ACP-EU Trade and Aid Co-operation Post-Lomé IV:

Chapter 2. The Export Record of the ACP Countries and the Role of Trade Preferences

Paper prepared for the Summit of ACP Heads of State and Government

Libreville, Gabon, 6-7 November 1997

SUMMARY

The share of exports in GDP and the level of economic development

ACP Exports to the EU

Export Diversification

Export Diversification to the EU

Implications for a New Agreement

SUMMARY

The small size of population and the generally low level of incomes means that the growth of most of the ACP countries will be constrained by the small size of the domestic market and that they must participate in international trade to a greater extent than larger countries at a comparable level of development. It is often claimed that the ACP countries have failed to respond to this challenge and that falling shares in world commodity markets and in EU imports are evidence of the failure of domestic polices. Also, the continued dependence on exports of primary products is seen as evidence that the ACP have failed to take advantage of Lomé preference to diversify the product composition of their exports. A high dependence by the ACP Group on the EU market is also interpreted as indicating a lack of market diversification compared to other developing countries, with adverse effects on the growth of exports. This chapter examines these claims and emphasises both the fundamental problems faced by the ACP countries in world commodity markets and, contrary to the pessimistic view, the increasing extent of export diversification by the ACP countries. The problem is not one of an absence of structural change, but that this needs to be both broadened and deepened. Preferences have been of some use in stimulating export diversification, but will be of limited value in the next century and a new agreement needs to concentrate on a co-ordinated strategy to reduce the variability of existing production for export and support further export diversification by product and by market.

5.1 The share of exports in GDP and the level of economic development

The share of exports in GDP tends to increase the higher the level of economic development of a country because the increasing diversity of the economy and efficiency of the operation of factor and product markets enables the country to become more competitive in world markets. This general tendency, however, is subject to a significant degree of variation since it depends on the initial resource endowments of the economy and, in particular, on the size of the population of a country. A small size of population combined with a relatively low level of per capita income (only 7 ACP countries, with a population of more than 1 million are classified by the World Banks as having a per/c GNP greater than $725) produces a small size of domestic market and to overcome this constraint small economies must enter into international trade at an earlier stage of development than large economies (Chenery and Syrquin, 1975; Perkins and Syrquin, 1989). Empirical work usually divides countries into very large (population over 50 million), large (15-20 million) and small (under 15 million) economies (see Gillis, Perkins, Roemer, Snodgrass, 1996, p.49). On this basis, 61 of the 70 ACP States can be classified as small, and of this nearly two-thirds have population of less than 5 million.

Table 2.1Size Distribution of ACP Populations

Source: World Development Report, 1996, World Bank

The two largest ACP countries are Nigeria (108 million) and Ethiopia (55 million), while the seven countries in the 'large category' are Zaire (41 million), Tanzania (28.8 million), Sudan (27 million), Kenya (26 million), Uganda (18.8 million), Ghana (16.6 million) and Mozambique (15.5 million). The competitiveness and growth of the export sector is therefore of even greater importance to the growth of the ACP countries than for developing countries as a whole.

2.2 ACP Exports to the EU

The alleged failure to take advantage of preferences

The generally accepted view of the record of ACP exports to the EU since the first Lomé Convention in 1975 is that, despite being at the apex of the EU's 'pyramid of privilege' providing free access (subject to rules of origin) for 95% of their exports, the ACP countries have failed to take advantage of their preferential status and, indeed, performed poorly in comparison with other developing countries. Moss and Ravenhill (1987) state that 'the ACP were markedly less successful than other developing countries in maintaining their share of non-oil exports to the Community: whereas non-oil exporting ACP states lost 40% of their share of the EEC market from 1970 to 1984, the market share of other non-oil exporting countries remained relatively stable. As a result, the share of non-oil ACP States in EEC imports from the non-oil-exporting LDCs declined substantially' (p.112). A recent survey by Davenport, Hewitt and Koning (1995) of ACP trade performance makes a similar point in emphasising that, despite the number of Lomé beneficiaries increasing from 46 countries to 69 countries over the period 1975-92, the share of ACP non-oil-exports in EU imports declined from 6.1% to 2.9%, and 'if ACP exports to the EU are compared with those of other developing countries which enjoyed less preferential access to the European market, the deterioration in the ACP performance becomes even more apparent.' (p.5).

Whilst such general statistics are useful in focusing attention on the marginalisation of the ACP in the EU's external trade relations, they run the risk of leading to the conclusion that the Lomé Conventions have therefore been of no practical value in increasing the exports of the ACP. To reach such a conclusion on the basis of this evidence would, however, be misleading. The Convention covers more than just trade preferences, the impact of preferences has complex dynamic as well as static effects which are not captured by such simple comparisons, while 'ACP exports' are dominated by a few countries and products, most of which are primary commodities subject to low growth and instability. Comparisons with the non-oil developing countries on the basis that they are 'less preferred' are also misleading because EU imports from this group are dominated by the NICs and such large and industrially developed economies as Brazil and China, none of which are comparable to the ACP countries. An analysis of 'ACP exports' must therefore be conducted at a more disaggregated level, both by product and country.

ACP exports depend on the performance of a few countries and products

The ACP comprise a heterogeneous group of countries with a widely different export experience. The performance of the 'ACP group' however depends upon the performance of a very limited number of countries and range of products. Over the period 1990-94, Nigeria accounted for around one-fifth of ACP exports to the EU and 12 countries were responsible for a further 50% of EU imports from the ACP group. The share of oil and oil related products in exports to the EU varies according to the world price of oil but has averaged around 30%. A further nine products (defined at the 4 digit CN level - diamonds, cocoa beans, coffee, sugar, wood in the rough, aircraft, bananas, refined copper, sawn wood) accounted for 43% and twenty-six products account for over 70% of the ACP non-oil exports to the EU. Table 2.2 lists the 12 ACP countries responsible for 72% of ACP exports to the EU in 1994 and shows their high export dependence on one or two product categories. Variations in the performance of a few key ACP countries and/or in a small number of products will therefore largely determine the performance of 'ACP exports'.

Table 2.2Main Exporters to the EU

Value exports 1994 Ecu 18.2 billion.

Sources: External Trade of the EU with the ACP Countries and OCT's

Column 4, Davenport, Hewitt, Koning, 1995.

Problems of low growth and instability

Exports of primary commodities are generally subject to very low and declining income elasticities of demand, producing low growth rates in the volume of exports. They are also subject to a long term decline in real commodity prices, with considerable instability around this downward trend.

Table 2.3Trends in Commodity Prices 1984-94

Sources: Commodity Yearbook 1995, UNCTAD

Eurostat, 1995

Notes: 1 % deviation from exponential trend value

2 Current $ divided by index of export unit value of manufactured goods exported by developed market economies.

Table 2.3 analyses the $ price change of twenty-two primary products over the period 1984-94 which together accounted for 42% of ACP non-oil exports to the EU in 1994. Nearly half of these commodities experienced a downward trend in their current $ price and all but four products (tropical hardwoods and plywood, copper, manganese) experienced declining real product prices which, in the case of coffee and cocoa, averaged -10.4% and -11.5% pa over the ten year period. The data also demonstrates that not only have ACP exporters been faced with declining prices, but there has also been a high degree of instability around this adverse trend with price variations often being a large multiple of the trend rate of change. The implications of this for ACP commodity exports to the EU in recent years is illustrated in Table 2.4.

For example, in the case of cocoa beans, the ACP increased the volume of exports to the EU by 7.7% pa over the period 1988-94 (as a result, in part of the implementation of structural adjustment programmes in Ghana and Cote d'Ivoire) and EU import share from 78% to 86% but the falling $ world prices of cocoa beans (-45% 1988-92 before rising by 130% 1992-94) meant that the Ecu value of export receipts increased by only 0.2% pa. Similarly, whilst the volume of coffee exports declined by an average of -3.3% pa, export receipts (in Ecu) fell by -9.7% pa and in the case of cotton, a decline in volume of -6.5% pa translated into a -10.1% pa in value. Only in the case of tropical hardwoods do we observe a rise in Ecu unit values sufficient to offset the rise in the price of imported manufactured goods. Also, it is only in the case of sawn hardwoods that variations in export volume (ie export supply) are greater than that for export earnings, in all other cases the variation in export values is greater than volume, suggesting that in recent years, instability in export earnings in these products has been more a price (measured in Ecu, which, given that 55% of Sub-Saharan Africa's imports come from Europe, is probably the most appropriate numéraire) problem than supply-side problem.

A declining share of world commodity markets

Taking a longer run perspective, the declining share of ACP countries in world commodity trade is often taken as evidence of a failure to compete effectively in world markets.

Table 2.4 Exports of Traditional Products to the EU

Source: Computed from Eurostat (1995)

1. Av growth rate uses the formula log X = a + b(t)

2. Variation in the growth rate is the standard error of the coefficient b and shows the typical % deviation around the estimated trend growth rate

This decline over the period 1975-93 has certainly been dramatic, as shown in Table 2.5, with a 54% fall for Sub-Saharan Africa and a 33% fall for the Pacific. This can be contrasted with a much smaller decrease for Developing America and for the developing countries as a whole; while Asia has increased its share by 39%. The substantial increase in the share of Asia has been largely due to the rapid growth of primary commodity exports up to the second half of the 1980s by Turkey, China, Hong Kong, Taiwan, South Korea, Indonesia, Malaysia and Thailand, and there are, without doubt, important lessons to be learnt by the ACP countries from the success of the Asian countries. When, however, we compare in Table 2.6 the shares in the volume of world production of the main ACP commodities listed in Table 2.4, then a somewhat different picture emerges.

Table 2.5Share in World Primary Commodity Trade

(excluding fuels) by Value

Source: Computed from UNCTAD Commodity Y.B. 1995

Table 2.6Share in World Production by Sub-Saharan Africa

of Major Commodities

Source: Computed from UNCTAD Commodity Y.B. 1995

The largest decrease in the share of world production over the period 1975-93 was in palm oil, but this was despite a 41% increase in the volume of production. The declining share was produced by the exceptional increases in production by Indonesia (762%) and Malaysia (488%). Similarly, the decreased share in cocoa beans was due to very large increases in production by Indonesia (4 m.t. to 239 m.t.) and Malaysia (13 m.t. to 2126 m.t.) which far outweighed the 32% increase in production by Sub-Saharan Africa (which itself comprised large increases in production by Cote d'Ivoire and substantial decreases in production by Ghana and Nigeria). The 63% increase in world production from 1.56 billion metric tonnes in 1975 to a peak production of 2.54 billion metric tonnes in 1989 produced the sharp decline and instability in world prices previously noted. The decline in production and exports by Ghana and Nigeria were probably due to inappropriate economic policies, but falling prices and instability also serve as an illustration of the 'fallacy of composition' (Cline 1985). This states that what may be a sensible strategy for one country, when considered in isolation, may not be sensible for all developing countries simultaneously. The sharp decline in the ACP countries' share of particular world commodity markets certainly deserves detailed study but it would be wrong to ascribe this exclusively to a 'failure of policy', nor does this necessarily lead to the conclusion that the ACP should seek to re-establish their share of world commodity markets where world production has exceeded world demand and which are characterised by declining and unstable prices.

ACP exports and trade preferences

The previous sections have emphasised the significance of country and commodity concentration in ACP exports. A third factor affecting an analysis of the trade effects of the Convention is that if Lomé preferences are to have a trade stimulating effect, then there must be a significant margin of preference in the form of exemption from tariffs, import quotas, minimum import prices and other barriers to trade relective to competing countries. Most ACP exports are composed of primary products, and the EU Commission has calculated that 63% of their exports would have entered the EU duty-free under MFN or GSP treatment in 1993 (since the EU's final offer bringing the tariff rate on cocoa beans down from 3% to zero was implemented in 1995, we can infer that this proportion has increased to over 70%), while a large proportion of the remaining products have been subject to tariff preferences of a trivial level of 5% or less. As a result, Lomé preferences could not have had any quantifiable trade stimulating effect on most ACP exports.

The combination of the dependence of ACP exports on commodities with little or no margin of preference and on the performance of a few countries leads to the conclusions that a discussion about the declining share of the ACP countries in EU imports despite Lomé preferences, while perhaps being useful in highlighting the plight of the ACP, fundamentally suffers from the error of misplaced aggregation. The appropriate counterfactual is 'what would the performance of the ACP have been without the Lomé Convention?'. One view (McQueen and Stevens, 1989) is that, given the internal situation in many countries and external world market difficulties faced by the ACP countries, it is difficult to imagine that the situation would have been better in the absence of the Convention. Furthermore, while preferences can never be a sufficient condition for export growth and diversification, they may be of significant help to small, infant economies. The alternative view is that the elimination of trade preferences would break a certain psychological dependency on the EU market and make them more aggressive in searching for new markets (Davenport, 1992). It is also argued that preferences create a 'false comparative advantage for recipient countries which leads to a misallocation of resources and lower growth. These issues concerning export growth and diversification by market and product are clearly of importance to a new ACP-EU agreement and are therefore considered in some detail.

2.3 Export Diversification

Why is export diversification so important?

The objectives of export diversification are to reduce dependence on a few primary products subject to fluctuations in export volume and falling and unstable prices. Diversification is seen as increasing both the growth and stability of export earnings, reducing the risk and uncertainty of investment in the economy in general, and thus accelerating growth. A related form of export diversification is geographical diversification, based on the assumption that cyclical fluctuations in demand in different countries are not closely correlated and so diversification will increase the stability of export earnings. Also, in the longer run, markets will grow at different rates as economies develop different structural characteristics and unless exporters respond to shifts in market conditions, export opportunities will be lost and earnings stagnate or decline.

Geographical distribution of ACP exports

A striking characteristic of African ACP exports is a much higher degree of dependence (58% total exports) on a single market, the EU, compared to either developing countries as a whole, or to any other regional grouping of developing countries and developed countries (e.g. Latin America and South and South East Asia depends on the USA for 23% to 25% of their exports). Such a high degree of dependence on the EU market may increase the instability of export earnings since variations in the growth of the EU market will be transmitted as fluctuations in the derived demand for African goods.

A further difficulty is that EU imports from the developing countries of food, agricultural products, ores and metals have, in a number of important respects, been growing less rapidly than imports by other developed countries or intra-developing country trade. These two characteristics of market dependence and growth are explored in more detail in Table 2.7 for non-oil exports and from which the following conclusions can be drawn:

1. Africa's dependence on the EU market for exports of food, agricultural raw materials and manufactured goods has increased to between 56% to 60% over the last two decades, while the dependence of developing countries as a whole on the EU has declined to around 20% to 29%.

2. The growth of Africa's exports of foodstuffs to the EU has been lower than that to Japan and to the developing countries.

3. The growth of Africa's exports of agricultural raw materials to the EU has been lower than that to the developing countries.

4. The growth of Africas exports of ores and metals to the EU has been lower than that to the USA, Japan and the developing countries.

Table 2.7Exports by Destination

Source: UNCTAD Handbook of International Trade and Development Statistics 1994

5. The growth of Africa's exports of manufactured goods to the EU has been lower than to the USA.

6. The growth of the developing countries exports of food, ores and metals, to the EU has been significantly lower than that to Japan and intra-developing country trade.

7. The growth of the developing countries exports of agricultural raw materials to the EU has been substantially lower than that to the USA and intra-developing country trade.

These observations suggest that Africa's concentration on the EU market may have led to it losing out on the globalisation of markets and contributed to the lower growth performance of its exports. Diversification of existing exports to more rapidly growing markets combined with a long run strategy of a much greater diversification into processed and manufactured goods is clearly essential if the ACP countries are to break out of their present vicious circle of a low growth of exports producing a low growth of output, and the low growth of productive capacity limiting the ability to increase exports.

On the other hand, it is important to recognise the diversity of experience among the ACP countries. Of the 33 African ACP countries listed in Table 2.8, nineteen depend on the EU market for more than 50% of their total exports, but only ten of these countries increased their dependence on the EU over the period 1970-92/3. Of the remaining fourteen African ACP countries where the EU accounts for less than half their exports, only four countries (Angola, Ethiopia, Mali, Madagascar) increased the share of exports to the EU. Similarly, more than half (18) of the African ACP countries have increased the share of exports to other developing countries. In the case of the Caribbean countries, only Dominica, Grenada and Surinam have historically been substantially dependent on the EU market and, with the exception of the Dominican Republic (which recorded a small increased share of exports to the EU), this dependence has fallen. Conversely, most of the Caribbean countries have increased the share of exports to other developing countries, in some cases (Barbados, Belize, Dominica, Grenada, Trinidad and Tobago) to

Table 2.8Geographical Distribution of ACP Exports

Notes: 1 1970

Source: UNCTAD Handbook of International Trade Statistics (various years)

quite substantial levels. The Pacific countries rely on the EU for around quarter or less of their exports but the major change is a sharp decreased dependence on the EU and an increased share accounted for by other developed market economies and other developing countries.

Thus a picture of ACP-EU trade relations being one of an unaltered and high dependence by the ACP on the EU would be an oversimplification of what is a more complex reality of considerable variation in dependence both between ACP countries and over time and lead us to reject the hypothesis that Lomé has had a systematic tendency to maintain a high degree of dependence on the EU market. Nevertheless, while the EU market will continue to be of importance to many ACP countries, it is also clear that many ACP countries could increase the growth of their exports by increasingly looking to non-EU markets and it is essential that a new ACP-EU Agreement helps to develop ACP exports to world markets and not just to the EU.

Export diversification by product

Diversification may be horizontal into new products within the same broad sector of production (eg new agricultural products), vertical into the processing of domestically produced raw materials, and diagonal into the processing or manufacturing of imported inputs.

Table 2.9(a) examines changes in the overall structure of ACP exports over the period 1970-93. Contrary to the perception that there has been no structural change in ACP merchandise exports, this data shows that even when we exclude the main exports of fuel and minerals (where large shifts in the price and volume would exaggerate the change in the share of 'other primary products'), the dependence of the remaining 16 ACP countries on exports of primary products, excluding fuels, ore and metals, has declined from an average share of 83% to 60% of total exports, while the share of manufactures has increased from 7% to 24%.

Table 2.9(a)Structure of Merchandise Exports

Source: Adapted from World Development Report, 1995.

Note: 1 Excluding Sierra Leone, Nigeria, Mauritania, Cameroon, Trinidad & Tobago, Gabon.

Table 2.9(b) provides a comparison for 1980-92 with similar results, in particular a doubling of the share of manufactures as a result of diversification by Benin, Central African republic, Barbados and Fiji.

Table 2.9(b) Structure of Merchandise Exports

Source: Small states: Economic Review and Basic Statistics, Commonwealth Secretariat, 1995

Care must be taken in interpreting these statistics. First, the instability around a falling trend in most commodity prices may affect a comparison of the share of primary products in total exports, with falling prices exaggerating the 'true' degree of export diversification. Second, there is a large degree of variation around the average share and this is particularly true of 'other manufactures' where the standard deviation has almost doubled and is almost as large as the mean value. Nevertheless the shifts, particularly in the share of manufactures, are such as to demonstrate that an increasing number of ACP countries have been able to engage in vertical and diagonal diversification.

Table 2.10Proportion of Manufactures in Non-Oil Exports

Source: UNCTAD Handbook of Trade and Development Statistics - various years.

Table 2.10 presents this information in a slightly different way for 1970-1992 and shows exports of manufactures as a proportion of non-oil exports for 39 ACP countries. Of these, 22 countries recorded an increased share of more than 5% and in almost all of these cases, manufactures accounted for over 20% of non-oil exports.

A more complete picture of export diversification is revealed by changes in the overall export concentration index. Table 2.11 provides details of the Hirschmann concentration index for 1970, 1980 and 1992 calculated for the 239 three-digit SITC Revision 2 classification of products. As before, the statistics have to be interpreted with care to avoid a 'false' decrease in export concentration simply because of falling world prices of commodities and/or a temporary decrease in export volume. In addition, while the three-digit level of classification is generally used to distinguish between different products, in the case of some products, differences in three digit headings occur for closely related products. In this case it may be questioned whether this is 'true' export diversification in the sense of a wider portfolio of products leading to greater stability in export earnings. For example, Mauritius records a two-thirds fall in the level of concentration of exports. However, almost all of these new exports are clothing, where differences between three-digit classification are finely drawn (for example, SITC 841 and 842 cover, respectively, male and female garments not knitted or crocheted, whereas 843 and 844 cover knitted and crocheted garments, while 845 covers knitted or crocheted garments for either men or women),. Thus, while Mauritius has been the most successful ACP country in engaging in diagonal diversification out of a dependence on in this case sugar, the degree of diversification is somewhat exaggerated by the change in the concentration index.

Table 2.11Concentration of ACP Exports 1970-94

Source: UNCTAD Handbook of International Trade and Development Statistics - various years.

Analysis of changes in the concentration index and the number of products exported over the period 1970-94 produces a rather less optimistic picture to that previously outlined on the basis of shares of particular groups of products. Countries such as Colombia, El Salvador, Sri Lanka, Guatemala, Tunisia, Uruguay and Morocco all have concentration ratios less than that of Tanzania (with the lowest ratio of the ACP countries of 0.248) and considerably less than the median value for the ACP countries listed in Table 2.11 of 0.491. Also, of these 42 ACP countries, only 17 countries have recorded a decrease of 20% or more in the concentration ratio over the period 1970-94 or 1980-94. This is partly due to the increased importance of oil and minerals increasing the ratio for eight ACP States (Nigeria, Angola, Gabon, Burkina-Faso, Sierra Leone, Cameroon, Papua New Guinea) and the increased importance of tobacco exports in the case of Malawi. However, it is also noticeable that only in the case of Papua New Guinea has the number of products exported increased; in all other cases the number of products exported has been more or less constant, or decreased (Nigeria, Angola, Malawi, Burkina Faso, Cameroon). Further research would be required to identify the causes of this but from a theoretical perspective we could expect this to result, for example, from 'Dutch Disease' effects and a progressive devalued increase in the CFA Franc, causing an appreciation of the real exchange rate and a consequent decline in the competitiveness of tradable products. Of the 17 countries recording a decrease in export concentration, only 10 countries recorded a significant decrease in the ratio over the period 1980-94 (Congo, Uganda, Vanuatu, Ghana, Dominica, Trinidad and Tobago, Fiji, Jamaica, Mauritius and Grenada). It is also worth noting that 11 of the 17 countries are very small economies with populations of less than 3 million and it might be argued that it is easier for these economies to record a decrease in the concentration ratio because it involves a comparatively small absolute increase in the export of 'new' products.

The conclusions we can draw from this analysis of export diversification by the ACP States over the period 1970-94 is that, contrary to the image of no structural change and unchanged dependence on traditional primary products, a substantial number of countries have diversified their exports. However, only 8 of the 47 ACP countries (for which there is adequate data) have experienced a sufficient degree of export diversification as to record a concentration ratio in 1994 which was at least 20% less than in 1970 and to have sustained that decrease over the more recent period of 1980-94. A wealth of empirical evidence strongly suggests that 'for successfully developing economies, export concentration ratios decline over time' (Colman and Nixson, 1994, page 145) and the central challenge facing a future Convention is help to create the conditions in which the ACP countries can be integrated more fully into the international economy and to nurture and develop the nascent export industries which have developed in recent years.

2.4 Export Diversification to the EU

Diversification 1975-88

It has been emphasised that an analysis of the extent to which the ACP countries have utilised preferences available under the Lomé Convention cannot be based on total exports but on a much more closely defined group of products in which Lomé provides a competitive advantage to the ACP through exemptions from EU tariff and non-tariff barriers to trade. The theory of preferences would then predict that, other things being equal, this should lead to the export of 'new' products (as well as increased exports of existing products for which preferences were available) to the EU. This form of analysis was first carried out by Stevens and Weston (1984) for the period 1975-80 and subsequently by McQueen and Stevens (1989) for the period 1976-87. The latter study initially reviewed 179 items at the 4 and 6 digit Nimexe level of classification with a total value of Ecu 956 million in 1987. From this the study focused on 70 items grouped under 8 product categories - canned tuna, leather and leather products, wood products, yarns and fabric, clothing, processed tropical products, fresh flowers, vegetables. Altogether, 28 ACP States recorded some new exports over the period 1976-87 and although many exported only a small number of products, one-third exported six or more new products and five countries exported more than 15 products. Furthermore, the range of ACP countries exporting these products included both more developed countries, such as Kenya, Zimbabwe, Cote d'Ivoire, Mauritius and Cameroon, and poorer countries at lower levels of economic development such as Ethiopia, Sudan and Ghana. It was also noted that although the total value of the 179 products represented only 7.9% of total ACP non-oil exports to the EC, this was equivalent to almost three times the level of EDF 6 disbursements in 1987 and 26% greater than the combined disbursements from EDF 5 and 6, emphasising that the trade regime of Lomé is infinitely more important as a source of foreign exchange than the aid provisions. McQueen and Stevens emphasised that these were often nascent exports, but drew the tentative conclusion that, in contrast to the pessimistic conclusions of most other studies, these results indicated a basis for future diversification.

Diversification 1988-94

Table 2.12 provides a summary of exports to the EU from 1988-94 for the products covered by the 1989 study. The total value of these products in 1994 was Ecu 1,856 million or 13.5% of ACP non-oil exports to the EU, representing a substantial increase in value and a significant increase in the processing of primary products and diversification into 'new' products. The number of countries has also increased, for example, 34 ACP countries exported more than Ecu 100,000 of clothing to the EU (see Davenport, Hewitt and Koning, 1995, Table A4). Fisheries are now the most important group of products, with clothing second in importance, and these two groups together account for 60% of the products listed in Table 2.12. The highest growth rates have been recorded for fishery products (with the exception of shrimps and prawns, and frozen yellowfin tuna), clothing, cotton yarn, fruit and vegetable products (excluding pineapples) and parquet flooring. It is interesting to note that processed traditional primary products, however, have generally not performed well, with low or negative growth rates for cocoa butter, cocoa paste, coffee extracts, veneers and plywood, laminated wood and woven fabrics of cotton. This serves to emphasise that although domestic value added can be significantly increased by vertical integration into processing and manufacturing (for example, the implicit unit price in 1994 for cocoa beans was Ecu 1.15 per Kg, while the price of cocoa butter was Ecu 2.51 per Kg; coffee beans were Ecu 1.86 per Kg, while the price of coffee concentrates, extracts etc. were Ecu 6.60 per Kg), it may simply produce a once-and-for-all gain and not necessarily increase either the long term growth or stability of export earnings. For example, the growth in the Ecu unit price over the period 1988-94 of cocoa beans was -7.5% pa and of cocoa butter was -4.4% pa; for coffee beans it was -6.4% pa and for coffee concentrates, etc., was 17.2% pa. The coefficients of variation (which may be regarded as a measure of instability) in both volume and value were similar (around 2.0) for cocoa beans and cocoa butter, but was substantially higher for coffee concentrates, etc. The latter may reflect production difficulties in the ACP countries rather than market instability, but this does not alter the fact that the costs and benefits of utilising resources in vertical integration must be carefully examined in exactly the same way as for any other investment project.

Table 2.12: Exports of Processed and Non-Traditional Products to the EU

Source: Computed from Eurostat (1995)

1 Growth rate computed from the formula log x = a + bt

2 Variation in the growth rate is the standard error the coefficient b and shows the typical % deviation + around the estimated trend growth rate.

There should be no automatic assumption that because an ACP State exports a raw material, it is necessarily in its best interests to diversify into the processing and manufacturing of that raw material. Horizontal diversification may offer a better use of resources. For example, a recent report (UNCTAD, 1995) draws attention to the fact that non-traditional agricultural exports from Sub-Saharan Africa were now third in value behind coffee and cocoa, while Zimbabwe had increased its exports of non-traditional agricultural products from less than $1 million in 1985 to more than $50 million in 1994. Production of high value crops can also generate a substantial amount of employment, both in production (the UNCTAD study reports that a joint venture between Kenyan and French firms to export canned green beans generated employment for 20,000 poor farmers in West Kenya) and in packaging, transport and distribution, and in associated suppliers of goods and services.

Diversification is often on a fragile base and the margin of preferences is falling

The variation in the export experience of non-traditional products is reflected in the change in EU import share. Table 2.12 shows a substantial increase in share by the ACP countries for veneers, cut flowers, fresh and chilled fish, lobster, with a smaller increase for fresh fruit. In terms of the value of exports, however, this has been outweighed by declining shares for tuna, coffee concentrates, oil extracts, cocoa butter, fresh and chilled pineapples, and frozen shrimps and prawns. With the exception of tuna, these decreases have been associated with zero or negative growth rates and underline the fact that preferences by themselves do not guarantee the success of ACP exports and should be considered simply as one of a number of factors assisting export diversification. They also illustrate the vulnerability of new ACP exports and this, together with the often significant degree of variation in export volume, and dependence on one or two ACP countries for a large proportion of these products, demonstrates the need for concentrated action to strengthen ACP export capacity and competitiveness.

Equally, although preferences do not guarantee success in exporting to the EU, their presence can be important in offsetting the cost/competitive disadvantages of the ACP countries. This is most clearly illustrated in the case of the textiles and clothing (see the case study on Mauritius, below), where the ACP are exempt from the EU's quantitative restrictions of the Multi-Fibre Arrangement (MFA) limiting the exports of most non-ACP countries. These import quotas will be phased out in stages ending in 2005, and will mean a substantial loss of preference, since 67% of the ACP exports of textiles and clothing are in MFA covered products (see Davenport, Hewitt and Koning, 1995, Table A3). To some extent, this may be offset by increased access to other, protected, markets such as the US, but this will have to be achieved in competition with other highly competitive countries whose exports have been constrained by quotas. Also, there is no guarantee that the industrialised countries will not, at least to a limited extent, replace the protection of the MFA with contingent protection measures. EU preferences on industrial products will then be limited to tariff preferences and although tariffs have been reduced to an average of 4% (a decrease of 38% on pre-UR rates), tariffs on sensitive products such as footwear, clothing, leather goods and some simple manufactures have generally been decreased by much smaller proportions to a range of 17% for footwear (previously 20%), through 12% for textiles and clothing (previously 13% to 14%) to 9% for leather goods and cutlery.

The effective margin of preferences for the ACP countries, however, should be evaluated in terms of the EUs other preference schemes. In terms of the GSP, this will reduce tariffs on most sensitive products by 15% to 30% (for example, to 12% on footwear and 10% on textiles and clothing) and by 65% in the case of cutlery (ie to 3%). Also, for the 9 non-ACP least developed countries and the 10 Latin American countries covered by the EUs enhanced GSP scheme, tariffs are, in most cases, zero. In addition, the EU has signed 26 free trade agreements since 1990 (Greenidge, 1996) which provide for duty-free entry for industrial products. In practice, therefore, the effective margin of preference on industrial products available to the ACP countries is at best less than 10% on a limited range of products and very low or 0% on a wide range of goods and 0% in relation to the 45 non-ACP countries with which the EU has preferential arrangements which go beyond the standard GSP.

Changes in the complex system of EU agricultural protection are discussed in detail in chapter 7. As has been the case for sensitive industrial products, reductions in tariffs have been less than the average of 38% for all products and have typically been of the order of 20% or less (to between 8% to 14%). The exception, as for all ACP exports are concerned, are flowers, where both peak rate and off-peak rate tariffs have been reduced by 50% (to between 8.5% to 12%). Also, agricultural and fisheries products are subject to only limited preferences in the GSP, both in terms of product coverage and are subject to only a 15% reduction in the MFN tariff rate. The exception (of relevance to Mauritius) is orchids, where the post UR rate of 8.5% (pre-UR rate of 17%) received a GSP reduction to 3%. The EUs enhanced GSP scheme for agricultural products is discussed in chapter 7 and, while it is difficult to generalise, preferences are generally less generous than those under the Lomé Convention.

Objective assessments of the role of preferences, however, invariably lead to a qualified acceptance of their role in increasing exports. Where domestic production conditions encourage exports, then preferences can act as a useful additional incentive to sourcing imports from a preferences receiving country. But more important determinants of demand, as compared to the margin of preference, for both industrial goods and agricultural and fisheries products, are non-price factors such as the quality of the product and the reliability of suppliers. Similarly, analysis shows only a small empirical positive relationship between the growth of ACP exports and the margin of preferences. A more important factor than the margin of preference is invariably the growth of the market. This equally implies that the erosion of ACP preferences should have only a limited adverse effect (though this may be significant for particular ACP countries) which could be more than offset by the growth of world trade, and that ACP negotiation resources for a new agreement should be much less concerned with trying to obtain new or increased preferences and much more concerned with radically improving export capacity.

2.5 Implications for a New Agreement

The ACP countries have recorded a declining share of world commodity markets and of EU imports from the non-oil developing countries and this has been interpreted as resulting from the failure of ACP trade policies. Closer examination of the trade statistics, however, reveals a large degree of variation between ACP States and a more complex picture underlying broad aggregates of trade flows. Declining shares of world markets in traditional products such as cocoa beans, palm oil and palm kernels can largely be explained by very substantial increases in production by Malaysia and Indonesia. The excess of world production in relation to demand has led to large decreases in prices. Also, variations in ACP export earnings of primary products have been greater than variations in export volumes (with the exception of sawn hardwoods) indicating that price instability is a larger problem than variations in production. ACP countries must therefore decide whether increased production of primary products is in their long term best interests, or whether it would simply be 'investing in decline. Conversely, downstream production into the processing of commodities, where unit values are twice to three times those of the raw material, could be of substantial potential benefit to the ACP countries. However, although in some cases the price of processed products has been rising (in contrast to falling commodity prices), in a number of other cases, for example, cocoa butter, prices have been falling. Also, this study has found that export instability in both volume and value has been similar or higher for processed products as for the raw material. The causes of this require further research, but in view of the limited extent of downstream production and the potential importance of processing, marketing, transport and distribution to many ACP countries, it would seem sensible to target this as a priority for a future agreement rather than simply, as in the present Lomé Convention, mentioning PMTD as an objective among many other objectives of the Convention.

This report has examined the apparently high level of dependence of the Sub-Saharan African countries on the EU market and found that while this is true in general terms, a more detailed analysis of individual ACP countries shows substantial variation in the level and trend of export market dependence. Nevertheless, SSA exports to the EU have grown more slowly than exports to other markets for all categories of goods and the erosion of ACP preferences as a result of the Uruguay Round and the extension of EU special preferences to a wider range of countries can be expected, if anything, to increase this trend. This underlines the importance of the ACP countries diversifying their exports to non-EU markets. Trade liberalisation, by removing the bias against exports, should assist this process, but it is important to ensure that financial and technical assistance from the EU to increase the export capacity of the ACP countries (the key elements of which are analysed in chapters 4 and 5) is directed at global markets and not simply at the EU market.

The large concentration of ACP exports in primary products has limited the value of preferences which ACP countries have been able to receive from the EU, and the anti-export bias of past trade policies of many ACP countries has retarded their ability to diversify into products where there has been a significant margin of preference. Nevertheless, it would be quite wrong to assume that no progress has been made. In 1987 just under 8% of ACP non-oil exports to the EU were in non-traditional produ