This Selected Issues paper highlights that aided by the easing of political uncertainties after the national elections in early 1994 and by the cautious stance of policies adopted by the new administration, economic performance and investor sentiment in South Africa strengthened markedly. Nonagricultural value-added grew by 4 percent in 1995, led by a sharp increase in real gross private fixed investment. In contrast, developments in 1996 were characterized by a shift in investor sentiment and unrest in the foreign exchange markets.


This Selected Issues paper highlights that aided by the easing of political uncertainties after the national elections in early 1994 and by the cautious stance of policies adopted by the new administration, economic performance and investor sentiment in South Africa strengthened markedly. Nonagricultural value-added grew by 4 percent in 1995, led by a sharp increase in real gross private fixed investment. In contrast, developments in 1996 were characterized by a shift in investor sentiment and unrest in the foreign exchange markets.

South Africa: Basic Data

article image
Sources: Data provided by the South African authorities; IMF International Financial Statistics; and Fund staff estimates.

Excludes statistical discrepancy.

Nonagricultural sector.

Includes errors and omissions of current as well as capital account.

Fiscal year beginning April 1, cash basis, excluding special receipts and payments.


A. Introduction

1. Aided by the easing of political uncertainties after the national elections in early 1994 and by the cautious stance of policies adopted by the new administration,1 economic performance and investor sentiment strengthened markedly. Nonagricultural value-added grew by 4 percent in 1995—led by a sharp increase in real gross private fixed investment—inflation, long-term yield differentials, and the net open forward position of the South African Reserve Bank (NOFP)fell considerably,2 and the rand appreciated by 5 percent in real effective terms in the second half of the year. However, private nonagricultural employment plateaued in mid-year and fell thereafter (Figure 1).



Citation: IMF Staff Country Reports 1997, 082; 10.5089/9781451840933.002.A001

Source: South African Reserve Bank.1/ Increase means appreciation.2/ Consumer prices excluding VAT, food, and mortgage interest payments.

2. In contrast, developments in 1996 were characterized by a shift in investor sentiment and unrest in the foreign exchange markets. These were brought about initially by unfounded rumors about the President’s health and speculation regarding a wholesale removal of exchange controls on residents. It was subsequently reinforced by growing concerns about the course of financial and structural policies. The turnaround in investor sentiment was reflected in the depreciation of the rand—by some 21 percent in nominal effective terms and 15 percent in real effective terms over the twelve-month period to end–December 1996 (Box 1 and Figure 2). In this context, economic growth slowed, inflation rebounded, and the fall in private nonagricultural employment continued. In an effort to rebuild confidence, the authorities announced the Growth, Employment and Redistribution (GEAR) strategy in June 1996 (Box 2).



Citation: IMF Staff Country Reports 1997, 082; 10.5089/9781451840933.002.A001

Source: Reuters news service.

B. Real Sector Developments

3. Real GDP grew by 3.1 percent in 1996, buoyed in the main by a rebound in agricultural output (see below), moderately below the 3.4 percent growth in 1995. The expansion in real GDP exceeded the annual population growth rate, which is about 2 percent, for the third consecutive year. All components of domestic demand registered relatively robust expansions, except for inventories which fell markedly (Appendix Table 1). Real public consumption grew by 5 percent in 1996, compared with 0.3 percent in 1995, reflecting increases in both public sector employment and wages. In contrast, the increase in real private consumption slowed to 3.8 percent in 1996, down from 4.7 percent in 1995, perhaps because the uncertainties that prevailed in the exchange market led to a more cautious attitude among consumers. In particular, real growth in consumption of services and non-durables slowed considerably in the second half of 1996, whereas consumption of durables remained robust throughout the year.

The Depreciation of the Rand During 1996

The depreciation of the rand during 1996, after trading in a range of 3.60–3.65 per U.S. dollar for two years, was episodic, with periods of stability interspersed with sharp declines (Figure 2).

The perceived inconsistency between the real effective appreciation of the rand in the latter part of 1995 and the rapid growth of the monetary and credit aggregates, coupled with signs of a widening external current account deficit gave rise to concerns that the rand had become overvalued. A tightening of conditions in international financial markets and periodic speculation of accelerated exchange control liberalization on residents compounded these pressures both ahead of and following the initial depreciation. Political uncertainties also contributed to the initial fall, as well as the subsequent declines in April and early May: in mid-February, the President’s health was (erroneously) thought to be in doubt; in late March, an ANC member was appointed Minister of Finance for the first time; this was rapidly followed by a sharply polarized debate about economic policy between representatives of labor and business; and in May, it emerged that the National Party would leave the Government coalition. But at each stage after these particular uncertainties were resolved, the rand failed to fully recover, suggesting that more fundamental factors may also have been at work.

Pressures on the rand were reinforced by growing market skepticism about the authorities’ intention to halt the depreciation through a forceful tightening of financial policies. Although the bank rate was raised by 1 percentage point to 16 percent in late April, following a particularly sharp decline in the rand, the primary policy response was heavy official intervention in the forward and spot exchange markets.

The presentation of the Government’s medium-term framework of “market- friendly” policies in mid-June, the Growth, Employment, and Redistribution strategy (GEAR), provided a brief respite. The rand fell again from mid-July, when apparent inactivity to implement the GEAR resuscitated doubts about policy, and it fell again in late October, when continued evidence of unexpectedly strong monetary and credit expansion overwhelmed the more positive evidence that the government budget deficit for 1996/97 remained on track. Neither further heavy intervention at that point, nor an additional 1 percentage point increase in the bank rate to 17 percent in November, were sufficient to stem or reverse these pressures decisively. By the end of 1996, the NOFP rose to US$22 billion, about US$15 billion higher than in mid-February 1996.

In early 1997, however, the rand recovered strongly as confidence in the authorities’ fiscal and privatization plans rebounded, evidence of an improvement in the external current account deficit emerged, and capital flows to emerging markets strengthened.

The Growth, Employment, and Redistribution Strategy (GEAR)

The GEAR was presented to Parliament on June 14, 1996. It firmly reasserted the authorities’ market-oriented and consensual approach to policy. Referring to projections reflecting current policy, it underscored the urgency of further action to raise labor-intensive economic growth to address inherited social inequities and the “unemployment crisis.” It rejected the introduction of a national minimum wage and an expansionary fiscal policy to achieve these goals, and saw the real depreciation of the rand as growth- enhancing, if sustained by a firm anti-inflationary monetary policy. The GEAR targeted real GDP growth of 6 percent in 2000 and the creation of more than 400,000 jobs in that year, and offered five new sets of policy proposals—in addition to existing commitments to privatization and the reorientation of public spending to the poor—to achieve those goals:

  • Faster fiscal consolidation, targeting a deficit of 4 percent of GDP in 1997/98, falling to 3 percent of GDP in 1999/2000—½ percentage point of GDP lower than previously targeted for each year;

  • A labor market policy of “regulated flexibility,” strengthening regulatory protection for workers, scaling down wage floors for trainees and to reflect regional labor market conditions and firm size, broadening official powers to withhold the extension of wage agreements negotiated at sector level, and introducing a payroll levy to fund training;

  • Accelerated tariff reductions to reduce imported input and consumption goods prices, while avoiding job losses in sensitive sectors;

  • An accelerated depreciation scheme for new investments in manufacturing, and a tax holiday for new projects in approved areas and industries, based on the share of remuneration in value-added;

  • Further exchange control liberalization in the context of a gradual process: the cap on foreign assets obtained through swaps was raised from 5 percent to 10 percent of total assets for eligible institutions, institutional investors were allowed to place 3 percent of net inflows during 1995 in foreign assets, and restrictions on foreign investor access to domestic credit were eased.

The GEAR’s projections reflecting strengthened policies are as follows:

article image

4. Real gross fixed investment increased by 6.8 percent in 1996, compared with 10.3 percent in 1995 (Appendix Table 2). Private investment grew by 6.7 percent—about half of the increase realized in 1995, as activity was adversely affected by economic uncertainty and slower growth in sales of manufactured goods. Investment by public corporations also moderated in 1996, largely owing to a scaling down of capital programs. However, investment activity by public authorities witnessed a strong recovery from a decline of 7.7 percent in 1995 to an increase of 4.4 percent in 1996, as expenditure on projects under the Reconstruction and Development Program (RDP) picked up (see Section I.C). With the exception of the agricultural sector, the buildup in inventories slackened markedly across all sectors, possibly reflecting the large accumulation that had taken place in previous years and higher interest rates. Meanwhile, as part of the GEAR, the authorities instituted measures with effect from July 1996, that are expected to boost private investment. These measures include an acceleration of depreciation allowances for income tax purposes on plant and equipment and on buildings intended for manufacturing activities, and a tax holiday of up to six years for new projects with assets in excess of R 3 million (see Section I.C).

5. Notwithstanding an improvement in government savings, gross domestic savings fell from 16.8 percent of GDP in 1995 to 16.5 percent in 1996, as the savings performance of both individuals and corporations deteriorated. Net foreign savings contracted from 2.1 percent of GDP in 1995 to 1.6 percent in 1996 (Appendix Table 3), reflecting in part a strong performance of the export sector, which benefited from the increased output of the “mega investment projects” (see Section I.E), the rebound in agricultural production, and the depreciation of the rand. At the same time, the slowdown in manufacturing activity resulted in a slower expansion in imports of intermediate goods.

6. On the supply side of the economy, the main source of growth in 1996 was the recovery in agricultural production. Real value added in this sector, which had contracted by 15 percent in 1995 on account of a severe drought, grew by 26 percent in 1996 aided by exceptionally good weather conditions (Appendix Table 5). Although the agricultural sector contributes only 5 percent of total value added, the sector accounted for nearly half of the growth in real GDP at factor cost.

7. The expansion of nonagricultural production registered a marked slowdown in 1996. Real output in the mining sector continued to fall, as the increase in diamond and coal production was more than offset by a decline in gold production, which was affected by outbreaks of ethnic violence, underground fires, and seismic activities in some of the larger mines. Manufacturing output grew by only 0.4 percent in 1996, compared with 7.6 percent in 1995, owing in part to the slowdown in private consumption and cutbacks in inventory accumulation. The tertiary sector also expanded in 1996 at a slower pace than in 1995. In particular, growth of value added in the commerce sector slackened from 6.5 percent in 1995 to 3.7 percent in 1996, possibly reflecting the increased cost of credit, and growth weakened in the transportation and communication sector as well, notwithstanding increased demand for transportation services arising from the good agricultural season. However, the expansion in the financial services sector remained strong, boosted by the high activity in the foreign exchange, equity, and bond markets.

8. Notwithstanding the further expansion in economic activity in 1996, total employment in the formal nonagricultural sector declined in the first three quarters of the year, after a modest 0.7 percent advance in 1995 (Appendix Table 8). Formal nonagricultural employment fell by some 80,000 (1.7 percent) bringing the total loss in jobs since the peak in 1989 to more than 400,000.3 The loss in jobs in 1996 came entirely from the private sector, which registered a decline of more than 100,000 positions, despite a 2 percent gain in employment (about 6,700 positions) in the services sector. Manufacturing—which is the largest provider of private sector employment—lost some 55,000 positions, and the continued contraction of the mining industry led to a loss of about 19,500 jobs. The drop in employment in 1996 reflected in part the slowdown in growth of nonagricultural output and the persistent rise in real wages (see below). In contrast, public sector employment rose by some 22,000 positions, despite the separation of about 40,000 civil servants at the central and provincial government levels in the context of the three-year civil service rationalization program.

9. The new Labor Relations Act came into effect in November 1996, although the bill was passed by Parliament a year earlier. Its implementation was delayed by the need to put in place the supporting institutions. The Act establishes the basis for a centralized bargaining process on matters pertaining to conditions of service of employees, while protecting them from dismissal in the event of strike action. Under the Act, a Commission for Conciliation, Mediation and Arbitration (CCMA) and a Labor Council have been established. Since then, the CCMA has reportedly contributed significantly to a reduction in the number of man-hours lost through strikes. In the first two months of its existence, 601 disputes were referred to the body for conciliation; 401 of them were settled outright and the balance required additional review work. Meanwhile, the government released in early 1997 the Basic Conditions of Employment bill, which is intended to replace the current Basic Conditions of Employment Act. The bill also seeks to cover the conditions of service for nonunionized workers in the agriculture and domestic service sectors, and part-time workers, who are not effectively covered by the Labor Relations Act. The proposed basic conditions of employment include a reduction in the maximum hours worked per week from 46 hours to 45 hours, an increase in the overtime pay rate from 1⅓ to 1½ times the basic rate, and a three-weeks paid leave. The bill will be discussed at the National Economic Development and Labor Council (NEDLAC) by representatives of labor, business, and government before its presentation to Parliament.

10. After moderating in 1995, the increase in nominal remuneration per worker accelerated in 1996, influenced by the pick up in inflationary pressures, an improvement in average labor productivity, and high gross operating surpluses of business enterprises. Average nominal wages in the nonagricultural sectors rose by 10.4 percent (on a 12-month basis) through September 1996, compared with a 9.4 percent increase for 1995 as a whole (Appendix Table 9), while real wages increased on an annual basis by about 1 percent in both 1995 and the first three quarters of 1996. Real wages in the public sector have fallen over the last 2 years, whereas real wages in the private sector increased by 2.4 percent in 1995 and by 1.2 percent (on a 12-month basis) through September 1996. Despite the increase in real remuneration per worker, real unit labor costs continued to fall during 1995 and the first three quarters of 1996, as the general upward trend in real wages was accompanied by higher labor productivity: the annual growth in real output per worker was 3.3 percent in 1995 and 2.3 percent in the first three quarters of 1996. The productivity growth has largely been the result of the continued shift to capital intensive technologies, and the attendant fall in employment.

11. Following annual inflation rates close to 10 percent a year during 1992–94, inflation fell markedly in 1995 and early 1996, declining to 5.4 percent by April 1996. However, consumer price increases accelerated again in the second half of 1996 and early in 1997, and the 12-month inflation rate rose to 9.4 percent by the end of 1996 and to 9.9 percent by April 1997. The upward trend in inflation over the last year can be attributed to a number of factors. First, the depreciation of the rand contributed to increases in the prices of imported goods, although the 9.7 percent rise in these prices during 1996 was significantly below the 21 percent nominal effective depreciation of the rand that took place during the period, as many importers shaved margins to protect their market share. Second, as indicated below, despite steps taken during the year to tighten credit conditions, broad money and bank credit to the private sector continued to expand strongly. Third, the surge in average nominal remuneration led to a resumption in the upward movement in nominal unit labor costs in the second half of 1995, which persisted in 1996.

C. Public Finance Developments

Developments through 1995/96

12. After peaking at 8.6 percent of GDP in 1992/93, the deficit of the central government declined to 6.7 percent in 1993/94, falling further to around 6 percent in 1994/95 and 1995/96 (Appendix Table 11).4 Government ordinary revenue increased by a cumulative 0.9 percent of GDP during 1994/95 and 1995/96, to 25.3 percent in 1995/96 (Figure 3; Appendix Tables 12 and 13). Virtually all of this increase can be attributed to larger collections of personal and corporate income taxes, as customs receipts rose only marginally—offsetting the decline and eventual elimination of the import surcharge—and revenue from the VAT, excises, and the fuel levy remained relatively constant in terms of GDP. Over the same period, government expenditure fell by 0.4 percent of GDP to 30.5 percent of GDP in 1995/96. An increase in interest payments was more than offset by a large decline in capital expenditure to just 1 percent of GDP in 1995/96. This decline reflected in part a slow rate of expenditure under the Reconstruction and Development Program (RDP), and was the major reason for the increase in the stock of “rollovers” from R 1.5 billion at March 1993 (0.4 percent of GDP) to R 8.9 billion by March 1996 (1.8 percent of GDP; the latest date for which data are available).5 Transfers to the provinces were around 14½ percent of GDP, representing 60 percent of non-interest central government expenditure.

Figure 3
Figure 3


(In percent of GDP)

Citation: IMF Staff Country Reports 1997, 082; 10.5089/9781451840933.002.A001

Sources: Department of Finance; South African Reserve Bank; and Fund staff estimates.1/ Fiscal year ending March 31. Data for 1990/92 are preliminary, budget data for 1997/98.2/ National expenditure including transfers to provinces, but excluding extraordinary transfers.3/ Central government only; including revenue which Is subsequently transferred to the provinces; excludes revenue from sale of strategic clocks of oil.4/ Includes secondary tax on corporations and corporate tax on gold mining.5/ Total expenditure by national and provincial governments, including stocks of rollover funds.6/ Data before 1992/93 not available. Actual data to 1995/98, and budget data for 1998/97–1997/98.7/ Excludes Wages, interest payments, and business subsidies.

13. The RDP fund was designed to mobilize resources for special initiatives aimed at redressing inherited inequities. The fund was financed by a reduction of planned allocations to spending departments, and funding was to increase by R 2.5 billion annually from 1994/95 through 1998/99. In the event, significant capacity constraints led to large increases in rollovers. These constraints, which are being addressed, included weak management capacity, cumbersome tender procedures, lack of experience with private-public partnerships, and insufficient coordination. The RDP Office was abolished in mid-1996, and the various projects became part of the regular ministry budgets; the RDP fund is now used only for channeling international financial assistance to earmarked projects and programs.

14. Data on economic and functional classifications are only available for total expenditure of the national and provincial governments (Appendix Tables 14, 15, and 16), and include that financed by own revenue of the provincial governments (which is small and is mainly derived from motor vehicle and other licenses). While these data need to be interpreted with care,6 they indicate that total cash expenditure of the central and provincial governments declined by a cumulative 0.9 percent of GDP during 1994/95 and 1995/96 to 31.3 percent in 1995/96.

Developments in 1996/97

15. The “headline” deficit of the central government for 1996/97 rose to 5.3 percent of GDP from 4.9 percent in 1995/96. This was 0.2 percentage points of GDP above the budget target, as higher-than-projected revenue was more than offset by higher spending, which accelerated sharply toward the end of the fiscal year. When this “headline” estimate is adjusted to exclude extraordinary receipts such as the proceeds from the sales of strategic oil stocks, the deficit outturn for 1996/97 is 5.6 percent of GDP, and when accrued rather than cash interest payments are included, the deficit was 6.2 percent of GDP, compared with 5.8 percent in 1995/96. The higher-than-budgeted deficit largely reflected the much needed increase in RDP expenditure.

16. Central government ordinary revenue was 0.8 percent of GDP higher than in 1995/96 and 0.6 percent of GDP higher than budgeted. This increase was mainly due to a surge in corporate income taxes, reflecting the rise in corporate profitability and unanticipated higher receipts from the Secondary Tax on Corporations, following the halving of its rate to 12.5 percent. Personal income taxes were buoyed by collections under the third and final tax amnesty,7 while collections under the new 17 percent tax on retirement income were somewhat below target. Higher customs revenue from imports of motor vehicles was offset by lower VAT receipts because arrear collections were less than anticipated. However, the overall target of 0.3 percent of GDP for collections of all types of tax arrears through improvements in tax administration was exceeded by a wide margin. Receipts were also boosted by nontax revenue, especially from interest earnings on government deposits, owing to higher interest rates than had been expected. Extraordinary revenue, resulting form the sale of the strategic oil stocks was in line with the budget.

17. Central government expenditure increased by 1.2 percent of GDP in 1996/97 to 31.7 percent of GDP, or 0.8 percent of GDP higher than budgeted. The increase in interest payments by 0.6 percent of GDP was in line with projections, and reflected in part the take over of the former debts of the four TBVC homelands by the central government. Other current expenditure fell by 0.9 percent of GDP compared with an anticipated fall in current expenditure of 1.5 percent of GDP in the budget; transfers to the provinces declined by 0.2 percent of GDP or half of what had been anticipated in the budget because some provinces overran their original allocations and were provided with additional transfers in the Adjustments Budget. Declines in the central government’s own current expenditure made up the balance of the adjustment (a disaggregation by type of expenditure is not yet available). Capital expenditure rebounded—rising from 1.0 percent of GDP in 1995/96 to 2.6 percent in 1996/97—reflecting in part the pick up in RDP expenditure.

18. As noted above, the economic and functional classifications of the 1996/97 Adjustments Budget are not yet available. Comparing instead the 1996/97 budget with the 1995/96 budget, the data for the central and provincial governments indicate that outlays on wages and salaries increased sharply, with some offset in current transfers to businesses, households, and other units of government. The increase in wage and salary spending reflected the three-year agreement on civil service conditions that became effective in July 1996, which included a rise in wages of salaries of R 8.5 billion (1.5 percent of GDP) in the first year of the agreement and R 6.5 billion in each of the following two years. In addition, the agreement provided that any savings that arise from the “rightsizing” program would be used for further wage increases for those that remain in the civil service (see Section III).

19. Two tax incentive schemes were introduced during 1996 in the context of the GEAR. The accelerated depreciation allowance provides for straightline depreciation for new manufacturing plants and equipment over three years (rather than five) and for buildings over ten years (rather than 20) that are installed between July 1996 and September 1999. The income tax holiday can be granted for two years for each of three criteria that are met by new manufacturing investments valued at more than R 3 million starting from October 1996. The criteria are labor absorption (based on the value added devoted to wages and salaries), regional location (based on the assessment of regional potential and need for development, as determined by the Department of Trade and Industry), and industrial policy (based on priority industries that are determined by the DTI, using criteria such as labor intensity, industries with strong economic linkages, critical industries affected by the tariff reform, and global potential).

20. As in previous years, the 1996/97 deficit was in the main financed through the issue of long term bonds (Appendix Table 17). These bonds were sold with an average coupon rate of 12.83 percent, compared with a weighted average yield of 15.87 percent. The consequent discount on the sale of bonds amounted to R 6.7 billion.8

21. Total debt of the central government (at face value) has increased rapidly in recent years, rising from 39.5 percent of GDP in 1991/92 to 55 percent in 1996/97 (Appendix Table 18). Of this increase, 10.4 percentage points reflect the discount on the sale of government bonds, as described above (3.4 percent), the takeover of the overdraft debts of the former TBVC homelands that had been guaranteed by the central government and that are being gradually switched to Treasury bill debt (2.4 percent), the recapitalization of part of the actuarial underfunding of the public pension fund (1.7 percent),9 and the transfer to the South African Reserve Bank (SARB) to cover the losses in forward exchange operations (2.9 percent).10

22. As at March 31, 1996, contingent liabilities of the central government stood at R 134.8 billion (27 percent of GDP in 1995/96). The largest components were the underfunding of the government pension fund (R. 54.4 billion) and the government guarantees (R. 66.5 billion). Of the latter, R 55.5 billion were guarantees on public enterprise debts, mainly for the transport company Transnet (R 25.2 billion), the telephone company Telkom (R 10.0 billion), and the electricity company Eskom (R 6.2 billion). Revised guidelines for the issuance of new guarantees were implemented in October 1996; since then, the government is charging a fee to discourage the use of such guarantees.

The 1997/98 budget

23. On March 12, 1997, the Minister of Finance presented to Parliament the budget for fiscal year 1997/98. In line with the government’s commitments under the GEAR, the budget targets a decline in the overall deficit from 5.6 percent of GDP in 1996/97 to 4 percent in 1997/98 (excluding extraordinary revenue and the difference between accrued and cash interest payments in both years). The budget is based on the assumptions that real GDP growth will be 2.5 percent and that the GDP deflator will increase by 8.5 percent in 1997/98, compared with 3.7 percent and 9 percent, respectively, in 1996/97. Given that ordinary revenue has been held virtually constant at 26 percent of GDP, the whole adjustment effort is to come from a decrease in expenditure.

24. To achieve the revenue targets, the budget includes the following measures: (i) modifications to the personal income tax—including a change in the threshold rate, an increase in the nominal value of the tax bands, and a reduction in the number of tax bands from 8 to 7—which on balance avoid fiscal drag on low income earners (those below R 30,000 a year) but allows some drag on all other income earners; (ii) an increase in the taxation of fringe benefits such as company cars, travel allowances, and housing and holiday accommodation; (iii) the broadening of the coverage of the 17 percent tax on the gross interest and net rental income of retirement funds, which was introduced last year, to unit trusts and certain interest-bearing instruments; (iv) the unification of the taxation of lump sum payments from public and private sector pension funds (from March 1, 1998; acquired rights were grandfathered) and the reduction of the tax advantage of the transfer of public sector pension fund balances to provident funds; (v) the halving of the market securities tax to 0.25 percent, while removing the exemption for purchases by foreigners; and (vi) increases in specific excises (especially for tobacco and alcohol) and stamp duties, and a halving in the ad valorem excise rate of many commodities to a maximum of 15 percent.11 In addition, further improvements in tax administration as a result of the merger in April 1996 of the Inland and Customs Revenue Services into the South African Revenue Service (SARS), the cash bonus scheme introduced at the SARS, and the forthcoming move to make the SARS autonomous from the civil service, are projected to yield R 2.5 billion (0.4 percent of GDP).

25. On the expenditure side, noninterest current expenditure of the central government is budgeted to decline by 1.3 percent of GDP—most of it as a result of a decrease in central government transfers to provincial governments—while capital expenditure of the central government is to decline by 0.5 percent of GDP. On an economic basis—where the data relate to all spending by the central and provincial governments and are subject to the caveats noted above—expenditure on goods and services is projected to decline by 0.5 percent of GDP to 15.8 percent, reflecting a slight decline in the wage bill, to 11.9 percent of GDP, and a reduction in transfers to business and other units of general government. On a functional basis, expenditure declines are forecast in the areas of defense, education (due to the rightsizing program; see Section III), and economic services, while increases are allocated to housing and welfare. With regard to business subsidies, recently the government shifted industrial promotion from the export subsidy scheme (GEIS) to various supply side measures, such as the promotion of industrial clusters and corridor developments, and the tax incentives program described above.

26. As previously, the deficit is to be financed almost wholly through the issue of government securities in the domestic market; the government will again issue debt at a discount with a coupon rate of some 2 percentage points below the expected yield. The lower deficit and the proceeds from privatization that accrue to the central government and those from the sale of strategic oil stocks will help keep the ratio of government debt to GDP at 55 percent, after including the planned takeover of old debts of Namibia to South Africa (0.2 percent of GDP).

27. The significant progress made by South Africa on a variety of structural reforms toward strengthening budgetary operations and tax administration is outlined in Section III and the new intergovernmental relations are outlined in Section IV.

D. Monetary Developments and Policies

28. Developments in the monetary area during 1995 were marked by an unprecedented volume of capital inflows to South Africa, which in turn presented the authorities with a new policy challenge. While they represented an increased external demand for South African assets arising from the strengthening of investor sentiment, they resulted in a large monetization which could lead to even higher inflationary pressures and a loss of competitiveness. The SARB firmed monetary policy in 1995, raising the bank rate from 12 percent to 15 percent in the course of the year, and sterilized a substantial fraction of the inflows. At the same time the SARB was able to reduce its net open forward position (NOFP) considerably. Nonetheless, the expansion of monetary and credit aggregates continued vigorously. The broadest money supply definition (M3) grew at rates exceeding 15 percent a year for most of 1995 (Appendix Table 19), well beyond the guideline range of 6–10 percent set by the SARB for the year, and the growth of bank credit to the private sector surpassed 18 percent on average for the year (Figure 4).

Figure 4
Figure 4


Citation: IMF Staff Country Reports 1997, 082; 10.5089/9781451840933.002.A001

Source: South Africa Reserve Bank.

29. In contrast to the relatively stable environment in 1995, investor sentiment shifted markedly in early 1996 and monetary developments were dominated by unrest in the foreign exchange markets. As in other international episodes of currency weakness, adjustment in South Africa to an unsettled foreign exchange market came through a mix of exchange rate depreciation, loss of international reserves, and interest rate increases. The monetary authorities attempted to tighten the stance of monetary policy to support the exchange rate and contain the inflationary pressures arising from the currency depreciation, and reiterated their guidelines for growth of M3 in the 6–10 percent range for 1996 as a signal of their commitment to a firm monetary policy. The bank rate was increased twice in 1996 (April and November) to reach 17 percent, and liquidity in the money market was continuously squeezed. At the same time, the authorities intervened heavily in the spot and forward exchange markets, and thus the SARB’s net holdings of gold and foreign exchange reserves dropped from US$4.2 billion by end-1995 to US$2.1 billion a year later, while the NOFP rose to about US$22 billion by end-1996.

30. Notwithstanding the drain of liquidity implied by the loss of foreign assets and the increase in interest rates, the South African economy continued to monetize vigorously during 1996 and early 1997. The expansion of monetary and credit aggregates remained buoyant during most of 1996—with incipient evidence of tapering off by the end of the year before resuming their rapid pace in 1997. The 12-month growth rate of M3, which was 15.2 percent in December 1995, rose to 16 percent in October 1996, before declining to 13.6 percent by December 1996. Thus, the growth in M3 exceeded by far its guidelines and the fall in the income velocity of circulation—which started in 1993—continued in 1996 (Appendix Table 20). The expansion of narrow monetary aggregates was even more notorious (with M1 growing 30.9 percent in the year to December 1996) as the public in general, and institutional investors in particular, revealed an increased preference for liquid assets in the context of volatile financial and exchange markets. Furthermore, the demand for narrow aggregates was also stimulated by the attractive returns offered by banks on demand deposits in order to fund themselves in an environment of tight money market conditions.

31. As can be inferred from the above developments, the counterpart to the R 38.3 billion increase in M3 during 1996 was a contrasting behavior between its foreign and domestic sources (Appendix Table 20). While the net foreign asset position of the consolidated banking system declined by R 14.5 billion as a consequence of the unrest in foreign markets,12 its net domestic asset position increased by R 52.8 billion in 1996. Given that most of the rise in domestic credit corresponded to an increase in the banking system’s claims on the private sector, the 12-month rate of growth of bank credit to the private sector rose from 17.6 percent in December 1995 to a peak of 18.9 percent in October 1996, before moderating to 16.1 percent by the end of the year (Appendix Table 22). The increase in credit to the private sector during 1996 (R 41.2 billion) was more or less evenly distributed between households and corporations—56 percent and 44 percent respectively. Almost 36 percent of the total increase in credit to the private sector was in the form of mortgage advances to households.

32. Developments in foreign exchange markets and the increase in the inflation rate also impacted the level and structure of interest rates. The average yield on long-term government stock rose from 13.8 percent in January 1996 to 16.5 percent in May, and then declined to 16.2 percent by December 1996 (Appendix Table 21). A similar movement was registered in interest rates prevailing in money markets. The representative rate on 3-month Treasury bills rose from 14 percent in January 1996 to 15.9 percent in May. This rate fell somewhat during the second half of the year, but rose again to that level by end-1996, as money market conditions tightened. The prime overdraft rate charged by banks followed closely the behavior of the bank rate, and increased to 20.25 percent in November 1996. It should be noted, however, that real interest rates prevailing at end-1996 were not significantly different from those at end-1995.

33. Conditions in the money market reflected the general tightness of the availability of liquidity. The “shortage” in the market (i.e. the amount the SARB accommodates) rose from an average daily value of R 4.9 billion in January 1996 to R 9.9 billion in December 1996, leading to an increased use of the second tier accommodation at the discount window in the last quarter of 1996. However, the SARB reduced its penalty rate for second tier accommodation from 1.5 percent above the bank rate to 0.75 percent in June 1996 to ease somewhat conditions in the money market.

34. Another marked change in investor sentiment occurred in early 1997, when confidence and stability returned to South African financial markets. The favorable sentiment in turn translated into capital flows returning to South Africa. These inflows—together with the proceeds (US$1.3 billion) of the sale of 30 percent of the shares of Telkom to a foreign consortium—permitted the replenishment of the SARB’s net international reserves, which rose to US$3.7 billion by May 1997, a strengthening of the rand, which appreciated almost 9 percent in nominal effective terms in the first quarter of 1997, and a decline of long-term yields to below 15 percent. However, the rate of growth of money and credit aggregates rose in 1997, and both M3 and bank credit to the private sector expanded again at rates exceeding 16 percent a year by March 1997, well above the renewed monetary growth guideline of 6–10 percent for 1997.

E. External Sector Developments

The Balance of Payments

35. The external current account registered a deficit of 0.3 percent of GDP in 1994—the first since 1984—owing to the rapid growth in imports, reflecting the pick up in economic activity, and a slowdown in exports. The deficit widened to 2.1 percent of GDP in 1995, as imports continued to grow at a fast pace (Appendix Table 23), and despite an increase in the volume of nongold exports and little change in the international terms of trade. Imports of capital goods were especially pronounced, suggesting that, as in 1994, investment was the most important factor behind the fast import growth. The current account deficit declined to 1.6 percent of GDP in 1996, with a similar improvement in the trade balance reflecting a continued strong export performance and a marked slowdown in import growth.

36. Exports picked up momentum in the second half of 1996 as a result of a number of factors, including the completion of some of the export-oriented “mega investment projects” (Alusaf, Columbus Steel, and Namakwa Sands), which contributed significantly to the strong export performance of the manufacturing sector; the rebound in agricultural production owing to favorable weather conditions, which led to the rapid growth of agricultural exports; and the depreciation of the rand. Moreover, access to new markets, relatively strong growth in the main industrial countries, and high growth in the emerging economies also contributed to the firm demand for South African exports. While the volume of nongold exports increased by 11 percent, compared with 14 percent in 1995 (Appendix Table 26), the increase was a modest 2.6 percent in U.S. dollar terms, compared with 22.5 percent in 1995, reflecting lower export prices, as exporters seem to have used the depreciation of the rand to increase market shares. Accordingly, while the rand depreciated in nominal effective terms by 21 percent in 1996, the unit value of nongold exports only increased by 6 percent in rand terms.

37. The value of gold exports declined in U.S. dollar terms by 1.5 percent in 1996, owing to a fall in production that more than offset the increase in prices. The average U.S. dollar price of gold increased by less than 1 percent in 1996, as prices remained firm through the first half of the year but fell in the last quarter to US$369.2 an ounce by December 1996 (US$344.5 an ounce in May 1997), compared with a peak of US$404.8 an ounce in February 1996. The share of gold in total merchandise exports declined from more than 28 percent in 1992 to 21 percent in 1996.

38. The growth of imports fluctuated markedly in 1996. It moderated in the first quarter, but picked up in the second and third quarter, reflecting some strengthening in aggregate domestic demand and in anticipation of a further depreciation of the rand, and turned negative in the fourth quarter largely owing to a decline in domestic demand. For the year as a whole, imports remained flat in 1996, after growing by 26 percent in 1995. Import growth in volume terms slowed to 7.6 percent in 1996, compared with about 16 percent in the previous two years. Intermediate goods and capital equipment continued to account for a large share in total imports. The pass-through of the depreciation of the rand on import prices was limited as the unit value of imports increased by 7.7 percent in 1996, at about the pace of increase in the previous two years, when the value of the rand remained broadly stable.

39. The deficit in the services account declined marginally from 3.3 percent of GDP in 1995 to 3.1 percent of GDP in 1996. This was the result of a marked further increase in tourist receipts, partially offset by higher payments for services associated with trade—mainly freight and insurance—and higher interest payments owing to the rise in South Africa’s foreign debt.

40. Following the strong capital inflow that began in the second half of 1994, the surplus in the capital account declined from US$5.3 billion in 1995 to US$0.9 billion in 1996 (Appendix Table 28). This reflected a marked change in investor sentiment about South African assets. In particular, net inflows of long-term capital declined from US$4.2 billion in 1995 to US$1.1 billion in 1996, while net flows of short-term capital (net of errors and omissions) became slightly negative. The net inflow of long-term capital largely reflected an increase in borrowing by the public sector that exceeded the levels of 1995 and more than offset the long-term outflows by the private sector. Government borrowing comprised the Euro-sterling issue of £100 million in January 1996, the Eurobond issue of DM 500 million—which was primarily aimed at refinancing a maturing bond issue—and a bond issue of US$300 million in the U.S. market. Public corporations were also net borrowers. By contrast, private long-term capital inflows declined throughout the year and became substantially negative in the fourth quarter, reflecting the repayment of foreign debt obligations, net sales by non-residents of securities on the Johannesburg Stock Exchange, and an increase in assets held abroad by South African companies and nonbank financial institutions.

41. Total assets swaps (see Section V for a description of this mechanism) authorized since the inception amounted to the equivalent of US$5.9 billion by end of 1996, but only US$3.7 billion had been utilized; of this amount US$2 billion was swapped in 1995 and US$1.7 billion in 1996. Notwithstanding the increase in the ceiling in 1996 on swap operations, the rather low use of this mechanism indicates that the existing limits are not binding. This may reflect a limited interest by foreigners in such operations during 1996, perhaps owing to the volatility in the rand, as suggested by the decline in swap operations from US$663 million in the first quarter of 1996 to an average of US$342 million in the subsequent three quarters.

42. After a substantial drop in 1993, and a further decline in the first half of 1994, net international reserves recovered with the surge in capital inflows in the second half of 1994 and in 1995.13 Net international reserves peaked at US$3.5 billion in the fourth quarter of 1995, or about 1.3 months of imports of goods and nonfactor services (Appendix Table 30), after having repaid US$0.7 billion in reserve-related external liabilities. However, reserves came under pressure during 1996, declining by more than half by the third quarter of 1996, and were US$1.6 billion by the end of the year. (Notwithstanding the pressures on reserves, the SARB did not make use of its external lines of credit through 1996.) In addition to intervening in the spot exchange market, the SARB intervened heavily in the forward foreign exchange market, raising the NOFP from nearly US$7 billion in mid-February 1996 to about US$22 billion in December 1996—an increase equivalent to some 6 months of merchandise imports in that period.

43. As indicated above, there was a marked strengthening of sentiment in early 1997, which resulted in large capital inflows and an increase in net international reserves to a record level (US$3.7 billion) in May 1997. In a departure of previous policy, the SARB made use of the external lines of credit by the equivalent to R 2.5 billion through April 1997, but repaid R 0.7 billion in May. The NOFP has gradually declined to US$17.8 billion in May 1997.

44. In line with the large capital inflow experienced in 1995, South Africa’s total external debt (denominated in foreign currency) increased by nearly US$4 billion to US$22.3 billion, or 24 percent of GDP (Appendix Table 29). By end-June 1996, this stock had fallen to just below US$21 billion as a result of the sharp reversal in capital flows and a reduction in long-term private debt—from US$12.5 billion at end-1995 to US$11.2 billion by June 1996; total private sector debt declined by US$1.7 billion between 1995 and mid-1996, to about 57 percent of South Africa’s external debt. At the same time, the share of short-term debt in total debt increased from less than 40 percent in 1994 to about 46 percent in June 1996, following the large short-term capital inflows in 1995.

Trade liberalization and trade agreements

45. Considerable progress has been made in rationalizing the external trade regime over the last few years. A trade reform initiative was launched in 1994 based on the government’s application to the Uruguay Round of Trade Negotiations. The reform aims to eliminate quantitative import restrictions; standardize tariff rates at six levels; reduce the number of tariff lines to the six-digit Harmonized Code level by the end of 1999; reduce the maximum tariff rates to 30 percent for industrialized goods and 36 percent for agricultural goods, with “sensitive” industries—such as clothing, textiles, and motor vehicles—maintaining the higher tariff rates and longer adjustment periods.

46. Under the reform, South Africa replaced during 1994–95 most quantitative restrictions and formula duties with ad valorem tariffs for most manufactured products and a number of agricultural products, and adopted tariff bindings covering 98 percent of imports to become effective in 1999. Nominal rates of protection were reduced—even though most tariffs were already well below the agreed WTO tariff binding rates when the Uruguay Round negotiations were concluded. The tariff structure was simplified through the near halving of the number of tariff lines to about 5000 and through the reduction in the number of ad valorem rates. Reforms of the trade and marketing of agricultural products were accelerated in 1995, with the deregulation of prices, the abolition or reform of agricultural control boards, the conversion of import controls to ad valorem duties. Finally, the last remaining import surcharge, on luxury goods and household appliances, was abolished in October 1995.

47. During 1996, the schedules for the reduction of tariffs over 5–7 years were finalized, and a number of tariffs on commodities that were either not produced in South Africa or which were imported inputs were abolished. Overall, the import weighted average tariff rate fell from 11 percent in 1995 to 10 percent in 1996, but the reduction in tariffs on inputs ahead of those on final goods resulted in an increase in effective protection, which raised the existing biases against exports, labor absorption, and growth. Finally, the three-year process leading to the elimination of the primary export subsidy, GEIS, initiated in 1995, was brought to an end with the abolition of the export subsidy program to take effect from mid-1997, six months ahead of the original schedule.

48. Negotiations on regional free trade agreements with the South. African Development Community (SADC) 14 and the European Union (EU) continued during 1996. SADC was originally established in 1980; its agenda was refocused in 1992 toward deepening economic integration, including in monetary and fiscal policies, exchange rate and trade regime, and mobility of capital and labor. The Trade Facilitation Protocol, drafted in February 1996, envisages a Free Trade Agreement (FTA) among SADC member countries with reciprocal tariff reductions. The proposed FTA would be phased-in over an eight-year period, and South Africa was considering accelerating the elimination of tariffs on imports from SADC before the reciprocal SADC tariffs on South African exports were eliminated under the FTA. The review of the revenue-sharing formula for the Southern African Custom Union (SACU) is nearly completed;15 it aims at helping avoid sharp reductions in South Africa’s transfers to the other members as a result of its unilateral tariff reduction.

49. The negotiations with the EU have proceeded slowly. After some delay, access by South Africa to the Lome convention, except for its trade preference provisions, was formalized in April 1997. The main benefits to South Africa are the right by South African companies to tender for contracts funded under the EU development fund and to participate in the joint bodies of the African Caribbean and Pacific (ACP) states and the EU—the joint assembly and ministerial meetings—where negotiations on the future of Lome are now taking place. Regarding the FTA proposed by the EU, which envisaged the asymmetrical phase-in of a free trade area over 10–12 years, the South African authorities are concerned that the EU offer has not adequately addressed South Africa’s trading relations with the region and the fiscal implications for SACU, that it excludes 40 percent of South African agricultural exports to the EU, and that it would expose local farmers to subsidized European agricultural products by removing the current tariff protection.16 As a consequence, discussions have continued at a slow pace.

Capital controls liberalization

50. A major step in liberalizing exchange controls was taken with the elimination of the financial rand mechanism in March 1995, and hence the removal of all exchange restrictions on nonresidents. Progress was also made in relaxing controls on residents in various steps, underscoring the authorities’ commitment to the gradual elimination of exchange controls on residents while ensuring that these steps were not disruptive to financial markets. For a description of these steps, see Section V.

51. It is important to note that the authorities’ approach was not simply to liberalize controls in a sequence of steps that allowed different types of capital transactions one at a time, but to open a number of new windows for resident capital transactions, each subject to a cap. In this way, an increasing array of forms of resident capital transactions have been permitted, but the risks of major aggregate outflows were limited by the caps applied to each type of transaction. This feature of the step-by-step approach allowed the authorities to gradually increase the range of authorized capital transactions, as well as the scope of each of these transactions by raising the caps. The authorities intend to continue with this approach until all types of resident capital outflows are permitted and until the caps are no longer binding, and the risk of major capital outflows minimized.

The Eurorand market

52. Beginning in September 1995, a series of rand-denominated bonds were issued in the European market by non-South African entities, including Merrill Lynch, EBRD, and the Government of Sweden. Total issues amounted to the equivalent of US$1.8 billion by mid- February 1996. The coupon rate fell from 15 percent on the initial issues to as low as 13 percent on issues in late January and the first half of February 1996. Activity in the Eurorand market slowed thereafter, but picked up markedly in early 1997, with total issues exceeding the equivalent of US$10 billion by end-May 1997. In contrast to the previous issues, which had been short- or medium-term, a number of the issues in 1997 have carried maturities of 10 years or longer, with one issue by the EBRD having a maturity of 30 years. The coupon rate has generally been above 14 percent in 1997.

53. The development of the Eurorand market is an important step in the development of rand capital markets. Thus far, these issues appear to be largely held in the stable retail market, and there was no evidence of large sales during 1996. The market allows high credit quality international issuers to lower their borrowing cost in the Eurorand market vis-à-vis the sovereign issuer in the domestic rand market. Issuers swap the rand with a top-rated bank or investment bank in London, who then hedges the acquired rand exposure with the outright purchase of domestic rand gilts, frequently as part of an asset swap. The impact of these operations on reserves is generally thought to be positive or neutral.


These policies consisted of phased reforms—based on social consensus and the constitutional independence of the Reserve Bank—including annual half-a-percentage point of GDP reductions in the fiscal deficit, an eight-year program of trade liberalization, step-by-step liberalization of foreign exchange controls, a four-year program to reform labor law and promote worker training, and public expenditure reprioritization towards the poor.


The NOFP is the net oversold forward book of the Reserve Bank, less net international reserves.


See Section II for a discussion on unemployment and employment issues.


The fiscal year begins April 1. Data prior to 1994/95 are not comparable with those thereafter because the 1994/95 budget changed the methodology concerning the treatment of four of the former homelands (Transkei, Bophuthatswana, Venda, Ciskei). Central government expenditure includes transfers to the provinces and the accrued rather than the cash interest charges; accrued interest data is around ½ percent of GDP higher than cash interest payments (this difference has been rising over time) owing to the sale of government bonds with a lower coupon rate than the prevailing market yield. The data here exclude extraordinary transfers to the Pension Fund and the Gold and Foreign Exchange Contingency Reserve Account, and revenue from the sale of strategic oil stocks.


Rollovers result from underspending, mainly on capital projects, relative to the original budget allocations. The Minister of Finance approves the amounts that may be rolled over on a case by case basis. Generally, rollovers must be spent by the end of the following fiscal year, although rollovers associated with the RDP could in some circumstances be spent over several years (a provision that has now lapsed).


The data show the economic and functional classifications of the Budget plus the changes in the appropriations introduced in the Adjustments Budget, which is presented to Parliament the following February, and includes a distribution by spending ministry of the stock of rollovers from the previous fiscal year. Because the budget data do not include a provision for the expected expenditure of rollovers in the current fiscal year, the economic and functional data that are based on the Budget estimates are not strictly comparable to those based on the preliminary outcome presented in the Adjustments Budget. Further, the economic and functional data that are based on the Adjustments Budget include adjustments for the cash that was transferred to spending ministries, but which was subsequently not spent and was surrendered back to the exchequer account. Accordingly, the data in Appendix Tables 14, 15, and 16 have been adjusted for both surrenders and rollovers to derive the total cash expenditure of the central and provincial governments. Finally, at this time the economic and functional classifications of the 1996/97 budget are available, but not those for the 1996/97 Adjustments Budget.


This amnesty, covering tax arrears prior to March 1994, netted around R 1 billion (0.2 percent of GDP) over the period November 1996 to February 1997.


Given the practice of issuing bonds with a coupon rate that is lower than the market yield (which began in 1980), the interests bill is lower than the cash interest payment. Adjusting for this factor, the annual interest bill (and consequently the cash budget deficit) in 1996/97 would have been 0.6 percent of GDP higher than was recorded, similarly, adjusting the stock of government bonds for the corresponding discount, the debt to GDP ratio at March 1997 would have been 7 percentage points of GDP lower.


As of March 31, 1996, actuarial underfunding of the state pension fund was estimated at around 40 percent of contingent government liabilities, or 10 percent of 1995/96 GDP.


Following the realization of these losses, the balance of the Gold and Foreign Exchange Contingency Reserve Account held at the South African Reserve Bank is increased as is the total debt of the central government. Periodically, this obligation of the central government is regularized through the issue of zero coupon paper to the Bank. The Bank has the right to request conversion of this paper into central government coupon-bearing stock, should the latter be needed for monetary management purposes.


For a description of the current tax system, see Annex I.


The fall in the consolidated banking system’s net foreign assets reflected the loss of foreign assets of the SARB and the considerable accumulation of foreign liabilities by commercial banks.


The CCFF purchase from the Fund in December 1993 is considered a reserve liability.


The members are: Angola, Botswana, Lesotho, Malawi, Mauritius, Mozambique, Namibia, Tanzania, Swaziland, Zambia, Zimbabwe, and South Africa.


South Africa, together with Botswana, Lesotho, Namibia, and Swaziland, form SACU, which provides for the free movement of goods and the right of transit among member countries. Under SACU, South Africa administers duty collection and distributes shares of the common revenue pool to the four countries according to a revenue-sharing formula.


Anti-dumping measures could not be applied against European subsidized exports under the Uruguay Round agreement.

South Africa: Selected Issues
Author: International Monetary Fund