Document(s) 5 of 11

Buying insurance

An apocryphal story about Fidel Castro and Che Guevara was popular in South African opposition circles in the 1970s and resurfaced in the early 1990s. According to legend, when Castro and the Cuban revolutionaries took power, Castro asked at an early planning meeting: 'Who is an economist here?' Thinking that Castro had asked, 'Who is a communist here?' Che put up his hand. 'OK,' said Fidel. 'You run the central bank.'

Unlike the Cuba of the legend, South Africa had a number of leaders of the newly dominant political party, the ANC, who were skilled in economics, though obviously not experienced in managing government. Some had post-graduate degrees in economics; some had post-graduate training in economic policy management; some had been given the opportunity to be special interns in investment banks such as Goldman Sachs. The second rank in the ANC was often more schooled formally, though less experienced managerially.

That is why many in the ANC were surprised when the two key economic positions in the new government were retained by the incumbents who had been appointed by the apartheid government. One was Chris Stals, the Governor of the South African Reserve Bank (SARB); the other was the Minister of Finance, Derek Keys.

President P.W. Botha appointed Stals in the midst of South Africa's politically leveraged debt crisis. Before his appointment, Stals had run the Treasury.

Stals was widely known as the key architect of South Africa's debt standstill agreement, reached with international bankers early in 1986. Chase Manhattan led a flight of banks and credit lines from South Africa in mid-1985 after President P.W. Botha made a defiant speech against reform. With a plunging oil price that led to a plunging gold price and then a plunging rand, a disproportionately large proportion of short-term international bank-to-bank loans precipitated a debt crisis. Stals negotiated a well-structured deal after Botha, under the threat of sanctions, made an announcement about reforming the influx control laws. When the ailing M.H. de Kock retired as SARB Governor in 1988 with a mediocre record as central bank manager, South Africa was lucky to get Stals.

Keys was appointed Minister of Economic Affairs (including Trade and Industry) by the reforming President F.W. de Klerk in 1992. Shortly afterwards, Keys also took over as Minister of Finance from Barend du Plessis, a former schoolteacher and a political ally of P.W. Botha. Keys was one of several National Party (NP) nominees in the post-1994 Cabinet which, although it was dominated by the ANC, was constituted as a government of national unity (GNU) according to an agreement embodied in the interim constitution. (There were also Inkatha Freedom Party members of the GNU Cabinet.) Keys had previously been CEO of the large, Afrikaner-based mining conglomerate Gencor, which he helped to transform into a world-class player.

While still under De Klerk's government, Keys had begun to bail out a very leaky ship. On the trade and industry side, he helped forge a new partnership between business, labour and government in the National Economic Forum (NEF). In finance he began to stem the growing budget deficit – the gap between government revenue and its expenditure, which his predecessor had managed to double in a few short years.

After presenting a sensible first budget for the new government (his second budget), Keys resigned late in September 1994, for personal reasons. It later transpired that he had been offered a top position in Gencor's operation in the United Kingdom. Instead of replacing Keys with an ANC politician, President Mandela appointed Chris Liebenberg as Finance Minister. Liebenberg, a professional banker, had been chief executive of the Nedcor group, one of South Africa's big four banking houses. He had distinguished himself in the apartheid era, during the transition towards democracy, by giving early and visible support to the Consultative Business Movement (CBM). The CBM was a pressure group set up by some relatively liberal business leaders who sought a meaningful interaction with the ANC, when many members of the business community were still afraid to be seen talking to the ANC, who had been demonised as terrorists for decades.

While he was somewhat liberal politically, Liebenberg was known as a steady, predictable banker. Mandela announced Liebenberg's appointment at a hastily convened press conference early in July after rumours of Keys' resignation started to affect financial markets. He made the announcement, as he put it, because 'the boys from the stock exchange and elsewhere seem to be very jittery'. He also felt compelled to assure journalists that there had been 'no change whatsoever in government policy' (Cape Times 6 July 1994).

The haste with which Mandela was forced to respond to the rumours of Keys' resignation indicates the kind of pressure and suspicion that the new South African government laboured under. This was a new, untried government, said the markets. This is a black African government, they added. It will have to prove itself over and over again, realised Mandela and the ANC Cabinet. Until proof was on the table, Mandela was forced to make gestures such as retaining Stals and appointing Liebenberg, though there were several ministers, deputy ministers and frontbenchers that could have stepped into Keys' shoes had competence been the only criterion.

When President Mandela appointed the respectable Liebenberg, he not only had to engineer a change in the constitution to allow for a person who was not an elected member of Parliament to serve in the Cabinet, he also had to adjust the Cabinet to compensate the NP for losing a seat in the GNU. Liebenberg did not represent any political party.

The overwhelming condition for macroeconomic policy for the first ANC government was that whatever the evidence they received to the contrary, the markets expected the Cabinet to make mistakes. If mistakes were made or even suspected, they were fed to the rumour mill that buttered the bread of currency and securities dealers. Once Mandela was recognised as a responsible and careful leader, his future, as if everything rested on it, became the main object of speculation. Periodically, throughout his term of office, rumours about Mandela's health, for example, fed currency speculation. Frustration with these tendencies led Trevor Manuel, after being appointed Minister of Finance in 1996, to once angrily refer to the market as 'amorphous'. Though this was essentially true, the media and the market punished him for his tactlessness.

Though many in the financial world may have hoped for the best for South Africa, they expected the worst. Stories about South Africa continued to take on the dark shades of the apartheid era. The South African story had long been associated with drama and tumult. Few outsiders dared to tell the South African story in any other terms. Mandela's government was not wholly successful in its struggles to counter a racially tinged pessimism about Africa, and to encourage the world to accept the South African 'miracle' as a carefully constructed compromise. This, at least partly, explains the preoccupation of Mandela's successor, Thabo Mbeki, with the theme of 'African Renewal', but this is a story we will come back to later.

Because the ANC operated, at least from 1992, under the assumption that financial markets would second-guess them whenever given an opening, it was forced to be more conservative than it would have been had it been the government in a well-established democracy. This was the main reason that the ANC, before and after the elections, publicly committed itself to economic policies that would not antagonise influential world economic players. The three most obvious policies falling into this category were the intention to reduce the government deficit, the general commitment to remove restrictions on exchange control, and the agreement to include a strong form of central bank independence in the constitution.

Another motivation for fiscal and monetary conservatism was the ANC's fear of giving up South Africa's limited economic sovereignty to the International Monetary Fund (IMF) and/or the World Bank as the result of a financial crisis. As ANC parliamentarian Ben Turok put it: 'A new democratic South Africa will need to defend its interests against the predatory actions of international capital and institutions like the International Monetary Fund [and] the World Bank...' (Sunday Times: Business Times 13 March 1994).

The irony was that in order to stave off the power of international finance, the ANC committed itself to policies approved by the same financiers. In order not to get too indebted to those who could turn their debt against them, they had to be conservative and pander to some of their prejudices.

And yet, the electoral platform of the ANC was the Reconstruction and Development Programme (RDP). The legitimacy of the ANC rested on its ability to deliver an improved life for its constituents – poorer South Africans excluded from power and privilege under apartheid. Moreover, it had to do this inheriting an economy in severe decline, with huge, growing government debt and interest burdens, low or negative growth, declining per capita income, capital outflows, and poor competitiveness in most of the economy. In short, the ANC government had to undertake a general restructuring of the economy and a reorientation of the economy towards the historically excluded masses at the same time.

Is it easier to confront growth and distribution issues at the same time, or is it easier to attend to one first then the other? South Africa's experience suggests that it is easier and more sensible to address the two challenges simultaneously. Restructuring for growth is politically difficult – it encounters the problem that those who lose from restructuring usually have more powerful voices than those who will gain in the future but for now have no voice. Therefore, to engage in restructuring with a powerful constituency with a great deal to gain through redistribution should make it easier to do the two reforms together. Moreover, some key aspects of restructuring for growth entail a good measure of redistribution of resources, certainly in the South African case – human resource development and small business development are two obvious examples.

This view is supported by comparative examinations of political and economic reform. When Joan Nelson and her colleagues compared transitions in several Latin American and eastern European societies, they found that it was easier to carry out effective economic reform in some of the countries of Eastern Europe, where the political change was more radical. As Nelson put it, 'where democratic reforms marginalize old political elites and actors, and where public opinion is prepared for fundamental economic changes... the launching of [economic] reforms is facilitated' (Nelson 1994: 13). This statement covers much of the South African case.

Mending a leaky tank

In the run-up to the transition, the government's budget was haemorrhaging. The budget deficit had reached a record high of 8.5% of GDP (gross domestic product) in 1992–93 (SARB 1998b).1 But to compare the country's finances in the apartheid era, when it was split into several allegedly independent countries, the fiscal accounts of the central government and the 'independent homelands' should be combined. If this is done, the consolidated deficit reached 9.5% in 1992–93, easing off to just under 8% in 1993–94 (estimated figures) in Keys' first budget (Fallon and Perreira da Silva 1994: table 2.6). From 1978–91 the fiscal deficit had averaged between 4 and 5%, but in 1991–92 fiscal restraint disintegrated. This led some ANC supporters to speculate that the old regime had deliberately set a debt trap to constrain the actions of the ANC government. The new government, which came into power in April 1994, had no spare cash and was locked into debt obligations for years to come. Whether or not there was a deliberate debt trap strategy, the new government certainly was weighed down by its obligations.

Part of the story was the rising cost of paying people to defend apartheid. The percentage of government expenditure that went to the payment of its employees, rose from 26.5% of government spending in 1978–82 to 35.5% of spending in 1992–93. From 1981–92 the number of employees in the central government and the provinces grew by more than a third, while the number of employees of the 'independent homelands' more than doubled (Fallon and Perreira da Silva 1994: 109).2 The apartheid regime was shoring up its defences by selling patronage, but at the long-term expense of the health of its finances.

Change in the composition of government expenditure itself contributed to lower growth. As Table 3.1 shows, the amount spent on recurrent expenses – that is, not invested for the future – rose very rapidly during the 1980s. From 70.3% of spending in the late 1970s, recurrent expenses rose to nearly 86% of government expenditure in 1993. The main reason was the rising government

wage bill. Another reason was the mounting government debt which, combined with rising interest rates, meant that the interest paid by government to finance its debt nearly doubled as a proportion of government expenditure between 1978–93. Rudolf Gouws, a respected banking economist, told the new Parliament's finance committee that the government had inherited a 'fiscal mess' (Cape Times 17 August 1994).

Steering a careful course

This was not a promising moment for the new government to attempt to implement new policies and programmes. The state apparatus was under-funded and over-borrowed, the apparatus for revenue collection was in a dire state, and the economy was limping along, having only recently entered positive growth territory for the first time in four years. Facing this, the government seemed to have a number of options: it could raise tax levels, either/or both income tax and a 14% value-added sales tax; it could increase government borrowing; or it could borrow heavily from abroad, as new sources of funds were available to the new government. At least some of these actions seemed necessary if the ANC government was to deliver on its election promises or, more importantly, shore up the fledgling democracy with real benefits for the mass of the people.

But it did not do any of those things. Instead, the ANC committed itself to reducing the budget deficit as a percentage of GDP, and to bringing down the overall government debt burden eventually. It was thought that the debt burden was crowding out private sector borrowing – what was certain was that the debt to GDP ratio raised fears of macroeconomic instability that could drive away private investment capital. So, the government came down heavily on over-expenditure. Alec Erwin, then Deputy Minister of Finance, reported to Parliament that a 'crash-team' was taking drastic steps to correct all cases of over-expenditure (Business Day 14 September 1994). In addition, the government planned to improve revenue collection, and reallocate funds from low-priority areas to high-priority areas. Before and after the elections, ANC leaders were quoted on several occasions saying that they would fund new social programmes through funds reallocated from downgraded or redundant functions (see, for example, Cape Times 16 March 1994).

What did budget reallocation mean in practice? If we compare 1990/91 expenditures with those of 1995/96 we will see that the major losers were defence, economic services and general government services. Defence was cut from 16% of the budget to 9%; general government services were cut from 8 to 7%; while subsidies that encouraged manufacturers to locate in remote homelands were cut from 2.5 to 1.3% of the budget. Most of the reallocated funds went to social services, including: education up 1.6 percentage points to 26.2%; health up half a percentage point to 12.3%; housing up half a percentage point to 2.3%; and social security and welfare, up 3.5 percentage points to 11.9% (South Africa Foundation 1996: 54). Also, a RDP fund was set up for special projects, starting with R2.5 billion – about 2% of the budget.

All of these increases were intended to contribute to the equalisation of services provided to the historically disadvantaged black population, though in some cases the overall standard could not be as high as it had been for privileged whites in the past. Up to 1994, social services for whites were hugely more generous than those provided to other 'race groups', especially black South Africans. For example, the ratio of government spending per school pupil, black to white, was one to seven in the mid-1980s (Wilson and Ramphele 1989: 141) and these figures had been even worse in the 1960s and 1970s.

All in all, not an extremely dramatic shifting of funds. The huge backlog in the provision of public facilities such as roads and water, and services such as health and education might take longer than the RDP originally envisaged.

By late 1994 the government realised that reallocation could not address the fiscal challenges confronting it, and it announced a 'six-pack' belt-tightening strategy. The key new elements were: a voluntary cut in salary for parliamentarians, ministers, and the President; the redeployment of and cutting back on civil servants (initially through voluntary retrenchment packages); a commitment to the full or partial privatisation of state assets and enterprises to reduce debt or supplement the RDP; and the reorganisation of fiscal relations between the central government and the provinces (Cape Times 31 October 1994).

The government was not keen to raise tax levels. South Africa was already, at least on paper, a heavily taxed society. Company and personal marginal income tax rates were over 40% of income. Though the basic company tax was 35%, there was a secondary tax on distributed profits that could raise the rate to 42.5%. Derek Keys had cut company taxes, but they remained high by international standards, as did personal income taxes. The ANC felt raising income taxes could scare off potential investors, and held off. There was a once-off levy in the 1994 budget to help pay the costs of the transition, particularly the elections, but this made it more difficult to justify any other additional tax. The major indirect tax, the value-added tax (VAT) of 14% on all purchases of goods and services, was seen by the ANC and its supporters as a regressive tax. It penalised the poor, who spent almost all their income on VAT-taxable goods and services. Raising it would damage the credibility of the ANC in its own backyard. Instead, the new government decided to focus on improving the performance of the revenue collection machinery under the existing tax dispensation.

The ANC was equally reluctant to borrow heavily from abroad. A World Bank report published almost simultaneously with the 1994 election suggested that higher levels of government investment expenditure was one of five key strategies for faster growth and redistribution in South Africa (Fallon and Perreira da Silva 1994: chapters 5 and 6). Implicit in the report is the notion that a growth promoting budget deficit could be financed through foreign borrowing at low enough levels not to worry the new government. More precisely, the World Bank economists felt that the government should urgently underpin growth with substantial investments in South Africa's social and physical infrastructure that aimed to redistribute facilities, resources and opportunities (Fallon 1993).

Tony O'Reilly, the Irish-born Heinz chair, president, CEO, and the major shareholder in the Independent newspaper group, which bought control of South Africa's biggest newspaper group in 1994, strongly urged the government to borrow heavily to get the economy going. O'Reilly, whose purchase of the Argus newspaper group was supported by Mandela and the ANC, said in July 1994, 'We have got to kick-start this economy, and it will not be kick-started by world business' (Cape Times 6 July 1994). Months later, after South Africa made progress in its credit ratings, he reiterated his message. He warned of the perils of under-borrowing, and urged the government to 'take appropriate risks to jump-start this economy, and jump start it tomorrow' (Cape Times 10 November 1994). This in an economy where the growth rate had doubled from 1.3% to 2.7% from 1993–94. Brian Kantor, a senior macroeconomist at the University of Cape Town, also called for a more ambitious borrowing programme (Cape Times 25 August 1994).

Nevertheless, the South African government refused to bite. Though some foreign bonds were issued, their main purposes were to refinance existing credit, to restructure the debt portfolio and its maturity profile, and to set benchmarks for other South African borrowers (CREFSA 1997).

Why did the government not take greater advantage of loan offers from private and public international lenders? One reason was undoubtedly the fear, noted earlier, that badly timed and badly managed borrowing could leave the government in the hands of the IMF. A second reason was that the government lacked confidence in its ability to invest the money effectively. The fiscal machinery was in a mess, and the capacity of most of the government departments to spend money effectively was far from proven. In retrospect, this concern was fully justified in 1994. A third and related reason was that the fiscal squeeze was a good rationale for a radical shake-up of the public service, something urgently needed in almost all South African government departments and organisations in the early phase of the transition to democracy.

The new government preferred to err on the side of caution. The new government leaders had little experience of managing a country's finances and suspected that the civil service managers they inherited were not necessarily competent or trustworthy. Their own handpicked staff was only beginning to enter government, particularly in the Department of Finance, which did not have an ANC minister or director-general until 1996, and even more slowly in the SARB. They feared that the long-term costs of making a mistake were potentially greater than the costs of missing an opportunity. This new government was able to think in longer terms than most democratic governments, partly because of the extraordinary vision of the new leadership, and partly because it had good reason to be confident of being returned to government in the second democratic elections in 1999.

In any case, the Minister of Finance was preoccupied with critical issues that had to be addressed before he could turn his attention to expansionist strategies. Simply constructing the 1995/96 budget was a 'nightmare' according to new Minister of Finance, Chris Liebenberg, because of the complexity of incorporating former homeland finances into a single national account (Business Day 11 November 1994). Even then, many homeland offices were still running semi-autonomously. The Department of Finance was reluctant to transfer homeland finances to the relevant province – nine provinces had been created in the constitutional settlement. South Africa had to transform four existing provinces and nine homelands, of which several were 'self-governing' or 'independent', into nine provinces in one nation. It was at the homeland level that the nation's finances were most problematic, and some problems evident in 1994, such as redundant and untraceable civil servants (the term 'ghost workers' was invented to describe this problem), took a decade to resolve in some provinces.

A key fiscal target set by the government in November 1994 was to reduce the budget deficit from the 6.8 to 4.5% of GDP over a five-year period, with an expected GDP growth rate of 3% per annum on average. As Table 3.2 shows, the deficit target was easily accomplished, growth on average was close to 3%, but the government could not immediately reverse the trend towards higher consumption expenditure.

Hands on the brake

Article 224 of South Africa's Constitution is headed 'Primary Object' of the central bank, though it goes further than that. It reads:

(1) 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.

(2) The South African Reserve Bank, in pursuit of its primary object, must perform its functions independently and without fear, favour or prejudice, but there must be regular consultation between the Bank and the Cabinet member responsible for national financial matters (Constitution of the Republic of South Africa 1996).

In 1994, Stanley Fischer, soon to become number two in the IMF, suggested a terminology for describing central bank independence. Fischer used the term 'goal independence' to describe the extent to which the central bank could set its own goals (low inflation, high growth, low unemployment or whatever), and the term 'instrument independence' could cover the extent to which central banks were free to deploy whatever instruments they chose – trading government stocks, trading foreign exchange, manipulating the interest rate, or whatever (Fischer 1994). It has been widely believed that central bank independence, especially instrument independence, is 'a good thing' because it insulates the currency from short-term-oriented politicians. However, in a democratic society, goal independence would insulate the central bank from the values of the sovereign state, which is not desirable. The results of sufficient independence are expected to be low inflation and a sound foundation for growth.

Recent surveys show that central bank independence is not an absolute virtue. The evidence increasingly suggests a trade-off between high independence and low inflation, and moderate independence and higher growth (see Eijffinger and De Haan 1996; Blinder 1998; Forder 1998). Of course, no independence and uncontrolled inflation would still be disastrous.

Applying Fischer's terminology, the South African Constitution clearly enshrines 'instrument independence' (the choice of policy instruments through which the Bank achieves its objectives) in the phrase: 'must perform its functions independently and without fear, favour or prejudice'. And the Constitution does not allow 'goal independence'. The Constitution sets the goal of the SARB in the clause which reads: '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' (Constitution of the Republic of South Africa 1996). This is the interpretation of a South African Chamber of Business (SACOB) report, which noted that 'instrument independence' and 'goal dependence' are consistent with international practice and 'probably also supportive of international investor perceptions and business confidence' (SACOB c.1993: 27). On the left, the wording of the constitutional clauses on the SARB has been characterised as a victory for 'capital' or 'international finance capital' (Marais 1998).

But was SACOB right to think that the SARB does not have 'goal independence'? The constitutional clauses define the goal vaguely, and then insulate the SARB from outside interference. The SARB has some room to interpret the goals of monetary policy, giving weight to what Alan Blinder, former vice-chairman of the United States Federal Reserve Bank, called the central bank's 'interpretive role' (Blinder 1998: 54). What do 'balanced' and 'sustainable' actually mean in clause 224(1)? The Minister of Finance is given access to the Governor's ear for 'regular consultations', and since 2001 has established an inflation-targeting regime, but the autonomy of the bank remains considerable.

Stals often referred to the goal of preserving the 'internal and external value of the rand' (Stals 1994b: 27). (The rand was floated in the late 1970s.) He would have approved of the fact that there was no obligation on the SARB to consider economic growth in broad terms, nor was there any mention of an obligation to create jobs, such as the 'maximum employment' objective in the legal foundation of the Federal Reserve Bank of the United States. Stals' vision was always explicitly closer to the 'purist' approach of the Deutsche Bundesbank, which is directed by law simply to 'safeguard the currency' (Blinder 1998: 54).3

His job as central banker under the new government, felt Stals, was just the same as it had been before. 'We have spent the past five years establishing relative domestic financial stability. We must spend the next five years ensuring we maintain that stability,' he said shortly after President Mandela's new government was installed. 'Our contribution to the RDP', he added, 'is to ensure monetary stability.' He emphasised that the major problem he still had to deal with was the pressure on the balance of payments (The Star 27 May 1994).

What Stals failed to recognise was that the circumstances of the post-apartheid government were very different from those before 1994. International capital was now potentially available, in all shapes and forms. Also, the prospects for greater political stability and certainty under a credible, democratic government meant that capital could flow in relatively quickly.

Stals had two preoccupations: the first being inflation, and the second being the balance of payments. In a country such as South Africa, where imports are high at 20 to 30% of GDP, and where the inclination for imports to grow during periods of GDP growth is very high,4 one anti-inflation strategy is to prevent the price of imports rising by maintaining the external value of the currency. The two targets were closely linked. A high interest rate had the combined effects of strengthening the currency and slowing the economy. In the South African case, the effect on imports of the currency appreciation resulting from the higher interest rate was normally more than offset by the contraction of the domestic economy. So, while sucking in short-term capital, imports were constrained, tackling inflation and the balance of payments simultaneously.

Stals was not afraid to use the interest rate to protect the value of the currency and the balance of payments. As Table 3.3 shows, the real bank rate (the difference between the SARB rate and the consumer price index) varied between 1 and 5% in the early 1990s, and rocketed up to over 13% in the late 1990s. The real prime interest rate rose as high as 17.5%. As Paul Krugman reminds us, in the era beyond the fixed currencies of the Bretton Woods agreement, the 'iron law of international finance' is that it is impossible for countries to maintain stable exchange rates and currency convertibility, and still have the freedom to use interest rates to support economic expansion (Krugman 1991: 60–61).

The way that Stals tackled monetary stability was old-fashioned monetarism. In March 1994 he confirmed that the SARB felt that 'money supply targeting remains... the most sensible anchor for monetary policy...' For 1994, he announced, the acceptable range for the rate of increase in M3 was 6 to 9% (Stals 1994a: 25–26). M3 is the quantity of broadly defined money, including deposits and credit. As Table 3.3 shows, the results for M3 growth in 1994 (15.7%) were not even close to the SARB's acceptable range, and yet the rate of inflation continued to fall.

In 1995, with an upswing well under way, Stals had two choices. On the expectation that continued positive economic trends would encourage more and more long-term investment that would compensate for rising imports, he could have continued with a moderately tough monetary policy and allowed for a gradual depreciation in the currency with modest single digit inflation targets. Instead he embarked on a much tighter monetary policy: he raised the Bank rate to 15% (3% above the mid-1994 level), increased the reserve requirements for commercial banks, and issued guidelines that indicated maximum desired credit extension levels to the banks.

He explained his action by referring to the rapid rate of M3 growth, driven by an even greater rate of growth of credit extension to government and private borrowers – credit grew by 21% in all during 1994. Stals feared the credit expansion would overheat the economy and lead to inflation (Stals 1995). In fact, as Table 3.3 shows, while M3 continued to grow rapidly in 1995, 1996 and 1997, inflation continued to decline, except in 1997 when the 20% depreciation of the rand in 1996 had an adverse effect.

The tight money policy meant the real interest rate more than doubled from 3% in 1994 to more than 6% in 1995 and more than doubled again to 13.7% in 1998. This had two effects: firstly it started to choke off what was a private sector, investment-driven expansion; secondly it attracted excessive volumes of short-term capital inflows undermining the prospects of a stable currency.

Volatile short-term capital, particularly in the South African case, largely takes the forms of bank-to-bank loans, short-term loans to public corporations and authorities, liabilities related to reserves and, most volatile, private sector, short-term, non-direct investment. Intermediate-term capital would include long-term loans and portfolio investment, while the most stable form of investment is direct investment (for more detail, see Khatri et al. 1997; Kaskende et al. 1998).

Developing countries, particularly a country such as South Africa where the propensity to save has fallen very low,5 and where the early period of the growth cycle is characterised by higher rates of import growth than export growth (Kahn 1987), need significant injections of foreign capital to underwrite growth.

There are two ways for foreign capital inflows to underwrite growth in South Africa. The first is for high real interest rates and a relatively strong currency to attract mainly short-term capital. However, sooner or later the high interest rates will cause the currency to appreciate, reducing competitiveness, or it will cut off domestic expansion, or both. In any case, it leads to a repeated boom-and-bust cycle. The second alternative is for longer-term capital to be attracted by steady longer-term growth, at the expense of the gentle depreciation of the currency because of the lag between the deficit on the current account and the inflow on the capital account. A strategic government-borrowing programme aimed at investments to improve productivity could help to prevent the lag being a very long one.

The latter course, when accompanied by general fiscal and monetary discipline, is potentially more beneficial than the former course. An analytical study published in one of the SARB's Quarterly Bulletin's supports this argument (Smal 1996). Why take the former course, then, as Stals did? There are several possible explanations.

One is confidence. Stals had long worked in a country where confidence in the future was very low, leading to few long-term private investment projects, and expedient, short-term relations with the outside world. He may have felt that the lagged long-term direct investment and project-related loans would never materialise on a scale to allow the economy to move towards balanced external accounts, let alone growing reserves. There was no government-borrowing plan.

In addition, South Africa had just embarked on a programme of exchange control reform. Exchange control had started during a political crisis in 1960, was briefly lifted between 1983 and 1985, and then reimposed during the 1985 debt crisis. Both the ANC and Stals had talked about the need for liberalisation, though it was never quite clear what the end-point of that liberalisation should be – complete as in fully developed countries such as the United States and the United Kingdom, or partial. What Stals and the ANC agreed was that liberalisation should be gradual. The first step was to abandon the separation of the financial rand – a closed international capital market where every sale had to be matched by a purchase – and the commercial rand used for current transactions. The passage of exchange control reform is summarised in Box 3.1. Though a key motive for exchange liberalisation was to increase confidence in the South African economy, after the disastrous experience of 1983–85 it was an act of faith. Stals must have been worried about the unknown outcome of liberalisation.

In pushing up the bank rate, Stals might also have been overly influenced by the M3 and credit extension figures. Stals seemed convinced that credit and broad money supply expansion would lead to overheating and inflation, though there was no evidence from the economy that this was indeed the consequence. What Stals may have missed here was the rapid restructuring going on in South African society and the economy, with a more secure black working class and a fast-growing black middle class opening bank accounts and adopting new financial facilities, as well as the expansion of micro-credit after an amendment to the Usury Act in 1992.

He was clearly worried about the low levels of domestic saving. It is possible that he felt that bank savings would continue to shrink as long as the inflation beast was not yet dead and buried. South Africa had experienced negative interest rates for a very long period between the mid-1970s and 1988, during which it was rational to be a debtor, not a saver.

Another concern was his view of the South African economy's ability to respond to the new opportunities offered by the more competitive currency. He thought labour markets were not flexible enough to allow depreciation to have long-term positive effects on competitiveness. Workers would bid their real wages back up to where they were before the depreciation.

A fourth possible reason was the way that the SARB tended to stabilise the currency when it got wobbly – in addition to intervening in the spot market, it got involved with the forward market for foreign exchange. This began through the provision of forward cover because of the absence of a commercial forward cover market in South Africa. When South Africa ran short of foreign capital, as during South Africa's debt crisis from 1985–93, the SARB offered preferential long-term rates to encourage South African borrowers to seek funds abroad (Khatri et al. 1997: 33–38). For most of the 1990s the main form of SARB intervention in the forward market was through foreign exchange swaps. When it used swaps to intervene in the spot market, it usually resulted in the SARB being 'oversold' in foreign currency to prop up the domestic currency.

Frequently, the net oversold position was as much as six or seven times as high as the gross foreign reserves position (foreign reserves being the central bank's holdings of foreign currencies and gold). The gross foreign reserves position tended to oscillate between the very low levels of about two and three billion US dollars, often less than the value of two months of imports (Khatri et al. 1997: 35). So the reserves were insufficient for the SARB to use them effectively as a bulwark against currency fluctuations. Currency swaps were far more heavily used.

Because the SARB had to settle its accounts in foreign exchange, it may have had an interest in further propping up the domestic currency so that the final settlement was not too expensive – in other words, so that the difference between the price of the rand at time p and at time p+1 was not too large. Some commentators suggested that the use of the forward market to stabilise a falling rand might have led to a vicious circle, where the SARB was forced to expend more and more resources protecting the rand and its position. Jonathan Leape's team at the London School of Economics argued: '... the size of the oversold position at particular points in time has raised suspicions that the need to reduce the oversold position became an independent factor guiding exchange rate and even interest rate policy' (Khatri et al. 1997: 44).

Whatever its intentions, the strategy of the SARB regarding the exchange rate and interest rates influenced the composition of capital inflows. A tight money policy with high real interest rates and an exchange rate strategy that tends towards overvaluation encourages short-term capital inflows because of the interest to be earned and the expectation that depreciation will not eat these short-term gains away. However, by discouraging investment and consumption in South Africa's real economy through high interest rates, and by decreasing competitiveness through a relatively overvalued currency, this policy did not encourage private, long-term, direct investment. Moreover, short-term capital inflow bubbles can burst, as happened in 1996 and 1998 when South Africa experienced two rather brutal 'mornings after', during which the currency corrected itself downwards by around 20%. The currency shock in 2001, though similar in effect, had different causes.

Was there a smoother path?

Was another path open to the SARB? Analysis of Latin American experiences with capital flows and exchange rates implicitly suggests that the South African experience in the 1990s was hardly atypical. When developing countries begin serious market-oriented reform processes, the pre-existing external credit constraint will relax. It should result in a long-run sustainable volume of capital inflows, but it will also 'generate a short run overshooting in the inflow of capital into the country'. Short-run capital inflows will exceed their long-run equilibrium value until the market completes its adjustment. In some cases the adjustment process takes a few years (Edwards 1998: 16).

The capital inflow during the adjustment period will lead to an overly appreciated real exchange rate, which when the inflow adjustment tails off, will require a 'massive adjustment'. 'The dynamics of capital inflows and current account adjustment will require, then, that the equilibrium real exchange rate first appreciates and then depreciates', argued Sebastian Edwards (1998: 16).

Anticipating such events would seem to be no more than good common sense. In several Latin American countries and in Israel, as Edwards notes, such shocks have been anticipated and counter-measures have been adopted. Indeed, South African economists began writing about the potential negative effects of post-democratic capital inflows into South Africa well before the 1994 elections (Kahn 1992; Hirsch 1993). The obvious question is: why did the SARB not anticipate this major shock, and why did it not attempt to counteract its effects? Indeed, its interest rate policies almost seemed to be designed to exacerbate the effects of the adjustment to the removal of the external credit constraint, rather than ameliorate them – perhaps because of a conflict between the inflation target and the exchange rate target in the context of capital inflows. There are no indications that Stals anticipated negative effects from the should-have-been-expected capital inflow. In fact in 1995 he admitted that the volume of capital inflows surprised him. Was this simply a case of an old general fighting the last war (capital flight) instead of the current one (unemployment)?

If Stals had anticipated the adjustment shock, and if he was concerned about currency fluctuations and the loss of competitiveness, what could he have done differently? Can developing countries design their exchange control regimes to insulate themselves from some of the negative effects of volatile short-term capital flows? Can they influence the composition of capital inflows to encourage a higher proportion of the more stable form of inflows, particularly long-term loans and direct foreign investment?

Had Stals recognised that he was winning the fight against inflation, that the relationship between M3 expansion and inflation had weakened in a more open competitive economy, and that some of the M3 data was a result of structural rather than behavioural change, he could have softened interest rate policy, encouraging domestic growth and long-term capital inflows.

Secondly, he, or the Minister of Finance, might have intervened directly in South Africa's international capital movements. Though Edwards is sceptical about the long-term effects of capital controls on the external value of the currency and interest rates, he does accept that such controls do slow down exchange rate appreciation, and give the monetary authorities greater autonomy temporarily in interest rate policy. More importantly, Edwards accepts that a country that has imposed such controls, such as Chile, has effectively skewed its capital inflows towards long-term serious money, and away from volatile hot money (Edwards 1998: 28, 30, 36). Even the IMF's managing director, Michel Camdessus, urged that, while capital flow liberalisation was necessary, 'the last thing you must liberalise is very short-term capital movements' (Edwards 1998: 13).

In the absence of an international agreement to inhibit volatile capital flows, some international economists see no reason why individual countries should not take such a step themselves (Helleiner 1998a). The risk countries take is opening themselves to scorn and misinterpretation (like Malaysia in 1998). The usual method is some form of explicit tax inhibiting short-term capital flows to and from developing countries. Chile has used several forms, including non-remunerated deposits in local banks and taxes. Rudiger Dornbusch took a lead from John Maynard Keynes and James Tobin in suggesting a transactions tax (Dornbusch 1998).

Moving onto new terrain

In the months before the 1994 elections, while some whites were amassing stores of canned foods and candles, many white stockholders were inclined to dump their equities. The market was not excessively weak, but some local investors thought that institutional and foreign buyers had not realised that the ANC would win the April election, and that there would be a major fall in share prices in reaction.

As it turned out, selling stocks in 1994 was a bad strategy. When the elections were won without too much fuss, the threat of a white-right reaction dissolved. After nervousness during the first half of 1994, markets settled down quickly, reassured by the reappointment of Derek Keys as Mandela's first Finance Minister, and Chris Stals' agreement to a five-year contract as Governor of the SARB. Conservative whites moved their candles and canned foods back from hideouts to kitchens. Foreign capital flowed in fast, and certain sectors of the Johannesburg Stock Exchange (JSE), such as construction and information technology, surged upwards.

Capital was attracted by high interest rates, the resumption of growth in the South African economy, bank borrowing, and public sector gilt issues. In October 1995, Chris Stals admitted that capital inflows 'exceeded the most optimistic expectations of the preelection period'. Over the 18-month period ending in June 1995, there was net capital inflow of R18.6 billion, or over US$5 billion. By contrast, the previous 10-year period had seen net annual outflows of at least R5 billion (The Star 16 October 1995). Net foreign purchases of equities mounted to US$1.3 billion in 1995. With the JSE bottled up because South Africans still had very little freedom to invest abroad, the rise in demand for equities soon translated into rising share prices.

Days before Stals' statement, the IMF had mentioned South Africa in a list of developing countries (which also included Indonesia and Brazil) that had to pay more attention to the way they were managing the inflow of foreign capital. In the World Economic Outlook, the IMF noted that in general there was no need to reconsider liberalising financial markets because of the potential for market turmoil. But an IMF spokesman did say that South Africa should take special note of the recommendation to proceed cautiously along the road to free international capital markets (The Argus 5 October 1995). A clear implication of the statement was that the IMF was concerned about the composition of South Africa's capital inflows, which were heavily skewed towards short-term, potentially volatile capital.

Later in October, the Union Bank of Switzerland (UBS) issued a report praising the government's policies. But, noting that the foreign direct investment component was only US$700 million (less than 10%) of the capital inflow, the UBS indicated concern that the rand was vulnerable to shifts in short-term funds. 'Foreign perceptions are such that any negative developments in South Africa could easily trigger an outflow of short-term capital...' (Cape Times: Business Report 23 October 1995).

For a few months markets held firm. Then, early in 1996, the UBS issued another report indicating that the rand was overvalued by 7–10%. This was persuasive in the context of the surge in capital inflows and a still-rising real interest rate. Days later, rumours about President Mandela's health sent the rand tumbling. The excess 10% was shaved off the rand's value in less than two weeks, before the end of February 1996.

Again the markets seemed to settle. Then, in late March, shortly after Chris Liebenberg presented his second budget, Trevor Manuel, who had been Trade and Industry Minister, replaced Liebenberg as Minister of Finance. Liebenberg resigned, probably in terms of an agreement he had with Mandela when he accepted the position in 1994. Alec Erwin, who had been Deputy Minister of Finance, took over as Trade and Industry Minister, and the RDP was dissolved as a separate government entity. Its functions were relocated: financial management to the Department of Finance; and planning and co-ordination, in a low-key form, to the office of the Executive Deputy President, Thabo Mbeki.

The RDP Office had failed to spend much of the funds committed to the programme, and the government decided that it was duplicating the functions of individual departments. A special economic growth Cabinet committee under President Mandela had been wrestling with the need to improve economic policy coordination for months. It was preferable that the departments themselves were empowered to meet the challenges of reconstruction and development. Now with an ANC Minister, the Department of Finance was ready to co-ordinate this drive. Liebenberg stepped gracefully aside and returned to the private sector, having helped ease the transition to the first ANC Finance Minister.

If ever there was a baptism of fire, it was Trevor Manuel's, on his appointment as the first ANC and the first 'black' Minister of Finance in South Africa. Manuel had shown in his 22 months as Minister of Trade and Industry that he was a good manager and leader, and was as tough as teak. He showed this in pushing through an initially unpopular trade liberalisation programme, and he showed it in Cabinet, where he was not afraid of confrontation when he stood on firm ground.

Toughness was the quality Manuel needed most in his first half-year as Minister of Finance. The rand, which had stopped its descent during mid-March, began to plunge again after the Cabinet reshuffle was announced. One reason was Manuel's appointment. 'The markets' seemed concerned that there was an ANC Minister of Finance, that he was 'black', and that he was not a financier. Perhaps they suspected that he would buckle under the demands of his Cabinet colleagues; perhaps they thought he would make other mistakes – his bull-headed approach to trade liberalisation left some commentators concerned that he might try to liberalise exchange controls too fast (The Star 30 March 1996). Perhaps some traders were merely exploiting uncertainty and selling short.

Another reason for the weak rand was the blossoming of an intense debate about economic policy. During 1995, SACOB had said several times that it felt that economic policy needed stronger co-ordination. It felt that the RDP document and white paper really did not provide clear enough guidance. The comment was first made in response to a Cabinet decision to set up a special committee to focus on growth, led by the President (Sowetan 10 August 1995). Later SACOB's comment was echoed in the ANC, which stressed the need for the co-ordination of macroeconomic policy (Business Day 16 October 1995).

In 1996, four weeks before Manuel's appointment, the South Africa Foundation, a pressure group for the largest South African companies, brought out an economic report called Growth for All, which set out the economic options, as they saw it, with brutal frankness (Business Day 1 March 1996). One key argument was that South Africa's high unemployment was in large measure due to inflexible labour markets. The proposed solution: if jobs were to be created, the trade unions had to be circumvented through a dual labour market. One labour market would be for skilled and unionised workers, another for unskilled and non-unionised workers. Other remedies included anti-crime strategies, leaner government, brisker privatisation and export promotion. In some respects – monetary policy for example – Growth for All might have been a useful contribution to the economic debate in South Africa, but the report was clumsily composed and non-strategically dumped into the public arena.

The ANC shot back immediately through Tito Mboweni, the Minister of Labour, in his capacity as chair of the ANC's economic policy committee. Mboweni lambasted proposals to cut the budget deficit and hoist VAT, and rejected the suggestion that privatisation could fund poverty relief, as naοve. But he reserved his most withering comments for the 'most ridiculous of all' proposals – the two-tier labour market – which he described as an 'affront to democracy' (Sunday Times 10 March 1996). Mboweni had spent exhausting months finding a compromise between labour and business for the new Labour Relations Act, and big business seemed to be trying to undo the agreement.

Before long, the ANC found itself attacked again, this time from the left. Vella Pillay, who had worked for the Bank of China in London, and was a long-time adviser to the ANC who had led MERG (the Macroeconomic Research Group project – see Chapter Two), was now the director of MERG's descendent – the National Institute for Economic Policy (NIEP). NIEP had been set up with foreign financial assistance to aid the ANC in policy formulation. Having failed to find a way to do this effectively, it found itself developing strategies and critiques from a step to the left of the ANC mainstream. Pillay ridiculed the debate between the ANC, the unions, and the South Africa Foundation, calling it a 'debate for simpletons'. He criticised the ANC's fiscal policy, suggesting that instead of cutting the budget deficit the government should be expanding the deficit to 7 or 8% of GDP (Cape Times: Business Report 12 March 1996).

Days after Manuel's appointment, the labour movement dipped its own oar in the water with a policy document called 'Social equity and job creation' (COSATU et al. 1996). The document was issued on behalf of the three major union federations. Like the business document, the labour document made many sound points, in this case from a left-Keynesian position. Had the two documents been the basis for an academic debate rather than a struggle for influence, it would not have been too hard to find common ground. But, following the publication of Growth for All, the tone of the debate was too strident, as if some participants saw it as a struggle for the soul of the ANC.

The South African Communist Party (SACP), a member of the ANC alliance, along with the Congress of South African Trade Unions (COSATU), also hit out at big business's now exposed underbelly. Jeremy Cronin, imprisoned for underground ANC work in the 1970s, a political philosopher and a talented poet, was Deputy Secretary-General of the SACP. In an op-ed piece in a national business daily, Cronin struck out at the underlying assumptions he saw reflected in Growth for All. He focused more on the subtext than the details. Growth for All, he felt, assumed that the market would solve South Africa's problems and that government was extraneous. He isolated unguarded comments that indicated insensitivity towards the long-oppressed sections of South African society and arrogance about the role of the business community. Growth for All, he implied, was out of place in the new South Africa (Business Day 24 April 1996). Gwede Mantashe, Assistant Secretary-General of the National Union of Metalworkers of South Africa (NUMSA), contrasted the approach of business in the Growth for All document with that of members of the business community then working in a Presidential labour market commission, trying to meet minds with government and labour (Business Day 22 April 1996).

In the midst of this rancorous debate, the currency plunged again, losing another 10% in value between late March and early May. Since the middle of February the rand had lost marginally over 20% (CREFSA 1996: figure 7). The government had to respond. It had to nail its economic colours to the mast as there was now a politician, not a banker at the helm of the fiscus; it also had to show onlookers that it was prepared to take charge – that it would not allow itself to be buffeted from side to side by the country's most powerful social forces: organised business and labour.

Hands on the wheel

The government had been working on an economic policy document. Government earlier asked the Central Economic Advisory Service (CEAS), an apartheid-era government agency that fell under the RDP office, to prepare a macroeconomic policy statement. The resulting statement was based on the Normative Economic Model (NEM) prepared for the De Klerk government in the early 1990s. The NEM was based on a conservative and static set of assumptions, particularly regarding the balance of payments. Like Stals, it failed to recognise the potential for capital inflows. The new document was irrelevant in a rapidly changing society. It was never published, and the CEAS was dissolved within a year of this failure.

The next effort was a project of the RDP office itself. This was undertaken as an interdepartmental project, to be co-ordinated and drafted by the RDP. One problem with this effort was that it grew almost unstoppably as each agency of central government ensured that its interests were represented. Once the decentralised drafts were collected, the RDP office rewrote them to no one's satisfaction. A summary paper – the 'draft National Growth and Development Strategy' was released in February 1996, but was downplayed by government.

A group of officials, clustered around the Department of Finance and the Development Bank, and including senior officials from the Departments of Trade and Industry and Labour, started working late in 1995 on a more focused macroeconomic policy document. Several academics were enlisted to test macroeconomic models, and World Bank economists who had worked on the South African model used in the 1994 World Bank paper were also drawn into the project. After Trevor Manuel was appointed, he adopted the project and instructed the co-ordinators to speed it up so that he could present it before the end of the debate on the 1996/97 budget. He intended that it should answer the questions that were being asked of him, offer a clear and coherent set of economic policies and strategies, and show that government was prepared to take charge rather than bob around like a cork in the ocean. This, he and the Cabinet hoped, would restore stability to the currency and related markets, amongst others the short-term capital traders who had been getting out of South African gilts from the middle of February.

The result was the GEAR strategy – its full name being 'Growth, Employment and Redistribution: A Macro-Economic Strategy'. The GEAR approach linked greater fiscal prudence to the ability to ease off a still-tight monetary policy and encourage private local and foreign investment. Government dissaving was a key target. The policy group believed that as long as the government debt grew rapidly it would put pressure on interest rates; conversely, escaping the debt trap and reducing the debt burden would allow interest rates to soften, leading to higher investment levels. As Trevor Manuel put it in his speech introducing GEAR to Parliament on 14 June 1996, '... in the recent past, the shortcomings of policy coordination has placed an excessive burden on monetary policy as the major instrument to maintain macro balances. This has led to higher real interest rates' (Manuel 1996: 7). Manuel and his team expected interest rates to fall with inflation under control and the government-borrowing requirement decreasing.

To this end, Manuel made the deficit reduction target more stringent than that announced by Chris Liebenberg in 1994. Liebenberg had set the target for bringing the fiscal deficit down from 6.8% of GDP in 1994–95 to 4.5% of GDP in 1999–2000. The expected deficit for 1996–97 was about 5.4% of GDP. Manuel, realising that this would not reverse government dissaving, certainly at current interest rates, set the new targets at 4% in 1997–98, 3.5% in 1998–99, and 3% of GDP in 1999–2000. Other fiscal reforms would include 'public service restructuring', which implicitly included revisiting an overly generous deal struck with the public service unions in 1995, revamping the budgeting process into a three-year rolling budget, some tax restructuring, and a major drive to improve revenue collection.

Manuel expected inflation to rise in 1997 as a result of the effects of the 1996 currency depreciation on the cost of imports, and he expected the real bank rate to fall from 7% in 1996 to 5% in 1997, and eventually to 3% in 1999 and 2000. In wording agreed to by the SARB, GEAR aimed to 'maintain the current competitive advantage created by the depreciation of the rand... [through keeping] the real effective exchange rate of the rand at a competitive level' (Department of Finance 1996: 10). This would be underpinned by further import tariff and exchange control liberalisation.

There were other elements in the GEAR strategy: new trade and industry programmes aimed at encouraging new investment through tax breaks and credit facilities; industrial innovation; small business development and export growth through a range of programmes; and to give teeth to the competition authorities. In the labour market, the government indicated its intention to improve productivity through better industrial training, to improve competitiveness through 'more flexible' labour markets, and to seek a national agreement to link wage growth to productivity growth. The government also indicated its intention to speed up privatisation, and to support more government investment, especially in municipal infrastructure, when spending on public servants' salaries and wages came under control.

The final element of the GEAR strategy was perhaps the most important – co-ordination of economic policy and implementation within government, and between government and its 'social partners' (business and labour). In words that foreshadowed an inevitable audit, the GEAR authors wrote:

Within government, especially in the fields of monetary, fiscal, trade, industrial and labour policies, there is also a critical need for coordination. Inconsistent approaches in any of these areas have the potential to destabilise the credibility of the overall macroeconomic framework (Department of Finance 1996: 21).

GEAR was a serious attempt at a co-ordinated strategy. All the key government departments implicated in the strategy, as well as the SARB, participated in the analytical process that preceded the strategy, and endorsed its outcome. The Deputy President, Thabo Mbeki, made an introductory speech underlining its importance. 'This policy', he said, 'is the central compass which will guide all other growth and development programmes of the government.' He also indicated that the macroeconomic strategy would be situated within the broader 'National Growth and Development Strategy' (Mbeki 1996: 137–140). The intention to locate GEAR within a broader growth and development strategy receded into the background until 2003, and debates on GEAR tended to assume that it was the alpha and omega of the government's economic policy.

Grinding gears

GEAR was generally well received by local and international business and, when it became clear that Manuel was serious about implementing GEAR, markets began to settle. Before the end of 1996 capital flow signals were strongly positive again (no doubt helped by the escalating real interest rate).

The left was less sanguine. Labour leaders were very concerned about the conservative fiscal stance of the strategy, and also interpreted the monetary policy proposals as too conservative. COSATU complained that the document made too many concessions to foreign investors without any guarantee that they would respond by delivering their capital. Mbazima Shilowa, General Secretary of COSATU, commented that COSATU and the SACP were finding it very difficult to deal with the ANC. Implying that they had not been properly consulted, he said, at a SACP seminar: 'It means something has gone terribly wrong that such a document is able to emerge and be on the table' (Sunday Times 7 July 1996).

Manuel may have provoked criticism from the left in his speech introducing GEAR when he said: 'The parameters are not up for negotiation at this stage.' Though he did add that the implementation of the programme rested on close co-operation with social partners (business and labour) and that 'there is undoubtedly so much of the detail which we will have to work through together' (Manuel 1996: 14), what remained with the ANC's allies was the perception of a closed door on substantive negotiations.

With the rand still unstable, Manuel probably felt that to show weakness of resolve was to reopen the currency speculation season. Government needed to show its willingness to take on the responsibilities of government.

Manuel's twin departments of Finance and State Expenditure moved fast and with remarkable success considering their limited resources. Though the 1996–97 budget deficit was a little higher than expected at 5.6% of GDP, in 1997–98 it fell to 4.4%. This exceeded the GEAR target of 4.0%, but slower-than-expected GDP growth had reduced government revenues, while rising interest rates were adding to the debt-servicing burden. Other finance reforms were set in motion too – particularly the reconstruction of the government's budgeting process.

The Department of Trade and Industry (DTI) also moved quickly, with several of its new programmes in place before the end of 1996, and others following the next year. But, in light of the fact that the rand was beginning to appreciate again and interest rates were rising, the DTI had second thoughts about accelerating an already dramatic tariff liberalisation programme. The Labour Department moved steadily towards an improved national skills framework, but more slowly on labour market flexibility and a national wage/price agreement. Privatisation moved forward, though government was often held up by the trade unions that were concerned about job losses and the delivery of services to the poor. A national framework agreement three years later improved the climate for privatisation.

A more serious case of uncoordinated policy, which punctured the GEAR strategy, came from the SARB. By late 1996, with the real bank rate at 9.5% (nominal: 17%), the currency had stabilised and international capital flows were positive. Inflation, at 7.5% for the year, was more than 2% better than the GEAR target and the fiscal deficit was coming down. It was time for the SARB to play its role in getting GEAR going through gently relaxing the real interest rate. As Maria Ramos argued, shortly before being appointed Director-General of Finance, while monetary policy should remain firm, after GEAR it no longer had to take the strain of keeping a grip on the economy alone (The Star 15 June 1996).

But interest rates continued to rise. Early in 1997, Stals explained his refusal to cut interest rates: 'the growing deficit in the current account had to be reversed'; there was an imminent 'danger of escalating inflation'; and the 'escalating growth rates in the domestic monetary aggregates... had to be reversed' (Stals 1997: 32). Though the SARB had participated in preparing GEAR, it was not comfortable operating within the GEAR targets, based on the GEAR model.

Taking it further, in March 1998 Stals referred to an expected annual inflation rate of 0–5% for the world as an indicator of South Africa's objectives, although he qualified this by saying that the range could not for now 'be accepted as a formal commitment or firm inflation target for the Reserve Bank' (Stals 1998: 38).

One result of the SARB's excessively tight monetary policy, which saw an increase in the real bank rate from 3% in 1994 to 13.7% in 1998, was that it choked off South Africa's expansion. Domestic investment and consumption tailed off, and even the growth of exports slowed due to the appreciation of the currency during 1997. Homebuyers had to repay mortgages with real interest rates of at least 11% during 1997 and up to 18% in 1998, for example. GEAR had planned for a slight depreciation of the exchange rate in 1997; instead the currency appreciated by at least 0.8%. Most of this preceded the interest rate hikes provoked by the Asian, Russian and Brazilian crises in 1998.

The other result of the high interest rate was that, as in 1995, it skewed the composition of the new inflow of foreign capital towards short-term money. Foreign investors could expect good returns for a limited period of time, as long as they got their exit strategy right to beat the next wave of depreciation. But the kind of foreign investors that would come for the long ride, who would look for steady growth and political and economic stability, might be put off by Stals' rather short-term approach to monetary policy. So, the boom-and-bust cycle of 1994–96 seemed likely to repeat itself again, except that the boom was smaller.

Stals' preoccupation with low inflation and a strong currency seemed to override the commitment of the SARB to co-operate with the GEAR strategy. His fixation was never more evident than in the middle of 1998 when the panic emanating from the Russian crisis hit South Africa. Stals intervened massively to try to retain the value of the currency at close to five rand to the US dollar. He did this largely through swaps in the forward currency market. As soon as traders became aware of this, they bet heavily against the rand. Stals lost, the rand fell by 20% to R6.70 to the dollar, and the real bank rate was forced up to 14%.

Why did Stals behave in this way? By the end of 1998 there was speculation in the South African media that one of Stals' main concerns was to prevent losses or even gain a profit on the SARB's forward book (Financial Mail 11 December 1998). Jonathan Leape also expressed concern that the size of the oversold book could influence the exchange rate strategy (Khatri et al. 1997: 33–38). But it is hard to imagine that a seasoned central banker such as Stals could sacrifice so much in the economy at large only to avoid excessive losses in the forward market by the SARB.

Another reasonable explanation is that Stals failed to grasp the effects of structural change on money market aggregates and on prices in South Africa. The redistribution of wealth and power since the early 1990s had a huge impact on South Africa, which will be dealt with in more detail in Chapter Six. This has meant a significant extension of credit facilities and opening up of a range of banking accounts that may well not have had an equivalent effect on consumer expenditure. The integration of the South African economy, new payments technologies, and the introduction of new forms of credit have also influenced the credit extension figures (see IDC 1998: 9). Even if credit expansion boosted demand, it had become far more difficult for manufacturers, wholesalers and retailers to exploit constricted markets or to pass on price increases to consumers. The most important structural change in this respect was the very significant decline in import tariffs on consumer products since 1995, which was preceded by the reduction of import surcharges in the early 1990s. South Africa was behaving far more like an open competitive market today than it did in the early 1990s. Had these structural changes been better understood by the SARB, monetary policy might have contributed more effectively to GEAR.

The performance of GEAR

What were the results of GEAR? The strategy included a set of projections indicating desired macroeconomic policy indicators, such as the fiscal deficit and average tariff rates, and indicators of outcomes, such as job growth and real export growth. In some areas performance exceeded the expectations of GEAR, which was measured as a five-year programme, ending at the end of 2000. The GEAR model aimed for a fiscal deficit down to 3% of GDP, government consumption down to 18.1% of GDP, and inflation (CPI) down to 6% per annum. All of these targets were beaten, with figures of 2.2%, 18% and 5.4% respectively. Some of the economic targets were more or less met, such as a real manufacturing export growth rate of around 10%.

The two significant disappointments were job growth and investment. The target for job growth was 2.9% growth in formal non-agricultural employment per year. Jobs grew at about 2.5% per year in the period 1995–2003, but more than half of these new jobs were in the informal sector (Statistics South Africa Labour Force Survey March 2003; Bhorat 2003; private correspondence with Ingrid Woolard, senior researcher at the Human Sciences Research Council in 2003). Investment performance was even more disappointing, especially in the later GEAR period: in 1999 and 2000 investment by private, parastatal and public sectors all shrank, instead of growing at around 10% per annum. Growth, not surprisingly, was also disappointing, averaging 2.6% over the GEAR period, instead of the projected 4.2%.

The tight monetary policy of the late 1990s, which was partially in response to the Asian, Russian, and Brazilian crises, and a 20% depreciation of the rand, certainly contributed to low investment and growth. Fiscal policy was probably also an important factor. It may well be that the positive effect of government reducing its borrowing (which had been crowding out private investment by raising the cost of capital) was more than outweighed by the effect of reduced government investment with its potential to crowd in private investment (Davies and Van Seventer 2004).

Or it may simply have been a matter of timing. The reduction in government borrowing may have been slower to take effect than the reduction in government investment. In fact, overall government debt only began to fall significantly as a percentage of GDP after 1999. One reason for the significant improvement of public finances at the end of the 1990s was the tightfistedness of the Minister of Finance, Trevor Manuel. The other was the effectiveness of the South African Revenue Services (SARS) under Pravin Gordhan, a lawyer and leading anti-apartheid activist in South Africa in the 1980s, who took over SARS in 1999. Public revenue collection improved to such an extent that Manuel could for several years, especially the period 2000–03, increase real government spending, reduce the budget deficit and reduce government debt levels as a percentage of GDP, while significantly cutting income tax rates. In this sense, the fruits of GEAR were only reaped from 2000. It was also in the early 2000s that private companies began to raise capital in the form of bonds, signalling that government crowding out had ended.

It was at this point, in 2000 that the fiscal stance began to move into a more expansionary phase. Before 2000, the most significant developmental interventions through the fiscus were noteworthy increases in expenditure on social services, largely health, education and social welfare. Expenditure on social grants rose from R10 billion in 1994 to R38.4 billion in 2003, and the number of beneficiaries grew from 2.6 million to 6.8 million. Targeted grants, such as the child support grant, were introduced. The education budget rose to make up a quarter of all government expenditure, with significant increases in enrolment in early childhood development and in secondary schools, where enrolment increased from 70% in 1992 to 85% in 2001, while the pupil-teacher ratio declined from 43:1 to 38:1. In health, the focus was on primary health care and related programmes (Policy Co-ordination and Advisory Services 2003: 16–21).

The extension of infrastructure services was a significant intervention too, though some of these were financed by public corporations, rather than the fiscus itself. The result was that, for example, the proportion of households with access to clean water grew from 60% in 1996 to 85% in 2001, while the number of homes with access to electricity increased from 32% to 70% over the same period (Policy Co-ordination and Advisory Services 2003: 24).

Infrastructure expenditure began to grow more rapidly after 2000. Between 1998 and 2001 government fixed capital formation fell for four consecutive years, but this began to turn around in the early 2000s. Since 2001, government and public corporation infrastructure spending has increased at a rate of about 6% per year, and this is expected to continue at least until 2006. The strongest growth is in infrastructure development by the provinces, which grew from R11 billion in 2001 to R18 billion in 2003 and is expected to grow to R24 billion by 2006. This indicates another significant fiscal trend – a steadily growing proportion of funds flowing to the provinces and to municipalities, for the provision of social services and infrastructure investment.

The GEAR period, 1996–2000, was supposed to see a simultaneous improvement in the management of public finances and investment response that would raise growth levels. Only the former happened. There are probably four main reasons why the growth performance of the GEAR era disappointed. The first was excessively tight monetary policy, especially in the 1996–98 period. The second was the macroeconomic crises in Asia, Brazil and Russia in 1998, which pushed South Africa to an even more conservative monetary policy. The third factor was that the very tight fiscal policy, partly a consequence of the need to reorganise government at national, provincial and local levels, reduced government investment expenditure. Lower level and microeconomic reforms proved considerably more difficult than the reform of the Treasury and trade policy. Finally, the private sector, both nationally and internationally, remained cautious about South Africa's future.

The GEAR strategy was not implemented in an entirely consistent and co-ordinated way, especially regarding monetary policy. It was also unrealistic, perhaps, about the timing of positive investment responses, and about the ability of the new government to develop effective institutions, plans and investment modalities. But criticism about the fundamental design remains debatable. GEAR certainly set the stage for a stronger performance in the new millennium.

Notes

1 The South African fiscal year runs from 1 April to 31 March.

2 Many people were recruited and promoted in the 1980s in a vain attempt to shore up the apartheid system with a black bureaucratic middle class.

3 Blinder thinks that even the United States law is too vague.

4 'The income elasticity of the demand for non-oil merchandise imports is 1.47 ... When domestic income increases, the percentage change in imports is considerably larger than the change in domestic income' (Smal 1996: 30).

5 Gross domestic savings as a percentage of GDP fell from 25.3% in 1983 to a low of 14.3% in 2001 (SARB 1998b: column 6286J).

Document(s) 5 of 11