In many parts of the developing world just the mention of the Bretton Woods institutions, especially the International Monetary Fund (IMF), brings fear and loathing, much of which is fully justified. The manner in which the Fund in particular has dismantled economies and societies in borrowing countries in the name of fiscal discipline and liberalisation has been well known in Africa and South America since at least the 1980s. So, as the transition to democracy in South Africa unfolded in South Africa in the late 1980s and 1990s, there were many, including one of the authors of this briefing, who were quick to write about what these Washington institutions stood for, and warned against the dangers of falling into their grip. In a paper first presented at the York conference on the post-apartheid economy, Padayachee noted that ‘the IMF, a supposedly apolitical international agency, can very well form an integral part of the formidable weaponry available to the West in shaping the emergence, nature and development of a post-apartheid South Africa’ (Padayachee 1988, 202). That sentiment echoes the view of Ronnie Kasrils, one of the chief architects of the notion of a ‘Faustian pact’, by which the African National Congress (ANC) and South Africa were understood to have been caught in a trap, pushed into selling the country’s soul to the forces of global capitalism and became a ‘prisoner of the neo-liberal global economy’ (Kasrils 2014, xxii). Central to the mechanism employed by the IMF in shifting ANC economic policy thinking according to Kasrils and others (see for example Saul and Bond 2014, 147–148) was the US$850 million (about SDR614 million)1 Compensatory and Contingency Financing Facility (CCFF) extended by the IMF to South Africa once the Transitional Executive Committee was established on 7 December 1993. The ‘conditionalities’ attached to that loan, it is argued by such analysts, pushed the ANC towards acceptance of the Washington Consensus package of economic policies of what is now known as the neoliberal kind.
While our whole political history and anti-imperialist thinking is closely associated with such a critical approach to the IMF, we argue here that there are many good reasons to cast doubt about whether that loan, and the terms and context under which it was approved, served such an immediate purpose at that time in contrast to the views reported above. Rather, whilst we do not doubt the fact of a sudden and extensive neoliberal turn, which can be dated towards the end of 1993, the role of the IMF loan in the policy shift is at most symbolic, not causal (Padayachee and van Niekerk 2019, Chapter 4).
Many commentators – both academic and in civil society – have attributed the shift in ANC economic policy to the role and influence of the IMF and World Bank, and in particular they cite the impact of the 1993 IMF loan to South Africa on the eve of democratic elections as a central and not just symbolic factor. John Saul and Patrick Bond (2014, 148) argue that both institutions were central to the battle for hearts and minds over economic policy, and they refer specifically to the US$850 million IMF loan to South Africa in December 1993. In a news report based on his latest book, Dr Alan Boesak, an anti-apartheid activist, clergyman and United Democratic Front leader makes the following claim:
this negotiated settlement was sold to the ANC by Western economists and institutions such as the International Monetary Fund and the World Bank, which argued for the adoption of a neoliberal capitalist system and property ownership model, warning that the alternative was an outdated communist model which would damage the economy. (Kgosana 2017)
These lines of argument echo that of the respected Stellenbosch academic economist Sampie Terreblanche (2012, 64) who, in the IMF’s Letter of Intent, sees an embryo version of the 1996 Growth, Employment and Redistribution (GEAR) programme adopted in June 1996 by the ANC-led Government of National Unity (GNU). So the question which arises is: are these commentators correct in attributing such policy influence to the IMF in general and to the 1993 loan in particular?
There can be little doubt that the IMF and World Bank would have used moral suasion to argue against any sort of socialist or communist model for a post-apartheid economy (that is after all to be expected), and they had both already become very active in South Africa in the early 1990s after a long absence.2 However, we argue that neither institution had a suitable vehicle to force a neoliberal policy platform down the throats of the ANC unless the latter had itself come to that view through other means. As we argue here, a combination of internal forces (including local business, the late apartheid state’s residual power and the views of some top ANC economic leaders) was more influential in the ANC’s rightward drift to, and embrace of, neoliberalism.
Typically, since at least the early 1980s, the IMF and World Bank played two key roles in influencing economic policy in developing countries. First, they both acted to some extent as knowledge or ideas banks, based on their ideology, their research and comparative experiences. Second, they influenced borrowing countries through what is referred to as ‘conditionality’ – the policy measures with which such nations had to comply as a precondition of receipt of loans. The ideological underpinnings of the IMF and World Bank have been described as ‘neoliberal’ by many on the left – and its conditionalities have consisted of a package of market-driven policies favouring liberalisation of trade and financial flows, fiscal conservatism, flexible labour markets and the like.
Through the IMF, South Africa under apartheid received its first loan from the IMF in 1957/58 and its last in 1981, which was repaid in full in 1987. Table 1 displays the country’s position with the Fund on 31 October 1993.
| Membership Status: Joined 12/27/45; Article VIII | ||
|---|---|---|
| General Resources Account: | ||
| SDR million quota | 1,365.40 | |
| Fund holdings of currency | 1,365.36 | |
| Reserve Position in Funda | 0.5 | |
| SDR Department: | ||
| SDR million net cumulative allocation | 220.36 | |
| Holdings | 4.17 | |
| Outstanding Purchases and Loans: | None |
Source: IMF staff report (1993) (adapted).
a. IMF member countries are assigned a quota which they have to deposit at the Fund, part of which is payable in SDRs or other widely traded currencies, e.g. the dollar, and part in the member's own currency. The difference between a member's total quota and the IMF's holdings of its national currency is called a country's Reserve Position.
The Fund returns to South Africa after 1990 initially as a relatively quiet and junior player (compared to a somewhat ‘noisier’ World Bank crowd) and mainly plays its role as a knowledge and ideas bank through some policy position papers. In 1992 it publishes its first major report on the country (Lachman and Bercuson 1992), focusing on the growth policies it believed were needed in the country. The report was both predictable and pessimistic. It recommended that external trade and financial relations should be liberalised (exchange controls to be abolished, albeit gradually, as its later pronouncements made clear) as a strategy to eliminate balance of payments deficits (Padayachee 1997, 30ff). Later, once sufficient consensus had in its view been reached about a more inclusive and democratic government in the negotiations, the IMF received an application through the Transitional Executive Council (TEC) for a Compensatory and Contingency Financing Facility (CCFF) of US$850 million (about SDR614m, as noted above). South Africa owed the IMF nothing at that point (IMF 1993, 1). The IMF staff report observed that the South African request ‘will relate to a shortfall in merchandise export earnings combined with an excess in cereal import costs for the year ending June 1993’ (1). The drought was specifically referred to as an exogenous underlying cause. In an earlier paper, Padayachee (1997, 32ff) made the following observations on the loan; ‘the loan was designed (purportedly) to support South Africa’s balance of payments following the decline in agricultural exports and the increase in agricultural imports caused by the prolonged drought.’ However, as Patrick Bond (2000, 68) has correctly pointed out, the drought had ended 18 months before, raising concerns about the rationale for the application. Even the IMF staff report acknowledged the end of the drought and commented that, ‘in light of the end of the drought and more favorable gold prices, the current account surplus is expected to increase from 1% of GDP to about 1.5% in 1993’ (IMF 1993, 8).
The IMF approved the loan on condition that a ‘legitimate body, gave its undertaking that the economy was responsibly managed’, in our view a low- or no-conditionality approval. The Letter of Intent which accompanied the loan facility warned in general terms against the dangers of increases in real wages in the private and public sectors, stressed the importance of controlling inflation, supported trade and industrial liberalisation, and repeatedly espoused the virtues of market forces over regulatory interventions. Ostensibly drawn up by the borrowing government (in this case the TEC) but normally drafted by the IMF, the TEC also promised to consider the introduction of ‘inflation targeting’ and a cut in the budget deficit to 6% of the GDP ‘within a few years’ (Padayachee 1997, 32). These were far from the usual stringent and quantifiable IMF conditionalities on longer-term loans with targets for key macroeconomic balances,3 and, indeed, the loan being very much of a temporary nature (repayment within 30 days) was not subject to weighty conditions. All these factors lead to the inevitable conclusion that the loan came without typical conditionalities though there are differing opinions on the subject. Sampie Terreblanche (2012, 64) argues that ‘the Letter of Intent committed the TEC to the ideologies of neo-liberalism and market fundamentalism.’
Why do we differ? It has to be noted that, if a country borrows within its SDR quota, i.e. that is if the loan could be accessed from its own reserves at the Fund (a reserve tranche drawdown), there is normally minimal or no conditionality. That was the case here. South Africa borrowed only 45% of its quota4 and, as no other obligations were due in 1993 (as confirmed in the IMF staff report), there would have been no or very minimal conditions in terms of IMF policy, as confirmed by Dr Cyrus Rustomjee, former special advisor to the deputy minister of finance and former South Africa representative at the IMF (email exchange 11 December 2017). In an interview with Robbie van Niekerk, one of us (Ben Fine) had already hypothesised ‘that the loan came without conditions. People think, oh, they took the IMF loan and therefore they had to implement the program, [but] the IMF loan came without conditions’ (interview, London, 30 June 2015).
The IMF historian James Boughton (2012, 692) argues from a more political angle that ‘entering into a stand-by arrangement with the IMF, with all of the attendant policy conditions, would have been unthinkable for the authorities as well as for the IMF.’ This unusually sympathetic stance (based on political factors, and unthinkable in IMF history) is clear from the tone of the appraisal set out in the 1993 IMF staff report. Here is the key paragraph:
South Africa is at the beginning of a fundamental transformation that is both inspiring and fraught with difficulties. It is inspiring because, at last, one is able to envisage the resources of the country – its enormous natural endowment, its wealth of entrepreneurship, and its sophisticated economic and financial institutions – being put to work to better the lives of all the people. It is fraught with difficulties because the economy is still largely mired in recession and is scarred by eight years of sanctions and decades of apartheid. It will take time to ameliorate these conditions and popular aspirations may well be inconsistent with the pace dictated by fiscal and external constraints. (IMF 1993, 18–19)
Our interpretation is shared too by the then deputy head of the ANC’s Department of Economic Policy (DEP) Tito Mboweni, later South African Reserve Bank (SARB) governor, who strongly refutes the view that the ANC sold out by accepting a loan with specific conditionalities:
The end result was that we agreed to that request by the then government to approach the IMF for the required balance of payments support called the Compensatory and Contingency Financing Facility (CCFF). There were no conditionalities attached since this was a soft loan as it were. However, we had to provide the IMF with a ‘Statement on Economic Policy’. The statement was fairly simple: assure the IMF that the future government will pursue prudent macroeconomic policies. This was something that the ANC in particular had adopted as an approach as early as 1992 in a document entitled – Ready to Govern. We did not sell out! … I emphasize that any careful reading of the statement will find no contradiction with Ready to Govern. But that is a matter for political economists to sort out in due course. (Mboweni 2004, 1–2, our emphases)
Mboweni’s view is confirmed in the IMF staff report (1993, 20) where it is noted that, ‘it is the staff's view that implementation of the policies described in the authorities’ “Statement of Policies” would constitute the maintenance of a prudent macroeconomic stance.’
So what was the origin of and logic for the application? The Letter of Intent was clearly in preparation as early as November 1993 well before the TEC began its work. We are now aware that an IMF staff mission visited Pretoria, Johannesburg and Cape Town during October 7–18 1993, ‘to conduct the 1993 Article IV consultation discussions and to discuss prospective access to Fund resources under the compensatory and contingency financing facility [CCFF]’ (IMF 1993, 1). This lends credence to the view that it was the National Party government which pushed for the loan (Mutize 2017) using the multi-party Economic Technical Committee (ETC) established by Finance Minister Derek Keys sometime in 1992 after his appointment. This is confirmed in the IMF staff report (1993, 3). It should be no surprise that Keys was an important player here. Patti Waldmeir’s (1997, 258) account is somewhat simplistic but she does make the important point that, ‘[Derek] Keys’s all party economic team, the Economic Technical Committee’ played a central role in persuading the parties of the value of such an application to the IMF (our emphasis). The ANC had finally bought into the world consensus on good economic government, and was applauded internationally for so doing’. She argues that the De Klerk regime ‘could not believe their luck at the economic transformation of their adversaries (though Keys ensured that they judiciously avoided crowing about it, which might have jeopardized the conversion).’5
On 5 November 1993, DEP Head Trevor Manuel requested comment from the Macro-economic Research Group (MERG) office6 on the draft IMF Letter of Intent, and especially its quantitative dimensions. Manuel sent the draft Letter of Intent to MERG via the National Economic Forum. John Sender’s cover letter in reply to Manuel, following discussion held with some MERG researchers who were based in the Wits Economics Department, is worth quoting in full, as it has never been referred to by any source before. Sender replied as follows:
I think the ‘quantification’ is less important than some of the ideological biases, or the theoretical straightjacket that Keys/IMF would like to lock the next government into. If you accept, or make explicit, the sort of theory they are pushing, then a wide range of policy options become closed.
I have tried to eliminate the phrases which suggest acceptance of inadequate and failed theory, as well as those phrases likely to cause trouble with COSATU. I don’t think it is worth fussing too much about the quantification on page 4, since it is qualified by words like ‘aim’, ‘about’, ‘projected’. There are all sorts of legitimate reasons why government’s aims might be frustrated and approximate, projected targets not actually be realised.
Finally, the best [original emphasis] method of negotiating on these letters is to call upon the services of powerful people in Washington, rather than negotiate with Lipshitz or delegations to S.A. Have you considered approaching the Clinton administration directly, or lobbying the IMF representatives of the Scandinavian, Dutch countries who could put pressure on the Fund at the Executive Board level …
Signed John Sender (Cover letter from John Sender to Trevor Manuel, 5 November 1993, accessed in the private archive of Vishnu Padayachee)
With the letter is an attachment of one and a half pages setting out the changes, deletions and additions that John Sender proposed.
We do not know if Manuel followed Sender’s suggestion to lobby likely supportive allies (though we doubt whether he did), nor if any of the contents of the letter was debated at the DEP. What we know is that the Letter of Intent was kept a secret for months until it was leaked to South Africa’s national daily newspaper Business Day in March 1994, a month before the elections. It only then became clear that none of the even relatively modest and pragmatic changes proposed by Sender made it into that final approved Letter of Intent. So, either the proposed changes were presented by the ANC/TEC to the IMF and simply ignored by the latter, or they were never presented to the IMF by the South African side as the basis for the negotiations. The Letter of Intent was signed off by the IMF and on the South African side by Derek Keys on behalf of the TEC, Dawie De Villiers on behalf of the National Party, and Pravin Gordhan on behalf of the Congress alliance (NIC/ANC) on 3 December 1993, and later ratified on 7 December 1993 when the TEC formally came into effect.
Patrick Bond (2000, 155ff; and 2016) was working as the coordinator of the Reconstruction and Development Programme (RDP) at the time, and he remarks that the IMF’s Managing Director Michael Camdessus put intense pressure on the TEC to ensure the reappointment of Finance Minister Derek Keys and Reserve Bank governor Chris Stals. We know that the Mandela government acceded to this apparent pressure so, in this respect, the Fund may have got its way. However, in the course of other research (Padayachee and van Niekerk 2017 and 2019, forthcoming), we were informed by a very authoritative source on condition of strict anonymity that Mandela had expressed a wish that Stals stay on as governor of the SARB after the latter’s term ended in August 1993 and that this informal proposal occurred in the course of a chance meeting between Mandela and Stals as early as 1991.
If we are correct in our interpretation, it suggests that the 1993 IMF loan was not really needed for balance of payments reasons, but rather that it was a signal of the IMF’s ‘stamp of approval’ for South Africa’s democratic transition and its economic policies, in effect encouraging private international lending, trade and investment to the country. That softer line, so atypical of the IMF’s approach to other developing countries at the time, appears to have informed the IMF approach to Mandela’s ANC. At that critical stage in South Africa’s transition to democracy, such a stamp of approval must have been regarded as important to Keys and the late apartheid government, even to some of the incoming ANC leadership. James Boughton makes the point clearly:
Major creditor countries recognized the global political importance of aiding the transition to democracy, but they also knew they had to have some assurance that committing money to the country would help and would not be wasted. The IMF, with its technical expertise and its regular dialogue with the authorities, was the natural conduit for determining when and how the international community could safely resume lending to South Africa. (Boughton 2012, 692)
But the ANC concern about potential IMF and World Bank interference with its sovereignty comes to the fore again once this stage was over. Effectively in power after April 1994, the ANC leadership was able to push its own policy stance and resisted all early attempts by the IMF managing director Michael Camdessus to persuade the new government to accept more of its loan facilities. University of Cape Town Business School lecturer Misheck Mutize takes this view on the ANC’s stance:
When the African National Congress (ANC) came to power after the elections in April 1994 it walked away from the IMF offer. Its concern was mainly that the IMF would undermine the sovereignty of the newly established democracy by imposing inappropriate, policy choices that would have further harmed poor people. (Mutize 2017)
Boughton makes the same point as follows:
When the new government showed no inclination to seek the Fund’s help, [Camdessus] sent Alassane Ouattara (Deputy Managing Director)7 to Cape Town in February 1995 to discuss the matter with Mandela and other officials, but Ouattara found that much of the leadership was still hostile to the IMF. A visit by First Deputy Managing Director Stanley Fischer in December was no more productive. Bitter memories persisted of the Fund’s past support to the minority government, and many political leaders believed that IMF lending would come with unacceptable restraints on economic policies and would threaten the country’s sovereignty. (Boughton 2012, 695)
After the GEAR programme was in place, and following a visit by IMF MD Camdessus to Mandela’s home in October 1996, another IMF loan looked a better than even bet. Why would the MD himself visit Mandela unless something big was in the offing? But while the world awaited the formal announcement of an IMF loan to South Africa, the ANC revealed its hand by vetoing any new borrowings from the Fund (Boughton 2012, 698).
From Table 2 it appears that democratic South Africa repaid the CCFF loan in full in two equal instalments in 1997 and 1998 – which was clearly a delayed payment – as CCFF ‘repurchases’ (repayment in normal parlance) at the time had to be made within 30 days. Such delays are typically recorded as such, but no such ‘default’ record can be found in this case, clearly a case of quiet condonation.8
| General Resources Account | Poverty Reduction and Growth Trusta,b | Total | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Purchases | Loans | Purchases and loans | |||||||
| Year | Disbursements | Repurchases | Charges paid | Disbursements | Repayments | Interest paid | Disbursements | Repayments | Charges and interest paid |
| 1999 | 0 | 0 | 462,100 | 0 | 0 | 0 | 0 | 0 | 462,100 |
| 1998 | 0 | 307,215,000 | 11,333,206 | 0 | 0 | 0 | 0 | 307,215,000 | 11,333,206 |
| 1997 | 0 | 307,215,000 | 24,476,971 | 0 | 0 | 0 | 0 | 307,215,000 | 24,476,971 |
| 1996 | 0 | 0 | 27,295,661 | 0 | 0 | 0 | 0 | 0 | 27,295,661 |
| 1995 | 0 | 0 | 33,020,481 | 0 | 0 | 0 | 0 | 0 | 33,020,481 |
| 1994 | 0 | 0 | 26,723,328 | 0 | 0 | 0 | 0 | 0 | 26,723,328 |
| 1993 | 614,430,000 | 0 | 0 | 0 | 0 | 0 | 614,430,000 | 0 | 0 |
| 1988 | 0 | 0 | 86,579 | 0 | 0 | 0 | 0 | 0 | 86,579 |
| 1987 | 0 | 397,500,000 | 16,520,044 | 0 | 0 | 0 | 0 | 397,500,000 | 16,520,044 |
| 1986 | 0 | 347,500,000 | 42,213,452 | 0 | 0 | 0 | 0 | 347,500,000 | 42,213,452 |
| 1985 | 0 | 0 | 52,126,186 | 0 | 0 | 0 | 0 | 0 | 52,126,186 |
| 1984 | 0 | 0 | 38,308,471 | 0 | 0 | 0 | 0 | 0 | 38,308,471 |
Source: IMF n.d.
a. Includes loans under the Structural Adjustment Facility and Trust Fund.
b. Formerly Poverty Reduction and Growth Facility and Exogenous Shocks Facility Trust.
It has been a somewhat harder task to track down whether (and despite its concerns) the ANC-led government in fact drew down on the US$850m CCFF facility in support of strengthening its balance of payments and (if it did) in what ways did it utilise the loan. It has proven incredibly difficult to come to a definitive answer to this question. Cyrus Rustomjee argues that it may have been drawn down but he is not certain (email exchange 11 December 2107). Many officials at both the National Treasury and the SARB initially claimed (embarrassingly) not to remember whether it was drawn down or not. After further investigation and many leads and dead ends, we received an email on 10 December 2017 from Johan Krynauw, a long-standing official of the Treasury and programme manager at the National Treasury’s Public Debt Management division who confirmed our suspicion and hunch. Speaking of developments after 1994 he confirms that ‘I can’t remember a time where the National Treasury of South Africa ever drew down on an IMF facility.’ The loan was repaid in eight quarterly instalments over two years, totalling just over SDR307,000 in each year (1997, 1998), though the table shows two equal payments, which must have been simply an aggregation of the payments on an annual basis.
This official confirmation strengthens our argument that the IMF loan was no more than a stamp of approval, pushed by Keys and the National Party, arguably with good reasons from their point of view, that it came with no or very low conditionality, and that it had little or no effect on the economic policy positions of the ANC or the ANC-led GNU. This conclusion, if correct, punches a hole in the (Faustian pact) argument of those claiming that the ANC was pushed towards neoliberalism by the stringent ‘conditionalities’ imposed by the IMF on the 1993 CCFF (although this may have been conveyed to, and wrongly believed by, ANC participants in negotiations at the time, let alone as ex post justification for the, or their, neoliberal turn). In fact, there is a whole ‘industry’ of opinion that has blindly supported such a Faustian pact narrative about how ANC economic policy thinking was shaped in which the IMF and the 1993 loan are central.
So here are our somewhat ironic conclusions:
The ANC leadership did not take or make use of the loan, and certainly further loans, for fear of loss of sovereignty or of a more or less popular backlash against the leadership as well as for ideological reasons reflecting the past (even though the present was dominated by IMF [and World Bank] thinking and even personnel).
Yet the ANC leadership was persuaded (by internal forces including local capital and the late apartheid state) of the need for a (self-imposed) austerity programme, very much in the IMF tradition if not pressed hard by the IMF itself (Padayachee and van Niekerk 2017).
The end result was the worst of all possible worlds – IMF-type policies without IMF financial support!
In the latter respect, at least, South Africa might be seen as an early, extreme and leading example of what the IMF can be critically interpreted to mean by country ownership of policy-making (doing what the IMF would do without being told to do so and, in South Africa’s case, with the added bitter pill of not being paid to do so). Indeed, even at this early stage, more important than explicit or even implicit conditionalities and influence exerted by the IMF through access to, and use of, its funds, may have been its role in sustaining and consolidating centralised and authoritarian forms of economic policy-making, with Treasury (later complemented by the Presidency) to the fore. Corresponding modes of policy-making under the Mandela/Mbeki presidencies more than prepared the way for state capture characteristic of Jacob Zuma’s disastrous 10-year administration (Public Protector South Africa 2016; Kasrils 2017; and Bhorat et al. 2017).