Chapter 3

Industrial Strategy

84. Key aspects of labour market performance ­ most significantly levels of (un)employment and labour absorptive capacity, but also skill requirements and shortages and wage differentials ­ are powerfully influenced by the unusually high levels of capital intensity that characterise the South African economy.
85. The industrial strategy followed will significantly influence the composition of investment and, hence, its labour absorptive capacity. A strategy directed at encouraging labour absorbing investment is effectively one that seeks to promote economic growth through measures aimed at increasing the deployment of labour and raising the productivity of capital. Expressed otherwise, these measures encourage the extensive utilisation of general labour and the intensive utilisation of capital and skilled labour. This is, for obvious reasons, a highly appropriate general strategy in a labour-surplus and capital-constrained economy like South Africa's.
86. However, the economy-wide imperative to increase the productivity of capital should not be construed as one that tolerates, much less encourages, inefficiency in the utilisation of labour. While in the non-traded sectors of the economy there is a degree of leeway that would permit society to trade off efficiency in resource use for other objectives ­ additional employment opportunities for example these possibilities are increasingly constrained in the production of internationally tradable goods and services. Here, within the parameters of a given technology choice ­ and these choices are, in fact, highly circumscribed ­ all factor inputs, including labour, have to be deployed at maximum levels of efficiency. In any event, as will be elaborated below, increasing labour productivity is not necessarily incompatible with the requirement for employment growth.
87. This chapter is then concerned with policy interventions designed to influence the composition of investment in the direction of greater labour intensity, or greater capital productivity. In the language of economic theory we are, in this chapter, concerned with the promotion of allocative efficiency, with ensuring that investment choices express our particular endowment of the key factors of production, best characterised as a relative abundance of unskilled and semi-skilled labour and a relative shortage of capital and skilled labour.
88. Industrial policy is, however, concerned with more than promoting overall allocative efficiency. It is equally concerned with enhancing the capabilities of the various inputs into the productive process and with enhancing the efficiency with which they are combined. These are, in the economics literature, respectively referred to as dynamic efficiency and x-efficiency. The policy interventions designed to secure these micro-level efficiencies, most particularly those directly associated with the functioning of the labour market, are detailed in Chapter 5, the chapter dealing with productivity. We must emphasise that the overriding concern of industrial strategy is with enhancing productivity ­ in the first instance it is concerned with the overall deployment of the society's productive inputs; in the second instance it is concerned with enhancing the capabilities, the efficiencies, of the various productive factors.
89. We should emphasise that while this chapter and our discussion of productivity in Chapter 5 both focus primarily on manufacturing, the question of productivity enhancement is no less relevant in other areas of the economy. Furthermore, the Commission holds that while industrial policy must be designed with sectoral considerations in mind, the boundaries between economic sectors are in fact not always well defined, nor do the firms in different sectors exist in isolation from one another. Hence, industrial policy must also be designed in recognition of the backward and forward linkages between firms in various of the traditionally demarcated sectors, between firms and their sub-contractors and between large and small firms in the same industry. The success of industrial policy will depend in part on its ability to recognise, and enhance, these inter-connections.

The Composition of Economic Activity

90. A recent comparative survey found that South African manufacturing is dominated to an unusual degree by sectors characterised by their high capital intensities ­ the unusually significant representation in the industrial structure of the commodity chemicals industry, electricity generation, and non-ferrous metals best illustrates this argument. Moreover, within these broad sectors, South African manufacturing tends to be oriented towards the most capital-intensive end of the spectrum of activities. A breakdown of the chemicals sector, with its relatively high degree of commitment to petrochemicals is illustrative of this point. The services and agricultural sectors are also dominated by capital-intensive sub-sectors.
91. The reverse side of this coin is, of course, the relatively low level of commitment to labour-intensive sub-sectors of manufacturing, services and agriculture ­ those sub-sectors that generate significant levels of employment for each unit of capital invested, and, as important, that generate jobs whose formal skill requirements are often low relative to those demanded by the capital-intensive sectors:

• In manufacturing, significant examples of sectoral "underinvestment" ­ relative, that is, to international comparators ­ are the clothing and footwear sectors and the furniture sector.
• South Africa is under represented, again relative to international norms, in key segments of the services sector, inevitably those segments where labour intensities are significant ­ tourism is the outstanding example here. There is also a general underdevelopment of SMMEs within most sectors, although recent trends to outsource and subcontract have begun to reverse this.
• Intra-sectoral investment that bears a low level of commitment to labour-intensive production is further manifest in agriculture. Vine crops, for example, have an employment multiplier 50% greater than other horticultural crops, which, in turn, have an employment multiplier 50% greater than livestock, which in turn has an employment multiplier 50% greater than field crops. This varying degree of capital intensity in agriculture is indicative of past policy distortions that did not always direct resources where the profit or employment generated was the greatest.

92. This unusual sectoral composition underpins the relatively high degree of capital intensity that characterises the South African economy. The labour market consequences of this bias towards capital-intensive industries are dramatic. Within manufacturing, the R353 000 (in 1990 prices) required to generate a single job in the chemical industry provides a telling contrast with the mere R2 700 (in 1990 prices) required to generate a job in clothing. An additional R1 million rand of final demand will create 19 jobs in wheat compared with 153 jobs in vine crops. The burgeoning number of micro enterprises is evidence of the low entry costs at this end of the market in many sectors, especially in the retail sector. However the high failure rate of these ventures renders them unreliable as income generators.
93. It is more difficult to demonstrate whether, within each sector, South African producers choose to employ relatively capital-intensive techniques. It is not, as already noted, even clear that the range of technical choice is as wide as economic theory assumes, particularly in the traded goods sector. South African firms tend to be larger than those in other countries at a similar level of development. To the extent that larger concerns tend to greater degrees of capital intensity than smaller units, we may expect technology choice in South African manufacturing to exhibit this bias. Submissions to the Commission indicate that similar considerations apply to agriculture ­ unusually large farm sizes are conducive to capital-intensive technology choices.
94. However, the problem is easier to identify than the solutions. An extremely complex array of factors, aside from relative factor prices, account for the unusually capital-intensive composition of economic activity. Past industrial strategies and policies have played a significant role in influencing the overall structure and composition of our economy. Our natural resource base, particularly the requirements of deep-level gold mining ­ both in the mining operation itself and many of the key inputs required ­ also stands out among the factors that underpin these features of our economy. Over the years powerful institutions and social interests, in both the private and public sectors, have become deeply locked into our economic structure and will resist policy interventions that attempt to transform it.
95. Moreover, the policy interventions necessary to shift the composition of economic activity from its historically capital-absorbing to a more labour-absorbing path, are by no means clear-cut. It must be understood that a simple decision to structure incentives and rewards so as to privilege investments in labour-intensive sectors over capital-intensive sectors will not, on its own, be sufficient to achieve the desired outcomes. As already indicated, the divisions between sectors are sufficiently blurred to allow for the possibility that a highly capital-intensive investment may necessitate or, at least, render viable, ancillary investments that are relatively labour-intensive. By the same token, achieving competitiveness in, for example, the labour-intensive clothing sector may necessitate considerable new investment in the relatively capital-intensive textile sector. Also, established, average capital-labour ratios are not necessarily accurate guides to future outcomes ­ the next job in the petrochemicals sector will absorb considerably less capital than the average of all those created thus far.
96. These qualifications notwithstanding, the validity of our initial observation is unchallenged: the overall composition of economic activity in South Africa is unusually capital-intensive. This reflects, in part, a set of strategic choices made by the policymakers of previous governments. In the context of trade and financial sanctions, previous strategic choices emphasised national self-sufficiency in key products and access to secure foreign exchange sources. The upshot of these strategic decisions was an industrial strategy that effectively rewarded investment in highly capital-intensive sectors and processes. The imperatives of the present government dictate a different set of strategic choices. These privilege employment creation and dictate that policy be reoriented such that investment in labour-absorbing activities be encouraged. Given that current imperatives rest upon the productive employment of presently unemployed and abundant labour, and the husbanding of relatively scarce capital, this strategy involves no compromise with the efficiency-enhancing imperative of industrial strategy. On the contrary, in marked contrast with the objectives of previous industrial strategies, it is precisely concerned to ensure the efficient allocation of the country's productive factors ­ the extensive utilisation of the abundant factor (unskilled and semi-skilled labour) and the intensive deployment of capital and skilled labour (the relatively scarce factors of production).

Mega-Projects

97. While the strategic orientation proposed here demands a review of the effective support extended by previous governments to mega-projects, it does not preclude public support for all large-scale capital-intensive investments. There are a variety of reasons ­ indirect employment opportunities, skill enhancement and technological externalities are just three of the most important ­ why a particular project may be deemed worthy of public support even if its initial direct impact on new employment opportunities is limited.
98. There are a number of natural-resource-based mega-projects in the early phases of their development ­ some are in the planning and feasibility study phase, while others are on the verge of beginning the labour-intensive process of constructing the plant. Every effort must be made to maximise sustainable employment-creation opportunities from these investments. Clustering ancillary industry around a large plant employing several hundred highly paid workers appears to afford a promising prospect of employment growth. We will also suggest a human resource development strategy designed to enhance the future job prospects of those employed in the construction phase.
99. The government itself is committed to major investments in infrastructure. The Commission welcome this approach. Sound infrastructure ­ both social infrastructure like housing and commercial infrastructure such as transport ­ underpins growth and further investment. Again, however, we would urge specific attention to the creation of sustainable employment. Considerations similar to those outlined with respect to the mega-projects apply. In particular, there is a labour-intensive construction phase followed by a sharp decline in employment directly associated with such projects. The current focus on development corridors clearly represents an attempt to generate sustainable employment clustered around infrastructural development and is welcomed. A training component must also be built into these projects and is discussed below.

Supporting Labour-Intensive Production

100. The potential benefits flowing from our mega-projects notwithstanding, it remains our general contention, our overall proposal, that South Africa's industrial policymakers be required to account for their range of interventions ­ be these in the areas of trade policy, competition policy, technology policy, or the allocation of industrial development capital ­ with principal reference to their impact on the prospects for strengthening of labour-intensive production. We are, in other words, arguing for an industrial policy that actively supports and rewards labour-intensive investment, as long as such investments can be made sufficiently profitable to warrant directing scarce capital resources in their direction.
101. The industrial strategy of previous governments exhibits a clear bias in favour of capital-intensive sectors. The most important policy measures that supported this overall strategy were:

• monetary policies that maintained negative real interest rates for sustained periods;
• a bias in the deployment of industrial development finance, principally through the activities of the IDC but also through the Land Bank and other institutions in favour of larger and more capital-intensive activities. This bias is particularly evident in the case of "strategic" investments in synfuel production and petroleum and recent investments in aluminium smelting. In agriculture, the dominant form of government support ­ subsidised capital ­ has also heavily favoured capital-intensive sub-sectors and technologies;
• investment incentives such as accelerated depreciation and other tax-based allowances;
• the increasingly explicit market orientation of publicly funded science councils which biased their activities in favour of better resourced and more technologically sophisticated firms and
• a trade regime which has effectively raised the cost of key intermediate products thus militating against labour-intensive downstream activities. This applies with equal strength to agriculture as it does to manufacturing.

102. This policy bias in favour of large, capital-intensive investments was powerfully exacerbated by measures that effectively stifled the rise of dynamic small and medium scale enterprises. Draconian restrictions imposed upon black ownership are extensively documented and effectively poisoned the wellspring for dynamic small business growth. But the bias against small business was not only an outcome of racial discrimination. It was deeply embedded in the general conduct of industrial policy. The pillars of this bias against SMMEs were found in:

• a notoriously weak competition policy which had little or no impact on limiting economic concentration, including vertical integration arrangements particularly hostile to new market entrants;
• state purchasing arrangements that favoured large established firms at the expense of smaller firms and new entrants;
• export incentives that were heavily biased against small firms;
• poorly resourced institutional support ­ in the areas of finance, technology development, training and marketing ­ for SMMEs; and
• a skills training system focused exclusively on formal apprenticeships which prevented the small entrepreneur from either formally giving or receiving training.

103. Our capital-intensive economic structure is thus, to a significant extent, policy-induced. An inchoate combination of policies bears a large responsibility for our capital-intensive sectoral composition. Policy interventions ­ or their absence ­ also account for our weak SMME profile, further strengthening our capital-intensive bias.
104. Re-orienting South Africa's industrial policy in order to achieve more labour-intensive outcomes will clearly presuppose supportive interventions within the traditional realm of labour market policy. However, the labour market cannot, and should not, be expected to bear the full brunt of the adjustment alone. To put it bluntly, for as long as the prices of other key inputs ­ for example, finance and electricity ­ reflect historical and current subsidies, then for so long will labour appear to be relatively costly. Or, from another perspective, the labour market should not be made to bear the full burden of facilitating new entrants in product markets when it is the presence of particular institutional arrangements (and the absence of others) in those very product markets, as well as in capital markets, that underpin high levels of concentration in the first place. In short, South Africa's product and capital markets are highly distorted. To insist that labour market reforms alone compensate for the impact of distortions in other markets is simply to compound, rather than solve, the problem.
105. The submission of the Department of Trade and Industry (DTI) to the Commission demonstrates that South Africa's current policymakers are aware of the range of policy instruments at their disposal and, to some extent, are deploying them in support of labour-intensive industrial development. In particular, the Commission welcomes:

• the current commitment to a strengthened competition policy and the apparent determination to elevate it from its historically low status in the industrial policy armoury;
• the priority accorded by the Department of Trade and Industry to the development of SMMEs, and the comprehensive range of measures and resources deployed in support of this class of enterprise;
• the determination to expose local producers of key intermediates to international competition, the better to ensure competitively priced inputs for relatively labour-intensive downstream producers; and
• the efforts made to secure access by South African producers to international markets, and the incorporation of SMMEs and other non-traditional exporters or potential exporters in these efforts.

106. However, from the perspective of the Commission's particular concerns, we would want to see more determined action in the following areas:

A Refocus of IDC Support

107. We urge more effective deployment of the IDC's resources in pursuit of job creation. The Commission appreciates that refocusing an institution like the IDC may be akin to turning a supertanker; we are also cognisant that the lack of nimbleness of even supertankers is compensated for by other qualities. However, the IDC must be mandated to accord higher status in its investment decisions to their impact on employment creation. This is a first-order priority of the IDC's shareholder, the South African government. If it cannot meet this requirement, then it must be obliged to hand back more of its resources to its shareholder for deployment in support of job creation. We are persuaded that the IDC's activities in agriculture and horticulture establish its potential as a creator of jobs; however, a disproportionate share of its resources is devoted to very large, highly capital-intensive projects, generally in partnership with a predictable array of well-resourced corporations.
108. We should add that this recommendation does not imply compromise with the requirement that the IDC be self-financing and that commercial returns weigh heavily in the IDC's investment decisions. Our point is simply that just as the IDC supported investment in large, high-risk projects, by accepting longer pay-back periods and by securing policy interventions designed to reduce the level of risk ­ for example the Section 37e depreciation allowances ­ so too must it lead investors back into sectors and projects that generate significant amounts of employment. Again, as in the past, it will do this via concessionary financing, but it will also insist on policy backing that reduces its risk.

A Review of Commitment to Natural Resource Beneficiation

109. The Commission recommends a re-examination of the widespread commitment to natural resource beneficiation. Unfortunately the preoccupations for which the IDC is frequently ­ and, we believe, justifiably ­ criticised are widely shared both inside and outside government. We refer, in particular, to the conventional wisdom that supports investment in natural resource beneficiation. These projects consume large amounts of capital but generate very little employment per unit of capital. Unless downstream producers enjoy an unsubsidised price advantage over international competitors ­ and we would welcome interventions that attempted to secure this ­ there is no reason why a natural resource endowment should uniquely qualify South Africa to undertake further processing activities, particularly those that require such massive capital commitments.
110. We should emphasise that this does not imply that natural resource beneficiation is never an appropriate direction. It simply suggests that the possession of a particular mineral does not necessarily imply comparative advantage in its further processing. The costs of supporting a mineral beneficiation project should be evaluated in the same way that the costs of support for any other project are evaluated. For example, the possibility of developing a sustainable comparative advantage in mineral processing would be significantly enhanced were our firms not obliged to pay the world price plus an imputed transport cost.

Policy Support

111. Policy support should target labour-intensive production. An industrial strategy is essentially characterised by the range of sectors and firms that it seeks to strengthen. Numerous criteria have dominated these choices in other countries ­ entry into international markets has featured significantly in the case of the Asian new industrialised countries; the desire to derive technological externalities has dominated the industrial strategies of other countries; South Africa's invoking of "strategic" (read: "energy and military self-sufficiency") considerations in the 1970s and 80s, considerations which directed resources into the petrochemicals and armaments industries, represented another industrial strategy.
112. We are concerned at the present absence of an explicitly articulated industrial strategy. Currently, our implicit industrial strategy directs support at mineral beneficiation and, at the other end of the scale, SMMEs. The former is, as we have already outlined, inappropriate; the latter is insufficiently focused. Laudable though the decision to support SMMEs is, it does not substitute for an industrial strategy. An appropriate industrial strategy for South Africa is one which makes an efficient and effective use of scarce capital resources. Given the need to create more employment, industrial policy should be particularly sensitive to supporting labour-intensive production methods.
113. Support for SMMEs should then be principally directed at those within the broadly defined sectors and subsectors listed in the following paragraph, or those which demonstrate linkages to large enterprises and thus have a relatively stable demand for their goods and services. This support will, in part, be represented in a wage and productivity-enhancement strategy as outlined in this Report. Other supply side strategies ­ for example,

technology support programmes and access to credit ­ are also vital and are part of the DTI's current programme of support for SMMEs. A key aspect of the support provided should be measures designed to enhance the interaction between the SMMEs and the large firms within each sector.

114. Some broadly defined economic sectors (such as clothing, furniture, footwear, tourism, construction, metal fabrication and auto components) have, across the available range of production techniques, higher labour/capital ratios than other sectors (such as petrochemicals). South Africa has an established presence in each of these areas which employ large numbers of people and have the potential for international expansion. In each of these sectors there are firms that, by the most rigorous standards, are highly competitive: they export successfully and even in the absence of protective tariffs will easily hold their own against global competitors. But the performance of these competitive leaders stands in sharp contrast with the great many laggards in each of the sectors in question. There is thus a strong case for devising mechanisms to enhance the performance of the laggard firms, by diffusing more widely the capabilities that characterise the leading-edge firms. Factor prices must be appropriate to the character of the activity in question and to the level of productivity achieved. However, it is instructive that outstanding firms have performed successfully in international markets by raising productivity. Their performance must be the benchmark for our producers and for industrial policy makers intent upon supporting productivity enhancement.
115. We reiterate that the concept of a sector is often difficult to operationalise, and that the boundaries between sectors are inexact. For this reason we welcome the attempts by the DTI to identify competitive "clusters" of firms as the potential sites of policy support. It should also be noted that projects eligible for support may emerge in any sector. Moreover, successful sectoral targeting does demand considerable institutional sophistication.

Sequencing Policy Interventions

116. More effective sequencing of trade liberalisation and the introduction of supportive supply-side measures is required. It is also important to co-ordinate trade policy and exchange rate policy. The imperatives to limit short-term job loss consequent upon trade reform, to maintain broad support for the restructuring programme, and to retain the skills and discipline acquired through work experience, emphasise the importance of getting this sequence right.

Supporting Labour-Intensive Production: Labour Market Requirements

Wages and Productivity

117. Wage levels are a significant component of overall cost structures across the full range of these labour-intensive activities. Unless producers are simply able to pass wage increases on to their customers ­ a practice limited by trade reform and increased domestic competition ­ there will be a persistent pressure to limit rising wage costs in labour-intensive activities. In fact, as the South African economy becomes more open to international competition, firms will only be able to survive if their costs and productivity are comparable to those of their major international competitors. This has implications for trade and industrial policy as well as for labour market policy. At the very least, South African producers must be able to obtain key intermediate inputs at internationally competitive rates.
118. Firms must also ensure that the relationship between wages and productivity is comparable with that of their major international competitors. International comparisons of unit labour costs ­ effectively the combination of the level of wages and the level of productivity ­ are difficult to make. They are, of course, sensitive to exchange rate movements and this means that, in the recent past, as a result of the depreciation of the Rand, South African wages, and hence unit labour costs, will have received a significant competitive boost. One recent report, using composite wage and salary figures, concludes that by international standards South African wages (especially unskilled wages) are high relative to productivity, accounting then for high unit labour costs. Another study, taking a more disaggregated view of the economy and the labour force, draws more favourable conclusions with respect to our unskilled and semi-skilled wage rates.
119. However, there is clearly no room for complacency. Raising the productivity of South African workers through training programmes, work-place changes and industrial relations reform must be an integral and urgent part of South Africa's restructuring agenda. These, and a host of other productivity related measures, will enable South African firms to compete in those product ranges where quality and reliability are core competitive features. We have already noted the evidence of South African firms' ability to successfully compete in these markets. It is here that South Africa's sustainable competitive advantage lies and industrial policy must seek to shift the industrial structure towards these broad market segments. We should however emphasise that even in this broad product range, competitiveness remains enormously sensitive to the relationship between wages and productivity.
120. The long-term prospects for South African firms competing in lower value-added product ranges dependent upon very low wages are not good. However, even in these product ranges, productivity performance is important and industrial policy must not ignore the opportunity for enhancing performance here. This will offer a long-term solution to many of these firms. However, the wage determination system must be sensitive to firms currently competing in lower value-added product lines. This is elaborated in our wage determination chapter where the system of extensions of wage agreements, the structure of bargaining councils, and wages applicable to SMMEs are examined.
121. Given South Africa's vast supply of unskilled, unemployed labour, it is essential that the system of wage determination does not undermine firms attempting to compete in lower-wage, lower value-added product lines. Industrial policy must progressively encourage these firms to focus on improving their product and on competing in more demanding product ranges. Accordingly, just as the wage determination system should not seek to eliminate these firms, nor should it eliminate any incentive for these firms to locate themselves in those product segments that reflect South Africa's long-term comparative advantage.
122. The second set of measures designed to support a labour-intensive industrial strategy is dealt with in Chapter 5. These are the interventions designed to secure dynamic efficiency and x-efficiency. Many of these interventions are directly in the sphere of the labour market, others not. Most of these measures impact significantly on labour market processes and outcomes. In our discussion of these interventions we will draw attention to those efficiency-enhancing measures which engender positive employment outcomes. For example, the introduction of multiple shifts significantly increases capital productivity and creates employment opportunities. Furthermore, innovation directed at the development of new products is often more employment-friendly than innovation directed at the production process. These and other productivity-enhancing measures are discussed in Chapter 5.

Human Resource Development

123. Human resource development (HRD) ­ widely recognised as a pivotal policy issue ­ is the subject of intensive examination inside and outside of government. For this reason, it has not been included in the terms of reference of this Commission. However, it is a critical component of any industrial strategy including, the Commission believes, one designed to boost employment creation through labour-intensive investment. Policy development in the fields of education and training on the one hand, and economics on the other, tend to occupy different worlds despite the unanimous acceptance of their interconnectedness. The following paragraphs will tentatively examine the interface between the industrial strategy outlined in this chapter and human resource development and make preliminary proposals directed at strengthening this interface.
124. The combination of apartheid and an inward-oriented industrial strategy was characterised by human resource development policies which resulted in a deeply fractured national skills profile. The essence of these policies was tertiary education for the managerial and professional few, trade training at subsidised parastatals for young white men and perilously little else. "Sitting next to Nelly", or learning from experienced workers, was the best available to the majority. The underinvestment by firms in education and training is documented in the ILO Review. These policies underpinned work practices in the mining industry, in the capital-intensive, resource-based processing sectors, and in the highly protected, mass-production consumer goods sectors of manufacturing.
125. The education and training system was broadly designed to serve these ends. Sound general education for whites laid the basis for tertiary and trade learning, while the black majority were educated to little more than the level of ­ in Verwoerd's notorious description - "hewers of wood and drawers of water". Despite the growing demands of the manufacturing and service sectors for additional skilled labour, tariff protection, steady increases at the educational margin and immigration policy served to buffer the society from the imperative to effect systemic change.
126. For the vast majority of the population the heritage of this system is mass illiteracy combined with an uneven array of unrecognised skills built up from experience, ranging from trade equivalent for those who acted as artisan-aides, to the more limited and tacit skills of those who performed routine and repetitive tasks over decades. Those described as "semi-skilled" and "unskilled" fall within this range. Those who have never worked in the formal sector have an undescribed spectrum of traditional, street and home knowledges which seldom translate into labour market currency ­ hence these people are also designated "unskilled".
127. The magnitude of the task should not be underestimated. Apartheid's scars are particularly apparent in the skills profile of the workforce. A recent study showed, for example, that 45% of adult Africans cannot read or write, while 35% of the economically active population in 1991 were reported as functionally illiterate. This neglect is matched by the particularly low level of commitment by the private sector to training. In South Africa, an average of 1% of payroll is dedicated to training, while in the OECD countries the figure is between four and seven percent. We are however persuaded that we have the wherewithal to rapidly enhance the capacity of our workforce to participate successfully in the activities identified here.
128. Our industrial strategy is intended to secure existing employment opportunities for these "unskilled" and "semi-skilled" workers, and to generate new employment opportunities for the millions without work who fit this skills profile. To recap: our policy is intended to promote employment growth in three areas:

• the product segments and sectors identified above;
• the construction phase associated with investment in social and commercial infrastructure; and
• the construction phases of capital-intensive, resource processing plants, although this employment is of a limited duration.

129. The HRD strategy at the core of this growth and development trajectory must be directed at the requirements of the manufacturing and services sectors, those sectors where sustainable employment opportunities will be created. In those sectors where large numbers of short to medium-term job opportunities are created ­ for the most part, civil engineering and construction, but, in the natural resource-processing operations, also involving the large mining and manufacturing corporations ­ the focus must be on providing a growing cadre of workers suitably equipped to enter the manufacturing and services sectors.

Labour-Intensive Manufacturing and Services: the HRD Requirement

130. For three decades apartheid's education and training approach supported an industrial strategy that relied upon high-level (white) personnel, skilled (predominantly white) supervisors and artisans, and a large body of (black) "semi-skilled" and "unskilled" labourers. However, the system proved unsustainable, principally because of the industrial conflict for which it bore a considerable responsibility. In addition it was incapable of adjusting to meet the requirements of new technologies, small- and medium-scale producers, global competition and a changing industrial relations environment.
131. The HRD system was found wanting on all fronts:

• The ILO Review notes that the number of workers attaining artisan status fell from 12 933 in 1985 to 7 132 in 1990. There are a number of influences at play ­ economic downturn and the privatisation and commercialisation of some of the large parastatals, are clearly two of the most important. But more fundamental factors are at issue. Essentially, in the face of more integrated technologies, manufacturing requires more flexible and varied skills. Old trade demarcations are too rigid and are being fragmented on the one hand while being upgraded on the other. In short, the (structural, as well as cyclical) demand for old-style artisans has declined.
• Commitment to training semi-skilled and unskilled workers has only marginally increased from its extremely low base. The ILO Review's South African Enterprise Labour Flexibility Survey (SALFS) finds however that in no manufacturing firm did production workers account for more than 10% of those receiving training in the past year. Again there are multiple factors at play, particularly sustained economic downturn. However, the absence of an appropriate strategy or institutional framework for the funding and provision of training to these layers of the workforce is fundamentally important.

132. As already elaborated, our strategy rests in large part upon the ability of the manufacturing, construction and services sectors to generate secure, regulated and moderately remunerated employment. This requires steady, sustainable increases in productivity and this, in turn, is powerfully influenced by the quality of our human resources. Here follow some tentative proposals aimed at significantly upgrading our human resources.
133. Public resources should be allocated to promote investment in education and training which would not otherwise occur, and which enhances the quality and quantity of employment. Industry Training Boards (ITBs) have an important role to play as intermediaries in the development of a framework for career-pathing and quality assurance when agreements are reached by bargaining councils, or in cross-sectoral forums at national, regional, local, and firm level.
134. While the HRD requirements will differ between firms and sectors, the following are likely to be common features:

• All learning should, wherever possible, be credit-bearing and portable, to support dynamic career-pathing within and across firms and sectors.
• The HRD strategy for unskilled and semi-skilled workers should be one of multi-skilling rather than being directed at providing a narrow range of skills. The strategy should not reproduce the rigid and inflexible nature of task-specific training. The ILO's SALFS reports that some firms are indeed attempting to introduce multi-skilling at the workplace. The successful implementation of multi-skilling strategies generally presupposes complementary reforms of workplace organisation, reforms that move away from the hierarchical features that characterise traditional manufacturing practices based skilled artisan and supervisory labour on the one hand, and unskilled labour on the other.
• Adult basic education. HRD strategies directed at semi-skilled and unskilled workers are significantly limited by low levels of basic education, principally reflected in the alarmingly low levels of literacy and numeracy even among employed workers in the manufacturing sector. The ILO Review notes, for example, that the mean number of years of schooling of the workforce in the early 1990s was seven. The Commission is concerned at the apparent lack of attention to the provision of basic education.
• Training should, where possible, be provided on full pay in working time. The recently published Green Paper on employment standards devotes considerable attention to rearranging working time and to reducing working time. The Commission's view on this is elaborated in Chapter 5. Suffice it to say here that an arrangement that maintains the number of weekly working hours or reduces them only slowly, but that requires employers to provide a specified number of hours of basic and, where appropriate, more advanced, training and education in working time, seems to be a more urgent requirement and one that will rebound to the long run advantage of the recipients of the training;
• SMMEs are proliferating in the labour-intensive sectors that will benefit from our industrial strategy. This is a global response to demands for greater flexibility in the production process. The rise of labour-only subcontractors in the construction industry is one example; sub-contracting in deep-level gold mining indicates how far this phenomenon is capable of spreading. The apparent proliferation of home-work and sub-contracting into the informal sector in clothing, is another example. In some circumstances there are public policy considerations that require the strict regulation of this phenomenon. Occupational health and safety considerations justify such regulation. The new Mine Health and Safety Bill imposes duties on owners to train workers irrespective of whether they are employed by the mine owner or the sub-contractor. However, trade unions would be ill-advised to press for a general prohibition of sub-contracting or to insist on a rigid adherence to old regulatory structures. An overly rigid regulatory net will be torn asunder by market pressures as is strongly borne out in the construction industry. Nor, on the other hand, would the requirements of our industrial strategy be served by the proliferation of sweat shops at the margins of production: it would be better to draw these SMMEs closer to the mainstream of production by extending training facilities to these enterprises. Training here should include entrepreneurial training where contractual relations exist. Where contractual relations do exist, the principal responsibility for ensuring the training should rest with the principal firm in the contracting relationship. There is little doubt that dynamic, as opposed to survivalist and harshly exploitative, SMMEs are distinguished by the quality of their relationship with large established firms. Insisting ­ possibly with the help of public assistance ­ that the principal firm assumes responsibility for providing training to its subcontractor will proliferate skills and cement inter-firm relations. Where no such contractual relation to a larger firm exists, the restructured ITB should be encouraged to provide assistance.

135. Although our HRD strategy is focused on unskilled and semi-skilled workers, the requirement to generate higher-level skills should not be ignored. In particular, where high skill levels are in short supply, and this leads to high wage differentials, expanding the supply through immigration or education and training policy is desirable. This is further explored in our discussion of wage determination (Chapter 4) and migration (Chapter 9). As already noted, it is widely accepted that the rigid, narrow view of skills that characterises traditional approaches to artisan and task training is not appropriate in the context of rapidly changing technologies and markets. The Commission is encouraged by the recognition that this has already received in some potentially important sectors ­ for example, the furniture industry is including business skills as a requirement within their traditional trades. Their trade test evaluates skills required for self employment ­ estimating, costing, marketing and the like.
136. As part of an attempt to promote dynamic efficiency, equity, skill formation, work security and economic democracy within enterprises, it may be useful to construct a nationally negotiated Human Development Enterprise (HDE) Index. The purpose of the Index, as the ILO Review notes, would be to provide a framework within which the human development enterprise can be promoted. Firms would, within this framework, be rewarded for performing well on the HDE Index. At the same time though, the performance of firms in improving HRD at the workplace may be tracked. Possible details of the Index are suggested in the ILO Review. The Commission supports the introduction of such an Index through negotiation amongst the bargaining partners.

Capital-Intensive Projects: the HRD Requirements

137. Our industrial strategy advocates that government redirects its support, where appropriate, from highly capital-intensive, resource processing activities to labour-intensive sectors. However, several projects are at the beginning of their construction phase, and others are in the pipeline. South Africa's natural resource bounty suggests that projects of this scale will be a permanent part of the landscape. Such plants ultimately employ few people, and those employed are at relatively high levels of skill. Here we would restrict ourselves to the following observations:

• Companies should be encouraged to train workers as opposed to simply poaching them from other projects and companies. One way of ensuring this is via a flexible, case-by-case requirement that the projects, usually based outside of the metropolitan areas, hire a minimum proportion of local labour. This would inevitably presuppose the generation of new skills;
• The limitations of the artisan system ­ as outlined above ­ should be carefully scrutinised here. The companies involved in these large projects - for example ESKOM and ISCOR ­ are among the larger providers of skilled artisan training. Their participation in these large projects provides a golden opportunity for piloting a systematic review of the country's artisan training system.

138. The labour-intensive phase is in the construction of the plant. For example Saldanha Steel will employ 600 mostly skilled workers in the plant, while 9,000 people will be employed for most of the 18-month construction phase. Maximum advantage must be taken of this period to impart skills and create sustainable job opportunities. To this end, the companies must be required to train each worker involved in the construction phase to a higher, formally accredited, level than that at which he or she entered the project. One Commissioner argued that this was not a realistic proposal. Companies must be encouraged to participate with other local stakeholders in the development of a strategy aimed at contributing towards sustainable output and employment growth in the region ­ preferably as part of a social accord process of the kind outlined in Chapter 10. IDC participation in these projects should be conditional upon the company meeting these requirements.
139. The Department of Labour's Labour Centres and sectorally based ITBs should interact with District Development Plans and multi-stakeholder groupings at local and regional level to assist with career planning and skill enhancement and to ensure the quality of skill enhancement for affected workers.

Investment in Infrastructure

140. Identical considerations apply with respect to the provision of commercial infrastructure. The construction phase is labour-intensive and requirements should be imposed on those who tender for the provision of the facilities to impart accredited skills to those employed. The raison d'être for the provision of commercial infrastructure is the economic activity it generates. This, as suggested above, should be identified through the preparation of a detailed economic development strategy for the region served by the infrastructure.
141. In the provision of the infrastructure, government may, unlike in the traded goods sectors, choose production techniques on criteria other than cost. This can only be decided on a case-by-case basis and employment creation would obviously be a strong criterion determining choice of technique. This is examined in our discussion of public works programmes.

Funding Training

142. Alternative mechanisms for funding training are presently the subject of a major tripartite-managed study in NEDLAC. There is little point in the Commission attempting to pre-empt the study. We would simply note that the social and private returns ­ to both employer and employee ­ are considerable and there is a strong case for compelling those who benefit from training to pay at least a proportion of the costs involved.

---