, in one of the plans the need for a manufacturing plan was justified in terms of, inter alia, the levels of unemployment: ‘There are problems such as lack of investment, high levels of unemployment, absence of [an] integrated, focused and proactive manufacturing strategy … A need therefore arises to formulate a manufacturing strategy’ (Amajuba District Municipality, 2008:92). This specific manufacturing plan had thus also been developed without considering the risks involved in the centralisation of manufacturing activities across South Africa over the past two decades (Nel et al., 2006).

Furthermore, it was also not uncommon to find that economic opportunities for the manufacturing industry had been identified on the basis of location, availability of land, adequate infrastructure, ‘attractive’ rates and taxes, the availability of semi-skilled labour, the availability of a local market and/or the availability of certain agricultural products for agro-processing (Amajuba District Municipality, 2008). The discrepancy between these factors and those mentioned in the international literature – competence, capital, communication, creativity and culture – is evident. The silence regarding the aspects emphasised in the international literature probably indicates that plans were made without sufficient background understanding of the international trends.

There were also serious concerns about the quality of the business plans developed according to the LCF CAP framework. The first major concern was that none of the 14 projects had come to the conclusion that there was no business case. Such a situation is the direct result of a system that is either input or supply driven. Fundamentally, consultants driving these processes are mandated to argue for the business case, and if they succeed in procuring funds for implementation they could well end up managing the implementation process. Consequently, only 15 per cent of business plans included any form of risk assessment (these had assessed the potential risks of the project but, alas, not its quality); just over 50 per cent included an analysis of potential competitors, and one-third included an assessment of the existing value chain. One area manager summed up thus: ‘The project is service provider driven. It has a bad business plan’ (interview, Gijima Area Manager, Pietermaritzburg, 14 March 2008). The question is, who is to blame for the fact that the project is service provider driven? Obviously, the partnership groups cannot be exonerated; but the overall supply-driven approach, whereby money is made available for a product without a thorough evaluation of that product, is probably a far more significant contributing factor. Furthermore, the poor quality of plans can also be ascribed to the fact that the developers of LCF CAP business plans commonly assumed they would receive funding for implementation from the LCF IMP funds. The absence of appropriate risk assessments and other important attributes of sound business plans can thus be put down to the fact that these plans were developed for an input-driven grant-funding scheme, and not for a funder expecting a return on investment.

In summary, the biggest concern about the way the BEF and LCF CAP plans were developed is probably the over-simplified way opportunities or threats were identified. The question is, if the overall approach in the project design had been determined by a commercial bank, in what respects would the application process and outcomes have been different? The argument here is not in favour of one particular approach but rather to point out the need for a better balance. At the same time, the independent way the initial applications were evaluated should be commended since this ensured that the programme was managed outside the influence of local politics.

4.2 Sectoral choices

The international literature suggests that sectoral choices made in terms of LED have changed considerably over the past 40 years. Overall, there has been a shift away from a production-oriented economy towards a knowledge-based one. An assessment of the Gijima projects in sectoral terms suggests, in the first place, that the largest percentage of the LCF CAP projects was focused on agriculture (45 per cent), manufacturing (19 per cent) and tourism (16 per cent). There was little evidence to suggest that projects were focused on aspects of the knowledge economy in accordance with either the international trend or the Gijima guidelines. Considering the pro-poor nature of the projects, the limited extent of the linkage to the knowledge economy does not come as a surprise, and the focus on basic production is thus not inappropriate. An analysis of the BEF projects does not suggest any justification for coming to a different conclusion. In fact, only one of 87 allocated projects was being conducted in the ICT sector. Most followed sectoral policies that seldom considered the realities of the knowledge economy.

Concerning the divide between the knowledge economy and the relatively low skills levels of poor people, the question arises as to whether it would be possible to bridge the gap. Because of the low skills status of most of the poor, their involvement in the economy is usually at a low technological level, while their services and products are undifferentiated (e.g. the cultivation of sugar without the use of sophisticated farming practices). The problem is that because there is large-scale competition in the production of such products, the poor are generally unlikely to be able to compete on an equal basis, since they are seldom likely to reap the benefits of scale. An alternative to the knowledge economy approach is to focus on what Toerien (2005:72) refers to as ‘low technology but differentiated projects’. Two Gijima projects could be said to provide differentiated but low-technology products or services, namely Nkandla Oils (production of oil from herbs in an organic process) and the Ugu Small Growers (organic vegetables); yet being able to offer such products does not automatically ensure viability.

4.3 Partnerships

The supply-driven approach was also evident from the fact that partnerships were a requirement in the Gijima guidelines. This requirement meant that most applicants probably formed partnerships to comply with the guidelines rather than to open markets. Table 2 gives details of the partnerships.

Table 2: Overview of partnership arrangements in the 17 projects in the study sample

Main beneficiary	Kinds of partnerships
NGO (LCF CAP)	Two partnerships between NGOs (with the community/municipality as part of the process but not part of the partnership)
	One partnership between organised agriculture and emerging farmers
	One extensive partnership involving private sector, public sector and NGOs
	One NGO with public sector
Public sector (LCF CAP)	Two partnerships between municipality and Chamber of Commerce
Private sector (LCF CAP)	One partnership with community (to be established as a community legal entity), a private investor, and municipality
	One partnership with a larger producers' organisation (NGO)
	Two partnerships with community
	One partnership between cooperative and larger firms
	Two partnerships with a range of private-/public-sector partners and NGOs
Public sector (BEF)	Two partnerships with Chamber of Commerce
	One partnership; limited indications of a possible partnership

Although it is difficult to draw conclusions from this, some preliminary points should be noted. Public sector beneficiaries mainly created partnerships with organised business. These usually meant that representatives from organised business served on steering committees related to the specific project. However, in these cases there was seldom any indication of a partnership between an entrepreneur and a local municipality – a characteristic of most such partnerships in the global North (Nel, 2007). There are two main problems here: in the first place, chambers of commerce are seldom appropriate partners because they are usually some form of ‘social club’ for business people, and are moreover highly susceptible to institutional and leadership change; and secondly, the emphasis on complying with the requirement to form a partnership, rather than on establishing a partnership in order either to raise capital or to open markets, is another indication of the supply-driven approach that was followed in the Gijima Programme. This reality prompts two questions. First, to what extent does such an approach reflect an underlying assumption in a programme designed and funded by government or a donor? Second, would a similar project, managed by a commercial bank, not have required a partnership that emphasised procuring capital or ensuring market access? The fact that none of these projects where there was a partnership between the public sector and business non-governmental organisations (NGOs) were ultimately implemented (i.e. received money to implement the project and progressed from plan to action) is probably an indication of the problems that have been pointed out in this paragraph.

NGOs were more likely to establish partnerships with other NGOs, and of course by definition NGOs are more accustomed to working as partners. Interestingly enough, two of the NGO partnership projects, involving NGO partnerships only, were also implemented, but with additional funding from non-commercial sources. This suggests that NGOs usually have well-developed funding networks. However, more importantly, it might also suggest that NGOs usually think along similar lines, which makes partnerships of this kind more feasible. Thus, the question can be asked whether the Gijima Programme's supply-driven approach favoured NGOs. However, it should also be asked whether NGOs have longer-term funds available, and whether they have the ability, at scale, to act as agents of economic development.

The most far-reaching partnerships were between private sector partners, in the form of direct partnerships between rural communities, or between small and established producer organisations, or between emerging and established enterprises. Yet the reality is that only two of these projects culminated in some form of implementation.

Four further critical comments are relevant here. First, the impression received from the interviews was that the Gijima partnerships could at best be described as ad hoc. There was little evidence to suggest that these partnerships that had been initiated through a project-by-project approach had been further developed into longer-term arrangements – this in spite of the fact that assurances had been given that such partnerships would be continued.

Second, the existence of a partnership agreement did not reduce the risk of non-performance by some partners. Project contributions, especially from the public sector, were hampered by a lack of commitment on the part of senior management and by political turmoil at the municipal level. An LED manager involved in one of the projects aptly summarised the situation, saying ‘There is a need for municipalities to cooperate – and another challenge was that other senior role players did not commit’ (interview, Gijima Area Manager, 14 March 2008). There seems to be very little evidence that government-driven LED has been more successful than a partnerships approach to LED.

Third, interviewees expressed concern about when precisely private sector investors should be introduced during the project cycle, and to what extent projects are hampered by the public sector's inability to provide adequate support. In two of the tourism projects, private sector partners were involved from the outset but lost interest because of the lengthy process and political in-fighting. A manager of one of them expressed frustration thus:

In our opinion, the involvement of a private sector partner in a project which is still at the conceptual stage creates a number of difficulties. … the most suitable private sector partner should be selected after a competitive process, once the opportunity has been crystallised, the viability assessed and confirmed, and the opportunity suitably packaged.

The question to ask here is whether the partnership approach is conducive to private sector investment, or whether it is excessively focused on a project-by-project approach.

And fourth, with regard to regional linkages, there was very little evidence of partnerships between regions, while the exclusion of the eThekwini metropolitan area has in effect unlinked much of the hinterland of the province from its economic core. Such unlinking of the periphery from the metropolitan area is in direct contrast to the international tendency to work in functional economic spaces. The inability to conceptualise possible funding at the market in the metropolitan area, which could be beneficial to economic processes in the rural areas, once again points to a supply-driven approach. In geographical terms, this means that the money is allocated only to the production (supply) processes in rural areas.

4.4 Supply-driven versus demand-driven approaches to the pro-poor nature of projects

This section analyses the pro-poor nature of the projects and argues that this aspect was influenced by the supply-driven approach.

Although the Gijima guidelines emphasise the pro-poor intent of the programme, they do not clearly define what is meant by the term ‘pro-poor’. The mid-term review (DEDT, 2007) also mentions the lack of an adequate definition of pro-poor LED, and reports that this matter had been debated at length by various role players, but that in practice the actual conceptualisation of the notion seldom extended beyond poverty alleviation principles. The Project Coordination Unit had apparently come to the conclusion that a clear definition was not needed, and had thus asked project applicants to try to ‘come up with solutions as to what they believed constituted pro-poor LED’ (DEDT, 2007:12). Although pragmatic, this approach did not solve the problem of how to define the term.

There has been a small amount of research on pro-poor LED and on making markets work for the poor (Schirmer, 2005), but few attempts have as been made at least to classify the various approaches and key concepts. The last two decades have seen an increased emphasis on the concept of making markets work for the poor (Harper & Tanburn, 2005). This concept is based on three assumptions: first that the poor have seldom benefited from economic development initiatives and must be given greater market access (although the need for macroeconomic stability is recognised); second that, according to a ‘strength based’ view, the poor are already involved in the market economy, so interventions should focus on how to improve their access to this economy; and thirdly that, as De Soto (2000) argues, developing countries need more structural change and less development aid. De Soto's (2000) argument emphasises the importance of unleashing the assets that are available to benefit the poor – particularly property assets – in order to formalise the economies of poor countries.

Against this background, and to give a clearer idea what pro-poor LED is, the following discussion distinguishes two types of approach in the 14 LCF CAP projects, supply-driven and demand-driven, and identifies some characteristics of pro-poor LED. Projects may display more than one approach or characteristic.

The first supply-driven approach is infrastructural. It assumes that the main reason poor people are unable to participate in the economy is an inadequate economic infrastructure. For instance, the provision of tourism infrastructure (lodges, accommodation, etc.) can link tourists to arts and crafts produced by poor people. In agriculture, the construction of a dam enables communal farmers to irrigate crops. In other projects, project offices and manufacturing equipment are of crucial importance. Most of the projects evaluated in this study, and indeed those in the whole Gijima Programme, contained an element of this approach, and it usually dominated the resource allocation.

It was virtually impossible to judge the cost–benefit aspect of this infrastructure. However, care should be taken not to overemphasise the value of infrastructure per se, instead of trying to understand market processes and value chains, and not to fail to calculate the economic benefits of such infrastructure. At the same time, it should be acknowledged that economic infrastructure is also relevant in the context of normal business development – where return on investment is usually the key indicator. The more important questions here are how far the poor are able to finance this infrastructure and how it is actually financed. The evidence suggests that for many of the projects these were difficult questions to answer. Not only were business plans of poor quality – thus contributing to failure – but innovative financing systems were also absent. In addition, the possible return on investment in the infrastructure development was seldom assessed – another aspect that points to a supply-driven approach, rather than a demand-driven one, which would have taken account of the cost–benefit issue.

The second supply-driven approach entails providing technical support. Inability to provide technically sound products is usually one of the barriers to market access for the poor. Such technical expertise should not be regarded as synonymous with general training and capacity-building. Furthermore, it is directly related to a specific product or service. In reality, technical expertise is usually provided either by middlemen, by direct training, or by means of the partnerships that were created. Technical expertise is usually best provided by producer organisations or big businesses because they have adequate market knowledge. The projects that envisaged partnerships with either producer organisations or larger enterprises fall within this group, and include at least all of the agricultural projects and some of the tourism ones. However, in many cases technical support was offered in the context of a supply-driven approach, rather than taking a market perspective.

A third supply-driven approach entails the provision of a direct service through a project. The four business development projects are in this category. The degree to which they could actually be described as pro-poor depends on the particular practice in each case, but the overall approach of providing a service to the poor is essentially supply driven.

Although less prominent in respect of budget allocation, four demand-driven approaches were also encountered among the projects, as follows.

The first is the middleman approach. Many value chain studies focusing on pro-poor LED emphasise the exploitative role played by traders. However, a number of recent studies have indicated that the middleman can be an important source of quality control and market information whose potentially valuable contribution should therefore not be disregarded in the value chain. It is significant that the three tourism projects specifically assumed the role of the middleman. They purchased arts and crafts items from various sources and sold them at marked-up prices at tourism outlets. At the same time, the staff involved in these tourism projects were able to provide valuable information to the arts and crafts producers by letting them know what was required by the market and, more importantly, by providing information on market shifts. We did not detect any sign of the middleman approach in any of the other projects. What seems to be a matter of concern here is that only three of the 17 projects followed this route of effectively financing the market side to make provision for pro-poor economic participation.

The second demand-driven approach is based on the assumption that if markets were opened appropriately for poor people, making their enterprises more profitable, the poor would then invest in the ‘health’ of these enterprises. Although all of the 17 projects in the study sample had some kind of market orientation (which had, however, seldom been adequately conceptualised), at least one project, Nkandla Oils, explicitly stated (at its inception and in its application form) that it intended to follow the market-led approach, and was able to provide proof that it had already done so with some success. The Ugu Small Growers Project can also be classified under this approach, as the market for organic farming was the main selling point in its plan. It could well be argued that the business support programmes also fall within the ambit of this approach. However, such a classification would depend on the overall approach to business support, while all business support programmes have a large supply-driven component (training, provision of business plans, etc.). The nature of a market-led approach can thus only be determined once the relevant plans are available.

A third demand-driven approach can be seen in the partnerships with established enterprises. Like the enterprises that use the middleman approach, established enterprises can also be sources of market information. Such partnerships also help ensure the technical feasibility of products and services (as mentioned earlier), and they can potentially help minimise financial risks. Four of the agricultural projects had entered into partnerships, either with established enterprises that were linked to the market or to their product, or with established producer organisations. However, partnerships with businesses at the market end of the value chain were less common: most of these partnerships were linked to the provision of technical support.

The fourth kind of demand-driven approach was partnerships between small-scale entrepreneurs. These are often excluded from markets, purely because the scale of their operations is too small. The result is that their overheads are too high for them to access markets effectively (e.g. in terms of the cost of transport to the market). By forming producer organisations, for example, small-scale entrepreneurs are able to share overhead costs and reduce the cost per producer. Most of the agricultural projects in the study sample had either included such organisations, or envisaged including them. However, the business plans seldom indicated that the involvement of producer organisations was regarded as an operational advantage (especially in terms of saving costs). It should also be noted that when private sector organisations come together as a group with a common purpose, such an arrangement is called ‘sectoral clustering’ – an internationally acceptable concept (Da Silva, 2006). Thus, rather than competing, these firms work together to ensure international competitiveness. The ideas put forward by the cluster of furniture manufacturing projects could well be viewed in this light. International competitiveness is central to the operation of the cluster, but also simultaneously addresses equity concerns.

The above were LED approaches of the LCF CAP projects. Those followed by the BEF projects were generally different. Overall, they can be described as an institutionally led approach whereby municipalities, by virtue of their democratically elected councils, engage in LED activities. It should be noted that, in the plans of the three BEF projects in the study sample, there was very little evidence of the pro-poor LED approaches discussed above.

The above evidence suggests that – contrary to the findings of the mid-term review (DEDT, 2007) – LCF CAP projects are generally pro-poor. However, the underlying assumption in the projects is that a supply-driven approach would benefit the poor. In fact, this assumption resulted in some of the more demand-driven approaches not being well understood or effectively implemented.

5. Conclusion

The Gijima Programme reflects a renewed approach to addressing economic development beyond the metropolitan areas of South Africa. Essentially, Gijima differs from the LED Fund in that it involves a far more stringent process of allocation, it does not function only in government circles, it focuses on partnerships, and it allocates a significant proportion of the funds to developing business or strategic plans. It has also shown how LED can be managed outside the political sphere and that selective programmes do indeed have potential. It has furthermore provided a laboratory for experimental learning and has re-emphasised that LED practitioners could be much better prepared than they are currently. Despite these improvements, however, the level of actual implementation remains low, and there are still some problems and concerns.

This paper's fundamental argument was that the problems are the consequence of a supply-driven approach, in which project design and implementation were seriously affected by the impediments encountered in donor-driven and government-driven programmes. This is evident from the requirements with regard to partnerships, the excessive emphasis on the quality of the applications – especially compared with the lack of quality control in respect of the end product – and an inability to conceptualise the nature of pro-poor LED in advance.

Consequently, the programme shows very little evidence that some of the latest trends in LED have been absorbed or considered in either its design or its intended outcomes. Furthermore, project quality beyond the application process seems questionable, and seems frequently to have been driven by consultants who were able to complete the application processes adequately. Sectoral decisions fell short of the latest international trends, and there is very little evidence that alternative conceptual frameworks for the knowledge economy, such as differentiated goods, were incorporated into projects. Partnerships were, at best, ad hoc arrangements, and were dominated by either the public sector or private consultants. Partnerships formed with partners from the market side of the economy were virtually absent, and were inhibited by the fact that the eThekwini metropolitan area was excluded from the programme. There is also evidence that municipalities have limited capacity, and in many cases their staff seem to lack the ability to work in partnerships – especially across political boundaries and in functional economic spaces. The overall approach to pro-poor LED was mainly focused on infrastructure provision, or on the provision of technical skills. Yet certain approaches that were obviously more demand or market driven (at least to some extent) also came to the fore; and although the relevant projects were under-funded – and seldom correctly funded – the approaches followed require a more in-depth assessment, with a view to learning some valuable lessons.

Finally, although the jury is still out on the level of implementation of these projects in the study sample, and of the Gijima Programme as a whole, and of projects funded through the LCF IMP Fund, this paper argues that the overall approach can be said to be a supply-driven one – with very little emphasis on finding ways to open markets to the poor.

Acknowledgements

The research is based on an assessment conducted for the Learning, Monitoring and Research Facility in KwaZulu-Natal.