Production networks in the SADC region

South Africa is by far the largest and most industrialized country in the SADC area. The country accounts for about 70% of total income (GDP) in the SADC area, about 85% of manufacturing value added and about 65% of total exports (World Bank 2000). South Africa joined the SADC in 1994, and entered into a free trade agreement with the European Union in the year 2000. The patterns of trade and investment between South Africa and the SADC countries are very different from the patterns of trade and investment relations between South Africa and the EU. In fact South Africa in many ways plays a similar role in relations to the region as EU does in relation to South Africa as will be discussed below. Note, however, that the SADC countries outside South Africa have a higher trade to GDP ratio than South Africa has.

Reliable time series data on direction of trade in South Africa is not easy to come by because of the sanctions period. Nevertheless, it appears that the opening of regional and global markets for South African exports have to some extent shifted exports from developed to developing countries, and among developing countries, Africa has increased its relative importance as a destination for South African exports. The same trend is apparent looking at sources of imports. Developing countries have increased their market share in South Africa, but while Africa received about 14% of South African exports in 1997, the continent provided only 3.3% of South Africa’s imports (IMF 1999). These trade patterns are depicted in figure 4.1.

Figure 4.1. South Africa’s trade with SADC and EU

Source: IMF 1999

Involvement in regional or international production networks should be revealed in part by growing levels of intra-industry trade. For the region itself, the level of intra-industry trade is extremely low (approximately 27% in 1997). 19This figure was calculated at the 2-digit harmonized system level. Calculations at the one-digit level reveal a much higher degree of intra-industry trade – around 38%. Isemonger (2000) performs the calculation at the 4-digit level and shows an economy-wide intra-industry trade measure of 24%. A closer look at the trade data (table 4.1 below) shows that although South Africa exports a broad range of manufactured goods to the region, the region mainly supplies South Africa with agricultural and mineral products, and only a very limited amount of manufactured products. Further, the manufactured exports are mainly confined to beneficiation and light manufacturing products – particularly clothing and textiles, along with some processed foodstuffs. These trade patterns are indicative of a North-South trading relationship rather than a South-South one. A simulation of the impact of the SADC Free Trade Area (FTA) by Davies (1998) on the trade between South Africa and the rest of SADC suggests that the most significant output effect will be on light manufactures (especially clothing and textiles). It is in these products that the rest of SADC appears to have a relative comparative advantage and where intra-regional trade barriers are high. The current SA-SADC tariff level for clothing exceeds 40%, while those for textiles operate in the 15-75% range (Kalenga 2000).

Table 4.1. Share of intra-regional imports/exports by commodity (1997)

Source: World Trade Analyser

The limited extent of regional intra-industry trade can be explained by lack of industrial capacity (including capital, labor and managerial/technical capacity) but also high tariff levels. Whilst the poor transport and communication links have contributed to this lack of capacity, improving these networks alone may not contribute to their greater integration.

High intra-regional tariffs are being addressed by the SADC FTA, which leaves industrial capacity as one of the main issues. One means of establishing that capacity is to import it through foreign direct investment (FDI). In terms of FDI, South Africa is clearly the most important investor in all the SADC countries, except Angola where the multinational oil companies dominate. Nevertheless, as figure 4.4 below shows, the SADC area hosts only a small share of total South African outward FDI. Table 4.2 shows the distribution of South African investments among SADC countries during the 1990s. Mozambique is the most important recipient of South African investments, followed by Zambia, Malawi and Zimbabwe. It is worth noticing that about 60% of South African investment in the SADC area is greenfield investments or expansion of existing capacity (figure 4.5 below). This means that South African investments in the SADC area create new industrial capacity and jobs.

Table 4.2. South African investment in SADC

Source: Business Map

An analysis of the sectoral composition of South Africa’s FDI in the region reveals that it is primarily exploiting the natural resource base, with only limited investment in manufacturing (figure 4.2). The largest recipient sector by far is metal products. There is, in fact, one investment that accounts for almost half of total South African investment in SADC and that is an aluminum smelter plant in Mozambique. 20The Mozal smelter plant is a joint venture between the London-listed South African company Billiton and Japanese Mitshubitsi. See Nordås and Pretorius (2000) for a discussion. Low-cost supply of electricity and good infrastructure developed through the so-called Maputo development corridor are the major factors attracting this investment. The investment pattern also reflects the fundamentals of location theory for ‘footloose’ industries at work. With little to differentiate the various SADC countries in terms of labor costs and industrial capacity, South African investors are likely to choose those locations that minimize transaction costs (Botswana, Zimbabwe, Mozambique) while providing the other necessary ingredients such as social stability and minimal exchange rate risk. 21Though recently many of the SADC currencies have been more stable than the SA Rand.

Figure 4.2. South African FDI in SADC by sector

Source: Business Map

Recent studies exploring the regional trade and investment networks support these observations. Black and Muradzikwa (2000) explore the automotive industry in the region and conclude that "the integration of the automotive industry in SADC is at an early stage…but real complementarities based on regional specialization and the development of supply chains on a sub-continental scale are minimal"(p. 8). One of the prominent reasons for this is the minimal capacity in the rest of SADC relative to South Africa. In 1998, the rest of SADC produced 24,000 units 22Of this, 14,000 units are produced in Zimbabwe and the rest mostly in Zambia and Botswana. in comparison to the 313,000 produced in South Africa (ibid.). The SADC production has even decreased in subsequent years as the Botswana Hyundai plant closed down. The authors also note that even in automotive component manufacture, the SADC countries lack quality accreditation and are mainly geared towards the aftermarket. The result is that although high tariffs in each SADC country may have originally caused a fragmented regional production structure, removing these barriers is likely to concentrate production in South Africa even further – with the closure of the Botswana plant the first signs of this process. Aside from the capacity reasons, this would occur because there are still considerable economies of scale to be exploited in the South African location while congestion costs have not reached high levels yet. This indicates that as transport/communication costs come down, there is likely to be a more marked agglomeration in those sectors with considerable unexploited economies of scale. It also confers with the finding of Venables (1999), who suggests that FTA’s in developing countries may lead to greater agglomeration in the lead industrial centers. This is because of the small size of the total manufacturing sector in these regions, which implies low congestion costs and unexploited agglomeration opportunities.

Not all sectors suffer from large economies of scale and limited industrial capacity in the rest of SADC. One such sector is clothing and textiles. Visser (2000) reports and extends a study on the supply chain for clothing, e.g., fibers, yarn, textile and clothing. These sectors are where South African producers are under considerable competitive pressure from Asia as tariffs come down, providing a large incentive to relocate to SADC countries with lower labor costs. This study throws some light on the nature of trade and investment patterns and their interrelationship. The purpose of the study was to see whether South African investments in the region created backward linkages to local firms. She finds evidence of emerging regional intra-industry trade and supply chains managed by the multinational South African retailers, such as Woolworths and Pepcor. They have invested extensively in the SADC region, and they outsource "cut make and trim" contracts to local producers by presenting them with a specific design and cover. The study covers Malawi and Mozambique.

Clothing manufacturers in South Africa, Mauritius and Malawi largely obtain key inputs from Asian countries such as Hong Kong, Taiwan, Japan, India and South Korea. A more detailed study was undertaken for Malawi where it was found that about a third of inputs to the Malawian textile and clothing industry is sourced from South Africa and Zimbabwe and two thirds from the Far East (Visser 2000). The reason cited for sourcing such a large share from Asia, is relatively lower prices, good quality and the variety available. In addition, SADC, particularly Zimbabwean, producers of high quality fibers, yarn and textiles often fetch better prices for their produce in the European markets and the US, and therefore reserve only the lowest quality output for the local market. The choice of location for South African investments in the region reflects the role of transport costs once more. Mozambique is geographically close whilst Malawi has lowered its transaction costs through negotiating a preferential tariff with South Africa (Kalenga 2000).

A study of the local impact of South African retail chains’ establishment in Zambia (Kolala 2000), finds very little linkages to the local industry. The only significant sourcing of local products was fresh, perishable food. This, the South African retail stores have in common with local retail stores, which largely offer imported goods from overseas. Kolala further finds that the retailers emphasize distribution efficiency as one of the key determinants of the profitability of their operations in Zambia, and that many of them have acquired their own transport systems in Zambia. Benefits to the local economy from South African investments in the retail sector are identified as employment and training of local staff, including local management; opening supply bottlenecks, introducing modern inventory management, quality control and marketing technology to the local economy. Kolala finally finds that consumers are very quality-conscious and prefer branded international or South African goods to local goods. This is similar to findings from South African townships where particularly the young generation demands branded goods.

Global production networks

Firms in the region may be linked into global production networks rather that regional ones. In terms of manufacturing, South Africa is the dominant force in the region and so it is worth focusing on South African firms’ integration into the global economy to assess where the region is going. Again, trade and investment analysis is informative.

South African trade with industrial countries is typical of North-South trading patterns with a significant share of exports being either natural resources (minerals, agriculture) or beneficiated natural resources (steel, chemicals, agri-processing). However, the increasing trade in manufactured products and rising levels of intra-industry trade (27% of trade in 1997) suggest that South Africa is making some headway in terms of integrating into the global production chains. 23As with the SADC calculation, this figure was calculated at the 2-digit harmonised system level. Calculations at the one-digit level reveal a much higher degree of intra-industry trade – around 54%.

Table 4.3 Share of SA-Developed Country imports/exports by commodity (1997)

Source: World Trade Analyser

The flow of foreign direct investment lends support to the notion that South Africa is succeeding in integrating to some extent into the global networks. In contrast to South Africa’s investment in SADC, investment in South Africa is concentrated in manufacturing and services. In addition, while the other SADC countries have very little outward FDI, the stock of outward FDI by far overshadows the inward stock as far as South Africa is concerned. The structure of inward and outward FDI in South Africa is presented in figures 4.4- 4.6 below.

Figure 4.3. Inward stock of FDI in South Africa by sector end 1998

Source: South African Reserve Bank, 2000

The two dominant recipient sectors of FDI in South Africa are manufacturing (44%) and finance (32%). The strength of the South African minerals companies mean that there is minimal FDI of this type in the country – in stark contrast to most other African countries. In fact, the strength of the minerals sector is reflected in a substantial stock of outward FDI by the sector. Turning to the sources of FDI to South Africa and the destination of South African outward investment, we see from figure 4.4 first that South Africa has a large net outward stock of foreign investment. Second, as should be expected, outward investment is focused on fewer countries than the source of inward investment. The largest stock of outward investment is found in "Other Europe" and is due to the fact that the international arm of one of South Africa’s largest companies, Gencor, was located in Luxembourg. Switzerland is the location of another of South Africa’s largest companies, the diamond cartel De Beers. Figure 4.4 also depicts a strong link between South Africa and the UK, while Africa still is relatively small as far as outward South African FSI is concerned. Finally, the large stock of FDI from Malaysia is worth noticing. This is of more recent origin, as Malaysia has invested heavily in South Africa’s telecommunication and energy sectors since 1994.

Figure 4.4. South African FDI stock by country

Source: South African Reserve Bank 2000.

Finally, figure 4.5 shows that about 30% of inward FDI in South Africa is greenfield investment or expansion of existing capacity. A reason for this is the domestic market power of South African firms, making it costly to ‘buy’ market share in the country. The implications are that in contrast to the other SADC countries, most FDI in South Africa does not create further industrial capacity or jobs. At least not in the short run.

Figure 4.5. FDI by type

Source: Business Map

Case study evidence on the ability of South African firms to upgrade their performance and integrate into the global production chains is mixed. On a positive note, Black (2000) finds that the South African production plants of German auto manufacturers and their component suppliers are becoming increasingly integrated into the global network 24The Japanese and American auto manufacturing plants in South Africa have yet to take on an important role in the global production chain.. Prominent reasons for this were a) the incentive structures of the Motor Industry Development Programme (MIDP), b) the expansion of global production capacity, and c) a strategy amongst the German firms to locate the new capacity outside of an increasingly high-cost Germany. 25The USA firms have not taken advantage of this opportunity due to surplus capacity worldwide, while the Japanese-based firms are only partially owned by the parent companies and so they have less incentive to integrate the production into a global network. Under the MIDP, exports of vehicles or components gives the firm import credits, which can be used to import components or other models free of tariffs. South Africa is now used as a production base from which to supply a single model to either Asian or European markets for BMW, VW and Daimler-Chrysler. In addition, a BMW subsidiary, SA Trim, provides almost all of BMW’s auto leather requirements globally (Black 2000). Other large component exports include catalytic converters, tyres and silencers/exhaust pipes. The success of integration into the global networks is reflected in the fact that component exports have increased from R400m in 1990 to R9.6 billion in 1999, while vehicle exports have increased from R380m in 1990 to R5.1 billion in 1999. Considering the motor vehicle industry is the original JIT sector, these achievements reveal a depth of logistical and manufacturing capacity.

However, other studies reveal a less positive picture. A study by Kaplinsky and Morris (1999) of the clothing, textile and automobile sectors in KwaZulu-Natal, finds that South African companies have so far had difficulties in coping with the new competitive environment introduced after 1994. They find that the clothing industry is not capable of competing with low-price imports from Asia, and has therefore concentrated on the upmarket fashion segments. However, in this market quality, delivery reliability, innovation, speed of response, flexibility of order offtake and financial soundness are critical competitive factors in addition to price competitiveness. The study finds that the companies considered have responded poorly to these changes and retailers anticipated a substantial shift to foreign manufacturers (Kaplinsky and Morris 1999).

The global trend of shift towards more differentiated and customized products requires a flexible work organization, organized around circles with multi-skilled workers rather than assembly lines with long runs of identical products. South African firms in other words need to introduce elements of the Japanese management techniques we discussed in the introduction to this paper. Kaplinsky and Morris (1999), however, found that the change to more diversified outputs resulted in a substantial increase in cost, and little of the necessary change in work organization that allows low-cost flexible production. Inability to shift to a more flexible work organization stems from lack of skills both on the part of workers and management. Illiteracy is widespread among workers, while management has not been exposed to modern management techniques and work organization. Kaplinsky and Morris found that they did not even realize that there was a problem and very little resources were devoted to systematic training.

The KwaZulu-Natal study finds that the sectors studied lagged far behind international practice in supply chain management, and that little was being done to rectify the problem. On the other hand the study by Black (2000) paints a more positive picture of the performance of the automotive industry in terms of output and export growth, but does not study work organization and business practices in the industry. The conclusions of the two studies can, therefore, both be valid. Over time the FDI inflows, although most of them are mergers or acquisitions, will probably contribute to the transfer of international best practice in supply chain management. Some firms, particularly "new economy firms," already performs well in terms of productivity and management as will be further discussed in the next section.

South Africa as a regional engine of growth?

The evidence on production networks in the SADC region suggest that whilst South Africa is beginning to integrate into global production networks, the rest of SADC is having considerably less success. Poor infrastructure, a lack of industrial capacity and a lack of experience in supply chain management suggest that these countries will find it difficult to break into the global networks in the short/medium term. However, the SADC countries have an opportunity to integrate into South African production networks, which can form the basis for learning and establishing industrial capacity. South African firms have an interest in expanding the networks into the region to tap into the lower labor costs and natural resources. These firms are coming under increasing competitive pressure as the country liberalizes its trade regime, and a regional production network may well constitute a viable survival strategy.

This process has already begun in certain labor-intensive sectors where capacity exists in the region. However, the importance of relatively high wage costs in South Africa and relatively high transaction costs in other SADC countries may still go in South Africa’s favor, providing little incentive for South African firms to explore further integration. The closest neighbors to South Africa may be an exception to this as communication networks to South Africa are relatively well developed. The danger is that lower communication costs and trade liberalization, at least initially result in further agglomeration in South Africa and a loss of industrial capacity in the rest of SADC as predicted by Puga and Venables’ (1998) location theory. It is these fears that are delaying the implementation of the FTA by the poorer countries in the region. Another crucial barrier to bringing about the FTA is the loss of fiscal revenue for SADC countries. As a considerable share of their imports come from South Africa, tariff elimination on these imports will make a sizeable reduction in fiscal revenue. However, the FTA is a crucial step in lowering internal trade costs and inducing South African firms to create regional production networks. The FTA can also play a role in bringing about macroeconomic policy convergence in the region, lowering many of the financial risks that face long-term investments such as these. It is no surprise that most FDI entering the rest of SADC is in either location-bound natural-resource sectors or sectors that are in distress in South Africa (clothing/textiles) and so willing to overcome the risks and logistical problems of the region.

South Africa’s inward and outward investment patterns are similar to the capital-rich OECD countries, while its trade patterns with the OECD is more similar to the typical North-South trade patterns. In contrast, both trade and investment relations between South Africa and the rest of the SADC area exhibit the typical North-South patterns. These macro data suggests that South Africa might be in a good position to embark on a catch up growth path, and also serve as a vehicle for bridging the gap between the SADC area and the rest of the world. In particular, South Africa’s physical, financial and economic infrastructure is well developed and should be conducive to growth. There are, however problems at the micro-economic level and rigidities, particularly in the labor market that could still prevent the economy from reaching its potential. Insufficient infrastructure and political instability in several of the SADC countries in turn may constitute a barrier to global as well as regional production networks.

To conclude this section, trade and investment patterns in the SADC area largely reflect arms-length trade in homogenous products produced in the primary sector, which is also the sector that attracts most FDI outside South Africa and Mauritius. These commodities are still kept in just-in-case storages by specialized firms and in some cases government reserves. The increasing time-sensitiveness of international production networks will probably not erect significant new trade barriers to these industries. Furthermore, technological developments in telecommunications and liberalization of trade in telecommunication services will reduce the cost of information and most likely improve commodity-exporting countries’ terms of trade. This is because the relative price of information intensive goods decline when the cost of information declines (De Groot and Nordås 2000). For other industries, however, weak quality control and high transport costs are likely to increase in relative importance as time-sensitiveness increases, transport costs come down elsewhere, and product control units are reduced to a minimum in downstream industries. The combined effect of these developments may well be higher barriers to entry into international production networks.

Regional networks are not necessarily an alternative to global networks. As we say in section 3, it is intra-SADC land transport that accounts for most of the transport cost disadvantage of SADC firms. Thus, even in the clothing and textile industry where we see the emergence of a regional supply chain, still the larger part of intermediate inputs are imported from overseas. Finally, a simple correlation analysis finds that there is a stronger correlation between FDI and having a coastline than between FDI and having a border with South Africa among the SADC countries, albeit both correlation coefficients are small in absolute value. 26A correlation between percentage change in the stock of inward FDI between 1990 and 1999 and having a coastline was found to be 0.26, while the correlation between the same indicator of FDI and having a border with South Africa was 0.22.