Document(s) 6 of 11

Manuel stands his ground

Presenting his 1999 budget to Parliament, Trevor Manuel, Minister of Finance, looked back. 'Integrating South Africa into the world economy has been a major challenge,' he recalled. 'We inherited an uncompetitive, inward looking, protectionist economy. Since 1994,' he added, 'we have sought to open up the economy in a measured and sustainable way' (Manuel 1999: 1).

As Minister of Trade and Industry between May 1994 and March 1996, Manuel had to implement South Africa's trade reform programme. The reform schedule was written into the 1994 Marrakesh agreement of the General Agreement on Tariffs and Trade (GATT), which followed the Uruguay Round of negotiations.

South Africa was a founder member of GATT in 1948 when there was no 'special and differential status' for developing countries. When, during the 1960s, GATT allowed differentiation between developed and developing countries to permit the latter to receive more gentle treatment, South Africa did not consider changing its status. Its white rulers believed they lived like wealthy western European or American societies, and would have seen developing country status as an insult.

During 1993, before the new government was elected, the ANC and the Congress of South African Trade Unions (COSATU) suggested that the South African government should enquire about changing the country's status to 'developing country'. Stef Naudι, then Director-General of Trade and Industry, was told that the United States would ensure that this would never happen. Neither did Japan nor Western Europe offer any assistance; apart from anything else, they were far too preoccupied finalising the Uruguay Round of negotiations to think about how they might assist South Africa. Nevertheless, South Africa was assured that, in some respects, it would be allowed greater flexibility – similar to the 'economies in transition' of Eastern Europe.

By and large, though, the trade policy community in South Africa could accept the retention of 'developed country' status as far as the Uruguay Round was concerned. Developed country status meant the biggest cuts in protection, implemented at the quickest pace. The South African economic policy community, from the union federation COSATU through the ANC, big business and the De Klerk government, agreed that the economy had become overprotected, that building a new platform for growth would require restructuring the real economy, and that substantial trade policy reform was a necessary, if insufficient, element of such restructuring.

It was one thing to agree in principle, but another thing to carry the project through. Manuel soon learned the costs. Even before the 1994 elections, the South African Clothing and Textile Workers' Union (SACTWU) persuaded the ANC to improve the terms of the GATT agreement. This required the ANC leader, Nelson Mandela, to call United States President, Bill Clinton, to ask for a special dispensation for the clothing and textile industries in South Africa. Though he had not yet been elected as President, Mandela carried more weight internationally than President de Klerk. The union and the industry wanted to extend the down phasing of tariffs to a 12-year period, rather than the mandatory 5 years. South Africa was not party to the separate Multi-Fibre Agreement that covered many fabric and garment importers and exporters.

The South African textile and garment industries employed over 200 000 people living in poor communities. Behind tariffs higher than 100% for clothing and not far below 100% for textiles, the industries had fallen out of step with world production and marketing trends. Many workers in the clothing industry lived in the Western Cape province, which was a tight electoral race for the ANC. For these reasons, Mandela agreed. At little or no cost, Clinton agreed too. Nevertheless, the ANC lost the Western Cape to De Klerk's National Party (NP).

A year later Trevor Manuel found himself facing the clothing workers again. At a huge public conference organised by his adviser on small business, Alistair Ruiters, Manuel was participating in the finalisation of a new policy for small business. But the clothing workers used this occasion to protest against the implementation of the tariff phase-down programme. They blamed it for what they saw as ongoing job losses in the garment and textile industries. This may have been the first vocal public protest against the ANC in government by a member of the ANC's constituency – the union concerned, SACTWU, is a member of COSATU, which is one leg of the ANC's tripartite alliance. The third leg is the shrunken South African Communist Party (SACP).

As a young activist in Cape Town in the late 1970s, Manuel had been a voluntary literacy teacher amongst the poor. Some of his students were garment and textile workers, or members of their families. Now Manuel was faced with representatives of the same communities telling him that his policies were causing garment and textile workers to lose their jobs. Though he may have wished he could explain his policies to the workers in small groups like the literacy classes he once taught, he could not.

But he never wavered in his policy position. Manuel left no one in any doubt that he was firmly committed to the trade reform process, even though it might appear to hurt some of the communities he knew and loved, and depended on politically. And his colleagues in the ANC leadership backed him.

Contours/prospects of the economy

Why was the ANC so sure that trade policy reform was necessary, though they knew that the strategy might be politically costly in the short to medium term? Had it been brainwashed by the World Bank and the International Monetary Fund (IMF) or seduced by international bankers, as some of its critics believed?

To any economic observer in South Africa in the early 1990s, it was clear that the country had entered an economic cul-de-sac. The economy was shrinking. Its assets were being run down – gross fixed investment was negative for four consecutive years to 1994, and capital was in full flight. National income was stagnating, and per capita income had declined every year since 1982, except 1988. Government debt was rising to dangerous levels, with the general government fiscal deficit over 9% of gross domestic product (GDP) in 1993.

What were the potential sources of growth? What sectors could the country rely on to get the economy going again so that problems of poverty and unemployment could be addressed?

The traditional strength of the economy had been gold mining. But the physical volume of gold extracted from South Africa's deep-level mines began to fall, inexorably, after reaching peak production at 1 000 tons in 1970. Even the very high price paid for gold during the two oil crises of 1973–75 and 1979–81 failed to boost gold output. The quality of accessible ore was deteriorating – whereas a ton of ore produced 13.3 grams of gold in 1970, it only produced 5.3 grams of gold in 1987 (Freund 1991). Though the tonnage of ore processed rose, the output of gold fell. By 1998 gold production had fallen to 473 tons, which was the lowest gold yield since 1956. Meanwhile, in the United States, Canada, Australia, and elsewhere, gold production grew cheaper and more efficient in geological formations more amenable to modern mining methods.

South Africa's deep-level, low-yield mines depended more on cheap labour than innovative extraction methods to maintain a sufficient margin between the cost of mining gold and the world price. Wages for black mine workers began to rise in the 1970s when the mine-owners turned away from cheap migrant labour from Mozambique. South Africa's eastern neighbour had suddenly won independence from Portugal after a democratising military cabal overthrew the Portuguese dictatorship in 1974. The mine-owners feared the volatility of the new state of Mozambique, and may have feared that the Mozambican miners were infected with a new blend of socialism that FRELIMO developed for the country.

When black workers on South Africa's mines began to organise themselves successfully, their justifiable demands for improved wages and working conditions added to the difficulties of the mine-owners. The owners compounded their own difficulties by being extremely slow to innovate production methods, to improve training, and to develop more sophisticated labour relations. When gold lost its starring role as an international store of value in the wake of the liberalisation of international financial markets in the 1980s, South Africa's gold mining industry was slow to acknowledge the changed circumstances, and even slower to react.

Other metal and mineral products, such as platinum group metals, coal and basic chemicals, experienced some growth in output and exports during the 1980s. A major new export in the early 1980s was pulp and paper produced by new mills in eastern higher-rainfall regions. Some sugar cane lands were converted to wattle production to feed the mills, but the new plants had a small net effect on the economy as a whole. They did boost the profits of Sanlam and Anglo American, two of the large financial corporations that were becoming increasingly gargantuan in the context of a stunted economy.

Export growth in semi-manufactures such as dimensional metal products (rods, bars, plates and coils), basic chemicals, pulp and paper, as well as more fabricated metal products positively correlated with productivity growth and overall economic growth in the 1970s and 1980s, whereas before 1970 manufactured exports made no contribution to growth at all (Holden 1990). But the contribution of these sectors was far less than would have sufficiently compensated for the effects of declining gold output and the low gold price.

What about agriculture? Could it provide jobs and exports on which to base a new round of growth? One limitation on the use of agriculture as a source of growth is the small supply of arable land. Only 10% of South Africa receives more that 750 millimetres of rain per year. Much of the country is desert or semi-desert, and variations of rainfall are severe enough to result in frequent drought cycles. Indeed, variations in agricultural output can push the GDP growth rate up or down by as much as 1.5%. Agriculture's contribution to GDP varies, averaging around 4%. The combined contribution of agriculture, fish and timber to total South African exports declined from 19% in 1957 to 5% in 1985 (Cassim et al. 2003; Holden 1990).

A second limitation is the fact that much agricultural land is owned by large landowners who are white and undercapitalised and do not have incentives to increase employment or output. Productivity and yield growth in the agricultural sector ran aground after the early 1970s, after years of lavish government support for farm investments. The land reform programme, intended to return a significant proportion of land to black people, started slowly in the agricultural sector. Though it has accelerated since 1999, the link between land reform and agricultural development remains fairly weak.

There are undoubtedly opportunities for agricultural expansion in high-value products such as wine, berries, fruits, nuts, and processed agricultural products, but South Africa's distance from major markets has been a drawback. Many of South Africa's competitors have a close relationship with their major customers, such as that between Europe and North Africa, or between the United States, the Caribbean and Latin America. Before the conclusion of a 'free trade agreement' with the European Union (EU) in 1999, some South African fruits attracted tariffs of over 20% in the EU, whereas products from favoured rival producers outside of the EU attracted barely any tariffs at all. EU tariffs on South African canned pears added 23.2% to the price in 1997, while in the United States they added 16.6%, and in Japan, 14.4%. EU import tariffs on South African canned apricots, as another example, were 25.1%, with the United States at 32.4%, and Japan at 16% (Kaplan and Kaplinsky 1998: 7). Since the conclusion of the EU-South Africa free trade agreement, the situation has improved a little, but remains difficult. As the conflict that wracked the World Trade Organisation (WTO) trade negotiations since Cancun in 2003 show, the inclination of the EU, Japan and the United States to subsidise domestic agricultural production and exports, in addition to protecting against imports through tariff and non-tariff barriers, has barely diminished, and even 'free trade agreements' have extensive qualifications and limitations.

In short, while South Africa has several competitive agricultural products, some of which will contribute to growth and employment creation in the future, domestic climatic constraints and world market conditions mean that agriculture can never be a complete answer to South Africa's growth and employment challenges. This was evident to policy makers in the early 1990s.

There were possibilities of employment and export growth in the services sector. Indeed, exports of services grew at 5.7% per year between 1988 and 1996, whereas the value of manufactured exports grew by only 3.2% per year (in US$). However, exports of services amounted to one quarter the value of exported manufactures in 1996 (Cassim et al. 2002: table 3). Services were not generally highly protected, and seemed to require less of a major policy reform than promotion. It was assumed that the services sector would benefit from the removal of protection on other sectors. The key service sector to promote would be tourism, a point we will return to later. On reflection, South African policy circles

certainly underestimated the potential of the services sector for growth, exports and particularly employment while policy was being formed in the 1990s, and certainly failed to consider carefully the key drivers for the sector.

Debating industrial policy

For these reasons, the focus of economic development fell on the manufacturing sector. Actually there were other reasons too. One institutional reason for focusing on manufacturing was the influence of the trade unions in the policy process. The most sophisticated unions, as far as economic policy was concerned, included the very broad National Union of Metalworkers of South Africa (NUMSA), and the clothing and textile workers' union, SACTWU.

ANC economists themselves tended to have a soft spot for the manufacturing sector. Whether in the Marxist tradition or the structuralist school of development economics, or even in the mainstream of economic history, the dynamism of the industrial sector was the mythic life source.1 This view was reinforced by the Japanese success after 1945, for which the dominant explanation heavily emphasised the importance of a strong and outward-oriented manufacturing sector. Even the World Bank report, South Africa: Economic Performance and Policies, was heavily weighted towards an analysis of the constraints and potential of the manufacturing sector (Fallon and Perreira da Silva 1994).

Though there was agreement amongst most economists that the key area for policy intervention for growth and employment creation was in trade and industrial policy to restructure and strengthen the manufacturing sector, there was a wide range of suggestions about how best to do it.

Anxiety about dependence on a few agricultural and mineral exports began to preoccupy South African policy makers in the 1970s. Initially the emphasis was on diversifying away from gold, but by the later 1980s, strongly influenced by East Asian successes and international debates, the focus narrowed to manufacturing, narrowly defined.

Why is it necessary to define manufacturing 'narrowly'? Many of the strongest sectors in South Africa are in the business of producing intermediate or semi-manufactured goods. These include products such as basic metals, including iron, steel and aluminium; basic chemicals; wood pulp; and paper. Though there is value added, the production process is usually very capital intensive. The profitability of the output usually rests more on natural comparative advantages such as climate or bountiful mineral resources, than on acquired skills or expertise. The products are generally sold in standardised categories on international markets, with the main competitive levers being price and reliability. In short, these products are essentially commodities. In the standard industrial classification, these items often appear under the 'manufactured' column, but this is misleading. Many countries that export metals and minerals engage in the first stage of manufacturing near the mines in order to cut transport costs. For most intents and purposes, these activities are better seen as final stages in the extraction process, rather than an early stage of manufacturing (see Fine and Rustomjee 1996: 76–90).

The successes in East Asia – essentially Japan, South Korea, Taiwan, Singapore and Hong Kong – showed that there was a powerful connection between economic development and strong manufacturing capabilities. All five countries had grown rapidly and achieved much higher living standards largely on the basis of competitive exports of manufactured products.

Like elsewhere in the world, the interpretation of the East Asian experience, and later the South-East Asian experience, was the subject of a vigorous academic debate in South Africa. One faction explained the success as the result of classical outward-oriented policies by the Asian newly industrialised countries (NICs). These policies included the reduction of protection on trade flows and the implementation of conservative macroeconomic policies. A homegrown statement of what was then the IMF/World Bank orthodoxy came from the economics department of South Africa's Standard Bank:

In an industrial policy framework designed to encourage a trade orientated modern sector, a core measure must be the lowering of trade barriers to permit more foreign competition. In this way the market mechanism will put pressure on domestic manufacturers to improve their efficiency. The removal of barriers such as import tariffs would mean that local and international prices for goods would converge. There would be pressure to shift funds and human or other resources into producing goods for export and/or the local market, where there is a relative advantage (Standard Bank 1991: 2).

More nuanced orthodox views were put forward by the Industrial Development Corporation's (IDC's) economic unit, and by Merle Holden, a leading academic trade policy analyst (IDC 1990; Holden 1990).

A somewhat different approach was offered by those who believed that trade policy was simply a backdrop for industrial policy. A report compiled by the government's Board of Trade and Industry (BTI) suggested that South Africa ought to target industrial sectors according to suitable measures of existing and potential comparative advantage, and aid those sectors with specially designed programmes (BTI 1988). The proposed programmes were, however, rather complex for government to manage effectively, and vulnerable to special interests. Government never adopted the report.

Economists working with the ANC and COSATU, who knew about the BTI report and sympathised with its sentiments, would not necessarily have proposed the same programmes (Hirsch 1992b: 24). It was not simply a case of bashing down trade barriers and seeing what came out, or what survived; it was a case of designing the trade policy reform with certain industrial policy outcomes in mind, and developing suitable industrial policy tools to support this.

What exactly this meant, we did not necessarily know. The East Asian experience was not always enlightening or transferable. In Japan and Korea the state had substantial control over credit and the interest rate, and tailored both to support industrial development. But, in South Africa, financial markets were too complex and internationally integrated to allow the government as much leverage in this sphere. An exception was the IDC, a government-owned industrial bank, which could apply some lessons from Japan and Korea. The IDC's newly appointed board, in the mid-1990s, instructed the bank to deal with industrial sub-sectors in a more systematic way, rather than operating proposition by proposition.

Other East Asian lessons difficult to apply were the small- and medium-size company character of the Taiwanese system, and the extremely open character of the Hong Kong economy, which, after all, had special characteristics as a very small state. Also, South African policy makers felt constrained in the application of East Asian methods, as they believed that those societies had been both more authoritarian and more blessed with skilled managers than a democratic South Africa was likely to be.

Nevertheless, there was extensive interaction between South African policy makers and interpreters, and theorists and practitioners of the East Asian miracle.2 Key carry-over themes were the emphases on export promotion, education and training, innovation support, policy-driven economic resource allocation, industrial sector targeting combined with geographic targeting, and public/private partnerships of all kinds.

Within this policy grouping – that saw trade reform as a component of industrial policy that was broadly aimed at exports – there was one significant polarisation. This was whether to focus on extending existing comparative advantages, or to seek new comparative advantages. I will spell this point out briefly, at the risk of caricaturing some arguments. Some policy actors, most notably Zavareh Rustomjee (later to be the Director-General of the Department of Trade and Industry, and later a senior executive in BHP Billiton, a minerals and energy conglomerate) and Paul Jourdan (later Deputy Director-General of the same department, and now President of Mintek – a minerals technology research agency) felt that in order to grow, South Africa should exploit its proven strengths. These strengths, they believed, were largely in industries related to South Africa's natural wealth – mainly mining, cheap energy, and manufacturing of inputs and outputs. Rustomjee and British economist Ben Fine developed the term 'minerals energy complex' to describe what they saw as the beating heart of the economy (Fine and Rustomjee 1996: part II). The point was not necessarily to invest in more of the same mines or power stations though this was not necessarily a bad thing. The point was rather for the new government to manipulate the experience, power and capabilities of the minerals-energy complex to facilitate new, more competitive investments. Because it was dealing with big businesses, the government could plan very large projects, or large groups of projects.

At the opposite end of the spectrum, within the broad family of supporters of industrial policy for exports, was Brian Levy (see Levy 1992).3 Levy is a South African who did his post-graduate work in economics at Harvard, taught economics in the United States, and later joined the World Bank. Unusually, in the light of practice at the time, which did not encourage 'nationals' to work on their own countries, the World Bank assigned Levy to cover South Africa during the early 1990s. He had built a reputation analysing small and medium business development in East Asia. Levy's view suggested that the distortions of the apartheid era led to a privileging of large-scale, capital-intensive projects in the minerals-energy complex (though he does not use the term), and that a removal of these distortions, along with appropriate supportive strategies, could lead to labour-absorbing, outward-oriented manufacturing in sectors that did not appear strong before. Levy felt that the underlying potential of the economy to create jobs through exporting was hidden by a legacy of one-sided intervention. The policy implications would be to reduce trade barriers and implement programmes designed to lift the best of new or neglected labour-absorbing industries on a rising tide, especially, but not only, by small and medium firms. The programmes could allow firms to select themselves, as long as they conformed to key developmental criteria such as creating jobs and exports.

Though none of the analysts in the broad 'industrial policy' group discounted the importance of local market growth opportunities, most emphasised export opportunities. There were some writers, though, who focused on inward-oriented development. Patrick Bond, for example, argued that South Africa should focus on producing for basic needs such as 'houses, services (electricity, water, sewage), simple home appliances, clothing and food'. He felt that South Africa would fail if it relied on export markets in the light of an unpredictable and increasingly competitive international environment (Bond 1991: 83). His view of the Asian experience was that it would soon reach its limits, leading to a plunge in exports and prices. The desire to link industrial growth with a growth in demand stimulated by the redistribution of income and social and economic infrastructure was consistent with one variant of the ANC's 'growth through redistribution' slogan that I discussed in Chapter Two. In terms of this approach, growth and redistribution were two sides of the same coin. Other South African economic analysts more or less associated with this position include Trevor Bell and Stephen Gelb.

The 'inward-industrialisation' argument had little initial support in the economic policy community. It was widely accepted that growth in South Africa normally led to imports growing more rapidly than exports (see Kahn 1987; McCarthy 1988). In the absence of sufficient South African savings, it was far-fetched for South Africans to believe that investors would underwrite inward industrialisation, which would reach its limits as a cheap programme after existing industrial capacity was fully used, and manufacturers needed to buy new equipment, mainly from abroad. Inward industrialisation could not sustain growth for any significant period of time, except on the basis of huge multilateral loans, which the ANC would not consider because of fears of threats to South Africa's sovereignty.

Left-wing critics of government sometimes still draw on the 'inward industrialisation' or 'basic needs' approach in criticising government, sometimes even extending it to an argument for resurrecting trade barriers. COSATU, the SACP and left-wing community organisations occasionally fall back on this approach as an alternative to the ANC's approach.

The outcome of the industrial policy debate

So, there were four positions, broadly speaking, on the South African path to growth in the early 1990s. From right to left (economically speaking) they were:

• trade liberalisation and getting the prices right;

• an emphasis on industrial policies with trade reform as a necessary condition, favouring smaller businesses and prereform uncompetitive labour-absorbing sectors;

• a similar approach, but with policies designed to exploit existing comparative advantages and bigger projects; and

• 'inward industrialisation', which linked industrial growth directly to domestic demand growing out of redistribution.

The position actually adopted by the ANC was a kind of unresolved compromise between the middle two positions. It accepted that a tariff phase-down was needed; this was reflected in the rather vague references in the Ready to Govern document to trade policy reform, and in the Reconstruction and Development Programme (RDP), which refers to the necessity of 'painful adjustment in certain quarters' and the aim to 'reduce and share out the impact of that adjustment while at the same time promoting efficiency'. The RDP mandated the government to 'simplify the tariff structure and begin a process of reducing protection in ways that minimise disruption to employment and to sensitive socio-economic areas' (ANC 1994a: 87–90).

While the RDP refers to manufacturers creating jobs 'through meeting basic needs', it also includes the statement: 'In general, our objective is to enhance our technological capacity to ensure that as a part of the restructuring of industry, South Africa emerges as a significant exporter of manufactured goods' (ANC 1994a: 87).

One of the clearest versions of the ANC's view on industrial policy is found in an unpublished document, prepared by ANC experts in the weeks before the elections on 27 April 1994. The document, entitled 'ANC trade and industry policy guidelines', was developed as a policy reference manual for the new Minister of Trade and Industry and his or her staff, assuming that he or she was an ANC member (ANC 1994b).

The document was drawn up from a collection of contributions from ANC-affiliated experts, some of whom were working in the Department of Economic Planning (DEP) of the ANC, some of whom were trade unionists or politicians, and some of whom worked in universities or research centres. The final meeting of the group took place on the 22 April 1994, five days before the election. Although four future Cabinet ministers – Alec Erwin, Trevor Manuel, Tito Mboweni and Jay Naidoo – were present at that meeting, and though they had overseen the process, the coordinator was Zavareh Rustomjee, who was to become Director-General of the Department of Trade and Industry (DTI) within a few months.

Rustomjee approached his task with the precision of an engineer. He had trained as a chemical engineer, had acquired a Masters degree in industrial engineering, and had worked as an industrial engineer in the petrochemicals sub-sector of South Africa's 'minerals-energy complex'. As apartheid began to crumble, Rustomjee decided to interrogate the beast through economic analysis. So in the late 1980s and early 1990s he completed a Masters degree in development studies at the University of Sussex, and a Ph.D. in economics at the School of Oriental and African Studies in the University of London. He then returned to South Africa and took a challenging, if poorly paid, position in the ANC's DEP. His family had long been involved in the struggle against apartheid: his aunt, Frene Ginwala, who also earned a Ph.D. studying in the United Kingdom, was made the first Speaker of the first democratic Parliament. His mother, a medical doctor, was appointed South African ambassador to Italy, and his younger brother joined the Department of Finance.

In 1994, Rustomjee was 37. Like many of his friends in the ANC, he had decided that, if the opportunity came his way, he would commit at least the next five years of his life working in the ANC's economic agencies to build up the democratic state. Of the 23 people at the last meeting of the Trade and Industry Policy Workshop at a seedy hotel in the inner-city flatland of Berea, Johannesburg, only three did not end up either in the Cabinet, or working for a government department or agency.

Rustomjee's summary of the discussion on industrial policy read as follows:

Industrial policies will be directed at five objectives, all of which are underpinned by the need to create sustainable employment in vibrant and growing industries:

1) To retain and extend the internationally competitive edge held by heavy minerals-based manufacturing industries at the core of the economy.

2) To support those light (non-minerals-based) manufacturing sub-sectors that have the potential to emerge from existing protection with an independent and sustainable dynamic.

3) Policies should be directed to bridging the cleavage between the minerals-based core industries and the rest of manufacturing by encouraging forward linkages of further beneficiation as well as backward linkages with industries that can supply consumable and durable inputs to the core industries.

4) Growth of the agro-manufacturing sectors should be targeted as these are less capital intensive and can directly and indirectly create formal employment for the most disadvantaged sections of the population. Many of the infrastructure programmes of the RDP, particularly those in the rural areas, will directly facilitate this objective.

5) While pursuing the above objectives, policies within each sector should drive industrialisation towards higher value-added activities (ANC 1994b: 12).

The wording leaned towards the 'minerals and energy complex' view of industrial policy, but was broad enough for many to work within. As the implementation of industrial policy unfolded, the choice between focusing on the 'natural advantages' derived from South Africa's bounteous mineral and less bounteous agricultural wealth, and the orientation towards building comparative advantage in sectors experiencing rising international demand and rising prices, regardless of South Africa's apparent assets, evolved, unspoken into a phased approach. For the first few years of democracy, perhaps the first decade, South Africa would have to rely on and exploit its 'natural advantages' and would conserve its managerial resources by focusing on relatively big projects. But, during that phase, South Africa had to begin to build the capabilities in human resources, management skills and infrastructure, which would support diversification into light manufacturing and services aimed at expanding world markets.

The Marrakech Express: Trade policy reform

On 24 April 1994, Professor Kader Asmal, legal scholar, antiapartheid activist and an ANC candidate for the 27 April election, participated in the signing of the Marrakech Agreement, which concluded the Uruguay Round of GATT. Asmal was not yet a representative of any government, and yet he joined South Africa's Minister of Trade and Industry (and Finance), Derek Keys, in Marrakech.

Asmal was not in Morocco as a passive observer; he was there to signify the commitment of the ANC to the conclusion of the Uruguay Round, and to South Africa's accepted offer. The government and the ANC both knew that the credibility of South Africa's commitment to this hugely important international agreement would be very low three days before a general election that was expected to turn political relations in the country upside-down. One way to engender belief in South Africa's commitment to the Marrakech agreement was for the ANC to send a senior representative along with the official South African government representative, to endorse South Africa's position.

Professor Asmal was a suitable man for the job. Born in Stanger in the subtropical province of KwaZulu-Natal, Asmal spent 27 years in exile teaching at Trinity College, Dublin, where he specialised in human rights, constitutional and international law. In 1990, his return journey ended in Cape Town where he headed a new ANC-aligned legal research centre at the University of the Western Cape (Harber and Ludman 1995: 3). Though not particularly interested in trade issues, Asmal is a very well-informed and erudite man, with unquestioned credibility in the ANC leadership, and a very sound grasp of international relations. He was soon to be a surprise appointment as Minister of Water Affairs and Forestry in Mandela's first Cabinet, and to bring to that apparently mundane job an aura of competence and excitement.

Asmal and Keys showed the world at Marrakech that South Africans understood that it was important to demonstrate that political history would not stand in the way of national interest in the new South Africa. Their double act at Marrakech (echoing Mandela and De Klerk's double acts in Davos and Stockholm) provided another foretaste of the realism and pragmatism that would characterise the first ANC-led government of national unity (GNU).

But, on what basis was the ANC able to commit itself to the GATT agreement? Two key ingredients combined to create the possibility of effective policy making in the interregnum period of the early 1990s. The first was COSATU's determination to intervene in the policy process, and the second was Derek Keys' strategic vision. Both realised that policy development and implementation were dead during this period, unless a legitimate interim policy-legitimising structure was created. Soon after he was appointed in charge of Trade and Industry, Keys made a speech in which he referred to a golden triangle between labour, business and government. In it, he implicitly acknowledged that he was powerless in a lame duck government (before the 1994 elections) unless he had the support of the best-organised economic constituencies in civil society. After some discussions, all three constituencies endorsed the establishment of a National Economic Forum (NEF), formally as a policy advisory body, but in practice, an organ of policy endorsement.

Though formally the forum represented government, business and labour, the ANC was also present, both through union leaders who were also ANC leaders, such as Jay Naidoo and Alec Erwin, and through participants in the Forum who represented the ANC and no one else, such as Tito Mboweni and myself. But the fact that the ANC was a participant was never formally acknowledged, and the ANC representatives always sat amongst the COSATU delegates.

The NEF discussed a number of policy issues, mostly dealing with trade and industrial policy matters. For example, it endorsed a plan to improve the operations of a leaky customs and excise agency, although the plan failed to have much effect at the time as the organisation needed to be entirely renewed, as it eventually was. Another issue discussed at the NEF was how to phase out the export subsidy called the Generalised Export Incentive Scheme (GEIS) that had been introduced for a limited period in 1990. But the real meat of the NEF, which may have been Keys' intention all along, was South Africa's offer of trade reform in the Uruguay Round of the GATT.

South Africa had submitted an offer in 1990, but because other issues delayed the conclusion of the Uruguay Round, South Africa's offer found itself on the backburner for years. Keys realised that the offer would not stick if labour and the ANC did not support it, and the business community was also anxious as it felt that the consultation process preceding the 1990 offer was shallow and a sham. So Keys withdrew the offer and put the matter before the NEF.

Throughout the period between July and December 1993, members of the Forum toiled over endless reams of spreadsheets of base tariffs (tariff or equivalent in the base year of the Uruguay Round negotiations), applied (actual) tariffs and proposed tariff bindings (ceilings above which new tariffs could not rise). Some spreadsheets also indicated the number of workers employed, imports and exports, and the relative magnitudes and impacts of the proposed tariff cuts. The only government agency with the skills and computing capabilities to act as a secretariat for this process at that time was the IDC. When they met in the panelled IDC boardroom, the NEF members deliberated under the gaze of six stern, white, male, former chairpersons of the IDC board staring uncomprehendingly out of larger-than-life, oil-painted portraits.

All the participants agreed that as a somewhat small and isolated economy, South Africa relied on the multilateral system of trade law to protect itself; they generally agreed that quantitative restrictions should be removed, that the number of tariff lines should be reduced by at least 30% from about 12 800, and that the number of tariff rates should be cut sharply back from about 200. The maximum tariff had to be cut back even more sharply from 1 389%. Not all participants were equally enthusiastic about the 33% average phase-down on all industrial tariffs and the 21% cut on agricultural products. However, the way the IDC chose the base data and calculated the conversion from quantitative into tariff barriers, generally inflating the initial protection levels so that the real cuts were not as bad as they appeared, assured most that the cuts would not be too excessive when phased in over five years.

The industrial tariff was approached in a systematic way. The fact that there had been an industrial policy debate for several years meant that there was a common discourse. The key feature of the new tariff structure was that it would have a cascading structure whereby raw materials would be unprotected, capital goods and intermediates would be taxed no higher than 10 or 15%, and consumer goods would be protected to the extent of 20 or 30% of their value. Where protection was already lower, it would generally not be raised. All duties lower than 5% would be entirely scrapped (Hirsch 1995).

There were some exceptions, for sensitive products – products that had very high levels of protection and significant levels of employment. The main sensitive sectors were clothing and textiles, and motor vehicle assembly and component manufacture. In both broad sectors several hundred thousand people were employed (directly and indirectly), but the level of tariff protection was very high – over 100% in some cases if one included the effects of the import surcharge.

With Mandela's intervention with President Clinton (before South Africa's 1994 election), South Africa's industrial offer was accepted in spite of the fact that it required longer phase-down periods and higher terminal tariff ceilings for the two sensitive broad sectors. The industrial offer was exemplary in most other respects.

The same was not true of the agricultural sector offer. There had not been an agricultural policy debate similar to the industrial policy debate, though there had been liberalisation of the domestic agricultural sector since the 1980s. Instead of tailoring the offer in terms of an agricultural strategy, however broad, all that was done was for the base tariffs to be set at very high levels, exploiting the vagueness of the process of converting quantitative restrictions into notional ad valorem tariffs, and to make the minimum cuts reach the new tariff bindings. This indulged the South African agricultural sector's resistance to change where it could be avoided, it showed the closeness of the Department of Agriculture to the agricultural status quo, and it demonstrated COSATU and the ANC's uncertainty regarding agriculture. It also reflected the fact that the world market for agricultural products, plagued as it is by preferential systems and subsidies in the major agricultural importing countries, is much less transparent than the market for industrial products – South African negotiators were content to be cautious.

However significant South Africa's GATT tariff reform commitments, the biggest single jolt to the status quo was the final removal of all import surcharges in October 1995. The import surcharges had covered about 60% of tariff lines, and ranged up to 40% of import value for consumer products.

Trade reform continued: Export subsidies, guarantees and promotion

The core of the Uruguay Round was a 33% cut in industrial tariffs, a 36% cut in agricultural tariffs and a 21% cut in agricultural subsidies, each over a five- or six-year period. There were other important elements, such as the attempts to establish new world regimes for services trade and for the management of intellectual property rights that South Africa participated in, and an attempt to establish new rules for government procurement, a voluntary code that South Africa declined to join. But the key effect for South Africa of the 1994 GATT agreement, beyond tariff reform, was the end of GEIS.

GEIS was launched in April 1990 amidst a deep economic panic. Foreign capital inflows had practically dried up and commodity markets were in the doldrums, jointly threatening South Africa's balance of payments. Tight money policies and an import surcharge, instituted under the emergency balance of payments provision of GATT, sought to constrict domestic demand. But, the economy was set to plunge even deeper into recession unless a new source of growth and balance of payments strength could be found. It was the culmination of an escalating crisis that began with the slump in the gold price in 1983–85, and the departure of international banks and their credit lines after P.W. Botha's Rubicon speech in August 1985.

The government decided to subsidise exports that were not raw materials. Instead of a complex sector-specific system that the BTI proposed in its 1988 'structural adjustment' report, the government chose a simple export subsidy. The subsidy was a tax-free grant paid to all exporters in proportion to the value of their executed export order. It was divided into four categories, receiving different benefits: primary product exporters (e.g. logs or minerals); beneficiated primary products (e.g. saw logs, billets); material-intensive products (e.g. planed planks, sheet metal); and manufactured products (e.g. furniture, steel cabinets) (Holden and Gouws 1997). Manufactured products got a tax-free subsidy of about 21%, the primary producers got nothing, and the rest got between about 5 and 14%, all modified by a formula that included a factor for changes in the real exchange rate.

The administration of the programme was fraught with fraud and backlogs, and budgeting was always out by hundreds of millions of rand. In the end, perversely, the biggest beneficiaries were the manufacturers of intermediate products such as paper and steel, who now received huge tax-free windfalls on products they would have exported anyway. The beneficiary companies were usually large, capital-intensive processing firms such as Sappi and Mondi (paper and pulp), and Sentrachem and AECI (basic chemicals), which belonged to the giant conglomerates Anglo American and Sanlam, or they were parastatals or privatised parastatals such as Iscor (iron and steel) and Armscor (military weapons) (Fine and Rustomjee 1996: 232). Most analyses of GEIS indicated that it had little effect on exports, especially when compared with factors such as the real effective exchange rate and business cycles (see Holden and Gouws 1997).

GEIS was a headache for government. As soon as he could, early in 1993, Minister Derek Keys removed the tax exemption from GEIS payments. But in 1994 it was difficult for the new government to do away with GEIS precipitously. Firstly, it was concerned that a commitment by the state to continue the programme for a certain number of years (its duration was extended once) had to be honoured, otherwise measures offered by the new government might not have sufficient credibility to influence investment decisions. Also, although GEIS was not cost-effective, most analysts inside and outside of government believed that South African exporters, especially new exporters in small and medium companies, needed encouragement from government, and not much existed beyond GEIS. The new government did not want to send the signal that it did not strongly support manufactured exports.

So, the requirement of the Uruguay Round that unadulterated export subsidies should be removed before the end of 1997 came as a blessing for government. It could blame GATT for the termination of GEIS. Soon, Zavareh Rustomjee, as new Director-General of the DTI, felt that whatever positive influence GEIS had ever had on exports would dissipate with the decision to phase it out, so he accelerated the downsizing of the subsidy, with the very last GEIS commitments made in July 1997. All the same, the programme still cost government about R21 billion in nominal rands over its seven-year history, or about US$5 billion – a lot of money for South Africa. For several years it grabbed the lion's share of the budget of the DTI.

Was there a better and more cost-effective way of supporting new exports? The first new export support programme launched by the democratic government was a credit guarantee programme for small exporters: the first programme to be developed through the Japanese Grant Fund process (see Box 4.1). The scheme was loosely based on a much bigger Korean programme, and was developed in outline for South Africa by Korean consultants working with Yung Whee Rhee, a top Korean industrial policy expert based at the World Bank. The programme took off more quickly than most new South African government programmes aimed at smaller firms, and was fairly effective (DTI 1998a: 90, 1999: 32).

The Credit Guarantee Insurance Corporation built the short-term export credit guarantee for small firms on top of an existing government programme to provide export credits. Another important element in the export guarantee portfolio is the Export Finance Scheme for Capital Projects, which is also underwritten by government. This programme is important in supporting South Africa's effort to penetrate developing country markets, especially in Africa, with competitive manufactured goods and project services (DTI 1998a: 90–91).

A second new programme linked with the trade policy reform was the World Player Scheme developed and run by the IDC. The IDC recognised that numerous firms, including several of its own existing interests, could be detrimentally affected by the tariff down phasing programme launched in Marrakech. The IDC offered all firms in industries facing tariff cuts of 15 percentage points or more, discounted loans for projects designed to bring them up to a suitable level of competitiveness. Loans were available at a 3–6 percentage point discount on the normal IDC fixed rate, which itself was usually 3–6% below the prime interest rate. The IDC lent close to a billion rand (about US$250 million) through this programme from 1995–98. Nearly half went to the textile industry and about one seventh went to the motor vehicle industry (DTI 1998a: 73; IDC 1998: 21).

The other key trade policy reform was the major overhaul of an export marketing assistance scheme. When GEIS was finally abolished in 1997, the government and its social partners in Nedlac felt that there should be an additional compensatory stimulus to exports, particularly for smaller, more labour-intensive firms. Instead of developing an entirely new programme, the DTI decided to remodel and upgrade a rather low-level existing programme. The new Export Marketing and Investment Assistance Programme offered seven modes to support firms in their efforts to find markets abroad for their products, and two modes to find foreign investors interested in South African opportunities.

A major new component was the provision for government to support the establishment of sub-sector export councils that could co-ordinate marketing strategies for groups of related firms. The first export councils established, in 1999, were those for structural steel, for capital equipment, and for stainless steel (DTI 1999: 18–19, 31).

Accessing markets

The new government's role in promoting exports was not confined to tariff and incentive programmes – there was also the issue of accessing and securing markets for South African exporters. Even in a world where the WTO was increasingly influential, special relationships were very important, especially in trade relations between developed and developing countries, and in the trade of agricultural products. Many of South Africa's competitors had such relationships. For example: Canada, the United States, Mexico and Chile entered a combination of free trade agreements; southern European countries were members of the European Union; eastern European countries had special deals with Europe, as did Israel (also with the United States) and much of North Africa; while all African countries other than South Africa, and several Caribbean, Asian and Pacific nations were members of the Lomι Convention with the European Union, giving them quotas and tariff preferences, mainly in agricultural products.

One symptom of this problem was the loss of market share by South African canned fruit exporters during the 1990s that led to damaging job losses and factory closures in the Western Cape. The shift in market share was almost entirely attributable to European preferences and subsidies (Kaplan and Kaplinsky 1998). It was this shift and the fact that South Africa is not located geographically near to any of the major world markets, that persuaded South African politicians and civil servants to focus on the issue of market access – bearing in mind that it was not only market access that was at stake, but also the investment flows that usually accompany strong trading relationships.

South Africa was ill equipped to enter negotiations – the isolation of the previous regime meant there was no legacy of trade negotiation skills. But, negotiations could not be avoided and officials entered an intensive period of on-the-job training.

The highest priority was to secure and deepen relations with South Africa's neighbours in southern Africa collected in the Southern African Customs Union (SACU) and the Southern African Development Community (SADC). One reason was that the new South African leadership felt that South Africa would sink or swim with the region. South Africa's success, particularly as an investment destination, was in part a function of the economic success and political stability of its neighbours. A second reason was that, as sanctions declined, southern Africa (13% of exports) became, as a group, South Africa's third largest export destination after the EU (38%) and the North American Free Trade Agreement (NAFTA) (14%), and the most important market for South Africa's manufactures and services, as opposed to primary products (DTI 1998b). The South African trade surplus with the region grew large and needed regional co-operation to sustain it. Another reason for focusing on the region was that South Africa's policy makers believed that if it could strengthen ties with its neighbours, heading for a free trade area or customs union, both South Africa's and its neighbours' bargaining power with the rest of the world would be enhanced. It should be noted, though, that the sum of the GDPs of all of South Africa's partners in SADC comes to less than half of South Africa's GDP.

South Africa's expectation in reforming SACU was that the relationship between the partners – South Africa, Botswana, Lesotho, Swaziland and Namibia – would become more equal. SACU had a colonial form in that the formula for redistributing the customs revenue inflated the incomes of the poorer members in compensation for having no real control over the customs union. Decisions about adjusting tariffs were South African domestic policy. The other members of the union, which had to go along with the decisions, were paid more than a pro rata share of customs revenue as compensation. But the poorer countries depended excessively on the customs union revenue for their tax base – for Lesotho, customs revenue makes up more than half of the government's annual income – and proved resistant to a major change in the formula. Progress on the democratic form to be adopted by the new SACU was also much slower than expected.

While reforming SACU, South Africa set out to join a SADC initiative to move towards a free trade area – SACU is a subset of the members of SADC. First South Africa had to join SADC, which had originally been set up by South Africa's neighbours as a defence against the apartheid state. Today SADC consists of SACU, plus Angola, Malawi, Mauritius, Mozambique, Tanzania, Zambia, Zimbabwe and the Democratic Republic of the Congo.

Late in 1998, South Africa put the first proposal for a mutual tariff phase-down on the SADC table. It proposed a 12-year period during which first South Africa and then its SADC partners reduce their tariffs, culminating in a free trade area. Like all free trade areas there would be some exceptions for sensitive products but, in line with WTO unwritten guidelines, this could consist of no more than 15% of traded goods. Though a protocol was agreed, the process of ratification and implementation has been very slow. Other barriers to the movement of goods, services and people have also slowed progress in the integration of the region.

The second priority for South Africa was its economic relationship with the EU. The EU had long been South Africa's major trade and investment partner, with the British and German markets being the most important. The urgency of the matter lay in South Africa's concern that its relationships with traditional trading partners were eroding. The EU was becoming enlarged (Sweden, Austria, Finland, former East Germany and, imminently, Poland, Hungary and the Czech Republic), developing new free trade agreements with Morocco and Tunisia for example, and deepening through the European Monetary Union and later the Euro. In relative terms, with all the EU's special relationships, South Africa was losing ground fast.

The EU approached South Africa shortly before the democratic elections in 1994 with an offer of an 'association agreement'. South Africa wanted access to the EU market on the same terms as its neighbours, partly to encourage economic integration in southern Africa, which entailed membership of the Lomι Convention. The EU blocked this, partly because it considered the South African economy too strong for Lomι, and partly because it was trying to steer Lomι towards a set of regional free trade agreements. South Africa found itself engaged in negotiations over a trade and development agreement with the EU, centred on a free trade agreement, and concluded after six years of discussions and negotiations.

By the early 2000s, the focus of South Africa's trade discussions had shifted to the possibility of free trade agreements with India, Brazil/Mercosur, and the United States, and to the Doha Round of WTO-led multilateral trade negotiations where South Africa helped build a coalition called 'the group of 22', which pushed the EU, the United States and Japan towards more significant concessions for freer trade with developing countries.

Investment strategies

Between the Second World War and 1976, total investment grew strongly in South Africa. Most of the growth was attributable to local companies and the public sector, though direct foreign investors played a role in important sub-sectors of manufacturing and the service sector. Gross domestic fixed investment (GDFI) reached a peak at an average of 26% of GDP over the period 1971–76, which was a higher ratio than that of most middle-income countries, such as Mexico, Brazil and Chile, following an import-substitution policy at that time. It was comparable, rather, with countries following export-oriented growth strategies, such as Malaysia and South Korea. Only in the 1980s did these export-oriented economies surpass South Africa in relative size of investment effort (Fallon and Perreira da Silva 1994: 53).

This is part of the tragedy of South Africa – it had so much capacity to invest, and yet the investments reaped such poor rewards. Though private investment was strong, from the 1960s to the mid-1970s investment was led by the public sector. Massive investments were made in roads, dams, railway lines, electrical power, and in 'strategic' industries such as synthetic fuels, nuclear power, iron and steel, and armaments.

Public investment was directed in terms of the perceived needs and desires of the constituents of the minority regime, the whites – and white farmers and business-owners in particular. In addition, distortions in the economy such as negative real interest rates and exchange controls for much of the 1970s and 1980s, and the apartheid restrictions that inhibited investments in human capital, undermined the value of the investments and led to investments that were excessively capital intensive and absorbed little labour. The social and economic returns on these investments were generally low.4

With the radical fluctuation of many key prices in the 1970s, the world of the NP slowly came unstuck. Government investment starting falling in the late 1970s, and investments by parastatals followed soon afterwards. After 1985, with the evacuation of foreign banks and their credit lines, the disinvestments of many important foreign-owned businesses (such as IBM, Ford, General Motors and two major British banks), and the privatisation of some parastatal companies, investment fell apart. From 1983–93, GDFI fell from 26.8% of GDP to 15.5% of GDP, while in the period 1986–91, public sector investment growth declined on average by 6.5% per year (SARB 1998b; Fallon and Perreira da Silva 1994: 54).

The old government had several industrial investment promotion schemes. A major theme in the investment support programmes was 'industrial decentralisation'. The main motive was political: to keep black industrial workers out of the major urban centres where they were getting more and more powerful, and to build up a loyal black middle class in the Bantustans. Later still, industrial development in some of the Bantustans was intended to legitimise their putative independence.

The first decentralisation programme began in 1960, but only its successors became really effective in the late 1970s and 1980s. The government poured in subsidies of all kinds, but the vast majority of operations established were destined to survive no longer than the duration of the subsidy programme. Most of the investors were white South Africans, but there were also foreign investors from Europe, Taiwan and Israel.

By the end of the 1980s, the cost of the programme weighed heavily, especially in the light of the growing cost of export subsidies. The report of an enquiry commissioned by the government suggested that the programme should be scaled down with regards the range and depth of the subsidies, but that they should now extend to all new manufacturers anywhere except in the Pretoria-Witwatersrand-Vereeniging area (the PWV). The PWV was deemed not to be in need of such support.

In 1993, a special version of the Regional Industrial Development Programme (RIDP) was developed for smaller firms. This programme – the Simplified RIDP – was effective in attracting more labour-intensive projects than the RIDP, at about R50 000 per job, rather than about R300 000 per job.

In 1996 both programmes were reviewed, using the Nedlac-based Japanese Grant Fund programme. The objectives of the government had shifted, fundamentally. The apartheid government tried to use investment incentives to shape society along lines that would suit its system of political oppression. Investment programmes were essentially nice-to-have add-ons to already satisfactory levels of investment, as far as the rulers were concerned. The objectives of the ANC-led GNU were very different. The level of investment had to be raised from very low levels, to increase the country's wealth, and, equally important, jobs had to be created for the many millions of unemployed.

Also, there were parts of the country that were neglected or even avoided for investment during the apartheid era for political reasons. One example is the eastern Transvaal, now Mpumalanga province, which is located next to Mozambique. Because apartheid South Africa's policy towards Mozambique meant that the latter country's ports were not accessible for South African importers and exporters, the whole region suffered. Another example is the former Transkei littoral, which has tremendous tourist potential, but lacks infrastructure because the apartheid government chose to marginalise the predominantly black region.

The ANC believed that breaking down the politically erected obstacles to development would not be enough to counteract years of neglect. But the national leadership also wanted to avoid the kind of subsidies that could enter the currency of pork-barrel politics, and could further distort markets for no good reason. Other considerations in the ANC's approach to investment included a desire to encourage foreign investors, especially where they brought with them good technologies and/or markets, a desire to increase the 'competitive temperature' in South Africa by introducing new rivals that would challenge old oligopolies, a desire for more labour-absorbing types of investment, and a desire to build up a significant black capitalist class beyond the tentacles of the oligopolies, with real presence in the industrial sector of the economy (not only finance and commerce) (Hirsch 1992a; ANC 1997b: 41–50).

Following a 1995 consultant's report on foreign investment commissioned by the Japanese Grant Fund of Nedlac, the government decided to relocate the marketing programme to attract foreign investors outside of the DTI, where it had languished. There were two main reasons: the intention to instil an entrepreneurial spirit within this organisation, which is more difficult in a government department; and a wish to ensure that the provincial investment promotion agencies truly believed that their seat on the board of the central agency meant that they owned it. The provincial agencies would do most of the work in attracting potential investors and reaching an agreement, but the national agencies could exploit economies of scale in marketing and information management. Investment South Africa was publicly launched early in 1997. Later it took on export marketing responsibilities too, and still later it was reincorporated back into the DTI.

At the same time, the government worked on improving the range and quality of the incentives and instruments to encourage both foreign and domestic industrial investors. It was the firm conviction of the economic policy leadership that frills such as financial incentives were secondary; rather, the overall economic and political conditions were far more important when it came to attracting the right kind of investment (see, for example, ANC 1997a: chapter 1).

The most important government-controlled agency for the development of the real sector, after the DTI, was the IDC. The IDC was set up in 1940 with state funds as an industrial development bank, to invest in strategic industrial sectors. Since the early 1950s it has operated successfully without any further injection of state funds, though until the 1990s it still took advantage of state guarantees of loans to parastatal companies. The legal mandate of the IDC was simple: it should invest in industrial undertakings that benefit the country and also meet the criterion of economic viability. Economic viability really means that the firm invested in should ultimately be expected to make profits and stand on its own. Under the ANC, the IDC's mandate was extended to allow investments in other African countries. The main difference between the IDC and a private investment bank was that with only one shareholder, the government, the IDC was able to take a longer view of investments, and to take some risks that a privately owned bank might not. Nevertheless, the corporation developed a practice and culture of proposition evaluation and economic analysis that made its professionals and its judgements very highly regarded in the private sector.

With its assets worth close to US$3 billion in 1994, and its culture of professional competence, the new government was keen to exploit the virtues and expunge the sins of the IDC as quickly as possible. But what were those sins? The IDC had become quite deeply integrated into elements of what Rustomjee called the 'minerals-energy complex'. Most of the high level work in the IDC was directed towards new 'beneficiation' projects such as steel mills and aluminium smelters. Its integration with the minerals-energy complex also meant that it developed cosy relations with some of the biggest conglomerates, such as Sanlam and Gencor.

Another weakness of the IDC was that it did not devote sufficient resources to black economic empowerment (BEE) and small and medium business development, which are two overlapping but not identical issues. The organisation itself had relatively few black professionals, and none who were in senior positions. Another issue was that, outside of the minerals-energy complex industries, the IDC did not approach investment propositions in the context of sectoral industrial development plans.

The IDC law seemed broad enough to accommodate new parameters. Government economic leaders approached the challenge of reforming the IDC by reconstituting its board and allowing the board to focus on policy issues, and the appointment of senior management. In 1996, the new board members found a new CEO, Khaya Nqula, who had worked for IBM and South African Breweries in marketing, and ran what was then South Africa's fastest-growing unit trust (mutual fund) business. Nqula agreed to a five-year contract, indicating that he had no intention of staying on after that.

Once in position, Nqula moved quickly, first to restore the morale of an anxious organisation, and then to make a string of new appointments, bringing black and female staff into executive management for the first time in the organisation's history. The IDC was turned around, without having to be dismantled and rebuilt from scratch. This was the way the new government preferred to operate: to retain existing assets, in this case the financial assets and the intellectual capital of the IDC, but at the same time turn the agency into an effective organ of reconstruction in the era of democracy.

The RIDP was cancelled after the Nedlac review. The evidence suggested that other methods should be tried to attract significant industrial investments. However, the modified version of the RIDP developed for small and medium firms was retained in a restructured form as the Small and Medium Manufacturing Development Programme (SMMDP). Later this was broadened beyond the manufacturing sector and renamed the Small and Medium Enterprise Development Programme. Its bait is an annual cash grant for investors for up to three years, based on the size of the investment and the audited performance of the firm. It is weighted toward smaller investments and can be used anywhere in South Africa where an enterprise can be established. When the SMMDP was reviewed after 18 months of operation, it had created approximately 26 392 new jobs. At R73 000 (or less than US$15 000) per job, these investments were clearly in labour-intensive sectors (DTI 1999: 15–17).

But for larger industrial investors who had easier access to funds and expected high profits, reduced taxes seemed like the best option. At the time of the RIDP review, in 1996, the government was short of revenue and could not make overall tax reduction commitments. So, a tax holiday programme for larger industrial investments, which was supported in the review, was announced in the GEAR statement in June and launched later in the year. The government wanted to launch the programme as quickly as possible because it wanted to instil confidence that GEAR was being implemented, and because it detected a tailing-off of private sector investment. The tax holiday programme ended up being difficult to implement due to a relatively rushed process of consensus building among all the constituents – in addition to Parliament, these included business, labour, the provincial governments and the government's tax commission. In order to qualify for six consecutive tax-free years beginning in the first profitable year, the enterprise had to be financially discrete (to avoid leakages), had to be located within one of a large number of industrial districts (not excluding the PWV), and it had to indicate a sufficient level of job creation.

Between the middle of 1997 and the end of 1998, 106 tax holiday projects were approved, representing a total investment of R3.1 billion, and 8 854 new jobs (at R350 000, or US$70 000 per job) (DTI 1999: 15–16). The programme had not worked as effectively as hoped, and, in February 1999, before a mid-term review could report, the Minister of Finance, Trevor Manuel announced that the tax holiday window would close, as planned, in September 1999. Simultaneously, Manuel announced a general reduction in the rate of company tax from 35 to 30% of profits (Manuel 1999).

When the DTI pressed for more investment support, two new initiatives were introduced in 2002. The Strategic Investment Programme is an investment allowance provided to large investments that meet a certain set of criteria. So far it seems to be more effective than the tax holiday programme. The Critical Infrastructure Programme provides a subsidy to businesses or public agencies that invest in infrastructure related to a planned private investment project; a railway siding, a harbour terminal or a power line, for example.

Perhaps the most original investment programme launched by the first democratic South African government became known as the Spatial Development Programme. In economic terms, the objective of the programme was to internalise the externalities of public and private investments in targeted regions. Public and private investments could reap higher social and economic returns if they were co-ordinated. For example, if the government was building houses, roads and schools, it would be more sensible to do this where firms were planning to create job opportunities. Some of the risk is removed from the government projects. And vice versa: the public investment could very well lower the cost of the private investment by reducing a firm's outlay on social and economic infrastructure.

In advanced economies, such as the United States, state and city governments would undertake such co-ordination. In South Africa, though, most provincial and city governments did not have the skills or financial resources to make this work. National government and national agency involvement is still essential in South Africa at this stage.

The first Spatial Development Initiative was a kind of experimental accident. South African government officials – particularly Paul Jourdan at the DTI – realised that the political settlement made possible significant new investments in the Mpumalanga region and across the border in southern Mozambique. After negotiations and arm-twisting, the South African government agreed to rebuild and extend both the road and rail links to Mozambique's Maputo harbour. Maputo is the closest harbour to much of the mineral rich interior of South Africa. The Mozambican government agreed to work on its road and rail links, and to upgrade the harbour. In the meantime, firms were being recruited to consider the region for investment, with the active support of the IDC. However, government decided that it would not go ahead unless the private sector saw the whole project as viable, and this signal required public/private partnerships in some of the infrastructure projects. The positive response to the Department of Transport's tender for a built, operate and transfer (BOT) toll road was the key signal. Public and private investments have since flooded into the region, which is seen as a major example of three key partnerships: between national and provincial/local government; between the public and private sectors; and between three countries – South Africa, Mozambique, and Swaziland (which shares a corner of the region).

As evidence emerged of the success of the Maputo Corridor Spatial Development Initiative, the methodology was extended to other regions that had a high development potential and need. Whereas the Maputo initiative is based on the extraction and processing of minerals, though also on agriculture and tourism, the Wild Coast initiative (in the former Transkei) and the Lebombo initiative (northern KwaZulu-Natal) are primarily tourism-oriented. The West Coast initiative in the Western Cape combines industrial, agricultural and tourist elements, weighted more equally. The key ingredients in successful Spatial Development Initiatives are effective co-ordination at a technical level, and at a political level. A special Cabinet Investment Cluster, which linked Ministers from key ministries such as Transport and Water, chaired by the then Minister of Trade and Industry, Alec Erwin, helped ensure that technical initiatives would get political support. It became an element of the strategy to ensure that a sufficiently prominent national politician, working with a sufficiently prominent regional politician, led each Spatial Development Initiative.

In some cases, no additional government funds were needed beyond the normal budget, though funds were diverted to Spatial Development Initiatives as a temporary priority. In one case, as in the need to build a new port near Port Elizabeth, a bigger public commitment was required. But this was rare. Mostly, no new financial commitments were required, beyond the cost of running the Spatial Development Initiative teams.

The key weakness of the Spatial Development Initiatives is the inevitability, in a democracy, that there will be too many initiatives and not enough effective public and private management skills. All the leadership can hope for under these circumstances is that the best prospective Spatial Development Initiatives get the 'A-team' treatment. In some cases the provincial government has taken over the initiative, such as in the case of Gauteng's Blue IQ initiative, which is essentially a special development plan with strong public support.

Two of the most important lessons gradually learned by the government during the first decade of freedom were that investment dynamics often were not national – rather they were frequently regional and local – and that the manufacturing sector was not ever likely to be a major supplier of jobs in South Africa, though it remained an important dynamo for growth. The result was that the location of investment incentive programmes began to shift away from their traditional location in the DTI, towards other departments, and provincial governments. The Department of Tourism and linked national and local agencies developed an extensive system of incentives for investment in the tourism sector. Gauteng's Blue IQ programme of infrastructure development, aimed at investment in a range of key sectors identified by the province, is another example of this new generation of investment strategy. The eThekwini (Durban) municipality's support for the tourism and conference sectors is another example, and the Western Cape government's support for the development of a major film production centre in Cape Town is another.

Supporting industrial innovation

In general, South African manufacturing firms had no reputation for genuine product innovation. Engineers were known for their ability to modify products and processes, and to find cheap short cuts, but most South African manufactured products were and continue to be made under licence to intellectual property rights holders in the advanced industrial economies. This is not surprising as the nature of South Africa's protectionist regime encouraged licensing and copying for the domestic market, not world-class innovation.

There were a few exceptions in industries that were heavily supported by the government because of their strategic importance in the era of apartheid, sanctions and isolation. Some areas of effort were: oil from coal techniques; nuclear power; military specification electronics and other military products; communications technologies (military and state-run commercial); and systems integration capabilities. The main mode of support was military funding of long-term contracts that would allow for innovation in state-owned or private firms. The government supported science councils, especially the Council for Scientific and Industrial Research (CSIR), would also be contracted to assemble their expertise behind the projects. It was a focused marshalling of nearly all the existing technology innovation talent, and channelling such talent from the universities into a carefully chosen set of missions.

From the beginning of the 1990s, military funding fell rapidly, and the long-term communications contracts came to an end. The focus shifted to commercialisation and cost savings in the context of re-entering the global market place. One unfortunate result was that private sector and public sector innovation expenditure, specifically research and development, fell from just over 1% of GDP in 1991 to under 0.7% of GDP in 1997. By 2001 the effects of new policies and programmes came through and research and development recovered to 0.76% of GDP, and rose to 0.81% of GDP in 2003, but this is still relatively low in an international context (Department of Science and Technology 2005).

Key policy developments were a science and technology white paper in 1996 sketching out the model of the national system of innovation, and laying the framework for incremental policy reform. This was followed by a national research and development strategy in 2001 and specific strategies for biotechnology and advanced manufacturing. An important complementary policy was the 'Integrated Manufacturing Strategy', which emphasised the importance of the development, organisation and transmission of knowledge for the development of a competitive manufacturing sector (DACST 2001a, 2001b; DTI 2002; National Advisory Council on Innovation 2003).

The first new-model innovation support projects were developed in the early 1990s. The Support Programme for Industrial Innovation (SPII) was initially developed by the DTI for the electronics sector, which was seen to be in crisis as demand from the military and the state-owned telecommunications company, Telkom, plummeted simultaneously. Soon it was extended to other industrial sectors. It is a matching-grant programme that supports commercially oriented innovators.

The Technology and Human Resources for Industry Programme (THRIP) was developed at about the same time. Its objective was to link the world of tertiary education in science and engineering with the industrial world. The government added a 50% grant to the contribution by a private firm or consortium to a university- or technikon-based applied industrial research project that had to involve the training of students. THRIP got off to a very slow start, but expanded very rapidly after being remodelled, reorganised and augmented.

These new programmes were designed to combine the knowledge provided by the market with the knowledge developed in public and private research institutions. Several other such programmes were developed to fill evident cracks in the late 1990s. But this did not really solve the problem or challenge of refocusing the science councils.

South Africa's science councils are significant organisations. The CSIR is one of the largest – it receives about R400 million from the government annually and earns a little more than that through contracts with the government and the private sector. By South African standards, this is a lot of money for industrial innovation research. The CSIR is one of a group of 'science councils' that conduct research, develop technology, build information bases and/or develop standards in fields such as agriculture, mining and metals, geophysics, medicine and human sciences. The science councils are partly government funded, and partly funded through contracts and fees through the government and private clients.

In the early 1990s, the CSIR shifted strongly towards the market as it saw that government contracts were drying up, and the 'parliamentary grant' would not grow rapidly. Though this was a logical and sensible thing to do under the circumstances, it meant that the CSIR drew on its existing knowledge resources without building much new intellectual capital through basic and exploratory research. By the end of the decade it was recognised that the CSIR had to balance its commitment to commercial contracts with a strong commitment to technology platform development.

The dilemma over the role of the science councils has three main elements. The first is whether the state has a role to directly supply innovation and technology services, or whether the state should aid innovation through grants or tax concessions directly to the clients. The second is, if the state is going to support the supply side of science and technology, is the general 'council' suitable, or should it focus on smaller, more focused, closer-to-the-client laboratories? The third is, if the council system is maintained, how much of the councils' funding should come from dedicated state funding (the 'parliamentary grant')?

Decisions made so far try to find a sensible path between these choices. Because the science and technology community is relatively small in South Africa at the moment, the councils will be retained for their economies of scale. This allows for long-term investments in equipment and people without expecting instantaneous results. The system will use both supply-side support through the councils and technology stations in higher education institutions and manufacturing advice centres, and support to users through grants or tax concessions, which should rise considerably. And, within 'the science vote' the funds committed to the parliamentary grant slowed in real terms while an increasing proportion of funds were channelled through competitive funds, such as the Innovation Fund, for which the science councils have to compete, and have to form collaborative partnerships with private companies and higher education institutions. Other competitive channels emerged with the creation of Biotechnology Regional Innovation Centres, and are being considered in terms of the strategy for innovation in advanced manufacture, modelled on the CSIR's recently established Automotive Innovation and Development Centre near Pretoria.

However, the biggest underlying concern about the national system of innovation is the constrained supply of people with the skills to innovate. During the 1990s the indicators were negative: a stagnating or declining number of science and engineering graduates; a deterioration in South Africa's relative performance in patenting and publishing in scientific journals; and the average age of South Africa's most productive scientists was rising, and many of them were white males over 50. One of the key constraints was the relatively small number of high-school graduates with suitable maths and science qualifications. The total number of registered professional engineers fell during the 1990s, and though the number of engineering technologists grew, it grew off a low base. These indicators did not bode well for South Africa's future based on competitive innovation (National Advisory Council on Innovation 2004).

The number of registrations in science and engineering degrees began to grow quite rapidly in the 2000s, and, together with the rising percentage of GDP devoted to research and development, this could herald the beginning of a new era of innovation in South Africa.

Strategy development and information programmes

Government support measures should address market failures – where the interests of individual firms contradict their collective longer-term interests. For example, in an economy short of skilled labour, firms are often reluctant to train workers because they expect that competitive firms will poach their trained workers. Without the internalisation of such externalities, firms will often make decisions that are against their own long-term interests and certainly against the long-term interests of the economy as a whole. Government can try to address these conflicts by establishing suitable rules and/or offering appropriate support. In the case of the skilled labour shortage, for example, the government could subsidise firms that offer training to their workers, or it could develop government programmes to increase the supply of suitably skilled workers (see Chapter Five).

Of all markets, one of the most imperfect is the market for information. Imperfect access to information is one of the key factors re-enforcing the inequality between the developed world and the developing world.

In South Africa, political and economic isolation during the apartheid era compounded this problem. South African managers fell behind their counterparts abroad. However, with its sophisticated research, education and communications infrastructure, South Africa should be able to catch up fairly quickly in target areas, with sufficient effort. This was the implicit philosophy behind a series of support measures developed by the DTI, which can be grouped under the heading: strategic and information support.

The first of these initiatives to get a name was the 'cluster programme'. South Africa has a history of industry sector investigations undertaken on the premise that the outcome would consist of a recommendation to government. For example: the import tariff should be increased, and the government should provide additional funds to train workers. In the wake of the Industrial Strategy Project (ISP), the new government's industrial policy leaders believed that, very often, the answer to the challenge of competition was not in the government, but in the firms or group of firms itself: re-engineering, supply-chain management, investment in key inputs, or better training methods, for example. Often the firm or the group of firms did not know what to do because they did not have the capacity to analyse their circumstances; sometimes they knew the answer, but had to be persuaded to do something about it.

More or less simultaneously, influenced by the work of Michael Porter and the Monitor Group, several institutions began to support cluster investigations. In fact, the Monitor Group conducted a study of five South African clusters during 1994, at the invitation of the ANC. The Monitor Group had offered a pro bono study, but the ANC insisted it should be a proper contract. Its findings essentially confirmed those of the ISP – that firms and industries and relationships between firms and industries were inefficient, reflecting outmoded and weak management, poorly trained workers, and a lack of rivalry amongst South African firms. What captured the imagination of South African economic analysts was the Porter methodology, which focused on relationships in addition to the more conventional benchmarking exercises (Porter 1990).

By 1996, about 15 cluster studies were underway, some undertaken by the IDC, some by the DTI, and some under the Japanese Grant Fund of Nedlac. The DTI co-ordinated cluster analysis discussions and training of analysts by international experts from time to time. Not all cluster exercises were the same. The IDC worked mostly with bigger firms and focused on investment opportunities, while some DTI and Nedlac projects tried to bring in smaller firms and look for co-operation opportunities. Cluster studies became 'cluster initiatives,' in which the relationships established during the study phase became part of a process of trying to address the problems revealed by the studies. A constant concern of the DTI, though, was to try to ensure that the outcome of the studies or processes was not solely to provide more ammunition for business and labour to lobby government for further support.

The DTI was expected to play a leading role in each cluster process, as well as manage its own. This was unmanageable in view of the weakness of the industrial sector directorates in the DTI and their myriad of other responsibilities. A new mode for encouraging cluster processes was developed – the Sectoral Partnership Fund (SPF). This was a wholly homegrown South African concept, with its roots in the ISP report (Joffe et al. 1995).

The SPF is a scheme to encourage related firms to seek a common solution to a common problem. It is easiest to describe by giving actual examples. A group of South African industrial refrigeration firms recognised that their penetration of the African market was limited not by the quality of their products but by their inferior design – the solution was collaborate on an appropriate design training strategy. A group of wood furniture processors needed to strengthen their bargaining power in relation to timber suppliers, and agreed to set up a purchasing co-operative. The SPF can be used to get such programmes going, if they meet a set of explicit criteria. In the initial version, if the project was approved, the DTI would fund up to 65% of the cost of the project up to a maximum of R3 million.

The DTI launched a similar programme for individual firms called the Competitiveness Fund. Unlike the SPF, the Competitiveness Fund, which pays 50% of the approved cost of consultants, was generated out of a Japanese Grant Fund project, strongly influenced by the World Bank's experiences in Mauritius and Argentina. The two projects were initially financed out of a World Bank loan – the only World Bank loan to the new government. It was a small loan, of about US$25 million, issued in 1998, mainly to allow the South African government to test the waters.

Another 'strategic and information' type support programme was the Workplace Challenge, which was designed at Nedlac to facilitate the joint training of workers and managers in improving productivity in a firm. This was one of the more surprising, successful outcomes of negotiations in the Trade and Industry Chamber of Nedlac. Till then, many workers and unionists had believed that productivity was a synonym for worker exploitation. Many managers probably did too. But, in the context of trade policy reform, and after studying some benchmarking data that showed how poorly South African labour and capital were generally used, workers and managers agreed that they could no longer look away from the harsh glare of competition.

Sharpening the instruments

The debate over competition policy in South Africa is full of ironies. The white business community claimed before and after the 1994 elections to be the upholders of the 'free market' in South Africa. By contrast, the ANC with its socialist background was very sceptical about leaving economic development to market forces alone, especially after white colonists had used racist policies and violence to establish their economic power in the first place. And yet, when it came to the issue of laws about market structure, the ANC pushed for a tougher pro-competition position than the white capitalists.

Two issues are interwoven in the competition debate in South Africa. The first issue is about the efficient operation of markets. South African capitalism became one of the most highly concentrated and conglomerated forms of capitalism (Du Plessis 1979). The 'mining houses' were so immensely powerful by the 1950s and 1960s that when new gold mine prospects started to dim, they bought up most of the rest of the economy (see Innes 1984). At most, there were six mining houses; some consolidated, and they began to integrate with the major financial institutions. In the 1980s when companies from the United States, Europe and the United Kingdom disinvested their South African holdings, the only available buyers were the already huge South African financial/mining house conglomerates.

The second reason for the anti-monopoly stance of the ANC government is political. The conglomerate empires were built at a time when the majority of the population was excluded from the rights of ownership and wealth. The apartheid government was isolated and needed all the allies it could get, so it was not likely to stand in the way of the conglomerates. The nature of conglomeration in South Africa led to a degree of stagnation in the private sector, and certainly inhibited its response to the new opportunities of the 1990s. Many South Africans believed that there should be a rectification of the legacy of white monopolisation of the economy. In the absence of nationalisation, competition policy was a key tool in this regard.

So, the call to challenge monopolies and oligopolies was powerful rhetoric in the hands of the ANC. The early period of debate was, in retrospect, one of the new South Africa's genuine opportunities for political fun. Then Minister of Trade and Industry, Trevor Manuel, enjoyed baiting the conglomerates. The conglomerates almost invariably dug themselves deeper into a hole whenever they responded.

The chosen spokesperson/victim of big business was Michael Spicer, a senior executive in the Anglo American-De Beers conglomerate. Spicer had come into Anglo specialising in public affairs, and as the personal assistant to Anglo's chair during the 1980s, Gavin Relly. Educated at the elite St Johns College, he attended Rhodes University, unlike many of the Anglo senior executives for whom a spell at Oxford University was de rigueur. By the mid-1990s Spicer had graduated to several Anglo boards, but remained a spokesperson, especially on issues of business-to-government relations.

Conscious of the reversal of ideological roles, Manuel hit out confidently. 'Some of the people who call themselves capitalists in this country would function best in the planned economy of the Soviet Union after 1917,' he said at a business breakfast in Cape Town early in 1995. We have capitalists who don't like markets, capitalists who don't want to compete, capitalists who don't want to be capitalists' (The Argus 6 March 1995).

'We can't survive like this in the global economy with its high competition and single set of trade rules,' he warned. 'Unless we take an entirely new approach, what's left will die.'

He then promised to introduce a new competition law before the end of the year. The existing law would be 'scrapped and replaced with much stronger legislation, or so substantially amended that it will hardly be recognisable'. Then he emphasised, 'Competition policy must feature very highly in our new approach to the economy' (The Argus 6 March 1995).

Another irony was the role played by foreign business in the debate. Local big business argued throughout the early 1990s that South Africa had to implement market-friendly policies in order to attract foreign investment. By 1995, however, prospective foreign investors were telling government and whoever would listen that the key obstacle to direct foreign investment was the defensive behaviour of South Africa's monopolistic conglomerates (The Argus 8 April 1995). Manuel could not resist bringing this into the debate from time to time.

The response from big business focused on two main arguments. South Africa's largest firms were not very large by international standards, and South Africa needed big firms to compete internationally. Secondly, they were anxious that ANC politicians should not conflate the issues of economic efficiency and black economic empowerment (BEE) in the competition policy debate. Big business felt that these were two separate issues, and that only the issue of efficiency belonged in the competition policy debate. But even the experienced and articulate Spicer struggled to avoid these arguments sounding like special pleading.

In the event, Manuel could not deliver on his promise to present draft legislation before the end of 1995. Too many other issues preoccupied the DTI to allow it to take on such a major project. Key officials were preoccupied in 1995 and 1996 with implementing the new Small Business Act, and establishing new agencies to support small businesses (see Chapter Six).

Perhaps the delay was fortuitous. By the time the government released its policy document entitled 'Framework for Competition, Competitiveness and Development' late in 1997, much of the hot air had gone out of the debate. Trade and Industry Minister Alec Erwin, his staff, and some skilful consultants piloted a careful course between big business, black business, and labour (which, surprisingly, took up the cause of black business), and delivered legislation to Parliament in May 1998. With some improvements introduced while the bill was before Parliament, it was eventually passed and signed into law in October 1998.

The bulk of the law takes its line from modern competition law, drawing on recent British laws, and on those of some European countries, Canada and Australia, rather than American law. It included several major departures from the old law. It requires for mergers above a certain size to be pre-approved by government. Unlike the old law, it lists anti-competitive practices as possible 'abuses of dominance', which also requires a definition of 'dominance'. Another innovation was the introduction of the concept of 'restrictive vertical practices' in addition to the more conventional 'restrictive horizontal practices'. Exemption provisions allow discretion on the part of the Competition Commission on the grounds of industrial policy considerations, employment considerations, and BEE considerations (Competition Act No. 89 of 1998).

However, BEE is clearly identified as one of the purposes of the law. The other purposes include efficiency, global competitiveness and consumer concerns, employment and welfare considerations, and the desire to build small and medium businesses.

To the surprise of many, and perhaps to the disappointment of the media, the new law was universally acclaimed. In practice it has worked quite well. After some modifications, the merger notification process ran reasonably smoothly. However, those fighting for lower prices, especially for intermediate goods such as semi-processed metals, chemicals, and pulp and paper products, believe that the 'abuse of dominance' provisions of the Competition Act are not effective. The concern is that they fail to address the challenge of import parity pricing – the pricing of goods fractionally under the cost of imports, which allows significant margins when the product has a high weight- or volume-to-value ratio, which makes it relatively costly to transport.

Since 1994 critics of the government, especially on the left, have complained that government has no industrial policy. This is in spite of the numerous measures introduced or considerably modified by government, such as the various investment support, innovation support and strategic and information support programmes developed since 1994, as well as the small business programme discussed in Chapter Six and the competition strategy. The complaint continued even after the launch of the Integrated Manufacturing Strategy in 2002.

Government could easily point to numerous successful interventions – certainly some of the innovation and strategic and information programmes have passed the test of stringent policy reviews. In addition, the Motor Industries Development Programme and the concerted effort to strengthen the tourism sector were notable successes of the late 1990s and early 2000s.

Nonetheless, when compared with successful strategies in East and South-East Asia, South Africa's industrial strategy successes seem modest and few and far between. In the early years after 1994, policy managers felt that government did not have the capacity to engage in sophisticated Asian-style interventions. Government focused instead on broad-brush programmes intended to allow the cream to rise to the top – to reward competence and commitment. Occasionally, such as in the tourism and motor sectors, government did engage effectively with its social partners in efficient development programmes.

It is probably fair to say that government is haunted by the nagging feeling that we could and should be doing more. Surely we know enough to be able to identify some key sectors that can grow faster with effective policies – interventions such as focused innovation support, dedicated training programmes, concerted marketing programmes, or some form of investment support? Why have we not done more of this? Could we not have grown faster than 2.94% per annum in the first 10 years of democracy?

There are several reasons – lack of confidence, a shortage of skilled management, opposition to 'targeting' in some parts of government, lack of suitable modalities with business and labour in some sectors and a range of related institutional factors. The issue has come round again after 10 or more years of democracy as part of the question: how can we sustain a higher level of investment by the private sector? How can we afford not to focus resources on high potential growth and/or employment sectors? These are some questions about industrial policy being posed in the second decade of South Africa's democracy.

Notes

1 Development economists in this tradition would include Raoul Prebisch, Hans Singer, Albert O. Hirschman and Hollis B. Chenery, whereas the economic history evidence is usually traced to Alexander Geschenkron or Simon Kuznets. Singer had contact with numerous South African graduate students at the University of Sussex, whose alumni include many senior civil servants and politicians in South Africa, including the current President, Thabo Mbeki.

2 Some key interpreters or practitioners of the East Asian miracle who have worked with South African policy makers are: Alice Amsden, Sanjaya Lall, Ha-joon Chang, Lin-su Kim, Duck-woo Nam, Daim Zainuddin, Yung Whee Rhee, and economists directly linked to South Africa such as Martin Fransman, Raphael Kaplinsky and Brian Levy.

3 Levy was not alone. Anthony Black's work has a similar perspective. Most World Bank economists working on South African trade and industry followed this position, as did I, to an extent, in my own report 'Trading up: trade policy for industrialization in South Africa' (Hirsch 1993). The Industrial Strategy project synthesis shared key assumptions with the Levy approach (Joffe et al. 1995).

4 See Fallon and Perreira de Silva (1994: chapter 3) for a discussion on the quantity and quality of investment in South Africa until 1993; and Nattrass (1990a) on profits.

Document(s) 6 of 11