This Selected Issues paper highlights that cautious monetary and fiscal polices of South Africa during 1997 resulted in a return of financial investor confidence and capital inflows during 1997 and through April 1998. These policies helped the South African economy emerge successfully from the exchange market pressures of 1996 and weather the contagion from the East Asian crisis in the second half of 1997. Throughout 1997 and up until May 1998, inflation and market interest rates fell considerably, net international reserves increased, and the net open forward position of the Reserve Bank was reduced sharply.


This Selected Issues paper highlights that cautious monetary and fiscal polices of South Africa during 1997 resulted in a return of financial investor confidence and capital inflows during 1997 and through April 1998. These policies helped the South African economy emerge successfully from the exchange market pressures of 1996 and weather the contagion from the East Asian crisis in the second half of 1997. Throughout 1997 and up until May 1998, inflation and market interest rates fell considerably, net international reserves increased, and the net open forward position of the Reserve Bank was reduced sharply.

VII. Measures of External Competitiveness

A. Introduction

239. With the introduction of the Growth, Employment and Redistribution (GEAR) strategy in mid-1996, the authorities embarked on a program designed to transform South Africa into “a competitive, outward oriented economy.”1 The GEAR places considerable importance on “a powerful expansion by the tradeable goods sector” as a means of generating sustainable growth. Accordingly, one major policy challenge is to ensure that South Africa’s external competitiveness is sufficient to play an appropriate role in the desired economic transformation.

240. In its broadest sense, external competitiveness covers a wide range of economic factors, some of which are directly amenable to being influenced by policy. Moreover, some elements of competitiveness are not obviously reflected in changes in prices. For example, even if the nature and price of a product appears unchanged, better telecommunications or transport links can allow a supplier to provide more rapid and convenient service. This is clearly an improvement in the quality of the product—there has been a change in the hedonic characteristics of the good—but one which may not be not folly captured in economic statistics.

241. While recognizing the impossibility of reflecting all facets of competitiveness in a single indicator, some insights can be obtained by considering the more narrowly focussed issue of developments in international price competitiveness. The real exchange rate, e, is defined as EP/P*, where E is the nominal exchange rate (measured as foreign currency per unit of domestic currency) and P and P* are domestic and foreign price indices. E and P* are weighted averages of bilateral nominal exchange rates and foreign trading partner prices respectively. Since competitiveness is here defined in terms of the relative price of foreign goods in terms of domestic goods, a decrease in the indicator denotes an improvement in competitiveness.2 Such an improvement should be manifest through, inter alia, an improvement in export volume growth, everything else being equal.

B. Estimates of the Real Exchange Rate

242. On two commonly cited measures South Africa appears to have made significant competitiveness gains in recent years. First, as shown in Figure 16, from 1994 to mid-1996 nongold merchandise export volumes grew rapidly, and faster than growth in major trading partner imports.3 Second, by March 1998 the South African Reserve Bank’s estimate of the real effective exchange rate, SARB-REER4, was around 15 percent below that in 1990, with a particularly sharp improvement evident since late 1995 (Figure 17).

Figure 16.
Figure 16.

Nongold Merchandise Export Volumes-Actual and Implied by Partner Import Growth

(In millions of rand)

Citation: IMF Staff Country Reports 1998, 096; 10.5089/9781451840940.002.A007

Sources: IMF’s Direction of Trade Statistics; and Fund staff estimates.
Figure 17.
Figure 17.

Real Effective Exchange Rate

Citation: IMF Staff Country Reports 1998, 096; 10.5089/9781451840940.002.A007

Sources: South African Reserve Bank; and IMF’s Information Notice System and Direction of Trade Statistics.

243. The growth in nongold merchandise exports is broadly consistent with the growth in partner country non-oil merchandise imports between 1990 and 1994. Faster growth in South Africa’s exports than in partner imports—a rising market share for South Africa—from 1995 may be at least partly attributable to the depreciation in the rand in real effective terms and a series of structural reforms, including in the trade area. It is also possible that some of this growth is a consequence of the end of sanctions and would have occurred even in the absence of the depreciation and the reform program.5

244, If the effect of sanctions had been to impose a barrier to South African exports into particular markets, the removal of sanctions would provide a stimulus to export growth, as South Africa would recover its “natural” market share in previously closed markets. However, such a “sanctions-ending” boost to growth would be only temporary—growth in exports would eventually settle in line with partner market growth unless competitiveness continued to improve. In the transition period, though, export volumes could be expected to grow faster than the growth in world trade in the exported products. In contrast, if sanctions had no impact on South Africa’s ability to export, the very rapid growth in export volumes is more likely to be attributable to competitiveness improvements alone.

245. Caution also needs to be exercised when considering developments in the SARB real exchange rate index as it is likely that the measure overstates the extent of the improvement in competitiveness, particularly in recent years. The SARB measure is based on a comparison with only four trading partners—the United States (with a weight of 51.7 percent), the United Kingdom (20.2), Germany (17.2) and Japan (10,9). This is not necessarily a problem if these countries were responsible for the bulk of trade with South Africa. However, their total twoway trade was only around 40 percent of South Africa’s total trade in 1996, a share that has declined in recent years as can be seen in Figure 18.6

Figure 18.
Figure 18.

Direction of Trade

(in percent)

Citation: IMF Staff Country Reports 1998, 096; 10.5089/9781451840940.002.A007

Source: IMFs Direction of Trade Statistics

246. An alternative measure of the real exchange rate that covers a broader set of trading partners is available from the IMF’s Information Notice System (INS). The INS real eflfective exchange rate, INS-REER, covers trade with 16 countries.7 Weights are derived taking into account trade in manufactured goods and primary commodities, and reflect both direct and third market competition. (The weights are set out in Table 18).

Table 18.

Country Weights in Various Measures of South Africa’s Real Exchange Rate

article image
Sources: South African Reserve Bank; IMF’s Information Notice System and Direction of Trade Statistics.

247. Figure 17 shows that a significant wedge opened up between the two measures during the course of 1997 and early 1998, suggesting that competitiveness gains may have been more modest than implied by the SARB measure—around 5 percent relative to 1990 according the INS-REER measure. Figure 19 shows that the SARB nominal effective exchange rate, SARBNEER, fell more sharply than the INS-NEER. This reflects the strength of the U.S. dollar and the U.K. pound during 1997 relative to the currencies of the broader set of trading partners covered by the IMF measure. In terms of the REER, the large depreciation of the rand in nominal effective terms in 1996 translated into a significant depreciation in real effective terms on both measures but, according to the INS-REER measure, much of this was subsequently unwound through the first half of 1997 as the rand strengthened and South Africa’s inflation rate continued to run ahead of that of its trading partners.8 In addition, the broader IMF measure reflects the real depreciation vis-à-vis the U.S. dollar and the pound experienced by many of South Africa’s trading partners.

Figure 19.
Figure 19.

Nominal Effective Exchange Rate

Citation: IMF Staff Country Reports 1998, 096; 10.5089/9781451840940.002.A007

Sources: South African Reserve Bank; IMF’s Information Notice System and Direction of Trade Statistics.

248. While having a wider coverage of partner countries than the SARB measure, the INS measure is open to the criticism that its weights are based on trade flows for the period 1988–90 and are therefore outdated. As Figure 18 illustrates, there has been a noticeable shift in South Africa’s trading pattern since 1990, in part due to the lifting of trade sanctions.

249. An alternative to INS-REER can be calculated by replacing the existing INS weights with weights derived from the shares of 36 countries in South Africa’s total trade between 1994 and 1996 (Table 18). While increasing the coverage of trading partners and using more timely data on trading patterns relative to the INS measure, this alternative approach fails to capture the impact of competition in third markets.

250. The result of calculating this direction-of-trade-based measure, DOTS-REER, is also shown in Figure 17, While the time profile differs until late 1995, the INS and DOTS-based measures move closely thereafter, both suggesting that there was little net gain in competitiveness since 1990 and that the improvement of 1996 was largely unwound during the first half of 1997.

251. The relative price variable used in measures of international competitiveness should, theoretically, represent the ratio of the domestic and foreign price indices of tradeable goods. Moreover, as noted by Turner and Golub (1997), these indices should be exogenous with respect to the exchange rate and unaffected by very short-term pricing decisions. In practice, such measures do not exist and the more common measures of the real exchange rate tend to use indices of consumer (CPI) or producer prices (PPI), domestic production deflators or export and import unit values.

252. The SARB measure is based on relative producer price movements, while the INS and DOTS measures are based on movements in relative CPIs. As a practical matter, CPI-based measures have some advantages over PPI-based measures. Because the CPI is the most common means of measuring inflation, data on movements in CPIs tends to be more timely, and the design of CPI tends to be reasonably similar across countries. However, both the PPI and CPI baskets suffer from shortcomings. In particular, both are comprised of traded and nontraded goods and there is no reason why the prices of both types of goods should move together over time. To the extent that the prices of traded and nontraded goods diverge significantly, relative CPIs or PPIs may provide an inadequate picture of developments in competitiveness.

253. One way to minimize this problem is to calculate REERs based on movements in unit labor costs.9 Due to data limitations, these are usually calculated for the manufacturing sector. Measures based on ULCs have a number of advantages as indicators of international competitiveness: they capture cost developments in an important sector exposed to international competition; offer a reliable gauge to the relative profitability of traded goods; are convenient from a statistical standpoint as fairly comparable data on the manufacturing sector exist for a number of countries; and, by construction, they bring into focus the largest component of nontraded costs and value added. Thus, ULC-based REERs provide a useful adjunct to CPI-based measures.

254. Figure 20 shows the results of estimating a REER measure based on data on unit labor costs in manufacturing in South Africa and 16 industrialized economies for the period 1990 to the second quarter of 1997 (the latest available data).10 The results are consistent with previous measures—competitiveness improved in the 1990s, although the exclusion of South Africa’s increasingly important Asian trading partners may overstate the extent of the improvement in competitiveness. Figure 20 compares this ULC-REER measure with a CPI-REER measure derived from the 1994–96 direction of trade data and normalized for the smaller group of 16 trading partners for whom the ULC data are available. While the two measures tend to move fairly closely together, the DOT measure for the 36 countries will only approximate a broader ULC measure if manufacturing unit labor costs move in line with changes in the CPI in the countries for which ULC data are unavailable.

Figure 20.
Figure 20.


Citation: IMF Staff Country Reports 1998, 096; 10.5089/9781451840940.002.A007

Sources: IMF’s Information Notice System and Direction of Trade Statistics.

255. While, all measures of the real effective exchange rate cited above point to some improvement in external competitiveness, it is not possible to infer from this anything about the appropriateness of the level of the real exchange rate. Time series data on the real exchange rate per se do not provide sufficient information to determine whether movements reflect changes in economic fundamentals or misalignments of the nominal exchange rate—this requires some idea of developments in the equilibrium real exchange rate. However, estimating movements in the equilibrium real exchange rate requires a lengthy time series of data. Given the structural changes that have occurred in South Africa since 1994, many of which have only occurred in the last year or two (so that the full extent of their impact in many cases is yet to be seen in the data), it would be difficult to estimate the impact of developments with precision.11 Recent research has concentrated on characterizing the influences on the equilibrium real exchange rate using simulations rather than presenting estimates of the extent, if any, of any real exchange rate misalignment (see Aron et. al. (1997)).

256. After very rapid growth, nongold merchandise export volumes, while volatile, have grown by an average of only 0.1 percent per quarter in the five quarters to December 1997. While data on export volumes by country of destination or by industrial sector are unavailable, Table 19 shows that the most rapid growth in the value of nongold exports has occurred in manufactures—under plausible assumptions, it is reasonable to assume this also reflects developments in export volumes. Within this sector, the most rapid growth—and the major contribution to the increase—has occurred in exports to Asia and to Africa. In Asia, exports to Hong Kong, Korea, Singapore, and Taiwan accounted for more than 60 percent of manufactured exports to Asia in 1996, with exports to Indonesia and Malaysia growing very rapidly.12 The aftermath of the Asian financial crisis and the increasing signs of a regional recession suggests that prospects for continuing rapid export growth to the region have been diminished, at least temporarily. The rapid growth of exports to the rest of Africa is a welcome sign of increased regional integration but again the scope for sustained double-digit growth in exports will be constrained by the capacity for growth in other regional economies, particularly as Zimbabwe has been South Africa’s largest and fastest growing market for manufactures. Combined with increased competition in commodity export markets (and the continuing trend decline in gold exports), this weakening in the outlook for manufactured exports suggests caution in extrapolating the continuation of past growth rates in nongold export volumes.13 Given the emphasis placed in the GEAR on the role of export-led growth in transforming South Africa’s economic prospects, this suggests that renewed efforts to boost South Africa’s international competitiveness through faster implementation of structural reforms may be required. The discussion in Section III and recent experience suggest that competitiveness cannot be durably improved as a result of the depreciation of the rand because it is likely to be eventually dissipated through higher domestic inflation (see Appendix).

Table 19.

Composition and Destination of South Africa’s Exports 1/

article image
Source: “South Africa’s Manufacturing Exports: Recent Market Trends” by Pieter Laubscher in Trade and Industry Monitor, December 1997.

Data apply to Southern African Customs Union members, i.e., Botswana, Lesotho, Namibia, South Africa, and Swaziland and are only available for the period 1992-1996. However, South Africa’s share of the total SACU export basket is estimated to exceed 90 percent and the data presented here exclude “other manufacturing” which is dominated by diamond exports from other SACU members.

C. Conclusions

257. Despite South Africa’s growing reliance on exports as a source of economic growth in recent years, it is unclear to what extent this can be attributed to developments in competitiveness. It is possible that nongold export volumes have been temporarily boosted by the catch-up effects of the removal of sanctions, leading to a step adjustment in export growth rates, although they have also reflected some improvement in competitiveness.

258. Existing measures of competitiveness, however, do not provide a guide to the appropriateness of the level of the real exchange rate. The extent of structural changes in recent years suggests that estimates of the equilibrium real exchange rate are likely to be of limited value until these changes have become fully embedded into South Africa’s economic and industrial structure.

259. Given the importance of increased integration with the world economy for South Africa’s economic transformation, sustained strong export growth seems imperative. In that regard, the performance of nongold exports in volume terms over the last two years is disturbing. Further stimulus to exports will require competitiveness enhancing measures such as labor market reform that promotes employment and skill accumulation; trade reform that reduces the anti-export bias of current policies and enhances efficiency; and with continuing efforts to reduce South Africa’s inflation rate to levels comparable with those of its major trading partners.


  • Aron, J., I. Elbadawi, and B. Kahn, 1997, “Determinants of the Real Exchange Rate in South Africa,Centre for Study of African Economies Working Paper, WPS/97-16, University of Oxford.

    • Search Google Scholar
    • Export Citation
  • Turner, A.G., and S.S. Golub, 1997, “Towards a System of Multilateral Unit Labor Cost-Based Competitiveness Indicators for Advanced, Developing and Transition Economies, IMF Working Paper, WP/97/151.

    • Search Google Scholar
    • Export Citation
  • Zanello, A., and D. Desruelle,A Primer on the IMF’s Information Notice System,IMF Working Paper, WP/97/71.

APPENDIX: Pass-through from Exchange Rates to Domestic Prices

Earlier analysis by the staff drew attention to the relatively limited response of import and consumer prices to the large depreciation of the rand in 1996. Extending the analysis over a longer time period, from 1992 to 1997, and taking into account the trade liberalization that has occurred during this period, the following conclusions emerge:

  • Between 1992 and 1997, the pass through from foreign prices to imported prices was between 77 percent and 85 percent (see figure below), depending on how foreign prices are measured.

  • When a lower bound estimate for tariff reductions during this period is incorporated, the pass-through was larger, between 84 percent and 92 percent. Tariff reductions are captured through changes in the ratio of duties collected to imports. This lower bound impact of trade liberalization between 1992 and 1997 is equivalent to a decline of about 3.5 percentage points in import prices; it is a lower bound because insofar as liberalization has entailed the replacement of quantitative restrictions (QRs) by equivalent tariffs, measured duty collections will be higher, biasing downwards the impact of tariff reduction. If it is assumed that the elimination of QRs has contributed as much as tariff cuts, the pass-through will be between 90 percent and 100 percent.

  • However, the wedge between changes in foreign prices and in the imported component of producer prices remains high, with only about 56 percent of the foreign price increase being reflected in imported producer prices.

Figure 21.
Figure 21.

Pass Through, 1992-1997


Citation: IMF Staff Country Reports 1998, 096; 10.5089/9781451840940.002.A007

Thus, over the period 1992-1997, exchange rate changes are almost fully reflected in domestic prices, once a sufficiently long period is considered to allow for lagged impacts, and the impact of trade liberalization on prices is taken into account. This is consistent with the conclusion in last year’s staff report that over time fuller pass-through (than was evident in 1996) would make itself manifest. However, it is puzzling why producer price increases are so much lower than increases in other prices.


Growth, Employment and Redistribution: A Macroeconomic Strategy, 1996, page 3.


Issues associated with the practical specification of real exchange rate measures are discussed in detail in Turner and Golub (1997).


Implied partner country import growth is calculated from the actual growth in imports of goods (excluding oil) in countries which together account for more than 95 percent of trade with South Africa. Weights are based on each country’s share in South Africa’s exports.


Calculated as the nominal effective exchange rate deflated by relative producer prices.


Changes in taste and/or technology may also have triggered faster growth in demand for South Africa’s exports.


Over the period 1994-96, specific trading partners can be identified for only around 76 percent of South Africa’s total merchandise trade. For simplicity, that part of merchandise exports and imports for which a partner could not be identified has been pro-rated across all partners in accordance with their share in identified trade.


Australia, Belgium, Brazil, Canada, France, Germany, Hong Kong, Italy, Japan, the Netherlands, Spain, Switzerland, Sweden, Taiwan, United Kingdom, and the United States.


It remains to be seen whether the large nominal depreciation of the rand since the latter part of 1997 will translate into a sustainable real depreciation.


For a detailed discussion of the relative merits of different approaches to estimating the real exchange rate, see Turner and Golub (1997) and Zanello and Desruelle (1997).


Data on unit labor costs are only available for 16 countries. Weights are calculated by dividing each country’s DOT weight by the sum of the DOT weights of the 16 countries.


In addition to the lifting of sanctions in 1994, major easings in exchange controls on capital transactions occurred since 1995, import duties were reduced significantly with the abolition of the import surcharge in 1995, most remaining quantitative restrictions on imports were replaced by tariffs by 1996, and the price of gold fell sharply in 1997.


Exports to Asia appear to have been boosted in part by the start-up of the so-called “mega projects,” and by some diversion of nonferrous metal exports from traditional European markets to Asia.


Progress on free trade agreements with the EU, and within SADC, would likely provide a fillip to South Africa’s exports in the medium term but at this stage the final form of these agreements remain unknown.

South Africa: Selected Issues
Author: International Monetary Fund