Abstract: This study yields a number of insights into the analysis of agricultural export trade in SSA. Statistical evidence reveals that SSA countries have significantly liberalized their trade regimes over the past two decades or thereabout. This is evident in the significant reduction of tariffs and non-tariff barriers, removal of exchange rate control and elimination of export marketing boards among other trade reform measures. While the study was able to confirm the distortions caused by trade barriers on export performance, there is not much evidence that the adoption of trade policy reforms, since the mid 1980's have produced a significant agricultural export response. Declining infrastructure investment, particularly transport infrastructure (air/sea port, railways and roads) have generated substantial transaction costs and dampened the response of agricultural exports to the more favourable trade environment being witnessed, since the mid 1980's. In addition, factors external to an individual country such as relative prices and income of trading partners appears to be more important determinants of exports than a country’s own structural adjustment policies. This however does not imply that at the margin, structural adjustment policy reform is not beneficial. The simple point is that there are many factors other than structural policy that help explain the poor agricultural export performance of SSA countries. Domestic policies are necessary to reduce the various constraints on supply response increase transport and marketing efficiency and encouraging investment. To benefit from agricultural trade, SSA countries need to increase the flexibility and efficiency of resource use so that they can be competitive in global markets.

INTRODUCTION

The dismal macroeconomic performance of SSA countries after the adoption of Import Substitution Industrialization (ISI) strategy prompted the International Monetary Fund (IMF) and the World Bank (WB) to suggest an outward-oriented development strategy to SSA countries as part of their structural adjustment and reform programmes (Onafowora and Owoye, 1998). The adoption of this type of developmental strategy, they argued, would enable SSA countries to attain higher economic growth propelled by a dynamic and more robust external sector. As a result, many SSA countries from the mid 1980's adopted the Structural Adjustment Programme (SAP) aimed at liberalizing their economies particularly the external sector. The structural reforms focused on supply and its redirection in the context of export promotion strategies. This new development option which marked SSA total departure from the import substitution strategy, sought to redirect growth-promoting strategies towards the external market (Hammouda, 2004). Consequently, trade policies in most SSA countries went through major changes within the context of Structural Adjustment Programme (SAP) during this period. Foreign trade was liberalized through the reduction of tariffs and non-tariff barriers and reduction of import duties applied to imports in a large number of SSA countries. Import permits were abolished and duty rates as part of tariff liberalization were also lowered in many SSA countries. Currencies were devalued to encourage exporters with the aim of boosting exports and growth and fostering the integration of SSA into the global economy. A sizeable number of SSA countries virtually eliminated parallel market premiums with buying and selling of foreign exchange becoming market-based while abolishing previous restrictions on currency transactions. Thus, this new policy strategy attempted to promote greater openness in order to boost growth and encourage the competitive integration of the SSA economies into the globalizing world.

Although, the general consensus was on the need to design and implement economic reforms in SSA, it was and is still not certain if the agricultural exports of the region would be enhanced through the adoption of programmes that encourage more open economic policies. This is because despite significant trade policy reforms over the past two decades, export performance has remained dismal. Trade flows from SSA are still characterized by shrinking share of SSA’s in world trade. Trade structure of SSA countries also reveals high dependence of exports on unprocessed primary commodities and the direction of trade heavily leaning towards European trade partners.

Many SSA countries have also not put in place the policies necessary to raise living standards by improving export shares in traditional markets thereby encouraging rapid diversification. Consequently, the focus of this study is to examine the supply response of agricultural exports to adjustment reforms in SSA. Are there other existing barriers apart from trade policy barriers that limit the supply response of agricultural exports to structural adjustment reform measures in SSA? Is there a fundamental role for external demand and world price in SSA export performance? These questions will constitute the main research focus of this study.

MATERIALS AND METHODS

Trade liberalization in sub-Saharan Africa: Import Substitution Industrialization (ISI) strategies were at the heart of SSA development strategies during the 1960 and 1970's. The import-substitution strategies adopted were meant to produce locally the consumer goods which had previously been imported from developed countries, so as to promote the diversification of their economies. Many countries in SSA established industries to produce consumer goods, mostly intended for the new urban middle class. The results were initially seen in an average annual industrial growth rate of 5.5% during the 1970's but which fell to 2.5% between 1980 and 1984 and 0.4% between 1984 and 1987, respectively. Manufacturing as a proportion of GDP increased rapidly and there was a rise in industrial employment and in its share of overall employment (Hammouda, 2004). Nevertheless, this strategy soon ran into problems (Bruton, 1998).

The development of final goods production led to rapid increase in imports of intermediate and capital goods, leading to worsening trade imbalances and balance-of-payments deficits. Small domestic markets did not generate sufficient demand for the products of emerging industries, industries were prevented from reaping the advantage of economies of scale and import substitution was biased towards elite urban consumers to maintain political support while focusing on consumer goods for the middle class. It therefore, generated rent seeking behaviour by firms as they took advantage of insulation from international competition. The disappointing outcome of import substitution strategy in SSA was seen most starkly in the poor productivity performance of the new enterprises (Mayer, 1996). Import substitution delivered very poor results in terms of productivity improvement, structural transformation of the economy and export diversification.

The failure of the import substitution strategy and the debt crisis in the early 1980's led to a new consensus on the importance of trade policy reform and exports in growth strategies. This new consensus was the main focus of the reforms initiated by SSA countries and the developing world in general from the early 1980's within the framework of SAPs (Oyejide, 2004). As a result, the mid 1980's witnessed the formulation and implementation of wide-ranging trade policy reforms by most SSA countries with the support of the IMF and the World Bank (2004).

The end result is that starting from the mid 1980's, and especially in the 1990's, most SSA African countries liberalized their trade regime to some extent with many countries reducing trade barriers significantly more than others (especially restrictions on imports). These reforms were aimed at making it easier to import by reducing tariffs and non-tariff barriers and encouraging exports by eliminating export taxes and providing export incentives.

For example, there was a partial or complete conversion of quantitative restrictions to tariffs (except for moral, health security and environmental restrictions in some cases) in most SSA countries. A study by Oyejide et al. (1999) revealed that in line with this new strategy, Mauritius abolished import permits in 1991, Ghana in 1989, Tanzania by 1993, Zambia by 1992, Kenya by 1993 and South Africa reduced 85% of restrictions by 1990. Also, Mauritius compressed its tariff structure from 60-10 tariff categories, Kenya from 25-6, Cote d’Ivoire from 10-6 and Zambia and Tanzania compressed their categories to 3. Mostly, this involves a switch from a positive list of permitted imports to a small negative list of mostly prohibited items or items considered to be luxury goods for the country. The sequence was first to levy import surcharges and then to adjust the minimum and maximum tariff prior to abolishing restrictions.

Import permits were abolished in Mauritius, Ghana, Tanzania, Zambia and Kenya in the 1990s. In addition, duty rates as part of tariff liberalization were also lowered in some SSA countries. Mauritius reduced its rates from 250-100%; Tanzania from 200-60%; Zambia from 150-50% and in Kenya from 170-40%. In Zimbabwe and Ghana, the rates ranged from 5-30% and 10-40%, respectively. Tariffs are now the main trade policy instruments of most SSA countries. A broad picture of trade policy reform can be obtained by examining the trends in SSA countries tariffs level.

Table 1:	Average tariffs by country

A clear pattern emerges from the data shown in Table 1. Average tariffs have been reduced significantly, almost halved on average in SSA over the past 20 years. While the overall variation or spread in tariffs has been reduced, progress varies across the regions of SSA. A breakdown of the tariff rates of various regions is shown in Table 1. It reports average applied trade-weighted import tariff rates for primary products, ores and metals, manufactured products, chemical products and machinery and transport equipment for selected SSA countries.

The pattern of average tariff rates by sector in SSA shown in Table 2 is consistent with the evidence in Table 1. Tariff rates in virtually all the SSA countries were compressed. For example, tariff rates on primary products which averaged 34.4% in 1994 have gone down to 16.0% by 2004 in Kenya. Similarly, Ghana which recorded an average tariff rate of 14.1% on manufactured products now recorded a tariff rate of 12.5% in 2004. Although, tariff rates are a little bit higher in manufacturing than in primary products, the gap is very close and there is no country with average tariffs in primary products and manufactured products in excess of 20%.

The reduction in tariff rates could not be said to have negatively affected revenue collection from this source, partly because of the simultaneous expansion of the tax base made possible by larger programme funded imports and a steep rise of the local currency value of imports due to large currency depreciations. Moreover, revenue increase may have been induced by the tariffication of quantitative restrictions and increased compliance from taxpayers because of the lowering of the scheduled rates. As part of the structural adjustment, other measures were also adopted to reduce anti-export bias in most of the SSA countries. Export taxes and levies were either significantly reduced or totally eliminated in most of the SSA countries under the review.

For example Cameroon removed all export taxes while Mali abolished export levies and duties on most exports (the only export levies in force are the Service Provision Contribution (SPC) of 3% on the free on board (fob) value of gold and the tax of CFAF 7.5 per kg of fish), Ghana has no export quotas or voluntary export restraints. In a similar vein, Uganda replaced its export licensing requirements by a less restrictive export certification system in 1990 and also abolished export taxes. Most exportation in Botswana does not require permits. Significant reduction in the effective rates of protection was also achieved in most of the SSA countries. Countries such as Kenya, South Africa, Ghana, Mali, Tanzania, Zimbabwe and Cote d’Ivoire witnessed a significant reduction in their effective protection rates.

The remaining export prohibitions that still exist in some cases apply only to sensitive goods because of the need to ensure quality and for health and environmental reasons. Export Processing Zones (EPZs) were also established by the government in some of the SSA countries. For example, the free zones act was enacted in the Gambia. Incentives take the form of exemptions or reductions in duties and taxes. Free zone enterprises are required to export a substantial proportion of their production; the government is currently using an indicative benchmark of 70%. Similarly, Mali also created free trade zones as part of the measures adopted to boost export performance. Export Processing zones companies also account for the bulk of manufacturing exports in Mauritius which is dominated by textiles and clothing.

Exchange rate regimes in most of the SSA countries were also liberalized. A good number of SSA countries stopped fixing exchange rates and overvaluing their currencies in order to stimulate exports and make the economy more competitive. Kenya, Uganda, Ghana, Tanzania, Zambia, Nigeria and Cote d’Ivoire virtually eliminated exchange rate premiums where buying and selling of foreign exchange is now market-based and abolishing previous restrictions on current transactions. The system of multiple exchange rates was abolished in Burundi.

From 1996, Ethiopian currency, the Birr was allowed to float thereby resulting in the convergence of the official, auction and parallel market exchange rates. After liberalizing its external sector in 1990, Benin Republic’s currency was devalued and its black market premium averaged only 2% between 1990 and 1999.

Table 2:	Average tariff rates by sector in SSA countries

We can therefore, conclude that most SSA countries witnessed a significant relaxation of trade barriers. Import restrictions are now lower and export barriers have been significantly reduced.

Export performance of sub-Sahara African countries: The export share of the African continent in the world trade has been experiencing a gradual decline since 1960. The share of the Africa’s export in world export which stood at 5.52% in 1960 declined gradually to 2.87% in 2007 before increasing to 3.45% in 2008 (Table 2). This share in total world export is far below the share of the Asian continent in world export. Statistical evidence actually presents a conflicting scenario between Africa and Asia. While the share of the Asian continent in world trade has been enjoying an upward trend the share of the African continent has been declining. A similar comparison is also noted between the African continent and the developing America continent. Thus, the trends showed that Africa has been losing market shares in exports in relative terms. Perhaps, this dismal trend can be linked to the composition of Africa’s merchandise trade. There is also dismal performance in terms of agricultural raw materials exports (Table 3). Except for a country such as Benin that has over 50% of its exports in agricultural raw materials, a sizable number of the SSA countries share of the exports is below 20% (Table 4 and 5).

However, when the absolute performance of agricultural exports is viewed against these indices, it is glaring that agricultural exports have increased. Nevertheless, the growth rate value in Table 6 shows that a good number of SSA countries also recorded negative growth rates in their agricultural exports between 2001 and 2007.

Table 3:	Export share in world trade

Table 4:	Export structure of selected SSA countries by major commodity groups
UNCTAD (2008)

Table 5:	Total agricultural exports (FAO, current US$'000)
Africa Development Indicator (2008)

Table 6:	Annual growth in total agricultural exports for selected SSA countries
Africa Development Indicator (2008)

The growth rate figures reported also indicate a sluggish response in the movement of agricultural exports in most SSA countries.

RESULTS AND DISCUSSION

The discussion in the last section then bring us to the issue of what could have been responsible for the sluggish response of agricultural exports to adjustment reforms in most SSA countries. Some of them are highlighted:

Natural barriers to trade: Aside from the barriers (trade and exchange rate) that structural adjustment seeks to dismantle, there are other factors affecting agricultural export performance in SSA. Trade policy barriers may be only a part of the various barriers to trade among the numerous factors that increase the transactions costs of trade. These other factors include among others natural or geographic barriers such as those associated with distance being remote or landlocked, usually focusing on transport costs. Transport cost is one of the non-policy barriers to trade. This is a particular problem in SSA, not only for the many landlocked countries but also because many countries with sea coasts have large interiors that make transportation of agricultural exportable goods to the sea coasts costly. A standard proxy in the empirical literature for international transport costs is the ratio of the cost, insurance and freight (cif) price to the free on board (fob) price of imports. The ratio captures the significance of transport costs. Table 7 compares such ratios for the 4 regions in SSA in 1980, 1985, 1990 and 1994. Transport costs have been increasing in SSA except in 2000 when it declined slightly to 10% of unit values. However by 2003, it had risen to 30% of unit values. Transportation costs of this nature are a barrier to trade. They are equivalent to a tax on exports, making SSA economies less competitive.

The Central African region which comprises many landlocked countries faced the highest transport costs of over 96% of unit values in 2003 while the East Africa region faced the lowest transport costs. Transport costs declined in both the East and Southern African region from 19 and 14-5 and 8% of unit values between 1985 and 1990. It later rose to 12% of unit values for the East Africa region in 2000. The increase in transport cost for the West Africa region was largely due to Niger republic while Chad and Lesotho were responsible for the large increases recorded in the Central and Southern Africa region, respectively (Table 8). These statistics clearly suggest that international transport costs could have a significant adverse impact on the level of SSA agricultural exports.

Table 7:	Regional transportation costs in SSA

Table 8:	Transport Costs (CIF/FOB) by country, selected years
Researchers’s computation from statistics obtained in IMF IFS (2005)

Nevertheless, these international transport cost statistics do not incorporate the cost of inland transportation or port charges which may be very high for some countries. The importance of these charges should not be underestimated. For example, World Bank data show that port charges for clearing a 20 foot long container through Abidjan, Côte d’Ivoire and Dakar, Senegal were $1,100 and 910, respectively. In contrast, the ocean freight cost for shipping the container to Hamburg, Germany or Le Havre, France ranged between $1,350 and 1,430. In addition, the structure of sub-Saharan Africa’s transport costs appears to have an important adverse impact on the types of goods exported from the region. Specifically, nominal freight costs for many processed commodities (like cocoa powder and butter) are higher than those on the primary unprocessed component (cocoa beans). Similarly, some processed products like plywood and veneers are higher than those on rough or sawn logs. Sub-Saharan African countries may have many reasons for wanting to shift to exports of processed goods (greater price stability, job creation, increased levels of export earnings, etc.) but the structure of freight costs often works against local processing of domestically produced commodities. Thus, remoteness poor infrastructure and being a landlocked country can be a damaging factor to agricultural trade because they increase trade costs and such costs are a burden on many SSA countries.

Market access: Access to foreign markets is a critical determinant of export performance. It relates directly to the characteristics of the trading partner countries such as the size of their market and transport facilities and inversely to their own internal transport costs. It also depends positively on the size of the export basket and the number of differentiated items and their prices which in turn are affected by market entry conditions. Developing countries in sub-Saharan Africa receive one of two types of general preferences: those provided under the European Union’s Lome Convention and those provided by other OECD countries under the Generalized System of Preferences (GSP). Many GSP schemes further differentiate between developing countries in general and those the United Nations (UN) has designated as least developed countries (which are predominantly in sub-Saharan Africa). These schemes extend even lower preferential tariffs to the latter group of countries. Once an account is taken of these preferences, tariffs on SSA exports to all OECD markets are very low. As a result, SSA countries have enjoyed preferential market access to OECD markets, most especially the EU. For example, mauritius has significant exports of textile and clothing manufactures but has relied to a great extent on preferential access to the European Union. Other countries such as Lesotho and Kenya have increased clothing exports to benefit from preferential access to the United States (US) under the African Growth and Opportunity Act (AGOA). In general, preferential access to developed country markets has been an important feature of African exports.

The erosion of trade preferences may imply significant losses for some SSA countries in the case of manufactures such as textiles and agricultural exports which may increase the challenges facing SSA countries attempting to diversify exports beyond processing of commodities. This is because an advantage from a preference can only be available where the preference-giving party (for example the EU) applies some restrictions on its trade with third countries.

Therefore if the EU does not apply restrictions on its trade with third countries, there is no value in granting preferences. Thus while SSA countries might have a trade preference in the form of zero or low duty, the value of such preferences needs to be examined in the light of the duty facing other non-preferential exporters. In addition, aside from tariffs and import duties, industrial countries apply Non-Tariff Barriers (NTBs) such as quotas and restrictive licensing requirements to agricultural exports from SSA. Sandrey classifies-albeit arbitrarily three broad categories of NTBs: those measures put in place to protect the public health, safety and the environment. These include most Sanitary and Phyto-sanitary (SPS) measures and standards; an assortment of trade policy measures such as export assistance, export taxes, import licences, import quotas, production subsidies and the use of trade remedies (Anti-dumping (AD)), safeguard and countervailing duty measures and an array of administrative disincentives to export (which are not regulations per sec for example, customs clearance delays, lack of transparency and consistency in customs procedures, high transport charges and bureaucratic or arbitrary document requirements and their processing for shipments (Maduna, 2005).

Over the past quarter century, Anti-dumping (AD) has emerged as one of the most widespread impediments to international trade. The number of AD initiations per year more than doubled between the late 1980s and the late 1990's, reaching 366 in 2001 and decreasing to about 220 in 2004.

The traditional users (including Australia, Canada, the European Union and the United States of America) accounted for over 80% of total AD initiations in the 1980's and the first half of the 1990's (UNCTAD, 2005). An explanation for the large increase in AD is that it is relatively user-friendly.

Contrary to most other trade policy instruments such as tariff, quotas and voluntary export restraints, AD has not been brought under strict multilateral discipline through the GATT or WTO. This has led to an increasing gap between the legal definition of dumping and any economic notion of dumping: AD has less to do with combating unfair trade and more with improving the competitive position of the complainant against companies of countries against which complaint is lodged. In other words, AD in many cases has become a contemporary form of trade protection. This has important implications for the agricultural export prospects of SSA countries trying to upgrade export products which includes improving domestic contents or selling their own-brand products through independent distributors. In doing so, some of these countries often rely on a price policy involving a reduction in the retail price to make such products attractive in foreign markets. Even though, such pricing may simply reflect lower profit margins arising from avoidance of middlemen’s rent, it exposes the exporting country to the risk of being targeted for AD initiations. These barriers have serious implications for developing countries in terms of high compliance costs and potential or actual trade losses as an increasing number of their exports are being subjected to these barriers.

Price volatility: It is quite often argued that world demand for tropical agricultural products is fairly price inelastic and increases in SSA countries exports may inevitably lead to lower world market prices and possibly even to lower export revenues. However, this argument does not hold for all tropical export products for instance fruits and vegetables. In the food category, the price of wheat increased from $91.4 per metric ton in 1985 to $158.4 per metric ton in 1996.

Thereafter, it witnessed a steady declined but showed some recovery in 2001. The value increased to $126.8 per metric ton in 2003 from $108.2 per metric ton recorded in 2002. A similar pattern was recorded by maize and sugar between 1980 and 2003. They both witnessed a decline in their prices between 1996 and 2001 when they started increasing again. However in the case of sugar, it declined from 105.6 cents per pound in 2001 to 84.2 cents per pound in 2002. The commodity recorded another increase to 87 cents per pound in 2003 (Table 9). In the tropical beverages category, the price of cocoa declined from 254 cents per pound in 1985 to 189.2 cents per pound in 1998.

Table 9:	Trends in primary export prices (2000 = 100)
UNCTAD (2005)

Although, it increased to 200.3 cents per pound in 2002, it however declined again to 197.7 cents per pound in 2003. An almost similar pattern emerges for the price of coffee and tea.

A noteworthy point about the data in Table 8 is the existence of the variations in commodity prices from one year to the next which makes it quite difficult to forecast prices. This large variability in prices could be taken as the primary cause of instability of African export earnings and could be another reason why the market share of SSA export is falling at the world market.

Poor agricultural policies: Government agricultural policies also have been poor, providing only weak economic incentives to rural producers. Privatization and other structural adjustment policies led to an over-hasty withdrawal of the state from direct production. In the absence of a sound private sector, this caused severe dislocation of production, farm trade and farmer support services. Agriculture also has been starved of investment. Many African governments devote >1% of their budgets to agriculture. Not only have overall donor aid levels declined but donor priorities have simultaneously shifted away from agriculture toward other sectors. Worldwide, the amount of aid allocated to primary agriculture declined from $11 billion in 1990 to $7.4 billion in 1998.

The decline has been especially sharp in the case of the World Bank which provided 39% of its total lending to agriculture in 1987 but only 7% in 2000 (Harsch, 2004). Thus, the reasons for the stagnation in agricultural responses can be said to be multiple: continuing dependence on uncertain, nutritional deficiencies in Africa's soils, small and dispersed domestic markets, the instability and decline of world prices for African agricultural exports, the small size of most farms, farmers' frequent lack of organization, the lack of rural roads and additionally, the neglect of the particular needs of women farmers (who produce most of the continent's food).

World demand: The success of any structural adjustment policy depends very much on the prospects of world demand for SSA export commodities. If world demand for a specific commodity increases by 3-5% because of population growth and per capita income increases, SSA countries exports could grow by 3-5% without leading to increased world market shares (Koester et al., 1989). A continental market share analysis has revealed that SSAs have lost market shares for most commodities over time. In the light of SSA countries past export performance any structural adjustment policy would be termed moderately successful if negative export growth for most commodities were stopped and current market shares maintained. However, this is not the scenario as highlighted in Table 5 where the annual average export growth for most SSA countries is negative.

Market share: If SSA countries were to increase their exports of individual products significantly, this might have only insignificant effects on world markets in cases where their share of world markets is small. Real problems are expected for commodities for which SSA countries share in global exports is rather large. SSA countries hold high market shares for their two main export products, coffee and cocoa. Hence, one might expect that SSA countries export growth could affect the world market prices of these two commodities. Moreover, increased coffee and cocoa exports could result in negative marginal export revenues. However, there are other products such as palm kernels, sisal, groundnut oil and groundnuts where SSA countries are dominant exporters with high shares on world markets but the export values and their share in SSA countries total export revenue are fairly small. Hence, a decline in world market prices of these commodities might cause fewer adjustment problems than in the cases of cocoa and coffee. By way of illustrations, in terms of individual countries share of world exports, Cote d’Ivoire is one of the few SSA countries with over 30% of world market in cocoa exports. Similarly, Niger and Namibia also account for a good share of the world exports market in uranium and tobacco exports, respectively. Although, the primary commodity exports of virtually all SSA countries take a substantial part of each country’s exports, the market share of SSA countries in primary commodity exports in world exports cannot be said to be encouraging (Table 10).

Unprocessed nature of agricultural exports in SSA: Most SSA countries are small and thus each account for a small percentage of world exports or imports. The decline in demand for major export products which are of crucial importance to SSA and the reliance on a narrow range of unprocessed primary commodity exports have made SSA countries vulnerable to market fluctuations and weather problems (Table 11). In the late 1990's, about 39 African countries depended for more than half of their export earnings on just primary commodities. In addition, the collapse of world commodity prices in 1998 was equivalent to a real income loss of 2.6% of SSA GDP between 1997 and 1998 (UNCTAD, 2008). Thus, supply shocks with its attendant problem of price volatility and the resulting terms of trade losses could be the probable causes of the weak response of agricultural export to structural reforms.

Table 10:	SSA countries with a high percentage share of world exports of a primary commodity (2001-2002)
UNCTAD (2004)

Table 11:	Major exports of selected SSA countries
CIA World Fact Book (2009)

Policy issue: The core theoretical criticism of export-led growth is that it suffers from a fallacy of composition whereby it assumes that all countries can grow by relying on demand growth in other countries (Blecker, 2000). When pursued globally, there is a danger of a beggar-thy-neighbor outcome in which all try to grow on the back of demand expansion in other countries and the result is global excess supply and deflation. For individual countries export growth represents a way of growing demand. However if export growth, comes at the expense of foreign demand growth then it may just shift the country’s composition of growth without raising overall world economic growth. Developing countries, particularly SSA countries are rivals with each other and when one country manages to increase its exports it often does so by crowding out the exports of another developing country. This is the fallacy of composition as it applies to the developing world. Export-led development may work when adopted by one or even a few countries but it takes on a zero-sum dimension when adopted by all this, the proponents of SAPs refused to acknowledge.

A second issue concerns developing countries’ terms-of-trade. The export-led growth model prompts countries to shift ever more output onto global goods and commodity markets, thereby aggravating the long standing trend deterioration in developing countries terms-of-trade. This pattern partakes of a vicious cycle since falling export prices compel developing countries to export even more thereby compounding the downward price pressure. This vicious cycle has long been visible for producers of primary commodities (Prebisch, 1950; Singer, 1950).

Way forward: African agriculture is in crisis and such a situation requires a crisis response. Nevertheless, the findings from the study discussed in the last section presents various policy implications for SSA’s countries policy makers in their attempts to boost agricultural exports and fostering the integration of SSA countries into the global economy.

Strengthening of trade policy reform measures: Continuous and increased utilization of structural adjustment reform tools such as tariffs and non-tariff barriers should be used to reduce anti export bias. Although, tariff regimes have been significantly liberalized over the past two decades, further reduction and simplifications of the tariff regime are still possible. The reason is that the number of different tariff rates is still high for some SSA countries and considerable variation of tariff rates at the 8 digit HS level remains. While the ability to reform the tariff structure is being constrained by the numerous trade preference agreement that have been negotiated, it is still possible to simplify the Most Favoured Nation (MFN) schedule by unifying tariff levels within the 4 digits HS code and also reducing the number of tariff rates to 6.

This simplification is very important since future bilateral trade negotiations may entrench the existing complexity of the tariff structure. In addition, other restrictive measures such as import duties and taxes, import prohibitions, quantitative restrictions and licensing should be reduced to the barest minimum if not completely eliminated.

Infrastructural support: Infrastructural support by the government becomes even more crucial for trade policy reforms to be effective. Roads are in poor state, sea-bound transport link for landlocked countries are weak and the costs of transportation to market overseas are prohibitive. In addition, electricity supplies in many SSA countries are also epileptic. SSA countries could do much to raise their supply capacity by investing in infrastructural facilities and services such as telecommunication, transport infrastructure, power generation and distribution for better export performance. In addition, adequate infrastructural support (especially good roads and adequate power supply) will go a long way to reduce the perishable nature of agricultural outputs which are meant for export. As earlier mentioned, transportation of agricultural output is a problem in SSA, not only for the many landlocked countries but also because many countries with sea coasts have large interiors that make transportation of exportable goods to the sea coasts costly. A more active role of the government as well as private public partnership is required to stimulate and carry out investment in infrastructural facilities.

Roads and marketplaces: In Sikorola, a village in western Burkina Faso, farmers generally benefit from adequate rains and more fertile soils. But their efforts to expand output are hampered by the area is very poor physical infrastructure. We are ready to produce more maize and potatoes, says one member of the Siguizani family but there is no road to transport the crop. Sikorola is not unusual. Across Africa, paved rural roads scarcely exist. Much produce is taken to market by cart or bicycle over unpaved roads or by foot along narrow paths cut through the brush. Africa has the lowest density of paved roads of any world region. Out of 1.8 mn km of roads in sub-Saharan Africa only 16% are paved.

Moreover, many of Africa's paved roads have deteriorated badly from overuse and inadequate maintenance. Because of poor road quality, lorry drivers in rural Cameroon may charge an extra CFA1,000-2,000 (US$1.70-3.40) for just a short trip of 6 km. Higher transport costs raise the prices farmers must charge, reducing their competitiveness in both domestic and international markets.

Because farmers will not have much incentive to grow more without the roads, storage facilities and other physical infrastructure, they need to market their crops, the CAADP urges that more than half of the investments projected under the plan be directed toward rural infrastructure (not counting irrigation systems). In addition to roads and other hard infrastructure, the CAADP argues, farmers also need soft infrastructure: communications and accurate price and market information in order to take the best advantage of changing market opportunities.

External markets also are vital for many of Africa's producers of cotton, cocoa, coffee, tea and other export crops. Yet world market conditions have not been favourable to African farmers. Not only are international agricultural prices volatile but African exports face restrictions on access to Northern markets and are severely hurt by the high subsidies paid to rich farmers in the industrialized countries. It is an irony, notes the NEPAD agricultural plan that many African countries have largely internalized the perspective that a dynamic and sustainable agricultural economy cannot be based on subsidies yet their agricultural systems continue to be undermined by the subsidies paid out in precisely the developed countries that are the main proponents of liberalization (Agriculture: Africa's engine for growth, small-scale farmers hold the key, says NEPAD plan By Ernest Harsch (2004).

Market access issue: An important step in improving market access requires the continuous lowering of trade barriers for developing countries at all stages of development. Efforts to improve market access should include tackling high tariffs and tariff peaks and escalation facing items of agricultural export of interest to developing countries, the argument for the phasing out of export subsidies and substantial reduction in trade-distorting domestic support by developed countries. NTBs have gained prominence in more recent times through the application of discretionary measures by importing countries under certain WTO rules. These include Sanitary and Phyto-sanitary (SPS) measures, Technical Barriers to Trade (TBT) and Anti-dumping (AD) measures as well as evolving voluntary health, environmental and other standards set by the private sector operators, their associations and other Non-Governmental Organizations (NGOs). These barriers have serious implications for SSA countries in terms of high compliance costs and potential or actual trade losses as an increasing number of their agricultural exports are being subjected to them.

Actions to improve market access in agricultural and non-agricultural areas should be accompanied by measures to help SSA countries gain actual market entry. These should include disciplining and removing as appropriate, non-tariff barriers and evolving discretionary measures, particularly those related to technical regulations and standards, sanitary and phyto-sanitary measures, environmental conditions and anti-competitive market structures and practices. Anti-dumping in particular seems to have become the defence mechanism of choice and further disciplining on the use of such measures may be required if the gains from trade policy reform are to be realized. Also important are private sector measures and requirements such as voluntary standards. A major effort is required to ensure that these standards and measures are developed transparently with participation of SSA countries and applied in a non-discriminatory manner. Simultaneously, innovative measures, complemented by capacity building support are required in order to upgrade substantially developing countries technical levels and capacity, particularly in standards setting in accordance with relevant international standards and scientific criteria as well as helping developing countries to meet legitimate health and safety requirements.

World demand for SSA countries exports could be further stimulated if industrialized countries were to open up their markets to imports from developing countries. According to the market share analysis, the EU and other OECD countries are the most important regions for SSA countries exports. These countries have restricted access to their markets much more for processed agricultural imports than for raw products. Liberalization of agricultural imports would create additional demand for SSA countries exports. Moreover, removing the presence of tariff escalation could help SSACs to set up an export-oriented agro-industry without facing strong external demand constraints.

Strengthening of SSA productive capacity and diversification of exports: The productive capacity of most SSA countries needs to be strengthened to achieve the little gains from structural adjustment reforms. Increasing domestic supply capacity and enhancing international competitiveness should rank high among the strategic objectives of policies at the macro, sectoral and micro levels. Specific market failures, the lack of entrepreneurial base, imperfection in technology and capital markets as well as linkages and externalities among different sectors should be adequately factored into policy measures. Addressing the problem of SSA countries’ inability to diversify their exports from primary commodities into semi-processed should enhance agricultural export performance of the sub-region. Efforts should be made to diversify from exports of unprocessed low value-added primary commodities often exported unprocessed into those agricultural exports where world demand is growing. The extent of benefit from improved agricultural export performance depends to a considerable extent on the magnitude of domestic value added.

Vigorous export promotion strategies: SSA countries should adopt vigorous export promotion strategies. Export incentives such as export subsidies and finance, duty drawback and manufacture-in-bond schemes, export promotion and assistance should be pursued to reduce the anti-export bias resulting from the protection of domestic markets by tariffs and import prohibitions. Nevertheless, export incentive schemes require individual evaluation to determine the extent of the contribution of each scheme to export promotion goals with a view of expunging moribund and redundant schemes and replacing them with new relevant ones.

Need to diversify into semi-processed products: There is a good need for SSA countries to diversify into semi-processed agricultural goods in order to gain competitive edge in the global market. This will help solve of the constraints being faced by unprocessed primary products. Most of the exported primary commodities usually find their way back to SSA countries as semi-processed commodities. Consequently, SSA countries will do a lot better if they processed some of the agricultural outputs. A good example is raw cocoa beans which can be processed into cocoa butter and cocoa cake for making chocolate.

A case study of success story of the indonesian oil palm sector: The development of the oil palm sector in South East Asia started in 1848 when four seedlings were transported from Africa to the botanical garden in Buitenzorg (the presentday Bogor) in Java which was then under Dutch colonial control. The descendents of these four palms were transferred to Deli in Sumatra where they were first used for ornamental purposes only. The first large-scale Indonesian oil palm plantation was set up by Dutch traders in 1911 using the seed of these Deli-palms. Soon afterwards, British traders set up oil palm plantations in Malaysia as well Setting up state-owned oil palm plantations in Indonesia was made possible by the Agrarian adopted by the colonial government in 1870 which declared all land not under permanent cultivation to be waste land. Dutch developers were then offered as much land as they needed on 75 years renewable leases at nominal rent. Until the 1940's palm oil production developed at a moderate pace in both Malaysia and Indonesia as it was restricted mainly to use as a lubricant. A more rapid phase of expansion began in Malaysia in the 1950 and 1960's which turned Malaysia into the dominant oil palm producer in the world. Final processing industries in the industrialized world discovered that oil palm could be a cheap alternative to crude oil (for instance in detergents) to butter (margarine) and to other edible oils. World demand for palm oil therefore increased substantially. At the same time, global demand for rubber decreased, making it attractive to turn rubber plantations into oil palm plantations.

Post-colonial decline (1945-1968): After Indonesia had gained independence in 1945, the plantation system partly collapsed as Dutch plantation owners no longer had the backing of the colonial government and labour migration was no longer undertaken with government assistance. Furthermore, president Sukarno promoted an isolationist policy during the period (known as Guided democracy), antagonistic towards the entry of foreign capital or foreign loans. In 1957, the Dutch colonial plantations were nationalized and placed under the control of the New State Plantation Company (Perusahaan Perkebunan Negara Baru). Since, then they suffered a period of declining production. In 1967, the oil palm plantation sector covered no >106,000 ha including 65,573 ha of state-owned plantations.

First expansion phase (1968-1985): From 1968, president Suharto started to invest again in the Indonesian oil palm sector by making direct investment via state run companies called Perseroan Terbatas Perkebunan (PTPs). During this period, the area planted with oil palm on government estates grew from 65,573 ha in 1967 to 176,408 ha in 1979. Most of these plantations were found in Sumatra, primarily North Sumatra. However, the government began to expand stateowned plantations into Kalimantan and Irian Jaya in the late 1980's.

Since 1979, the development of private plantations and smallholder estates was stimulated by the government as well with some World bank aid. Under the so-called PIR/NES schemes (Perkebunan Inti Rakyat or Nucleus Estate and Smallholder Scheme) private developers (known as Inti or Nucleus) planted plots of land with oil palms on behalf of smallholders located nearby. Most of these smallholders were migrants from other areas.

As the oil palms matured, usually after 3-4 years, the plots were transferred to the smallholders (known as Plasma) who developed the plantations under the supervision of the Inti developers. Inti developers were then required to purchase the oil palm Fresh Fruit Bunches (FFB) from the smallholders, process them into CPO and sell this CPO on the market.

Since the PIR/NES scheme was initiated, smallholder plantations have further expanded under the PIR-Transmigration scheme (1986-1994) and the KKPA scheme (1995-1998) which both stimulated smallholder developments in transmigration areas. Non-existent in 1978, planted areas held by smallholders grew to a total of 1.1 million ha in 1999. Most smallholder estates are found in Riau, South Sumatra, North Sumatra, Jambi and West Kalimantan. During this first expansion phase, total acreage of the Indonesian oil palm plantation sector increased fivefold from 120,000 ha in 1968 to 600,000 ha in 1985.

Second expansion phase (1985-1998): As the world demand for oil palm continued to grow at a rapid pace, the Suharto regime recognized the possibilities of further developing the oil palm sector during the 1980's. Labour costs are much lower in Indonesia than in Malaysia and land is more abundantly available. The average cost of producing 1 ton of crude palm oil was calculated in 1998 at US$ 225.5 for Malaysia, US$ 296.1 for Colombia, US$ 298.4 for Ivory Coast and only US$ 206.8 for Indonesia. In the mid 1980's, the Indonesian government formulated a policy goal to overthrow Malaysia as the world’s largest palm oil producer.

To achieve this aim, large forest areas where handed out to the large Indonesian business groups and to foreign investors. Officially, the government reserved 5.5 million ha, mainly covered with forests to be converted into oil palm plantations. But the Indonesian consultancy data consult in 1996 calculated from records of the investment Coordinating Board (BKPM) that the government had actually allocated 9.13 million ha of land for oil palm plantations in the eastern part of the country alone including 5.56 million ha in Irian Jaya, 1.70 million ha in East Kalimantan and 1.80 million ha in Maluku.

Even this area looked insufficient to satisfy, the appetite of the oil palm plantation sector. According to some sources, private plantation companies around 1995 had applied for the conversion of an additional 20 million ha of forestland into oil palm plantations. Part of this tremendous demand was motivated by the search for cheap timber supplies, rather than serious investment plans.

Greater private sector involvement in the oil palm sector was also encouraged between 1986 and 1996 by granting access to credit at concessionary rates for estate development, new crop planting and crushing facilities.

Newly established companies could then draw on a loan from an executing bank at a rate of 11% during land preparation and establishment of the trees and 14% after the trees had begun to yield. In turn, the executing bank was eligible to borrow from the bank of Indonesia at a concessionary rate of 4%.

The interest subsidies were intended to help investors overcome risks and uncertainties associated with establishing estates involving smallholders. As a result of this expansion drive, the area planted with oil palm in Indonesia increased considerably from the mid 1980's. Starting from about 600,000 ha in 1985, the total area reached approximately 2.8 million ha in 1998 and 4.1 million ha in 2003.

Private plantations which covered just 145,000 ha in 1986, experienced the strongest growth during the 1990's and now cover 2.0 million ha. Greasy palms: European buyers of Indonesian palm oil Jan Willem van Gelder, March 2004.

Increased investment: Although, available financial resources should be managed better, the Comprehensive Africa Agriculture Development Programme (CAADP) acknowledges, overall investments in agriculture need to be drastically increased if Africa is to pull out of its agricultural crisis. The 4 pillars highlighted in the plan will require $251 billion in investments between 2002 and 2015 (Table 11). Although, an ambitious amount the annual average of $17.9 billion is less than the continent spends each year on food imports. The plan expects that Africa will be able to mobilize about half of the total amount from its own resources with its share gradually increasing as the continent strengthens its domestic capacities (Harsch, 2004).

Boosting yields: Much can be done to increase African farm yields through simple methods of improving soil fertility and better utilizing the continent's available water. The vast bulk of farmland is left to the vagaries of weather including insufficient rainfall and flooding that may strip away topsoils as well as unsustainable farming methods that gradually deplete soil nutrients. There is a need for an integrated approach that combines much greater use of organic matter and mineral fertilizers, higher-yielding hybrid seeds, small-scale irrigation and other methods. The plan notes that just 7% of arable land in Africa is under irrigation (with 40% of that land area in North Africa).

As with fertilizer use, Africa's utilization of irrigation pales in comparison with other regions: the share of irrigated farmland in South America stands at 10% in East and Southeast Asia at 29% and in South Asia at 41% (Harsch, 2004).

While there is a need to upgrade and rehabilitate existing large-scale irrigation schemes and to develop new ones, the importance of promoting small-scale efforts within the reach of poor farmers should be taken into consideration by African policy makers.

Technology for development: To succeed, SSA's efforts to boost agricultural output which can then be exported, it must also rely on greater use of science and technology, the plan's fourth, long-term pillar. Research has shown that improved varieties of millet, sorghum and other traditional African grains can also significantly boost yields. Beyond seeds, farmers need access to animal health remedies, safe pesticides and other inputs as well as training in agro-forestry and various skills. SSA's agricultural research institutes and extension services have very little capacity to engage in new scientific research or get existing technologies out into farmers fields. In part such problems can be overcome by finding new ways to generate and handle scientific knowledge.

Relaxing external demand constraints: For SSA countries, the issue is not whether to integrate or not. SSA countries need to and must integrate their economies. More than ever before, SSA countries need to pay more attention to their regional integration agenda. Regional integration efforts should now move beyond rhetoric into concrete plans and actions for effective integration of Markets. SSA is the most fragmented continent with about 165 national borders demarcating the region into some 53 countries 22 of which have a population of 5 million or less and 11 of which have a population <1 million. This does not augur well for industrialization as the national markets are small and fragmented.

SSA countries may face less severe external demand-side constraints if simultaneously implemented adjustment policies were to stimulate intra-African trade in agricultural products. Intra-African trade could be boosted because adjustment policies are supposed to reduce the barriers to trade and moreover are supposed to spur income growth. Growing intra-African trade could play a dominant role in contributing to the successful promotion of adjustment policies. The export pattern of African countries could become more diversified and the dependence on industrialized countries import demand could be relaxed.

However, growth in intra-African trade in agricultural products will only materialize if there is a potential for trade expansion. It is a well-known fact that African countries do not trade much with each other. Past experience seems to support the widely held belief that African countries are very similar and hence, there is only a small potential for intra-African trade in agricultural products. However, this view considers only natural factors such as climate and soil fertility as determinants of production costs. SSA countries are quite different with respect to other determinants of production costs for instance per capita resource endowment, level of development, infrastructure and institutional framework. Additional benefits from intra-SSA trade accrue in several ways. First, the expansion of a country export markets supports the development of comparative advantage in production. Second, intra-SSA countries trade leads to export diversification away from the traditional exports to industrial countries, i.e., coffee, cocoa, etc.

Third, intra-SSA trade is politically desirable, since it supports regional economic and political integration in sub-Saharan Africa. Of course, it has to be realized that trade expansion among African countries is not an easy task. It requires a liberalization of each country exchange rate system and a simultaneous liberalization effort across countries. In the African case, not much may be gained if only one country opens up its borders. Because of high transport costs to overseas markets and thus, large differences between import and export parity prices, trade effects due to unilateral liberalization may be insignificant. However, simultaneous liberalization might allow specialization with respect to comparative advantage. Simultaneous implementation of adjustment policies can be considered as a unique chance to integrate African markets and thus, relax dependence on traditional markets.

CONCLUSION

This study yields a number of insights into the analysis of export performance in SSA. Empirical evidence shows that SSA countries have significantly liberalized their trade regimes over the past two decades or thereabout. Although, the pace and patterns of trade policy reform varies among countries, the general trend tends towards lower barriers to trade.

This is evident in the significant reduction of tariffs and non-tariff barriers, removal of exchange rate control and elimination of export marketing boards among other trade reform measures. While the study was able to confirm the distortions caused by trade barriers on export performance, there is not much evidence that the adoption of trade policy reforms since, the mid 1980's have produced a significant agricultural export response. Declining infrastructure investment, particularly transport infrastructure (air/sea port, railways and roads) have generated substantial transaction costs and dampened the response of agricultural exports to the more favourable trade environment being witnessed, since the mid 1980's. In addition, factors external to an individual country such as relative prices and income of trading partners appears to be more important determinants of exports than a country own structural adjustment policies.

This however does not imply that, at the margin, structural adjustment policy reform is not beneficial. The simple point is that there are many factors other than structural policy that help explain the poor agricultural export performance of SSA countries. Domestic policies are necessary to reduce the various constraints on supply response, increase transport and marketing efficiency and encouraging investment. To benefit from agricultural trade, SSA countries need to increase the flexibility and efficiency of resource use so that they can be competitive in global markets.