ARTICLE FINANCING FASTER GROWTH IN SOUTH AFRICA: THE CASE FOR REFORMING THE FINANCIAL SECTOR Chris Edwards1 The need for faster growth in post-apartheid South Africa As is well known the problems faced by the South African economy are immense. The formal end to the apartheid system with the general elections in 1994 left a legacy of problems. At that time, three-quarters of the total population of 40 million were black Africans. They had the vote for the first time, but it has been estimated that... • a quarter of them (or about 7 million) lived in squatter camps or crude temporary housing • over 40 per cent of them (or about 12 million) had no access to clean water and that • over half of them (15 million) were illiterate. And of all rural African households, only about a tenth had access to basic sanitation and only 5 per cent had access to electricity (see Edwards 1995). Now, in 1997 and three years later, what is the position? The answer is that it is vastly improved in the case of access to electricity and water, but in one important respect, namely unemployment, it is probably worse. In 1995, according to the Central Statistical Services (CCS), the position was as set out in Table 1. Table 1 Employment, labour force and unemployment - South Africa, 1995 (millions) (%) Labour forca (economically active population) 14.3 100 Employment - formal sector 8.0 56 Employment - informal sector 1.6 11 Employment - total 9.6 67 Unemployed 4.7 33 Source; NEDLAC 1996, 4. TRANSFORMATION 35 (1998) 49 EDWARDS ARTICLE Other sources give a higher level of informal sector employment but also a larger labour force so that they come up with something like the same rate of unemployment (about 35 per cent). In 1996, the situation seems to have become even worse, for according to CCS figures, there was a loss of 60 000 jobs rather than the increase of 126 000 planned for by the Government's 1996 Growth, Employment and Redistribution ('GEAR') report (see F&T Weekly, 25 April, 1997: 9 and Financial Mail, 18 April, 1997: 48). Thus now, in 1997, an average of at least one in three adults is unemployed2 and it is probable that almost two-thirds of Africans aged between 15 and 24 do not have jobs (World Bank 1996: 4). In recent years, in South Africa, there has been a growing concern with a sharp rise in crime (what Nadine Gordimer has referred to as the 'return of the repressed' - see Guardian, 29 July, 1997:9) and although it is too simplistic to put the whole of the blame for this on unemployment, it is surely a major factor. In his opening speech to parliament in February 1996, President Mandela pointed to South Africa's slow economic growth, rising unemployment and persistent poverty and called on the public and private sectors to develop and implement a ''national vision to lift us out of the quagmire' (quoted in Nattrass, 1996,25). However to lift South Africa out of the quagmire will require a radical redistribution of income and wealth3 and/or a faster rate of economic growth than has been achieved since the elections. Focusing here on the rate of economic growth, it is important to note that the high level of unemployment persists in spite of a rate of economic growth since 1994 which (relative to the 1980s at least) has been quite strong - an average of about 3 per cent a year over the three years of 1994-96 - see NEDCOR 1996(4): 8. The rate of growth (and its composition) has not been anywhere near fast enough to keep pace with the increase in the labour force, let alone reduce the backlog of under- and unemployed. Thus even if the rate of economic growth experienced since 1994 is maintained, there is no hope of unemployment being much reduced, if at all. The labour force {after allowing for retirements) is estimated to be growing by about 2.5 per cent a year (see GEAR 1996:34) or by about 350 000 a year. If we assume that labour productivity grows by between 1 per cent and 2 per cent per annum, then real economic growth has to be something like 4 per cent a year before any dent is made in the backlog of unemployment. Thus even if we assume that a 6 per cent a year growth rate can be achieved and creates jobs at the rate of say 2 per cent a year, this means that unemployment will be eaten into at a rate of no more than about 160 000 a year. Even at this rate, it would take over ten years before unemployment is halved and it would be at least 20 years before 50 TRANSFORMATION 35 (1998) EDWARDS ARTICLE unemployment is brought down to the remotely acceptable level of 10 per cent of the labour force. So to begin to reduce unemployment (and its associated poverty) and lift South Africa out of the quagmire, the target real growth rate needs to be at least 6 per cent a year. This is well above the average that South Africa achieved through the period from 1974to 1994-as shown in Table 2. The table shows the relatively abysmal growth record of the South African economy over these 20 years and partly explains why there is the legacy of an enormous unemployment problem in post-apartheid South Africa. Table 2 Real GDP growth In South Africa compared with groups of countries 1974-94 Real GDP growth (% per annum) 1974-90 1991-94 South Africa 2.0 0.2 All Developing Countries 3.4 5.2 of which... ... Sub-Saharan Africa 2.1 1.5 ... South Asia 5.0 7.8 ...East Asia 7.3 ... Latin America and the Caribbean 2.5 3.4 Industrial Countries 3.3 1.9 Sources; World Bank, 1004,2,3 and IM= 1000, 147,140, 702 and 703. However, in this paper, I want to look to the future by asking; what is required to achieve a real growth rate of 6 per cent a year? In the next section, I look at rapidly-growing East Asia to see what, if any, lessons can be derived from that experience. Then in section 3,1 look at the level of investment required in South Africa to achieve a 6 per cent a year rate of growth and in sections 4, 5 and 6 at some of the proposals for achieving such a level of investment and a higher rate of growth. Finally in section 7,1 summarise the paper and its main proposals. How to get faster growth - lessons from East Asia? Table 2 shows that the growth record of East Asia has been quite spectacular compared to other regions of the world. This is true even if we extend the period back to the mid-1960s. As the World Bank pointed out in a prominent study published in 1993 (World Bank, 1993), over the 1965-90 period, the per capita GDP growth of East Asia was more than double that of any other region of the world. Furthermore, since that growth had been accompanied by a relatively TRANSFORMATION 35 (1998) 51 EDWARDS ARTICLE equal income distribution, the World Bank referred to this experience as a miracle and for once the term 'miracle' seemed to be justified for there had been rapid development (namely growth accompanied by a reasonably equal distribution of income) without any particularly good 'luck' (in the sense of the region experiencing an unusually favourable trend in its external terms of trade). The East Asian countries studied by the World Bank in its East Asian Miracle were the Eight High-Performing Asian Economies (8HPAEs) of Hong Kong, Indonesia, Japan, Malaysia, Singapore, South Korea, Taiwan and Thailand. It is beyond the scope of this paper to look in detail at the virulent debates around the East Asian Miracle (a flavour of these debates can be obtained from Amsden 1989, Fishlow et al 1992, Jomo et al 1997, Wade 1990, World Bank 1992 and World Development 1994). Here I want to limit myself to the lessons for South Africa.4 In two respects the East Asian experience can be said to hold no lessons for South Africa. Firstly, what has been referred to as the initial conditions are very different. Rodrik (1995) has highlighted the importance, for Taiwan and South Korea, of two initial conditions which were particularly conducive to economic growth. These were the level of education and the degree of equality. Indeed Rodrik argues that a very large part of the economic growth in South Korea and Taiwan can be' econometrically explained' by the initial levels of schooling and equality. Rodrik argues that the high level of income equality that predated the growth spurt in the period looked at by the World Bank was brought about by land reform. Why were these initial conditions important? The importance of education is self-evident but there are three reasons given for the importance of the level of equality. The relative equality reinforced (see Perkins in World Development 1994: 657) the emphasis on non-elitist (primary and secondary) education; it provided a larger domestic market which, with economies of scale, provided a launching pad for the simultaneous development of production for exports; and the land reform meant that there was no strong landlord pressure group to block the governments' drives for industrialisation. Clearly in terms of these initial conditions, there is an obvious difference between South Africa and these East Asian countries. In South Africa, education is poor for the majority and inequality is very high. The Gini coefficient5 in South Africa for 1995 is given as 0.61 (see Labour 1996:3) compared to the very much lower (more equal) coefficients for South Korea and Taiwan in 1960 of between 0.30 and 0.35 (see Rodrik 1995:76). A second sense in which the experience of East Asia is quite different from that of South Africa is in the geo-political context. The role of Japanese 52 TRANSFORMATION 35 (1998) EDWARDS ARTICLE investment and trade in the development of Malaysia has been emphasised in Jomo et al (1997). For the development of Japan itself it has been argued that the Korean War was a great stimulus and, for South Korea, the war in Vietnam was a stimulus (see Bagchi 1989: 35). But of course the geo-political, regional framework is important not just because of markets for war supplies but also because of the effect on trade and aid flows. Thus it is worth noting that economic aid to South Korea totalled almost US$6 billion between 1945 and 1978, almost as much as the total aid provided to all African countries over the same period (see Hewitt et al (eds) 1992: 106). Clearly in terms of initial conditions and the geo-political context, there are major differences between South Africa and the 8HPAEs. Are we left with any lessons! The answer is 'yes'. The obvious one is the role of state intervention which was both considerable (particularly in the northeast Asian economies of Japan, South Korea and Taiwan) and effective. Indeed the World Bank's East Asian Miracle (EAM) represented something of a U-turn in 'official' World Bank thinking in the sense that there was a major change in emphasis from previous high-level presentations from the Bank6. The East Asian Miracle recognised, to a greater extent than in previous reports, that state intervention had played a significant and positive role in the development of the Newly-Industrialising Countries (NICs). State intervention in East Asia was effective in both directing the allocation of resources but also in determining the accumulation of capital and labour. It is the latter - the accumulation factor - that has been advertised most loudly by Krugman in Foreign Affairs (Krugman, 1994) and which has been the subject of lively debate in the Economist (see December 9, 1995 and March 1, 1997). Krugman has not been alone in expressing this view (see also Akyuz and Gore in World Development 1996), but Krugman has put the argument more forcefully and bluntly than others. The essence of the argument is that the development of the East Asian countries owes more to 'perspiration than inspiration' (Krugman 1994: 70). Thus in Singapore between 1966 and 1990 the employed proportion of the population surged from 27 per cent to 51 per cent and over a similar period the ratio of investment to GDP increased in Malaysia from about 20 per cent to more than 40 per cent (see Economist, December 9,1995). In Taiwan and South Korea investment was raised from 'around 10 per cent of GDP in the late-1950s in both countries to 30 per cent in 1980' (Rodrik 1995: 59). In putting his argument, Krugman drew on a study of total factor productivity growth (TFP) by Alwyn Young which showed that over the 1970 to 1985 period, TFP growth in the 8HPAEs was low. There has been, and continues to be considerable controversy about measuring TFP growth7 but the broad TRANS FORMATI ON 35 (1998) 53 EDWARDS ARTICLE conclusions of Young's study have been backed up by Collins and Bosworth in a recent issue of Brookings Papers (1996: 139). Thus in the views of Krugman, Rodrik, Akyuz and Gore the secret of the rapid rate of growth of the 8HPAEs lies in the high rates of capital investment and absorption of labour. These authors all agree that there has been insufficient emphasis in the EAM debate on the sources of accumulation and on the nexus between state intervention and profits and between profits and investment. There are two specific criticisms of the EAM which are made: • firstly that the role of foreign loans and aid has been under-emphasised; in Korea and Taiwan in the form of aid (see above); in Singapore in the form of loans and FDI; in Hong Kong in the form of a transfer of funds from Overseas Chinese and in Japan through the US special procurement programme of the early-1950s - see Akyuz and Gore 1996: 464; • secondly the role of reinvestment from high profits has been under-emphasised. Thus ' ... available information suggests that what differentiates the East Asian NIEs, including the second tier ones, is not household but corporate savings. An exception is Singapore where household savings - mostly through the Central Providence Fund - are important' (Akyuz and Gore 1996: 465). Retentions and reinvestment were encouraged by the governments and high profit rates provided both the incentive to invest and the source of funds for investment (Akyuz and Gore 1996: 464). It is argued that by a series of interventions, the South Korean and Taiwan governments engineered an increase in the rate of profit for industrialising companies. In South Korea the rate of profit in manufacturing rose from 9 per cent in 1951-53, to 16 per cent in 1954-56, to 28 per cent in 1957-62 and to 35 per cent in 1963-70. In Taiwan the rate of profit also rose rapidly (see Rodrik 1995: 83). In Korea the main form of subsidy was the extension of credit to large business groups at negative real interest rates while in Taiwan the main form was tax concessions (Rodrik 1995: 85,87). The governments redirected profits into re-investment while making sure through directed credit, tax incentives and coordination of investments that high profits would be made thus ensuring further investment (Akyuz and Gore 1996:467). As pointed out, the high rate of investment was a major factor in generating the high growth rate with increases in total factor productivity being fairly unimportant, implying that much of the technology was embodied in capital inputs (see Rodrik 1995: 59). Thus high profits were reinvested and generated further growth and employment. The initial income equality has tended to fall, but quite slowly (see World Bank 1993: 4). 54 TRANSFORMATION 35 (1998) EDWARDS ARTICLE Thus the level of investment is important in explaining the high rate of economic growth in East Asia and the next section looks at the record of South Africa in this respect relative to groups of other countries. What level of investment is required to achieve a rate of growth of 6 per cent a year in South Africa? The real economic growth of an economy can be analysed quite simply using a model that was formulated in the context of the post-war reconstruction of Western Europe and which is attributed to two economists, Roy Harrod and Evesey Domar. The rate of economic growth can be broken down into the level of investment and its efficiency. The level of investment is here presented as investment (I) as a percentage of output (Gross Domestic Product) while the efficiency is represented by the extra growth in output (dGDP) that is produced by the investment. Dividing the percentage 'level' of investment (I/GDP per cent) by the 'efficiency' ratio (I/dGDP) gives us the percentage rate of economic growth (dGDP/GDP per cent). In Table 3 below, this break-down of the economic growth rate is shown for South Africa and groups of other countries. As the table shows, the record of South Africa in terms of investment 'level' is not too bad relative to groups of other countries but the record on 'investment efficiency' has been extraordinarily weak, with South Africa's ratio being very high at 14.6. Table 3 Investment level, investment efficiency and the rate of economic growth; South Africa compared with groups of other countries; 1986-95. Industrial LDCs countries Africa Asia Western South hemisphere Africa Period (1986-95) (1986-93) (1986-94) (1986-94) (1986-95) Investment level (I/GDP%) 20.7 19.3 27.1 20.9 18.3 Investment efficiency (I/dGDP) 8.2 8.6 3.7 8.3 14.6 Economic growth (dGDP/GDP%) 2.5 2.2 7.3 2.5 1.3 Sources; IMF, 1886, 146,147,164,155. TRANSFORMATION 35 (1998) 55 EDWARDS ARTICLE Given this record, the prospects of South Africa reaching a GDP growth rate of 6 per cent per annum look poor. However, if we look at periods of relatively fast economic growth, South Africa has achieved better investment efficiency (I/dGDP) ratios. The average for the 1966-75 period (when growth averaged 4.4 per cent a year) was 6.2 and the average efficiency ratio for 1994 and 1995 (when growth averaged 3 per cent a year) was also 6.2. However it is clear that even if we assume an optimistic efficiency (I/dGDP) ratio of 5, South Africa will have to invest 30 per cent of its GDP to achieve a real growth rate of 6 per cent per annum. This investment volume percentage of 30 compares with an average of under 20 in recent years (see table 3). The vast majority of this investment will have to come from domestic sources. It is true that, as we have seen, foreign financing (in the form of loans, aid or direct foreign investment) was important in the development of the 8HPAEs. It is also true that foreign financing has the virtue of plugging two gaps - the savings-investment gap and the foreign exchange gap,8 but in the 1990s, with aid budgets being cut in real terms, it is unlikely that foreign financing will be equivalent to much more than 5 per cent of South Africa's GDP. A sustained net inflow of more than 5 per cent is unlikely, even though South Africa's relatively low external debt (see Adelzadeh 1996: 78) gives considerable leeway for foreign borrowing. Thus I agree with Vishnu Padayachee when he says in referring to South Africa that 'there exists at the level of rhetoric at least, an exaggerated and uncritical view of the importance of foreign capital to South Africa's economic development' (Padayachee 1995: 175). It is worth noting in this context that gross inflows of Foreign Direct Investment (FDI) into Malaysia and Singapore (two of the biggest recipients of FDI among the LDCs) between 1967 and 1986 averaged 4.6 per cent and 6.8 per cent of GDP respectively. Thus foreign financing of more than 5 per cent of GDP is likely to be excessively optimistic. GEAR 1996 assumes that foreign financing will be equivalent to about 4 per cent of GDP and the ILO argues that foreign borrowing could be 5 per cent of GDP (Nattrass 1996: 38). Thus investment, equivalent to about 25 per cent of GDP, will have to be found from domestic sources if a rate of growth of 6 per cent a year is to be sustained. Clearly the task is formidable and the stakes are high. It is necessary to ask: what are the best policies for achieving this formidable target of a high investment level from domestic sources? 56 TRANSFORMATION 35 (1998) EDWARDS ARTICLE How to get higher investment and faster growth in South Africa - from de Kock through NEM and MERG to GEAR In the past decade or so, there have been various reports on the South African economy. For the purposes of this paper it is unnecessary to go through them in detail but a brief review will help the understanding of the later sections of this paper. Before the elections and under the previous apartheid government, two reports were produced which have been influential and important reference points on the South African economy. One was the de Kock Commission of Inquiry into the Monetary System in South Africa which reported in 1985 (see de Kock 1985) and which, as Pillay has argued '... still remains the essential intellectual framework for monetary policy formation by the present unreconstructed Reserve Bank and Ministry of Finance' (Pillay in Michie and Padayachee (eds) 1997: 107). Pillay argues that what is striking is the unreality of the Commission's analysis but what is significant for this paper is the Commission's belief that 'financial markets function best in the national interest if they are reasonably free, competitive, active and broad and if they produce realistic market-related interest rates' (quoted in Michie and Padayachee 1997: 106). In the year before the 1994 elections, another report was produced (which in terms of financial intermediation and markets) was something of an echo of the de Kock Commission's. This was the Normative Economic Model (NEM) of the Central Economic Advisory Service (CEAS 1993). Pillay argues that this was hardly a model in the correct understanding of that term - that is, with a clearly specified, logically consistent set of defined relationships between economic variables. 'Rather the NEM was little more than a series of assertions grounded in a largely subjective view of how the economy functions' (Pillay in Michie and Padayachee 1997: 120). However, again for my purposes, it is worth noting that as far as financial markets are concerned, the NEM argued for deregulation and favoured high real interest rates. In the meantime, in the early-1990s, the democratic movement (led by the African National Congress and COSATU) had begun to formulate its own analysis of the South African economy and in 1993 produced a report entitled Making Democracy Work: A Framework for macroeconomic policy in South Africa (see MERG 1993). This set out an economic programme covering the period from 1994 to the year 2004. The MERG report stood in stark contrast to the NEM. While the latter had set out its preference for growth with redistribution, the MERG report set out the case for growth through redistribution. That is it set out the need for state intervention to achieve TRANSFORMATION 35 (1998) 57 EDWARDS ARTICLE redistribution and through it a faster rate of growth. It set out to remove both the gross inequity and inefficiency of the apartheid system, inefficiency in the form of the costs of commuting vast distances from the townships to places of work and excess capacity in energy provision, education, and health inherent in a spatially fragmented system. The theoretical emphasis of the MERG approach was on the notion that state expenditure would encourage, rather than 'crowd out' private investment, the latter being the argument of neo-liberal economists. The MERG emphasis was on low real interest rates to encourage private investment and to generate the growth in real income from which savings will be generated. The 'structuralism' of the MERG report was in stark contrast with the neo-liberal approach of the NEM. These reports (the de Kock Commission and the NEM on the one hand and the MERG on the other) have been followed by a series of reports published in 1996 and which attempt to reinforce the case either for the neo-liberal free market approach or a structuralist, interventionist, approach to economic policy. In February 1996, Growth for All was published by the South African Foundation (SAF1996). This was followed in April by the Labour Movement's Social Equity and Job Creation (Labour 1996), in the middle of 1996 by the ILO country review (Standing et al 1996) and by a report entitled Growth, Employment and Redistribution - a Macroeconomic Strategy which was published by the government (GEAR 1996). These four reports from SAF to GEAR have been analysed in some detail by Nattrass (1996). Broadly speaking, both SAF and GEAR follow the NEM inasmuch as they emphasise and advocate free markets and the need for 'flexibility' in labour and financial markets, while the emphasis of the ILO and Labour reports is on state intervention, a more direct attack on poverty and the effecting of a faster rate of economic growth through the creation of a higher, state-promoted, level of effective demand. For a more detailed analysis of these reports, refer to Nattrass (1996) and Adelzadeh in Michie and Padayachee (1997). Here the focus is on the proposals for interest rates and the steps thought necessary to encourage a higher level of investment, as well as efficient investment. Thus, in the next section, the neo-liberal case for, and the structuralist case against, high interest rates are examined. How to increase growth - the case for and against high interest rates Neo-liberal economists argue for high interest rates, that is for interest rates that are positive in real terms. The argument is that an economy will suffer from 58 TRANSFORMATION 35 (1998) EDWARDS ARTICLE 'financial repression' if interest rates are maintained at a low level. The theory of 'financial repression' is generally associated with McKinnon (1973) and Shaw (1973) who argue that low interest rates generate a low level of savings together with a low efficiency of investment. In this view, the supply of savings is a function of the interest rate - the higher the interest rate, the higher the quantity of savings. Thus with a low real interest rate there will be a low level of savings and funds will have to be rationed between investors. The rationing means, in turn, that high-quality investors are likely to be squeezed out or 'repressed' and thus, with a low rate of interest, we have 'financial repression' (see McKinnon 1973:103-111). It is not surprising that this neo-liberal approach has been referred to as the 'prior-savings' approach (see Thirlwall 1994) since the direction of causation is from savings to investment to national output. The structuralist or Keynesian approach to savings and investment is in stark contrast to the prior-savings approach The Keynesian argument is that aggregate savings are largely a function of aggregate income and investment is an inverse function of the interest rate. Thus with low interest rates, we get higher investment which, through a multiplier effect, generates higher aggregate income. New factories generate income which in turn generates higher savings. Thus for the Keynesians and neo-Keynesians what is needed is an investment-led approach to the promotion of rapid economic growth through a low and stable cost of capital and a high expected growth rate in effective demand. The reference to stability means that the longer-term the finance, the more likely it is that investment will take place. In this view there is a strong case (see chapter 24 of The General Theory by Keynes (1976)) for the state directing funds into investment so that the investor is assured of borrowing at a low and stable rate of interest. Which approach is more realistic - the neo-liberal or the Keynesian? The evidence is undoubtedly on the Keynesian side, with the neo-liberal McKinnon-Shaw theory being subjected to severe criticism. The theory that higher interest rates generate higher aggregate savings has been particularly subjected to criticism. After two decades of research into the subject, this is the conclusion of Fry who acknowledges that 'those investigators looking for financial repression find it, while those expecting no influence find none. What is agreed however is that if the effect exists at all, it is relatively small' (Fry 1988: 132). A similar conclusion, namely that aggregate savings are not interest-elastic, was arrived at by the World Bank's EAM. It found that in the 8HPAEs, savings were not responsive to interest rates9 (World Bank 1993, chapter 5) and that governments used a variety of more interventionist measures to increase savings.10 Similarly a recent survey article by Schmidt-Hebbel et al TRANSFORMATION 35 (1998) 59 EDWARDS ARTICLE concluded that '...the evidence generally shows that interest rates ... have little or no effect on saving' and that' ... saving typically follows, rather than precedes, growth' (1996: 91, 101). The theory that rationed or directed credit is harmful has also been questioned. In its EAM study, the World Bank conceded that direct credit at low interest rates played a large role in the development of the region especially in Japan, Taiwan and South Korea (World Bank 1993: 235-9 and 358). It admitted that there was 'financial repression' in most of the 8HPAEs (World Bank 1993: 17) inasmuch as deposit and lending rates were below market clearing levels and that there was a resulting transfer of income from households to firms. However, it further (somewhat begrudgingly) admitted that 'while we cannot establish conclusively that mild repression of interest rates at positive real levels enhanced growth in northeast Asia, it apparently did not inhibit it' (World Bank 1993: 17)". Thus the empirical evidence does not in general support the neo-liberal theory that low real interest rates harm economic growth. Low interest rates are harmful neither in terms of reducing the availability of savings nor in terms of rationing credit. The research on South Africa provides similar conclusions on the savings effect. In an excellent study, Nick Wilkins found that 'The results of the econometric tests for financial repression ... do not provide any support for the hypothesis of positive interest-sensitivity of aggregate or financial savings'. Wilkins went on to conclude that; 'Contrary to the neo-classical and repressionist models, ex ante investment and saving are not determined simultaneously by the real interest rate in the financial market...' and that 'the Keynesian model of the financial system has greater validity than the neo-classical and repressionist models in the respect that investment (partly) determines aggregate income which in turn determines savings' (Wilkins 1993: 75). Thus the implication is that, even if real interest rates had been lower, financial repression would not have occurred. To quote Wilkins again; 'Financial repression is therefore unlikely in the South African financial system because in effect the real rate of interest responds to changes in the levels of economic activity and not vice versa' (1993: 75).12 The way forward for South Africa Thus the weight of empirical evidence suggests that higher interest rates do not generate higher aggregate savings and that directed credit has been (at least in the case of Northeast Asian development) beneficial. In spite of this, a number of reports on South Africa's economy continue to promote financial liberalisation and higher, real, interest rates. These TRANSFORMATION 35 (1998) EDWARDS ARTICLE recommendations have come not only from within the apartheid government - through the de Kock and NEM reports - and from big business (see SAF 1996), but also from the post-apartheid government in the shape of the 1996 GEAR report. GEAR's targets of raising (by the year 2000) the I/GDP ratio to 26 per cent and the savings/GDP ratio to 22 per cent (the 4 per cent difference being covered by capital inflows on the Balance of Payments) are too low and too slow, but the major bone of contention is on the mechanism for raising investment and savings rates. GEAR states that '... the main objective of monetary policy will continue to be the maintenance of financial stability and the reduction of the inflation rate' and that; 'monetary policy will also aim to maintain real interest rates at positive levels to encourage savings and investment' (GEAR 1996: 10). The GEAR report clearly goes against Wilkins' evidence that higher real interest rates will not encourage savings and yet it is GEAR's neo-liberal call for higher real interest rates that has been heeded in recent years in South Africa. Table 4 compares the average (for the three years 1993-95) real Central Bank discount rate for South Africa with those of other countries The table shows that the average real rate was higher - very much higher - in South Africa than in all the other countries. Table 4 A comparison of interest rates In South Africa and other countries; 1993-95 Nominal Central Inflation (rise in Real Central Bank discount rate consumer prices) Bank discount rate (%pa) (%pa) (%pa) South Africa 13.3 9.1 4.2 Germany 4.4 3.0 1.4 Japan 1.3 0.6 0.7 Korea, South 5.0 5.2 -0.2 Malaysia 5.4 4.2 1.2 USA 4.3 2.8 1.5 Source; IMF 1008, 05, 111. Note; all figures are averages for 1903 to 1995 inclusive. However South Africa's interest rates have not always been so high. Table 5 shows that the Central Bank discount rate in recent years (1993-95) has been very much higher than the averages for previous periods and it is clear that the recent high (real) rates have been the exception rather than the rule. TRANSFORMATION 35 (1998) 61 EDWARDS ARTICLE Table 5 Interest rates in South Africa; 1966-95 Nominal discount Inflation (% rise in Real discount rate consumer prices) rate (%pa) (%pa) (%pa) Average* - 1966-75 5.9 6.3 -0.5 1976-85 11.6 13.0 -1.4 1906-95 14.1 13.3 0.8 of which ... 1993-95 13.3 9.1 4.2 Source; IMF 1006, 700-703. These high interest rates are contrary to what South Africa needs. Instead the government should be following a regime of lower interest rates together with an expansionary programme of public sector investment and a programme of directed credit to promote private sector investment in South Africa. Low interest rates are required not so much to encourage private sector investment, since, as the research by Wilkins suggests, 'Planned investment does not... bear a simple relationship to the real rate of interest as in the Keynesian model; during phases of high business confidence, the demand for finance to fund investment and working capital is relatively insensitive to increases in interest rates until the latter reach very high levels' (Wilkins 1993: 75). Instead lower interest rates would be beneficial inasmuch as they would reduce the interest burden of the government's debt,13 would allow (for a given public sector deficit) greater public sector investment in infrastructure and basic needs and would therefore boost the level of effective demand in the economy. The GEAR report argues that (higher) fiscal deficits would lead to higher inflation and higher interest rates and its target is for a deficit of 3 per cent of GDP by 2000. In setting such a restrictive target, it is clear that GEAR is accepting the 'crowding-out' theory of neo-classical economics and yet 'there is no consensus among economists about the importance of 'crowding out' and no empirical evidence to suggest that it has ever occurred in South Africa' (Adelzadeh 1996: 74). Indeed, as Nattrass pointed out, one of the models which GEAR supposedly consulted - namely the Gibson and van Seventer Development Bank model - suggests that a contractionary fiscal policy would actually cause a/a// in income output and employment (Nattrass 1996: 37). Similarly, 'The World Bank's three policy scenarios for South Africa concluded that a rise in the fiscal deficit to 12 per cent of GDP is sustainable because the higher growth pattern will gradually generate more public savings such that by the year 2000 the country will experience fiscal surplus' (World Bank 1993a: 5) and 'World Bank research 62 TRANSFORMATION 35 (1998) EDWARDS ARTICLE suggest that the private sector responds positively to changes in public-sector-generated demand' (Fallon and de Silva quoted in Michie and Padayachee 1997: 181)14. This is not to say that the fiscal deficit should be widened by reducing taxes (there is some evidence that South Africa is 'undertaxed' given its stage of development - see Adelzadeh 1996: 77). If the budget deficit is to be widened it should be by expanding state expenditure particularly on housing. In the context of budget deficits, it is worth noting that between 1960 and 1992, the budget deficit of rapidly-growing Malaysia averaged 6.9 per cent of GDP (see Collins and Bosworth 1996: 175). Lower interest rates would not only be helpful in reducing the cost of servicing public sector debt but also in discouraging the inflow of short-term capital into the country. The latter is harmful because it exposes the economy to the possibility of a sudden outflow if confidence is reduced. This was a major problem in Mexico in 1984 and in South Africa in 1985 (see Kahn 1996:4). Thus with high interest rates, there are likely to be higher short-term capital inflows but precisely because they are short term, they are likely to generate greater exchange rate volatility as well as the maintenance of a higher average real effective (or trade-weighted) exchange rate. To reduce volatility, exchange controls (which in recent years have been relaxed15) need to be reinstated. Thus what must accompany lower real interest rates is a reinstatement and tightening, rather than the final abolition of exchange controls on residents. The controls should be strengthened through the surveillance of trade transactions to reduce transfer pricing, double-invoicing and other ways of exporting capital. It seems that illegal exports of capital between 1970 and 1988 were running at anything between 1 per cent and 6 per cent of GDP (see Kahn 1991). A major objection to exchange controls is that they cannot be rigidly enforced. This is undoubtedly so, but even if the exchange controls are not watertight, their existence imposes a tax or transaction cost on the export of capital and helps to make the exchange rate (and domestic interest rate) more 'sticky' than they would otherwise be. Thus the strident cries of papers like the Economist for the abolition of exchange controls should have been ignored.16 With the further relaxation or complete abolition of exchange controls, there is likely to be an increased volatility in the exchange rate. With a regime of lower interest rates, there is likely to be smaller inflow of speculative, short-term capital. Thus with lower interest rates, there is likely to be less volatility in the exchange rate as well as a lower real, effective (trade-weighted) exchange rate. It may be objected that a devaluation is not necessary since the rand is not overvalued with South Africa running a surplus on the current account of the balance of payments equivalent to about 2.6 per TRANSFORMATION 35 (1998) 63 EDWARDS ARTICLE cent of GDP between 1985 and 1993. However a major factor contributing to the current account surplus over the 1985-93 period was the slow rate of economic growth (an average of only 0.6 per cent a year). It is noteworthy that the higher economic growth rate since 1994 has been accompanied by a deficit on the current account of the balance of payments. With this deficit, the real effective exchange rate index declined by about 15 per cent between 1994 and the end of 1996. However, in spite of this devaluation, the rand can be said to remain overvalued, given the persistent labour surplus in the economy. Thus a further weakening in the real trade-weighted exchange rate would probably be beneficial to the economy in promoting the diversification of manufacturing production - for exports and import substitutes - and for promoting tourism. Certainly maintaining a high rate of interest to attract short-term capital inflows is a high price to pay to reduce the pressures for a further devaluation which may in any case be desirable. Clearly an objection to devaluation is that it may be inflationary, since imported prices have a weight of about 20 per cent in the producer price index (Kahn 1996: 6). But devaluation is only one factor in cost-push inflation and the extent of devaluation and inflation depend to a considerable extent on the composition of growth. Furthermore it should be noted that lower interest rates may help to reduce cost-push inflation. Nevertheless if there is a trade-off between lower interest rates on the one hand and the inflation associated with an expansionary economy on the other, then the price of inflation is worth paying to reduce unemployment. As Keynes put it in his Treatise on Money, 'It is worse in an impoverished world to provoke unemployment than to disappoint the rentier' (quoted in Thirlwall 1976: 9). It is likely that to bring about a regime of lower interest rates, institutional changes are needed, although this is far from clear. Vella Pillay has argued that 'for long, the South African Reserve Bank has been the creature of the government in power' (in Michie and Padayachee 1997: 107) and since a majority (including the Governor and three Deputy Governors) of the Board of the Reserve Bank are appointed by the State-President, it would seem that the Government has control over the Board's membership. However, Pillay goes on to imply that the Reserve Bank's policies are heavily constrained by a series of agreements and understandings with the Treasury over monetary policy (Pillay in Michie and Padayachee 1997: 107) and by its objectives as set out in the Reserve Bank Act No. 90 of 1989 and under the 1996 Constitution. The 1989 Act specifies that, 'In the exercise of its powers and the performance of its duties the Bank shall pursue as its primary objectives monetary stability and balanced economic growth in the Republic and in order to achieve these objectives, the 64 TRANSFORMATION 35 (1998) EDWARDS ARTICLE Bank shall influence the total monetary demand in the economy ...'(South African Reserve Bank Act 1989, paragraph 4) and the 1996 Constitution specifies that: 'the primary object of the South African Reserve Bank is to protect the value of the currency in the interest of balanced and sustainable economic growth in the Republic' (The Constitution of the Republic of South Africa, Act 108, 1996: paragraph 224,1). Thus if there is a trade-off between the stability of the currency (internally or externally) and fuller employment, it would seem that the Reserve Bank's statutory objectives would require it to put stability first. This is in contrast to the situation in the USA where the Federal Reserve Bank is independent but is required under the Full Employment and Balanced Growth Act of 1978 to aim at what have been called the dual targets of full employment and price stability. It may well be that the statutory structure, membership and objectives of the Reserve Bank all need changing to herald a change in priorities. Certainly there have been a number of calls for such a change. Thus the MERG report argued for making the Reserve Bank accountable and subordinate to the Ministry of Finance and subject to parliamentary scrutiny and it rejected the view that inflation is controllable only through the application of monetary policy. With the objectives of the Reserve Bank needing to be changed, it is probable that the Governor and Deputy Governors of the Bank need also to be changed. Certainly there is no case for a Central Bank in South Africa operating independently as it does at present under the Constitution.17 The evidence is that while independent Central Banks promote price stability, they have no measurable impact on real economic performance (Alesina and Summers 1993). Indeed 'recent studies on the costs of disinflation show that higher Central Bank independence can lead to costlier disinflations' (Jordan 1997). Certainly the stakes in South Africa (in terms of very high unemployment and its associated poverty) are too high to leave monetary policy to an independent Central Bank. However a change in the membership, structure and objectives of the Central Bank is not the only institutional change required in South Africa's financial structure. Other institutional reforms, required to promote savings, were set out in the 1993 MERG report. MERG argued that the state needs to go beyond low interest rates to encourage productive investment and to discourage speculation and that changes in the tax and regulatory structure are also required. Thus MERG advocated the adoption of a capital gains tax and prescribed asset ratios for pension funds and assurance companies. In addition MERG advocated steps which would encourage personal savings outside the core banking system. Thus MERG recommended the establishment of a 'People's Bank' to tap the savings of Africans particularly of those in the rural areas. In addition MERG TRANSFORMATION 35 (1998) 65 r EDWARDS ARTICLE recommended measures to encourage savings through institutions concerned with housing finance and through state-managed pensions. The main point here is that these MERG proposals were designed to change the quantity of savings and the quality of investment (as well as to achieve social goals) not through de-regulation but through institutional reform. However, the reforms required in the area of financing need to go beyond these MERG proposals which are thought necessary to increase savings through better access and through tax changes. State intervention is also needed to reallocate or direct savings into long-term approved investments. Superficially (as Harris 1994: 9 notes), the institutional structure of South Africa's banking system looks a bit like the British system. The principal banks are commercial banks with extensive national branch networks and asset holdings of marketable paper and short-term loans. This banking system has a high degree of concentration which has been brought about by a process of liberalisation over the past decade or more (both before and after the de Kock Commission's report in 1985). The financial liberalisation had been justified in terms of increasing competition, but Harris argues that the main effect of South Africa's liberalisation has been a wholesale restructuring of the banking sector in which banking groups and building societies have merged to form new monopolistic financial groups. Thus, by 1991, four banking groups controlled 95 per cent of all deposit-taking institutions' assets (Harris 1994: 5). However, Harris argues, the similarity with the UK system turns out to be somewhat superficial once the links between the commercial banks and other sectors of the economy are considered. For the banks in South Africa are not only concentrated but they are also highly integrated with mining, industrial and commercial activities. As Fine and Rustomjee have emphasised in a recently published book, the structure of private corporate capital in South Africa is represented by six, broad-based, organically linked 'axes' of capital each with varying interests in mining, manufacturing and the financial sector. The six are as follows; • the SA Mutual axis consisting of SA Mutual, Nedbank, UAL/Finance Bank, Perm Building Society, Barlow Rand, Rand Mines; " the Sanlam axis consisting of Sanlam, Trust/Santam bank, Senbank, (no building society), Gencor, Malbak; • the Anglo-American axis consisting of Southern Life, First National Bank, First Corporate Bank, (no building society), AMIC, Amgold/Amcoal/De Beers/JCI; • the Liberty/Standard axis consisting of Liberty Life, Standard Bank, Standard Merchant Bank (no building society, no direct mining/industrial arms); 66 TRANSFORMATION 35 (1998) EDWARDS ARTICLE • the Rembrandt/Volkskas axis consisting of Lifegro/Federated Life, Volkskas/Boland, Rand Merchant Bank, United Building Society, Remgro, GFSA; and finally, • Anglovaal operating relatively independently of financing with some Sanlam and SA Mutual influence in their mining and industrial holding companies (Fine and Rustomjee 1996: 108). This integration across sectors is quite different from that in the UK. Indeed what this highly concentrated and integrated structure would seem to signify is that there already exists in South Africa a basis for the long-term commitment of finance to industrial activity - precisely the sort of commitment that has been evident in Germany and in Japan and which reduces the problems of asymmetric information that is arguably found in the non-integrated systems of the UK and USA (see Harris 1994: 8 and 9). So far so good, but this immense concentration in what Fine and Rustomjee call the Minerals-Energy Complex (MEC) of South Africa also produces a highly segmented structure which cries out for state direction along Japanese lines. For although, within the MEC, there are many internal markets, outside these, there is likely to be a perception on the part of the banks of high systemic risk especially in the context of South Africa's volatile social and economic transformation. Thus Pillay argues that the rates of interest charged on loans outside the MEC network rise 'by leaps and bounds, with the highest rate applied to the small and medium-sized business community and in particular the black population at large' (Pillay in Michie and Padayachee 1997: 107). In these circumstances confidence is likely to surge and fall quite dramatically. Thus when things are going well the MEC core may do the job of investing well and even of diversifying a little. But if confidence ebbs, the investment job will not be done and diversification is likely to dry up. What does this institutional structure imply for financing policy? There is a range of choices. At the very minimum it means that the state must supervise closely the activities of the component parts of the MEC so as to maintain systemic confidence. Further deregulation is unlikely to achieve this. In addition to this supervision, there is a need for directed credit to help to ensure the implementation of the integrated and diversified industrial policy advocated by Fine in chapter 7 of Michie and Padayachee (1997) and by Fine and Rustomjee (1996) (see chapters 8 and 9). This direction of credit could conceivably be done through the Industrial Development Corporation (IDC) since the IDC played a major role in the integration of large-scale Afrikaner capital in the MEC economy (see Fine and TRANSFORMATION 35 (1998) 67 EDWARDS ARTICLE Rustomjee 1996). On the other hand precisely because the IDC is so identified with the policies of the apartheid regime, it may require a separate institution (such as a new Investment Bank) to direct the industrial finance. Thus, as well as reforming the financial system to ensure lower interest rates, it is also desirable for the state to take a more active role in directing funds out of (but integrated with) the MEC core of the economy. Thus lower real interest rates are only one tool, although they may go some way to reducing the economic power of the MEC since, as Pillay argues, high real interest rates favour the MEC core in so far as high rates dampen competition from small and medium enterprises outside the core (see Pillay in Michie and Padayachee 1997: 119). Additional steps are required, especially as the vast majority of savings in the economy accrue to the private corporate sector. The expansion of finance channelled through a state-directed Development Bank and directed at approved industrial investments is also needed. Summary and Conclusions This paper is about one of the legacies of the long period of apartheid in South Africa, namely the appallingly high level of unemployment and underemployment and the inequality associated with it. At present, not only is unemployment a third of the labour force or more but at the 3 per cent a year growth rate of the past three years or so, there is no hope of it being significantly reduced. And yet in South Africa there is some considerable complacency in official circles about the growth rate - no doubt because in historical terms it represents a reversal of the negative growth era. However, as the first section of this paper sets out, this 3 per cent a year growth rate needs to be at least doubled - to 6 per cent a year - if any significant inroad is to be made in the backlog of unemployment. As the second and third sections point out, one lesson which comes from the spectacular growth in East Asia over the past 30 years is a straightforward one - that investment in labour and capital have to be on a massive scale to bring about rapid economic growth. To achieve a growth rate in the real GDP of South Africa of 6 per cent a year requires that at least 30 per cent of South Africa's GDP be invested, compared to less than 20 per cent in recent years. But how can such a level of accumulation be brought about? In the past year or so, there have been a number of reports on the South African economy which, as section 4 shows, can be divided simply into those that favour a reliance on a neo-liberal, free market approach and those that rely on a structuralist, state-interventionist approach. As far as the financing of savings and investment in the South African economy is concerned - which is this paper's topic - the 68 TRANSFORMATION 35 (1998) EDWARDS ARTICLE emphasis of the neo-liberal approach is on high real interest rates and on market forces not regulated by the state. But the evidence from East Asia, South Africa and elsewhere (as discussed in section 5) shows that higher real interest rates do not encourage savings and instead, in discouraging investment by both the public and private sectors, are likely to lead to a stagnation of aggregate incomes and savings. Instead the evidence is that state intervention is invariably necessary to bring about the increase in investment which is likely to lead to higher incomes and higher savings to match the investment. The lesson, in brief, is that 'financial repression' is healthy and that the South African government should be following a regime of low interest rates. Instead', over the past three years, as discussed in section 6, South Africa has been following a regime of high rates - high, that is, relative to other countries and relative to South Africa's own history. This paper argues that lower interest rates would be beneficial not so much in directly encouraging private sector investment, but rather in reducing the debt burden for the public sector and in not encouraging speculative investment. The reduction in the debt service burden would allow public sector investment to be increased. The reduction in short-term capital inflows would reduce the volatility of the exchange rate as well as generating the further devaluation of the rand. The latter would be healthy insofar as it stimulates the production of exports and import substitutes. The downside is that it might lead to a faster rate of inflation. The issue is then a simple one - if inflation is a price which has to be paid to reduce the level of unemployment, then it is a price worth paying. To bring about a regime of lower interest rates in South Africa requires institutional changes. The objectives of the Reserve Bank as set out in the Reserve Bank Act of 1989 and in the 1996 Constitution need changing and the structure of the Reserve Bank also needs changing, with direct accountability to Parliament. These changes probably mean that it would be necessary to replace the present Governor (Dr Chris Stals) and his three Deputy-Governors. However these are not the only institutional changes that need to be made. In addition the institutional reforms set out in the MERG report, designed to improve both the composition of investment and the quantity of savings, should be vigorously pursued. Section 6 also argues that there is a case for more direction by the state of investment in approved industrial investments. At present the financial sector (indeed the whole economy) is dominated by six, broad-based 'axes' of capital which have been created around, what Fine and Rustomjee term, the Minerals-Energy Complex (MEC) core of the economy. These 'axes' grew as large-scale Afrikaner capital was promoted by the apartheid state with the Industrial Development Corporation (IDC) and state enterprises playing TRANSFORMATION 35 (1998) 69 EDWARDS ARTICLE prominent roles. Now the IDC or another state body needs to play the role of integrating the post-apartheid economy by directing credit into industrial investments which link the MEC core and the consumer goods sector by developing capital and intermediate goods industries. Thus as well as reforming the financial sector to ensure lower interest rates, it is also desirable for the state to take a more active role in generating savings from a hitherto untapped rural African market and in directing funds out of (but integrated with) the MEC core of the economy. Unfortunately the policies followed since the 1994 elections have a distinct neo-liberal flavour depicted by both a high-interest-rate regime and an emphasis on reducing the budget deficit of the public sector. Millward has argued that the battle between the neo-classical NEM and the Keynesian MERG was won by NEM (Millward 1996: 2). Adelzadeh has pointed to the transformation of the Keynesian RDP through the National Growth and Development Strategy of 1996 into the neo-classical GEAR (Adelzadeh 1996: 67). More generally, Kahn argues, 'if we compare the current thinking of the present government with ANC economic policy proposals at the time of its unbanning, the transformation in thinking is truly remarkable' (Kahn 1996: 1). In 1993, Laurence Harris argued that South Africa contained a strong basis for a socialist society (a socialist-inclined leadership in the ANC and South African Communist Party, underpinned by a large, militant and organised working class base, which was close to power in an economy far stronger than most of today's new industrial giants - South Korea, Taiwan or China - had 20 or 30 years ago). Paradoxically, he argued, these strong preconditions were accompanied by the dissolution of any effective commitment to a serious socialist strategy (Harris 1993: 101). This is equally true now, in 1997, as it was in 1993. At the present time, there is an urgent need to reform the South African financial system so as to finance expanded government expenditure at the lowest possible cost and to subordinate financial policy to the rapid growth of the real economy spearheaded by an integrated industrial policy. And yet it seems as though it is not so much the real economy that is being transformed as the ideas of the ANC leadership. The adoption of neo-liberal policies in South Africa is especially remarkable in view of; • the limited, even negative impact of such neo-liberal programmes, especially in southern Africa; • the lack of any leverage that the international financial institutions such as the IMF and World Bank need to have over South African policy makers given South Africa's low external debt; 70 TRANSFORMATION 35 (1998) EDWARDS ARTICLE • the lack of any dramatic changes in the economic and political environment to warrant such major shifts in policy orientation; and • the lack of a transparent and fully argued justification for the adoption of an entirely different policy framework (see Adelzadeh 1996: 67). Joe Slovo said in a 1976 article that there was a danger that 'the national struggle is stopped in its tracks and is satisfied with the co-option of a small black elite into the presently forbidden areas of economic and political power' (quoted in Harris 1993:102). This is what seems to be happening. And yet if South Africa is to achieve a growth rate of 6 per cent per annum and above and if a more equal society is to be created, the neo-liberal policies, which are currently in vogue among the leadership of the government, need to be consigned to the dustbin of history. In calling for reforms of the financial sector, this paper is a modest contribution to that aim. References i Adelzadeh, A (1996) 'From the RDP to GEAR: the gradual embracing of Neo-Liberalism in Economic Policy', Transformation 31, 66-95. Akyuz, Y and C Gore (1996) 'The Investment-Profits Nexus in East Asian Industrialisation', , World Development, 24(3), 461-470. > Alesina, A and LH Summers (1993) 'Central Bank Independence and Macroeconomic Performance: some comparative evidence', Journal of Money, Credit and Banking, 25(2), > May, 151-162. i Amsden, A (1989) Asia's Next Giant; South Korea and Late Industrialisation, Oxford University i Press. \ Bagchi, A (1989) 'Industrialization1, in J Eatwell et al (ed.) (1989) The NewPalgrave; A i Dictionary of Economic Development, MacMillan. Bell, T (1995) 'Improving Manufacturing Performance in South Africa: a contrary view', 1 Transformation 28. Bell, T (1996) 'Improving Manufacturing Performance in South Africa: a rejoinder'. Transformation 31. Brixen, P and F Tarp (1996) 'South Africa; Macroeconomic Perspectives for the Medium Term', World Development, 24(6), 989-1001. CEAS (1993) The restructuring of the South African economy - a normative model approach, Pretoria, March. Chenery, H and A Strout (1966) 'Foreign Assistance and Economic Development', American Economic Review, September. Collins, S and B Bosworth (1996) 'Economic Growth in East Asia: accumulation versus assimilation', Brookings Papers on Economic Activity, 2, 135-203, Brookings Institution, Washington. CrefSA (1997) Quarterly Review of the Centre into Econoinics and Finance in Southern Africa (CrefS A), London School of Economics, London, April. TRANSFORMATION 35 (1998) 71 EDWARDS ARTICLE de Kock, JPC (1985) The Commission of Inquiry into the Monetary System and Monetary Policy in South Africa, Government Printer, Pretoria, March. Edwards, C (1995) 'Restructuring the New South Africa', International Review ofApplied Economics, Volume 9, November. Fine, B and Z Rustomjee (1996) The Political Economy of South Africa: from Minerals-Energy Complex to Industrialisation, Hurst and Company, London. Fishlow et al (1994) Miracle or Design?; Lessons from the East Asian Experience, Overseas Development Council, Washington DC. Fry, M ( 1988) Money, Interest and Banking in Economic Development, John Hopkins University Press. GEAR (1996) Growth, Employment and Redistribution: a Macroeconomic strategy. Department of Finance, Pretoria. Harris, L (1988) 'Financial reform and economic growth; a new interpretation of South Korea's experience', in L Harris (ed.) New Perspectives on the Financial System, Croom Helm. Harris, L (1994) 'Financing Post-Apartheid South Africa: Theoretical Aspects', SOAS Working Paper 46, University of London. Hewitt, T et al (eds.) (1992) Industrialisation and Development, OUP/Open University. IMF (1996) International Financial Statistics Yearbook, International Monetary Fund, Washington DC. Jomo, KS et al (1997) Southeast Asia's Misunderstood Miracle: industrial policy and economic development in Thailand, Malaysia and Indonesia, Westview Press. Jordan, TJ (1997) 'Disinflation costs, accelerating inflation gains and central bank independence', Weltwirtschaftliches Archiv, Vol 133 (1). Kahn, B (1991) Capital Flight and Exchange Controls in South Africa, Research Paper No 4, Centre for the Study of the South African Economy and International Finance, LSE, London. Kahn, B (1996) Exchange Control Liberalisation in South Africa, IDAS A, Public Information Series Number 3, Cape Town. Kaplan, D and D Lewis (1996) 'Improving Manufacturing Performance in South Africa; A Response to Trevor Bell' Transformation 30. Keynes, JM (1976) The General Theory of Employment, Interest and Money, MacMillan. Krugman, P (1994) 'The Myth of the Asian Miracle', Foreign Affairs, 73(6). Labour (1996) Social Equity and Job Creation: a key to a stable future, Nedlac Labour Caucus (COSATU, FEDSAL, NACTU). Lowitt, S (1994) 'Repression and Liberalisation; an application of the McKinnon model to South Africa', unpublished M Com. thesis, Witwatersrand University. McKinnon, R (1973) Money and Capital in Economic Development, Brookings. MERG (1993) Making Democracy Work: a framework for macroeconomic policy in South Africa, Macroeconomic Research Group (MERG), Centre for Development Studies, Bellville, Cape. Michie, J and V Padayachee (eds) (1997) The Political Economy of South Africa's Transition, Dryden Press, London. Millward, C (1996) Capital Movements in South Africa; Post April, 1994, paper presented to the second ACDESS Conference in Johannesburg. Nattrass, N (1996) 'Gambling on investment; competing economic strategies in South Africa', Transformation 31, 25-42. 72 TRANSFORMATION 35 (1996) EDWARDS ARTICLE NEDCOR (1996)(4) Guide to the Economy, Johannesburg. NEDLAC (1996) 'Report on Social and Economic Developments in South Africa', May. OECF (1991) 'Issues related to the World Bank's approach to structural adjustment - proposal from a major partner'. Occasional Paper No 1, October. Padayachee, V (1995) 'Foreign capital and economic development in South Africa; recent trends and post-apartheid prospects', World Development, 23(2), 163-177. Rodrik, D (1995) 'Getting interventions right: how South Korea and Taiwan grew rich'. Economic Policy, April. RSA (1996) Presentation notes by the Department of Trade and Industry, Pretoria, September. SAF (1996) Growth for All; an economic strategy for South Africa, Johannesburg; South African Foundation. Schmidt-Hebbel, K et al (1996) 'Saving and investment: paradigms, puzzles, policies', World Bank Research Observer, 11(1), February, pp.87-117. Shaw, E (1973) Financial Deepening in Economic Development, New York, OUP. Standing, G et al (1996) The South African Challenge: restructuring the labour market. An ILO Country Review, Geneva, International Labour Organisation. Thirlwall, AP (1976) Financing Economic Development, Macmillan. Thirlwall, AP (1994) Growth and Development, with special reference to Developing Economies, MacMillan. van Aardt, C (1994) The Future South Africa. Wade, R (1990) Governing the Market: economic theory and the role of government in East Asian industrialisation, Princeton UP. Wilkins, N (1993) Financial Repression: do real interest rates determine savings and investment in South Africa? Research Monograph No 8, Economic Research Unit, University of Natal, Durban. World Bank (1983) World Development Report, Washington DC. World Bank (1992) World Bank Support for Industrialisation in Korea, India and Indonesia, Operations Evaluation Department, Washington DC. World Bank (1993) The East Asian Miracle, Economic Growth and Public Policy, Oxford University Press. World Bank (1993a) Paths to Economic Growth, Washington DC. World Bank (1994) Global Economic Prospects and the Developing Countries, Washington DC. World Bank (1996) Higher Growth and Employment for South Africa: the options,' World Bank workshop paper, 14 February. World Bank (1996a) World Development Report, OUP for the World Bank. World Development (1994) Special issue on The East Asian Miracle (see especially articles by Kwon, Lall, Page, Perkins and Yanagihara). World Bank (1997) World Development Report, Washington DC. TRANSFORMATION 35 (1998) 73 EDWARDS ARTICLE NOTES 1. This is a revised version of a paper presented at a Conference in Durban in April 1997. The inspiration for many of the ideas of this paper come from Laurence Harris (1994), Ben Fine and Vella Pillay, but I do of course take total responsibility for the paper. I am grateful to Vishnu Padayachee and Peter Draper for comments on earlier drafts, to the British Council for financial support towards the cost of my visit in April 1997, to the University of Durban-Westville for inviting me to present the paper at the Conference and to an anonymous referee for comments. 2. Both the recent SAF (1996) and GEAR reports give figures of over 30 per cent (see Natfrass 1996: 26). It is worth noting that a recent ILO country review paints a more optimistic picture by arguing that South Africa's unemployment rate is closer to 20 per cent rather than 30 per cent. The review gives 16 reasons why official statistics understate employment (Standing etal 1996: 67-71). 3. The South African policy problem is often presented as a choice between redistribution and growth, but as Laurence Harris has pointed out, there is considerable evidence that countries with high growth also have low measures of inequality (see Harris in Michie and Padayachee (eds) 1997: 93). Thus, a more equal income and wealth redistribution in South Africa might not only be consistent with a rapid rate of economic growth but might also bring it about. 4. It is important to note that it is difficult to generalise across the experiences of all the 8HPAEs. This is one major criticism that Dwight Perkins made of the EAM in World Development, 1994. As he put it; 'one problem with this World Bank study is that, despite its frequent mention of the diversity of the HPAEs' experience, it keeps trying for generalisations that apply to all' (World Development, 1994: 658). In particular, Perkins suggested that there are at least three different groups of models within the 8HPAEs, namely the manufactured export-led model of Japan, Korea and Taiwan; the free port service, commerce-dominated model of Singapore and Hong Kong; and the natural resource-rich model of Thailand, Malaysia and Indonesia (World Development 1994: 655, 656). 5. The Gini coefficient is a measure of inequality, ranging from nil (perfect equality) to one (maximum inequality). 6. It is instructive to compare the change in 'World Bank thinking' between the World Development Report of 1983 (World Bank, 1983), the EAM of 1993 and the latest World Development Report (World Bank 1997). 7. Total factor productivity is the contribution to growth which is left over after measuring the contribution made by increases in the volumes of labour and capital. There are data problems of measuring the increases, as well as major problems of attaching weights to the increases (see the comments in Collins and Bosworth 1996, 193) and other, quite different estimates of TFP growth for East Asia have been produced (see Economist, 1 March, 1997: 24). For debates about the growth of TFP in the manufacturing sector of South Africa, see Bell (1995: 2-8); Bell (19%: 46); and Kaplan and Lewis (1996: 116-118). 8. For the two-gap model which was developed in the 1960s, see Chenery and Strout (1966). 9. Note that it is important to distinguish between total savings and savings in the formal financial sector only, since higher interest rates offered by the commercial banks may attract deposits away from the informal or 'kerb' market. See Harris (1988), where he points out that, in the 74 TRANSFORMATION 35 (1998) EDWARDS ARTICLE late- 1960s, higher interest rates attracted deposits into the state-owned 'commercial' banking sector in South Korea which facilitated the state direction of credit. 10. 'Singapore and Taiwan maintained unusually high public savings rates. Malaysia and Singapore compelled high private savings rates through mandatory provident fund schemes. Japan, Korea and Taiwan all imposed stringent controls and high interest rates on loans for consumer items and levied stiff taxes on so-called luxury consumption' (World Bank 1993: 16). 11. It is somewhat ironic that the World Bank's EAM study which came to this conclusion was itself carried out following an argument between the Japanese government and the World Bank over directed credit and financial repression. The argument between the World Bank and the Japanese government was based on Japanese criticism of aspects of the World Bank's structural adjustment lending and by Japanese concern that the World Bank should pay more attention to the actual experience of East Asian, and particularly Japanese, development. In the late-1980s, Japan had been criticised by the Bank for offering subsidised loans to assist private sector development in South East Asia. Then, in 1991, Japan issued a paper criticising the World Bank's Structural Adjustment Loans for over-emphasising macro-economic aspects and for under-emphasising the arguments for selective industrial policies and for direct and subsidised credit to compensate for market failures (see OECF 1994). Shortly after, Japan proposed that the World Bank should undertake a study of the East Asian development experience. The World Bank was reluctant but led by Masaki Shiratori, a former Japanese Executive Director, Japan's weight within the World Bank forced the issue. And so the project was launched from within the World Bank with the cost ($1.2 million) being largely financed by Japan but with most of the research staff being from within the Bank (see Fishlow et al 1994:4). 12. For a further study of financial repression in South Africa, see Lowitt (1994). 13. At the end of December 1996, government debt amounted to R304 bn., and for the 1997/98 fiscal year, nearly R40 billion, was earmarked for the servicing of this debt, compared to only R13 bn. earmarked for investment. In the 1997/98 Budget presented in March 1997, debt-servicing was estimated to be equivalent to 6.4 per cent of GDP and, without debt-servicing, there would be a surplus on the budget equivalent to 2.4 per cent of GDP. 14. The 'crowding-in' view is also put forward in a recent article by Brixen and Tarp (1996). They argue that not only is there room for increased public spending in real terms to help address South Africa's pressing social needs and that such expansion is possible without falling into a much feared debt trap but that if growth and real resource inflows falter, not even considerable moderation will be sufficient to maintain macroeconomic balance and avoid unsustainable public sector deficits (Brixen and Tarp 1996: 989). 15. Exchange controls on both residents and non-residents from and into South Africa existed in 1961. Until March 1995 control over non-residents' capital transactions were based mainly on the financial rand system. In March 1995 the Government abolished the financial rand (it had been abolished in 1983 and then reinstated in 1985 - see Kahn 1996: 3). Now there is a unified exchange rate but with some restrictions on movements of capital by South African residents, though these have been successively reduced. Thus, for example, since July 1995 i some financial institutions have been allowed to diversify parts of their portfolio overseas i (Kahn 1996; RSA 1996: 32,33), and, in the March 1997 Budget, further relaxations of exchange controls on resident individuals and on corporations based in South Africa were ; announced (see CrefSA, 1997: 26). TRANSFORMATION 35 (1998) 75 EDWARDS ARTICLE 16. GEAR commits itself to a gradual relaxation of exchange controls but without explaining why such a loosening is desirable (see Adelzadeh 1996: 80). 17. Under paragraph 224,2 of the 1996 Constitution, the South African Reserve Bank is required 'to perform its functions independently, and withoutfear, favour or prejudice, but there must be regular consultation between the Bank and the Cabinet member responsible for national financial matters'. 76 TRANSFORMATION 35 (1998)