Primary exports remain the most important link to the global economy for many countries in Sub-Saharan Africa. Arising from the emphasis on manufacturing, the importance of the primary sector exports has been underplayed. The de-emphasis on the primary in efforts to diversify the economies and partly stem the effects of the decline in primary commodity export prices when looking for lessons from South East Asia for other developing countries. Chibber and Leechor (1995) found that one of the key factors to the integration of Southeast Asia was the expansion of primary exports. This led to a surplus that was important not only in its own right but also as a basis for the subsequent upgrading and shift to manufacturing. The question is how have sub-Saharan African countries performed on both primary resource exports and manufactured exports over the last 30 years and what have been the factors that played a considerable influence on the direction of SSA export performance. The trend and factors that have affected export performance in South East Asia is the subject of another paper (Jomo, 1998).

The 1997 report to the UN Secretary General indicates that economic growth in Africa slowed to 3 percent from 4.4 percent in 1996. Declines in agricultural production and exports as well as in oil prices contributed to slow growth. Given the significance of agriculture, low output adversely affected incomes, consumption and the growth of the processing sectors. In some countries, political instability or civil strife also affected negatively economic growth in 1997 (UN, 1997). Growth in Africa could further be affected adversely if expected export growth is held back by the currency crisis and economic slowdown in East Asia which has become the fastest-growing trading partner for some African countries (for example, South Africa) in recent years.

In addition to the introduction, this paper is presented in five main sections. While section II presents an account of the export performance in SSA over the last three decades, in section III, we present an analysis of the factors that have influenced export performance over the relevant period. Section III deals mainly with sub-Saharan African countries’ resource exports (agricultural raw materials, fuel and numeral exports) and section IV deals with the factors affecting processed exports, mainly manufactured exports.

In section V we present a brief account of the institutional framework for export expansion and promotion while section VI presents some concluding remarks.

II. EXPORT PERFORMANCE OF SUB-SAHARAN COUNTRIES

In recent years, contrary to the period of the mid-1980's SSA countries’ exports have began to show some signs of recovery. This may be attributed partly to the structural adjustment and trade liberalization measures since the early 1980's. For example Africa’s exports grew by 12 per cent in 1995 in value terms, a considerable improvement compared to the 3 per cent of the previous year and the preceding three years of negative growth in value. However, Africa’s export growth continued to lag behind world trade in both value and volume terms. Yeats and Ng (1996) have shown that while world trade in all non fuel goods increased at a compound rate of 11.8 percent, the corresponding growth rate for 30 African products was more than six and half percentage points lower. Based on the available statistics, countries with growth in export exceeding 20 per cent included Angola, Central African Republic, Kenya, Reunion, Tanzania and Togo (WTO, 1996).

As observed from Table 1 merchandize exports for a selected number of SSA countries experienced fluctuations over the period 1984 to 1995, and except for the period between 1991 and 1995 the trend has generally been upwards. Commodity exports experienced much higher degrees of fluctuations in export earnings than mineral and oil exporters. Figure 1 shows that as a proportion of world exports SSA countries have experienced a general decline of all categories of resource based exports for the period 1970-1995. As a percentage of total world export SSA share dropped from 0.8 percent in 1970 to about 0.3 per cent in 1995 (see also table 10).

Table 2 indicates that most SSA countries are essentially primary commodity exporters where the share of primary commodity exports as a proportion of total merchandize exports have remained high and in most countries, except Mauritius and Zimbabwe, the share has remained above 70 per cent. However, Kenya and Seychelles experienced significant improvements in export diversification where the share of primary commodity exports in total merchandize exports declined from 93.8 per cent in 1980 to 70.9 per cent in 1993 for Kenya and from 96.9 per cent to 71.1 per cent for Seychelles.

Not only are most SSA countries dependent on primary commodity exports, but export earnings are also highly concentrated in a few primary products. Table 3 shows, for example, that some countries are dependent on three commodities for as much as 90 per cent of their total merchandize exports and for most of the countries above 50 per cent. African countries’ exports are also concentrated in a few markets, mainly the European Union, North America and Japan. This has implications for market access calling for both product and market diversification.

A question of considerable interest is which countries were primarily responsible for the erosion of Africa’s market share. Was the erosion broad based in terms of the competition, or was one or two groups of countries primarily responsible? Yeats and Ng have shown that the OECD countries themselves made the largest overall displacement of African exports. Specifically, while Africa’s trade shares fell by 11.1 percentage points for these products, OECD shares rose by 9.9 percentage points. Market shares for middle income Asia rose by over 4 percent, while those for other (non OECD) Europe and central Asia increased by almost the same amount. In contrast, Latin America’s trade shares dropped by about 4 percentage points which was about one- third the overall African losses. Perhaps the key point to note is that no other country group has experienced any general loss of competitive position which come close to matching that for Africa. Comparing Asia, Latin America and Africa, it has been observed that in both Asia and Latin America, non-traditional crops expanded while the composition of exports from SSA remained fairly stagnant, (Akiyama and Larson, 1994). In Latin America, export expansion included fruits, vegetables and oilseed production. In Asia on the other hand, production increases affected all sectors of agriculture with the largest gains coming from fruits and vegetables.

Yeats and Ng (1996) have shown that for a combined total of 30 African export products, Africa’s market share declined by over 11 percentage points (from 20.8 percent to 9.7 percent) which implies annual trade losses for the region of just under $11 billion which is equivalent to the total development assistance of $10.9 billion in 1991. Over the decade of the 1980s for example, we witnessed rapid declines in the share of world export for both Ghana and Zambia. Ghana’s merchandize export earnings declined from $1104 million in 1980 to $560 in 1984, recovered to $890 by 1990 but have experienced an upward trend since 1990 primarily due to the structural adjustment and trade liberalization measures undertaken since the early 1980s. Zambia on the other hand, had its merchandize export earnings drop from $1460 in 1980 to $550 in 1985 arising from the very rapid drop in world copper prices. Copper is Zambia’s main export product.. The trend, though fluctuating has been reversed since 1987. We have also witnesses sharp declines in the rates of growth of GDP and per capita GNP for both countries. Is this purely a case of lack of market access, terms of trade shock, domestic supply constraint or a combination of the three factors? In section III an attempt will be made to determine the main facors that have influenced SSA countries’ export supply.

Elbadawi (1998) has shown that the evolution of the share of manufacturing in total export has remained extremely low in most SSA countries, although some of them have made considerable progress in this respect. Most remarkable examples are Cote d’Ivoire and Mali in the CFA zone and Ghana, Madagascar, and Tanzania (during the 1990s) in the non CFA zone, which achieved a steady increase in the export share of manufacturing (Table 1 and Table 4). Mauritius has been particularly successful in promoting manufactured exports primarily arising from a policy of Export Processing Zones (EPZ). Starting in the 1970s with an export share of manufacturing lower than in Kenya, and with a share of MVA in GDP lower than in Senegal, manufactured exports in this country reached in the first half of the 1990s more than two-thirds of total exports (Elbadawi, 1998).

Notwithstanding efficiency consideration, the share of gross investment to GDP is a useful broad indicator of an economy’s potential to sustain high rates of export (as well as overall economic) growth. On this score Elbadawi (1998) has pointed out that most African countries lag behind the fast growing economies. Except for Mauritius and Tanzania--which have investment ratios comparable to those of Chile and Cost Rica, between 26 and 31%--virtually all the remaining African countries have investment rates lower than 20%. Few other countries such as South Africa, Mauritius and to some extent Zimbabwe have shown substantial capacity on these areas.

III. FACTORS AFFECTING EXPORT PERFORMANCE

SSA countries have become increasingly marginalized in the world economy in terms of their share in world trade and output (both agricultural and industrial). Heavy dependence on primary commodities is one important reason for their slow growth of exports and of their economies in general. However, most SSA have undertaken trade liberalization since the early 1980s in efforts to expand and diversity output and exports.

During the last decade, most of the economies in SSA have been liberalized from command economies and one party state to macro economic and trade liberalization and contested politics. State trading and marketing boards have been eliminated and in a majority of cases, export taxes have either been eliminated or reduced drastically, except in some countries in West Africa. With the changed policy environment, the micro level sources of export bias have to a large extent been reduced.

The extent of trade liberalization in SSA appears to have been quite impressive, but has it improved economic performance? To answer this question, one has to bear in mind what liberalization was supposed to achieve, namely an expansion of exports through a diversion of resources from the domestic to the export sector. Such export orientation would in turn, it is maintained, lead to faster growth of GDP (see, for example, Balassa, 1982). Moreover, by removing the traditional bias against exports and productions of manufactures, trade liberalization would lead to diversification of production and exports in favour of manufactures.

Recently completed AERC studies on trade liberalization in ten countries in SSA have indicated clearly that domestic supply constraint constitute a significant part of the anti-export bias observed over the last three decades.

A measure widely used by the AERC case studies to determine whether trade liberalization took place is the extent of reduction of the anti-export bias related to the application of both tariff and non-tariff barriers for protective purposes. The results for cumulative effects of the various liberalization episodes as summarized by Oyejide, Ndulu and Gunning (1997) confirm that anti-export bias has generally been on a downward trend particularly if one takes into account steep decline in exchange rate premia as quantitative restrictions were dismantled and a general downward trend of import tariff rates.

Oyejide, Ndulu and Gunning also point out that the switch to the use of exchange rate for clearing disequilibria in the market for foreign exchange has considerably reduced the need for using trade policy instruments for managing balance of payments pressures. What has been paused as the critical question is whether this momentum can be sustained in view of the fact that the process of liberalization to a dominant extent has been prompted by external pressure and sustained to a significant extent so far by large inflows of external resources to ameliorate adjustment costs along the way. This is an empirical question which we will not attempt to answer in this paper. Our aim is to present an apirical assessment of the export suply constraints that SSA countries have experiences over the last 30 years and their effect on export performance.

The ability of developing countries to take advantage of the emerging opportunities in world markets depends crucially on the ability to foster the development of internationally competitive industries which can meet strict standards of cost, quality, reliability and delivery schedules. Supply capabilities can be a major constraint on the ability to exploit the opportunity arising from globalization.

On the impacts of trade liberalization for example, the most widely cited one is that liberalization provides expanded market opportunities, which when coupled with reduced discrimination against exports, allow exploitation of comparative advantage, permits greater capacity utilization and exploitation of scale economies. Two is that liberalization through reducing anti export-bias stimulates export performance, particularly non traditional exports. Openness to international trade and complementary macroeconomic policies are regarded as key to successful export-led industrialization and rapid growth.

The AERC studies have also shown that trade liberalization has been associated with increased export orientation and higher trade shares in all countries in the study. Most of the studies though, identify real depreciation rather than trade policies per se as the key explanatory factor behind the improved export performance. It has however, been acknowledged that a fundamental problem lies in the attribution of observed post liberalization changes in economic performance to changes in trade policies.

Using a panel data of 22 SSA for the period 1970-1995 we have tried to determine the factors that influence export performance for the categories of resource exports of agricultural raw materials, fuels and mineral exports.

The regression equation is presented as:

Yit = BiXit + ei

where Yit is a vector of the dependent variable

Bi is a vector of coefficient

Xi represent a block metrix of the explanatory variables

ei is a vector of the random error terms

t = 1970-1995.

The key variables are the real exchange and government expenditure, as measures of macroeconomic stability, gross domestic investment as a measure of infrastructure development, and an economy’s potential to sustain high rates of export growth, corruption as a proxy for the cost of doing business and external tariff as a measure of market access. The results show that while the combined effect of all the factors is about 74% in the case of agriculture raw materials and about 86% in the case of minerals, ores and metal exports, their effect on fuel exports is negligible. In all three cases, the real exchange rate is significant but negative while the lag of the RER is significant and positive. If it represents variability then it has perverse influence on export supply. Elbadawi (1998) has pointed out that the absolute level of the RER (or its equilibrium level) is irrelevant to export performance. This observation arises from his econometric results which show that the levels of RER were not significantly related to exports. The variable government expenditure is significant in the case of fuels and mineral exports but not in the case of agricultural raw materials although it has the right sign. The variable corruption does not seem to constitute a significant cost of doing business in SSA. The lag of gross domestic investment is important and significant in the case of agricultural raw material export supply. This is consistent with the postulation that gross domestic investment represents infrustructural development which is very essential for sustained export growth particularly in the movement and storage of agricultural exports in SSA. We have observed in some countries for example, that during periods of significant output increases, some of the output increases were wasted as a result of poor transport and storage facilities. External tariffs were not significant in the case of agricultural raw materials and mineral exports. Tariffs are, however, significant in the case of fuel exports. It should be noted, for example, that products of the ACP countries, the majority of which are in sub-Saharan Africa, that enter the EU market free of customs duty accounted for 99.3% of total EU imports from the ACP in 1996. Exceptions cover products under the Common Agricultural Policy. It should be also noted that ACPs share in extra-EU imports of agricultural products was 14% compared to a mere 3 percent in the case of industrial products in 1996. This not only shows the significant role of agricultural exports in total exports but that very few of such products face entry barriers. The regression results as presented in tables 6, 7, and 8 are very preliminary, and they would need to be refined to make them robust to be able to draw policy lessons.

IV. FACTORS AFFECTING PROCESS EXPORTS

Wood and Berge (1997) have shown that East Asia’s ‘miraculous’ development success (in terms of equity as well as growth) has been intimately associated with the export of manufactures. By contrast, countries whose exports still consist largely of primary products-most notably in Africa-have done far less well. This broad correlation between export composition and development performance raises some controversial questions in the development literature, both about the causes of economic progress and about the best policies for achieving it. The question that arises from this analysis is whether African countries can emulate the development experience of the East Asian countries.

Recent research (Wood and Berge, 1994, 1997) raises serious doubts, however, about the scope for other developing countries to follow East Asia down the road of export-oriented industrialization. The problem is that many of these countries, particularly those in Africa and to some extent Latin America, do not have a comparative advantage in manufacturing, because they have the wrong resource endowments. More specifically, they have too low a ratio of human resources to natural resources, or, in other words, of skill to land, which causes their comparative advantage to lie instead in primary exports.

This question overlaps with another one, long discussed in the literature, concerning the benefits to developing countries of further processing their primary exports. Notable contributions to this debate on "resource-based industrialization," which includes many case studies of particular products and countries, are Roemer (1977), Singer (1978), Wall (1987), Yeats (1991) and Londero and Teitel (1996).

Wood and Berge (1997) point to a minor distruction between primary processing and narrow manufacturing - as regards the importance of transport costs, and the volatility of primary commodity prices, for example. In general, however, most authors conclude that the similarities between these two sorts of manufacturing outweigh the differences. Primary processing, like narrow manufacturing, provides opportunities to acquire new technologies and learn new skills, and can be an important new source of export revenue. Growth of primary processing is constrained, like growth of narrow manufacturing, by protectionist policies in developed countries, and by shortages of skills and infrastructure in developing countries. Whether significant gains can be reaped from further processing of local raw materials thus varies, depending on the product and on the circumstance of the developing country concerned. Sub-Saharan African countries are faced with numerous supply constraints problems particularly those related to processing and manufacturing for export. Top on the list are infrastructure problems.

Some of the infrastructure related supply constraints include: frequent power cuts and water shortages which greatly affect industrial production, poor road network and in particular lack of all weather roads and feeder roads necessary for the transportation of agricultural produce from villages to major centers, insufficient rolling stock, and lack of refrigerated trucks and cold storage facilities for perishables. Some of the services necessary to support production, such as the provision of adequate finance or marketing services may be lacking or often inaccessible.

It has been argued (Yeats, Amjadi, and Ng 1997) that many African countries adopted anti-competitive cargo reservation policies to foster the development of material fleet and to conserve foreign exchange but without any success. They cite, for example, that in 1990/91 Sub-Saharan Africa’s net freight and insurance payments were about $3.9 billion, or roughly 15 percent of the value of the region’s exports, compared with 11 percent in 1970 and for a third of the countries, more than 25 percent of the value of exports exceeding 70 percent for Somalia and Uganda.

Yeats et al (1997), have also shown that Africa is at a transport cost disadvantage relative to its competitors. For example, half the minimal vessel freight rates for middle income West Africa are about 2 percent points higher than those paid by other exporters of the same goods.

The cost of doing business in Africa relative to other parts of the world is further complicated by the very low of access to information systems. While everybody is moving towards the information superhighway, most African countries still control the airwaves which makes it extremely expensive to get direct satellite connection since the controlling agencies extract monopoly rents which they are not readily willing to let go.

Africa’s transport and telecommunication policies and international freight costs have a major negative impact on the promotion and diversification of exports.

To build export supply capabilities and encourage product diversification, specific measures and incentives are necessary with regard to investment (including foreign direct investment), technology acquisition, and human resources development, as well as direct fiscal and financial export incentives. While appropriate exchange rate policy is central to a success export promotion strategy, an export-promoting exchange rate policy cannot be sustained unless monetary and fiscal policies are fully consistent with it. In many developing countries, mismanagement of macroeconomic and trade policies led to real exchange rate misalignment - that is to a substantially overvalued RER with respect to its market clearing level. Real exchange rate misalignment is damaging to economic performance - and especially to manufactured exports, as it decreases the profitability of production of tradables. All successful East and Southeast Asian countries have kept the RER close to its market clearing level, while Sub-Saharan Africa and Latin America countries experienced serious RER overvaluation.

Moreover, inconsistence macroeconomic, trade, and exchange rate policies increase the variability of the real exchange rate. In turn, higher RER volatility sends conflicting signals to economic agents and increases uncertainty of long-term investments as well as of the profitability of producing tradable goods. The negative influence of RER variability on economic performance of SSA countries has been demonstrated by Ghura and Grennes (1993). Its negative impact on manufactured exports has also been established by Grobar (1993) on a panel of ten developing countries excluding Sub-Saharan Africa (.................).

Using panel data representing 16 countries over the period 1970 to 1995 we estimated a manufactured (representing processed exports) export supply equation using the real exchange rate, gross domestic investment, government expenditure, freight costs and external tariffs as explanatory variables. The regression equation is as earlier specified. The real exchange rate and government expenditure are used as proxies for macroeconomic stability which is conducive to manufactured export expansion. Manufactured exports are also likely to be associated with investment in capital and intermediate inputs and these are represented by the variable government expenditure. Gross domestic investment is used as a measure of infrastructural development necessary for export development and export promotion. It has been argued (Yeats, Amjadi and Ng 1977) that sub-Saharan African exports face disproportionately higher transport costs when compared to exports from other parts of the world to the EU, USA and Japan. We have used an index of freight costs to reflect transport costs faced by SSA manufactured exports as well as tariff faced by manufactures exports as a measure of the barriers to entry. However, this may not be a good indicator as tariffs on manufactures vary from one product to another and from one country to another. For example, Mauritius which has a highly developed textile sector faces higher tariffs on its textile than say, Tanzania.

The regression results are presented in Table 9. In interpreting the results, one should note that the period of estimation 1970-1995 represent three very distinct periods as far as export development and promotion are concerned. The period of the 1970s represent stable macroeconomic environment followed by a period of extreme macroeconomic instability in 1980 and again followed by a period of relative stable macroeconomic environment of the late 1980's and early 1990's. It should also be noted that the sample of countries in SSA, represent both CFA and non-CFA countries with differing macroeconomic environment over the relevant period. Hence, the results show that real exchange rate is not significantly related to exports. However, the results also show that gross domestic investment is important and significantly related to exports. Investment including investment in human capital has been credited for the tremendous export-led growth of most of the South East Asian economies. While government expenditure is unimportant and insignificant in explaining SSA manufactured exports, freight charges negatively affects exports and is highly significant. This is consistent with earlier observations. Overall, the explanatory variables jointly accounts for about 83 per cent of manufactured exports in SSA.

V. THE INSTITUTIONAL FRAMEWORK FOR EXPORT EXPANSION

Getting prices right, though necessary, is not a sufficient condition to trigger the requisite supply response. Institutions do matter. In this section we will examine the institutional framework that is conducive for efficient market operations and enhanced exports.

Aron (1996) explain the failure of supply response in Africa after "getting the prices right" on the basis of cumulative institutional impoverishment. "Getting the prices right" may be useless in the absence of getting the institutions and the rules right. She argues that macroeconomic performance is positively related to the extent of development of the institutional framework and that rules and contentions incorporated in the institutional fabric of the society and the economy narrow the feasible policy space and reduce the scope for the discretionary action by decision makers. Consistent adherence to a set of macroeconomic rules plays a crucial role in providing a stable, predictable and credible macroeconomics environment and encourages a process of institutional change favourable to growth and poverty alleviation.

She argues further that a written constitution, if it is to be meaningful, must reflect the underlying values of the population. In turn the existence of a working constitution acts as both a guide to, and restraint on the actions of the state. It commits the Government to follow certain objectives and principles and helps provide a sense of continuity, consistency, and credibility to its actions.

The experience of the North Asia and South East Asian economies may shed some further light on the role of institutions. It has been demonstrated that while many nationalist regimes of the post colonial era in Africa and Latin America promoted industrialization directly through state enterprises, some regimes in North Asia used protection and support for import substituting industries to make them produce for export, thus raising the quality and efficiency to achieve international competitiveness. Government administrative competence has been cited as the single most important explanatory variable determining differing economic performance (Jomo, 1997).

A successful strategy for export expansion will need to be based on an interactive approach to supply, demand and export marketing in a framework of close cooperation between the government and the business community. The strategy will need to focus on export product and market diversification in a dynamic search for comparative-advantage in product and market niches.

To diversify markets and take full advantage of trading opportunities, attention needs to be given not only to demand conditions and market access opportunities on major world markets but also to regional markets including the opportunities for cross-border trade. The recently completed AERC studies have, however, shown that intra African trade has been minimal. The explanation has been found in the structure of production (primary and raw materials), marketing channels (North-South), poor infrastructure and lack of information on markets and the small size of the African economies.

Export marketing support and other trade-related services are crucial if small and medium-sized (SMEs) are to be able to export and compete on world markets. Given the relatively small size of developing-country exports, government-backed trade-related services are an essential component of any export expansion strategy. In this regard, key services include export financing schemes, quality control, marketing and distribution services and the trade promotion activities of trade promotion organizations (TPOs).

A major concern is how to make TPOs more effective. In the post Uruguay Round situation of increased global competition, TPOs are necessary because of a greater need for information, especially by SMEs, on market opportunities and trends in diverse markets around the world, as well as to make known the products of a country in various markets (UNCTAD,1995).

VI. CONCLUSION

In this paper we have shown that the export performance of SSA over the last three decades has not been encouraging. In fact, SSA has lost its share of world exports by over 250% over the last 30 years. All categories of exports for all sub-regions in SSA including South Africa have faced drastic falls in export earnings. SSA faced serious import compression particularly during the period of the 1980s,and since trade is the main vehicle for Africa’s participation in the global economy, its participation has declined and its place taken over by the fastest growing economies of East Asia and Latin America as well as the OECD countries.

Both institutional and policy induced factors have affected SSA’s export performance. Preliminary regression estimates for all categories of exports from SSA have shown that real exchange rate, government expenditure, gross investment and freight costs in the case of manufactured exports influence the supply of exports. It has not been easy to model institution, but evidence from the literature indicates that institutions do matter and the performance of trade promotion institutions is central in this regard. It should be noted that this analysis is very preliminary and subject to further revisions. Further refinement of the analysis should produce robust results from which policy implications may be drawn.

FIGURE 1 : SUB-SAHARAN AFRICAN EXPORTS BY CATEGORY AS A RATIO TO TOTAL WORLD EXPORTS

FIGURE 2 : REGIONAL AFRICAN EXPORTS BY CATEGORY AS A RATIO TO TOTAL WORLD EXPORTS