The Initial and Potential Impact of Preferential Access to the U.S. Market under the African Growth and Opportunity Act1 Paul Brenton and Takako Ikezuki International Trade Department The World Bank MSN MC2-201 1818 H Street St. NW Washington DC 20433 Tel: (202) 473-4255, Fax: (202) 522-7551 E-mail: pbrenton@worldbank.org Abstract The ability to export clothing products under preferences with liberal rules of origin is the key factor currently determining whether the African Growth and Opportunity Act (AGOA) has a significant impact on non-oil exporting African countries. At present only a small number of countries receive substantial benefits and Least Developed Countries (LDCs) that do not receive preferences for clothing have yet to see an impact of AGOA on their overall exports. However, the benefits from exporting clothing under AGOA appear fragile in the face of the removal of quotas in the United States on major suppliers, such as China, at the end of 2004, and the planned removal of the liberal rules of origin that allow for the global sourcing of fabrics from least-cost locations. To entrench and enhance the benefits of AGOA, it is important that the scheme be extended over a much longer period, if not made permanent, and the special liberal rules of origin for clothing products be extended considerably beyond 2004. The effective inclusion of textile products and a number of high-duty agricultural products would also help to broaden the range of opportunities for African exporters in the U.S. market. Nevertheless it is important that the opportunities created by AGOA are integrated into a broader framework for promoting trade and that it be recognized that if the opportunities offered by more open trade are to be exploited there must be concerted efforts to improve the environment for investment in countries covered by AGOA. Keywords: trade preferences, African Growth and Opportunity Act, Least Developed Countries JEL: F13, O19, O24 World Bank Policy Research Working Paper 3262, April 2004 The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Policy Research Working Papers are available online at http://econ.worldbank.org. 1We are very grateful for the comments of Ndiame Diop, Phil English, Larry Hinkle, Jeff Lewis, Caglar Ozden and Claire Thirriot. 1 1. Introduction This paper provides an initial assessment of the impact of the African Growth and Opportunity Act (AGOA) on trade and seeks to highlight the key stylized facts relating to this programme of trade preferences. It is clear that discussion of aggregate exports from African countries to the United States under AGOA is not very useful due to the preferences on oil and other energy products. In 2002 more than three-quarters of trade covered by AGOA was accounted for by two oil exporting countries, Gabon and Nigeria. Hence, the aim of this paper is to derive the magnitude of the benefits for individual beneficiaries. The paper shows that there is a wide variation in the value of the trade preferences across countries covered by AGOA. Deeper preferences for some countries than for others; more liberal rules of origin for some African suppliers than for others for key products, such as clothing; and differences in the structure of exports in tandem with variations across products in the size of external protection ensure that the impact of AGOA is not evenly distributed among Sub-Saharan African countries. The paper starts by describing the impact of AGOA in terms of the number of tariff lines liberalized, the overall amount of trade affected by these new preferences and those already existing under the Generalised System of Preferences (GSP), together with an analysis of the average duties on liberalized products. This shows that there are important differences between key groups of beneficiaries and in particular between those that have access to preferences on clothing products and those which do not. AGOA offers very little improvement in market access for Least Developed Countries (LDCs) that do not have preferences on clothing products. The paper then proceeds to look at the trade coverage of AGOA preferences and the utilization of those preferences, and provides a simple estimate of the economic value of the preferences for each AGOA beneficiary. This shows that currently most of the benefits go to very few of the countries. Preferences for clothing products lead to significant transfers to a small group of beneficiaries but for most countries the overall impact of AGOA preferences is at present likely to amount to no more than one-tenth of 1 percent of GDP. This reflects that the value of preferential treatment amounts to a small 2 proportion of exports to the United States and that exports to the United States are typically a small fraction of total exports. Seven of the beneficiaries account for 96 percent of the estimated transfer under AGOA in 2002, with the remaining 31 beneficiaries receiving very little. However, it should be noted that the number of countries being granted preferences on clothing is increasing and that there may have been some very recent impacts of AGOA on certain countries that are not adequately captured in the available data. Nevertheless, maintaining and broadening the benefits of AGOA is a key issue, but something which is unlikely to be achieved if the current liberal rules of origin for clothing for least developed countries are replaced by much more restrictive requirements to use U.S. or regionally produced fabrics, as planned in October 2004. 2. Summary of AGOA Provisions In this paper we concentrate on the trade provisions of AGOA, which are the key mechanisms for increasing the integration of the developing country beneficiaries into the global market. However, it is recognized that there are also other important provisions in AGOA which may have longer term indirect impacts on trade, for example, via enhanced domestic reform in AGOA countries, through increases in foreign direct investment by U.S. multinationals and through technical assistance to increase trade capacity. The key changes introduced by AGOA with regard to access to the U.S. market are as follows: AGOA entrenches the current preferences available under the GSP by guaranteeing benefits until September 2008. The current GSP expires at the end of 2006. Elimination of the "competitive need limitation" of the GSP.2 This is of benefit only to non-LDCs since LDCs are already excluded under the GSP. The designation of a range of products normally excluded from GSP preferences as eligible for duty-free preferences under AGOA. As discussed below this is of most relevance to non-LDC countries since LDCs were already able to export the majority of these products duty-free under the GSP. 2Under the GSP there are limitations in terms of both the share of total US imports and a fixed monetary value after which preferences for a particular country for that product are withdrawn. 3 o For these products the standard rules of origin under the GSP are amended to allow 1) for cumulation across all AGOA beneficiaries in meeting the standard 35 percent value-added requirement. Under the GSP, cumulation is possible only between specified countries in particular regional groupings.3 2) for the use of imported US inputs, up to a maximum of 15 percent of the product's value, to be counted towards the value-added requirement. o However, it is very important to note that for agricultural products which are subject to tariff rate quotas, the duty exemption only applies to in- quota imports. Once the quota is exhausted, products from AGOA beneficiaries must pay the full MFN duty. As we shall show, how one interprets the treatment of these tariff quota products is crucial to the conclusions that are drawn concerning the extent to which liberalization under AGOA is comprehensive. The removal of quotas on textile and clothing (apparel) products. In practice, however, only two countries, Mauritius and Kenya, were previously subject to quotas and only a subset of those on Mauritius could be considered as binding (Mattoo et al (2003)).4 The granting of duty-free preferences to clothing products 1. Assembled from fabrics and yarns formed and cut in the United States 2. Assembled from fabrics formed in one or more of the AGOA beneficiaries from U.S. or regional yarns, subject to quantitative limits 3. Assembled in LDCs from any fabric or yarn.5 This provision expires at the end of September 2004. 3 For example, under the GSP cumulation is possible between three members of SADC, Botswana, Mauritius and Tanzania. 4However, the imposition of the quotas may have contributed to the decline on the Kenyan industry during the 1990s, see Mwega and Muga (1999). 5The definition of an LDC with regard to the clothing provisions of AGOA differs from that which defines and LDC under the GSP. The GSP group is linked to the UN definition which includes an annual income per head of less than $900. The special rules of origin for clothing under AGOA are open to countries with an annual GNP per head of under $1500. There is also a degree of flexibility, since in the revision of AGOA in 2002, both Botswana and Namibia were redefined as LDCs for the purposes of AGOA. 4 Clothing assembled from non-U.S. fabrics (categories 2 and 3 above) is subject to quantitative restrictions with a limit of 4.23 percent of total U.S. imports of clothing in 2003 rising to 7 percent in 2008. Within this there is a sub-limit on imports under the special rule of origin which allows for global sourcing of fabrics (category 3 above) of 2.06 percent of total U.S. clothing imports. For the year October 2002 to end of September 2003 the overall quota was 36 percent filled. Within this the limit on products subject to liberal rules of origin was 62 percent utilized. Whilst the quota on products assembled from regional fabric was less than 10 percent filled. As we shall argue below these differences in utilization rates reflect in large part the differences in the restrictiveness of the rules of origin Access to preferences on clothing products is not automatic for AGOA beneficiaries. Countries must apply for these benefits and there are requirements regarding measures to prevent illegal transhipment including an effective visa system for clothing products. However, these requirements are unlikely to be a barrier to the granting of clothing preferences in many countries and technical assistance in meeting the requirements is available. 3. Recent Export Performance and AGOA Ex-ante simulations of the impact of AGOA suggested very small gains to Sub-Saharan African countries as a group. Ianchovichina et al (2002) using a general equilibrium simulation model suggest an increase of 0.6 percent (or $192 million) in non-oil exports for Sub-Saharan African countries as a group from preferential access to the U.S. clothing market. Lederman and Ozden (2003), using a gravity model, find that in aggregate AGOA countries export between 20 to 40 percent more than excluded countries of similar economic and geographic characteristics. In practice exports from certain members of this group of countries have increased substantially while exports of others have not increased or stagnated. This reflects in part that to date only a subset of countries have been granted access to preferences on clothing products. 5 This is crudely shown in Table 1 which looks at recent trends in exports from key groups of AGOA beneficiaries to the United States and compares these recent changes with the increase in total exports. The choice of base year is fairly arbitrary although by going back to 1999 it does capture any changes immediately prior to the implementation of AGOA in anticipation of that policy change. A more detailed table with information for each country is provided in the Table in the Appendix. These tables show that, apart from the oil exporting countries, the United States is not currently a major destination for AGOA countries exports. For none of the non-oil exporting country groups does the share of the United States in total exports exceed 15 percent. The U.S. share of total exports exceeds 20 percent only for Madagascar and Swaziland together with the clear outlier, Lesotho, which sends 90 percent of total exports to the United States. Table 1: The Importance of the U.S. as a Destination for AGOA Countries Exports and Trends in Exports over 1999-2002. Share of U.S. in Total Growth of Total Growth of Exports Exports: 2002 Exports: 1999-2002 to U.S.: 1999-2002 AGOA Countries LDCs Without apparel benefits 6.4% 2.6% -30.2% With apparel benefits 13.7% 19.5% 80.1% Non-LDCs Oil Exporters 33.1% 25.1% 30.9% Non-LDCs Non-Oil Without apparel benefits 8.2% 15.4% -16.8% With apparel benefits Liberal rules of origin 6.6% 21.5% 38.0% Restrictive rules of origin 13.0% 11.1% 30.9% Other US GSP LDCs 33.5% 43.2% 41.4% Other UN LDCs6 8.0% 86.6% 57.0% Source: WITS Within the group of LDCs, exports to the United States from countries that have not been eligible for AGOA benefits on clothing products have fallen by 30% over the past four years while total exports of these countries have increased slightly. On the other hand, exports to the United States from LDCs with clothing benefits have increased by over 80 percent, much faster than their exports to the world. Similarly, amongst the non-LDCs, exports to the United States from those countries without clothing benefits has fallen, 6A number of UN classified LDCs are not eligible for US GSP preferences. These include Liberia and Sudan. 6 whilst exports to the U.S. from countries eligible for clothing benefits have grown faster than total exports. Of course, these simple trends will reflect much more than the impact of AGOA and indeed there may be a degree of self-selection, with those countries with the lowest capacity to increase exports and the most challenging business environment being those which have not sought preferences on clothing products. It is also interesting to note that the other group of LDCs, which have not been granted access to AGOA benefits, have also increased their exports to the US substantially, by more than 40 percent, over the past 4 years.7 These countries are on average much more dependent on the US market with over one-third of total exports going to the US. Strong growth in exports to the US is also a feature of the LDCs that are not eligible for GSP preferences in the US, although total exports have grown at a faster rate. Again, this suggests that not too much should be made of the impact of trade preferences. Such incentives may be exploited when domestic conditions are favourable but they are no means a panacea for export growth. We now proceed to look in more detail at the importance of AGOA preferences across countries and the utilization of the available preferences. 4. The Impact of AGOA 4.1 The Extent of Liberalization under AGOA Here we provide a detailed look at the impact of AGOA in terms of the number of tariff lines liberalized by the Act and the amount of trade affected. It is important to remember that AGOA follows on from an existing scheme of preferences for developing countries under the GSP that was enhanced for the LDCs in 1997. As we shall show below this means that AGOA currently provides for very little additional liberalization for a number of countries. 7This conclusion continues to apply if the oil exporters in this group, Angola and Equatorial Guinea, are excluded. 7 4.1.1 Agriculture Table 2 summarizes the effect of AGOA on agricultural products. At present the economies and the exports of most of the AGOA countries are dominated by the agricultural sector. Hence, what happens in agriculture is of crucial importance. Unfortunately, the nature of the US tariff schedule means that simple assessments of tariff liberalization under AGOA are not possible, with different interpretations of the way that certain products are treated being possible, which can lead to conflicting conclusions about the extent of overall AGOA liberalization. We focus on the extent to which all products in the tariff schedule are fully liberalized and conclude that a significant number of products remain effectively excluded from AGOA preferences. This differs from the approach of the office of the US Trade Representative who concentrate on the extent to which products currently exported from Sub-Saharan African countries to the US are liberalized under AGOA which leads to the conclusion that "substantially all imports from Sub-Saharan Africa are eligible to enter the United States duty free" (USTR (2003)). When looking at the US tariff schedule we first note that there are a number of lines that are shown as AGOA products which are likely to be economically meaningless. These are lines which refer to General Note 15 of the US tariff schedule, which provides for products that are subject to tariff rate quotas and to safeguards and that are imported by the US government, individuals in quantities of less than 5 kilograms, or imported as samples, for exhibitions or for display at trade fairs not to count against the in quota quantity for that product. If such products are imported from AGOA countries then they are eligible for zero duty access and do not count against the quantitative limit for that product. In the 2002 tariff schedule 85 agricultural lines designated as AGOA products referred to General Note 15. This amounted to 14 percent of the total number of AGOA designated agricultural tariff lines. Imports from AGOA countries were recorded in only one of these categories in 2002 and in this case the amount was negligible. For an accurate representation of the impact of AGOA we think these lines should be excluded from the analysis. 8 The key issue is that in the US tariff schedule there are a number of agricultural products which are subject to tariff rate quotas. As described in more detail in Box 1 the way that these products are treated is important in determining the conclusions that are drawn concerning the liberalization of agricultural products under AGOA. Our approach leads to the conclusion that a significant number of products are not fully liberalized. Box 1: The Treatment of Agricultural Products Subject to Tariff Rate Quotas Products which are subject to tariff rate quotas have different tariff lines and different tariff rates for imports within the specified quantity and for imports in excess of this amount. AGOA preferences are only available for the in-quota quantities. Once the quota is exceeded then the full (usually very high) MFN duty is applicable. This leads us to conclude that the preferences on tariff quota products may not be effective and may have little economic value. For these reasons we analyze tariff lines and treat the out of quota tariff lines as being excluded from AGOA preferences. Indeed these lines are not designated as eligible under AGOA. An alternative approach is to assume that if there are preferences on in quota quantities and the quota is not exceeded then the out of quota duties are not relevant. Of course, if these duties are not relevant it is very difficult to understand why these products have not been fully liberalized and the AGOA countries offered quota and duty free access to the US. It is important whether the quota is global (an amount which can be satisfied by any country) or bilateral, in which a limit is specified for a particular country or countries. Most of the quotas specify only a global quota. In 2002, imports from AGOA countries were recorded in only 7 of the 158 tariff quota products. The key sector here is sugar where 7 countries exported under bilateral quotas to the US in 2002. These preferences on raw sugar are however covered by the GSP. The key tariff rate quota product for which in quota quantities were liberalized under AGOA is tobacco for cigarettes, with Malawi being the main exporter in Sub-Saharan Africa. The fact that AGOA countries are not exporting most of the products subject to tariff quotas does not mean that these quotas are unimportant. Indeed these quotas could be an important factor constraining the impact of AGOA. The out of quota duties are typically very high. For products with bilateral limits the available quantity may not be sufficient to warrant investment in raising capacity. For products subject to a global quota, uncertainty concerning when the quota will be filled may deter any investment or indeed any attempt to commence exporting. For these reasons we think it important to reflect that AGOA does not liberalise these out of quota duties. Table 2 shows that the impact of AGOA in agriculture differs substantially between the LDCs and the non-LDCs. For the LDCs, AGOA liberalizes only an additional 26 agricultural tariff lines, equivalent in number to less than 2 percent of the total number of agricultural lines and just under 12 percent of the remaining dutiable lines. In the main 9 the products liberalized under AGOA are those that have already been liberalized for LDCs under the provisions of the GSP. For non-LDCs, AGOA adds 541 products to the 519 products already eligible for duty-free preferences for developing countries under the GSP. Hence, the potential impact on the non-LDCs is much greater. Table 2: Liberalization of Agricultural Products under AGOA ­ The Number of Tariff Lines Liberalized Non-LDCs LDCs Total Tariff Lines 1723 1723 Total GSP 519 1038 GSP 519 (38) 547 (158) GSP LDC ... 491 GSP LDC but not AGOA ... 4 AGOA 541 (120) 26 Duty-Free Lines 440 440 Lines Excluded from AGOA 223 219 Main sectors containing products excluded from preferences Meat, Dairy Products, Sugar, Chocolate, Prepared Food Products and Tobacco The numbers in brackets show the number of product lines relating to in-quota duty rates for products subject to TRQs. Table 2 also shows that in 2002 of the 541 agricultural products liberalized under AGOA 120 were subject to tariff quotas. A further 38 product lines under the GSP were subject to tariff quotas. For the LDCs, all of these product lines had previously been covered by the GSP. In Table 2 we show that there are over 200 agricultural tariff lines with no preference under AGOA. These amount to 17 percent of the total number of dutiable agricultural tariff lines in the US schedule. Of these lines more than 150 relate to the over quota rates for products subject to tariff rate quotas. The products which have not been fully liberalized include certain meat products, a large number of dairy products, many sugar products, chocolate, a large range of prepared food products, certain tobacco products and groundnuts, that latter being of particular importance to a number of African countries.8 If these lines are ignored then clearly it appears that AGOA covers nearly all agricultural products. 8The tariff quota on groundnuts is very small, with Argentina allowed to export 44 thousand tons and all other countries 9 thousand tons. The out of quota tariff on unshelled groundnuts is 163.8 percent whilst that on shelled groundnuts is 131.8 percent. 10 4.1.2 Manufacturing9 With regard to manufacturing, see Table 3, AGOA liberalises 1,249 tariff lines for non- LDCs on top of the 3,116 lines given duty free preferences under the GSP. These additional preferences amount to 14 percent of the total number of manufacturing lines. Again, the impact for LDCs is more limited due to the prior liberalization of many lines under the GSP. AGOA liberalises an additional 199 lines compared to the 4,223 lines already liberalized under the GSP. The third and fourth columns of the table show the importance of the liberalization of clothing products under AGOA which adds an additional 557 tariff lines. This creates a significant difference in the benefits available to Sub-Saharan African LDCs under AGOA. Table 3: Liberalization of Manufacturing Products under AGOA ­ The Number of Tariff Lines Liberalized Without apparel benefits With apparel benefits Non-LDC LDC Non-LDC LDC Total Lines 8660 8660 8660 8660 Total GSP 3116 4223 3116 4223 GSP 3116 3116 3116 3116 GSP LDC ... 1107 ... 1107 GSP LDC but not AGOA ... 57 ... 57 AGOA 1249 199 1806 756 Clothing (HS61, 62) ... ... 557 557 Duty-Free 2773 2773 2773 2775 Lines Excluded from AGOA 1522 1465 965 911 Main sectors containing products excluded from Preferences Leather products, Textile and textile articles, Headgear, Glass and glassware Data source: USITC dataweb For manufactures, about 16 percent of the dutiable lines are excluded from preferences for the countries that receive clothing benefits, while for those countries that have not been granted the clothing benefits duties must be paid on 25 percent of the dutiable lines in the US tariff schedule. The key products excluded from tariff preferences are textile 9It is important to note that cotton (which is part of chapter 52) is treated as an agricultural product by the WTO but is included under manufacturing in this paper. Raw cotton is subject to a tariff rate quota and safeguard measures in the US. This means that in chapter 52 there is a `General Note 15 product'. In fact there is a special clause in general note 15 such that imports of upland cotton do not count against the tariff quota, but this is irrelevant to the main cotton producers in Africa. Here again we believe that inclusion of this tariff line as an AGOA product is misleading so we exclude it. 11 products,10 certain glass products and certain headwear.11 The average duty on textiles is over 8 percent and the duty on certain textile products reaches almost 30 percent. In terms of its impact on total trade, AGOA adds 1790 lines of preferences to the 3,635 lines already liberalized under the GSP for non-LDCs.12 With eligibility for clothing preferences adding a further 557 lines. For the LDCs which are not eligible for clothing benefits only an additional 225 lines are liberalized under AGOA. For the LDCs without clothing benefits 23 percent of dutiable lines remain excluded from preferences, whilst for non-LDCs with clothing benefits 16 percent of dutiable lines remain without preferences. So, in terms of the number of tariff lines liberalized the principal impact of AGOA falls on the non-LDC Sub-Saharan African countries. This reflects that, with the exception of clothing, most of the products liberalized under AGOA had already been liberalized under the GSP for the LDCs. For the LDCs what matters is whether they are able to access the preferences on clothing products. We shall also discuss below that a key factor determining the ability of LDCs to exploit the preferences on clothing are the rules of origin. Current liberal rules of origin for many countries have helped to stimulate exports. However, much more restrictive rules of origin will be imposed in 2004 which will seriously constrain the ability of LDCs to benefit from AGOA. 10 AGOA does provide preferences for specific textile items which are hand loomed, handmade or a folklore article of a beneficiary country, but in practice this has had no impact on trade. No imports into the US have been recorded of such products. A number of countries have been approved to export such products but the impact on trade will be minor. 11Although there is a quota on exports of clothing under AGOA we do not add the out of quota rates to the group of lines excluded from preferences, as we did for the agricultural products subject to TRQs, for a number of reasons. First, the clothing quota is not specific to particular tariff lines but covers all clothing products. Second, the quota is applied to AGOA countries, whereas most of the agricultural TRQs have global quotas so that the quota can be filled by any country or countries. Finally, the clothing quota is specified as a (growing) share of (the expanding) value of total US imports of clothing rather than a more narrow and restrictive quantitative limit. 12Many publications which discuss AGOA, see USTR (2003) for example, suggest that the GSP covers 4650 lines. This is a very dated figure and does not capture that for a large number of tariff lines duties have been subsequently reduced to zero under the commitments entered into under the Uruguay Round and therefore there are now no GSP preferences. We also exclude the General Note 15 products. 12 4.2 The Amount of Trade Affected Tables 4a and 4b show separately for LDCs and non-LDCs the amount of trade in 2002 that entered the U.S. market with GSP or AGOA preferences as well as trade in zero duty products and exports on which the full MFN duty is paid. The table shows that in 2002 LDCs without clothing benefits did not export any products under AGOA. This reflects that AGOA led to very little additional market opening for these countries relative to the preferences already available under the GSP. It is worth noting that these countries are currently not exporting to the U.S. products that are excluded from preferences. Table 4a: Value of LDCs Exports to the U.S. under AGOA in 2002 ($mill) LDCs without apparel LDCs with apparel benefits benefits Total Value of Exports to US 280 683 Value for Exports Requesting Preferences 146 454 GSP 146 17 Total AGOA 0 437 AGOA apparel 0 406 Exports for which MFN rate is zero 122 198 Exports of products excluded from Preferences 1 0 Exports Eligible for Preferences but Preferences not Requested 11 30 Countries Benin, Central African Rep., Cape Verde, Ethiopia, Chad, Dem. Congo, Djibouti, Lesotho, Madagascar, Gambia, Guinea, Guinea- Malawi, Mozambique, Bissau, Mali, Mauritania, Uganda, Tanzania, Zambia Niger, Rwanda, Sao Tome&Prin., Sierra Leone Data source: USITC dataweb Note: The classification of countries in the table is that pertaining in 2002, to allow for a match with the available trade data. It should be noted that subsequently the Central African Republic has been removed whilst Angola has been added to the list of AGOA eligible countries. Benin (2004), Mali, Niger, and Rwanda (2003) have also become eligible for apparel benefits since 2002. In terms of development the key sectors here are the agricultural products subject to tariff rate quotas, textiles and clothing. Preferences on clothing products are available under AGOA. Thus, key issues for these countries are whether to take the necessary steps to request eligibility for clothing preferences and to consider how to improve the domestic investment climate to allow firms to better take advantage of existing opportunities created by AGOA and new opportunities for diversification that would arise with preferences for clothing. For these countries AGOA will be of value if in the future it provides for the export of a greater range of products. 13 For those LDCs that are eligible for clothing benefits a substantial proportion of exports to the United States request preferences, about two-thirds in 2002, and most of those preferences are for clothing products under AGOA. In fact 90 percent of products from these countries requesting preferential access to the United States in 2002 were clothing products. It is worth remembering that all of these countries are able to export clothing to the United States with the liberal rule of origin that allows for the global sourcing of fabrics. Products covered by the GSP are not important in the exports of these countries whilst 30 percent of the exports of these countries to the U.S. are products for which the MFN duty is zero. Table 4b: Value of Non-LDCs Exports to the U.S. under AGOA in 2002 ($mill) Non-LDCs without apparel Non-LDCs with apparel benefits benefits ALL Non-Oil Special Rules of Restrictive Rule Origin of origin Total Value of Exports 8050 627 666 4426 Value for Exports Requesting Preferences 6712 156 368 1465 GSP 26 25 25 561 Total AGOA 6686 131 342 896 AGOA Apparel 0 0 201 192 Exports for which MFN rate is zero 519 344 233 2532 Exports of products excluded from Preferences 1 0 2 102 Exports Eligible for Preferences but Preferences 819 127 63 326 not Requested Countries Rep. Congo, Cote Rep. Congo, Botswana, South Africa, d'Ivoire, Eritrea, Cote d'Ivoire, Cameroon, Ghana, Mauritius Gabon, Nigeria, Eritrea, Kenya, Namibia, Seychelles Seychelles Senegal, Swaziland Data source: USITC dataweb Note: Since 2002 Eritrea has been removed from the list of AGOA beneficiaries and Cote d'Ivore has become eligible for apparel benefits. Table 4b for the non-LDCs again distinguishes between those countries that are eligible to export clothing under preferences and from those that are not. The table first shows that for those countries not eligible for clothing benefits, AGOA preferences still affect a substantial proportion of trade ­ $6.6 billion out of total exports to the U.S. of $8 billion. However, the table then shows that almost all of this amount is accounted for by two oil exporting countries, Gabon and especially Nigeria. The next column of the table shows that for the non-oil exporters of this group, products exported to the U.S. under AGOA 14 accounted for only about 21 percent of total exports, whilst 54 percent of exports were those for which the MFN duty is zero and there are no preferences. The value of products for which the MFN duty is positive but there are no preferences is currently very small for this group. However, it is interesting that 20 percent of total exports of this group to the US are products eligible for preferences but do not request those preferences. This means that the utilization rate of available preferences is only 55 percent. A probable reason for this low uptake of preferences is that the costs of satisfying the origin rules and/or the costs of proving conformity with those rules outweigh the value of the available preferences. Table 4b then shows the amount of exports from non-LDCs eligible for preferences on clothing products that enter the US under AGOA. We further distinguish between those countries which have been granted liberal rules of origin until 2004 and the two countries which are eligible for clothing preferences but which have to satisfy much more restrictive rules of origin, Mauritius and South Africa. For the first group of countries with the liberal rules of origin AGOA has clearly had a substantial impact in aggregate. Products covered by GSP preferences are of little importance to these countries accounting for only 3.8 percent of total exports to the US in 2002. On the other hand, products covered by AGOA preferences, including those on clothing, accounted for over 51 percent of exports to the US and 80 percent of exports to which MFN duties are applicable. Almost 60 percent of exports from these countries requesting preferences under AGOA are clothing products, highlighting the importance of these products and the potential significance of the rules of origin. Just over one-third of the exports to the US of this group of countries are products for which the MFN duty is zero. A negligible amount of current exports are of products for which there are no preferences. More important are products which although eligible for preferences do not request preferential access to the US. These exports which forego preferences accounted for 10 percent of total exports to the US in 2002. 15 Taking South Africa and Mauritius together, more than 57 percent of their exports to the United States are products for which the MFN rate is zero. GSP products account for about 13 percent of total exports to the U.S. while exports under AGOA comprise 20 percent of the total. However, in contrast to the previous group of countries, clothing products do not dominate exports under AGOA accounting for only 21 percent of AGOA exports. This reflects the much broader industrial base of South Africa and the ability to exploit preferences on products such as cars. There is also a degree of under-utilization of preferences with almost 19 percent of exports eligible for preferences not requesting these benefits. Again this raises the issue of the rules of origin for clothing products to which we return later. 4.3 Average Duties on AGOA Covered and Excluded Products Table 5a, for agricultural products, and 5b for manufacturing, provide a very simple look at the average MFN duties that apply to each of the preferential groups of products. The numbers presented are simple unweighted averages of duties applied in 2003. This provides a crude view of the margins of preference available in the US. The tables show that AGOA covers products with higher duties than those covered by the GSP. For example, the average duty on agricultural products covered by the GSP for non-LDCs is 3.5 percent whilst the average duty on the additional products for which preferences are available for LDCs is 6.9 percent13, which is similar to the average duty on products liberalized under AGOA. For manufactures, the average duty on products covered by the GSP is 3.8 percent, whilst the products covered by the basic AGOA provisions are subject to an average duty of 6.1 percent and the clothing products on average pay a duty of more than 12 percent. 13As discussed earlier we only include the in quota duty rates for the TRQ products in the calculation of the average MFN duty on AGOA products. We then put the out of quota rates in the calculation of the average duties on lines not covered by AGOA 16 Table 5a: Unweighted Average Tariffs for Agriculture Tariff Lines under AGOA and GSP, and for those excluded from Preferences in 2003 Non-LDCs LDCs Total GSP 3.5% 5.2% GSP 3.5% 3.5% GSP LDC ... 6.8% Total AGOA 6.7% 7.7% Dutiable Lines ­ Products excluded from AGOA 30.7% 31.0% For specific tariffs data on 2002 calculated duties and customs value (based on 2002 US GSP Group) are used to derive ad valorem equivalents. When there are zero duties for the GSP group of countries we use data for total imports to calculate the ad valorem equivalent. Data source: USITC dataweb These tables also show that the products excluded from preferences are high-duty products. For agriculture the average duty on excluded products is over 30 percent, these include the out of quota duty rates for products affected by tariff rate quotas. This, at least in part, might explain why there may be so few exports of these products from Sub- Saharan African countries to the US, as shown in Table 2. Similarly for manufactures the average duty on products not eligible for preferences is high relative to the overall average duty and high relative to the products covered by the GSP. Hence, in terms of the average margin of preference, AGOA significantly enhances the preferences available to Sub-Saharan African countries, particularly for those countries eligible to receive benefits for clothing products. Nevertheless, a range of high average duty products, particularly in agriculture, remain excluded from preferences. Table 5b: Unweighted Average Tariffs for Manufactured Products Under AGOA and GSP, and for those excluded from Preferences in 2003 Without apparel benefits With apparel benefits Non-LDCs LDCs Non-LDCs LDCs Total GSP 3.8% 4.0% 3.8% 4.0% GSP 3.8% 3.8% 3.8% 3.8% GSP LDC ... 4.8% ... 4.8% Total AGOA 6.1% 12.7% 7.7% 12.3% AGOA (excluding GSP) 6.1% 12.7% 6.1% 12.7% AGOA apparel ... ... 12.1% 12.1% Dutiable Lines ­ Products excluded 9.9% 10.2% 8.7% 9.1% from AGOA 17 5. The Magnitude and Importance of Preferences under AGOA The importance of US trade preferences varies immensely across African countries. For certain countries AGOA (including GSP preferences) is of little or no relevance under the current structure of exports since there are no or few exports of products eligible for preferences, that is that are legally eligible to enter the U.S. market with lower duties than other non-preferred suppliers. Figure 1 shows that there are nine AGOA beneficiaries for whom less than 5 percent of current exports are eligible for preferential access to the US. In part this reflects that these countries are producing and exporting products for which the MFN duty in the U.S. is zero. In this case it follows that benefits may flow from AGOA only if these preferences encourage and permit diversification into a broader range of exports. Most of the 9 countries are only eligible at present for basic AGOA preferences. Uganda and Zambia are exceptions in that they have been granted eligibility for clothing preferences, although, by 2002 this had not been reflected in substantial exports of such products to the United States. However, the data that are available for 2003 suggest that the preferences on clothing for Uganda which were granted from October 2001 are having an effect with exports of clothing to the United States rising from $0 in 2002 to $819,000 in the first 8 months of 2003.14 Figure 1 then shows a further 14 countries, including South Africa, for whom the amount of exports eligible for preferential access to the U.S. market comprises less than 50 percent of total exports to that market. There are then 16 countries for whom AGOA preferences are more significant, with exports eligible for preferences amounting to more than 50 percent of the total. Within this group there are five countries for whom more than 90 percent of their current exports to the United States are eligible for preferences. Here it is interesting to look at Lesotho, where all exports to the U.S. are eligible for preferences, whereas in contrast only 28 percent of exports to the European Union in 2001 were eligible for preferences. In this case exports to the United States amounted to $321 million whilst exports to the European Union were only 14 million euro. This 14See New York Times, 14 November 2003, concerning the job creating impact in Uganda of AGOA preferences on clothing products. (http://www.nytimes.com/2003/11/14/international/africa/14AFRI.html) 18 reflects the very large amounts of clothing exported to the U.S. duty and quota free, at present, under liberal rules of origin. There is duty free access to the European Union for Lesotho but there are much more restrictive rules of origin. Nigeria and Gabon are different cases. For these countries the value of exports to the European Union and the United States are broadly similar yet exports eligible for preferences in the E.U. are small whilst most of exports to the U.S. can enjoy preferential access. This reflects the different treatment of mineral fuels, the dominant export for these two countries, for which the E.U. MFN duty rate is zero whilst it is positive in the U.S. Table 6 shows for all countries covered by AGOA the share of exports to the US which were legally eligible for preferences and the share of exports to the EU legally eligible for preferences. Interestingly there is a very low correlation between these two series (15 percent). Thus, it is apparent that for many developing countries the products that they currently export are often subject to different degrees of preference in different OECD markets. This reflects that: Preferences are available for a given product in one market, where there is a positive MFN duty, but not in another, where the MFN rate is zero. MFN rates are positive in all OECD markets but the product may be eligible for preferences in one market but excluded from preferences in another. Products are eligible for preferences in all markets but the margin of preference often varies across those markets due to differences in MFN rates and due to the size of the preference that is granted. Finally, the rules of origin for a given product differ across preferential schemes. This all implies difficulties for firms in developing countries in investing and developing capacity on the basis of serving a global market. The impact of preferences in developed countries would be enhanced by greater coordination and harmonization of rules of origin. Hence the segmented nature of preferential access to developed country markets 19 limits the value of those preferences as a mechanism for integrating developing countries into the world economy. Table 6: Share of Exports Eligible for Preferences in the E.U. (2001) and the U.S. (2002) U.S. E.U. Benin 0 33.9 Guinea-Bissau 0 41.5 Seychelles 0.3 86.2 Guinea 0.5 11.7 Chad 0.7 2.5 Uganda 1.2 52.7 Zambia 2.1 40.2 Rwanda 2.2 15.8 Sao Tome and Principe 3.8 33.1 Ethiopia 10.7 30.4 Tanzania 11.0 48.1 Central African Rep. 12.3 0.4 Namibia 14.0 50.7 Níger 17.0 2.5 Djibouti 18.0 6.7 Mali 19.3 18.0 Sierra Leone 23.0 21.9 Cote d'Ivore 25.5 37.1 Botswana 27.4 9.1 Senegal 28.3 71.7 Ghana 33.6 44.2 South Africa 36.8 Madagascar 43.6 82.3 Eritrea 55.0 46.3 Mauritania 63.5 33.5 Gambia 64.0 46.0 Mozambique 74.4 90.0 Kenya 76.3 52.6 Dem. Congo 81.3 3.9 Cameroon 82.3 20.5 Malawi 82.8 83.3 Rep. Congo 83.2 Swaziland 85.8 25.0 Mauritius 94.5 66.6 Gabon 94.9 6.2 Cape Verde 97.7 79.5 Nigeria 98.4 4.6 Lesotho 100.0 27.9 6. Current Utilization of AGOA Preferences The economic impact of available preferences depends on the magnitude of those preferences and the extent to which they are actually utilized. The magnitude of the preferences depends on the size of the MFN tariff, while the utilization rate will be 20 influenced by this margin of preference as well as the rules governing access to those preferences, especially the rules of origin. Figure 2 shows a wide variation in the utilization of AGOA and GSP preferences across countries. The rate of utilization is defined as the share of exports requesting preferences relative to the amount of exports eligible for preferences. The average rate of utilization for AGOA eligible countries in 2002 was over 80 percent. This is high relative to the utilization of GSP schemes and is comparable to the utilization rate of E.U. preferences for these countries under the Cotonou agreement. There is little difference in the average utilization rate once oil exporting countries are excluded. Nevertheless there are sixteen countries that utilise less than 50 percent of the available preferences. A key issue for the future is that at present many of the countries which are eligible for clothing preferences are able to export to the United States under very liberal rules of origin. This provision is due to expire in 2004 and given the experience of other preferential schemes with the more restrictive rules of origin that will be imposed from late 2004 it is certain that the utilization rates for many of these countries will fall. Mauritius and South Africa are the only two countries that are eligible for clothing preferences but which have not been granted liberal rules of origin. In 2002, 90 percent of exports from Mauritius to the U.S. were clothing products, yet only 41 percent of the available preferences for these products were taken up. Clothing only accounts for about 4 percent of South African exports to the U.S. although the absolute amount is similar to that exported by Mauritius. In 2002 only 47 percent of the available preferences for South African clothing products were actually utilized. The issue with the more restrictive rules of origin is not just the constraints that these rules impose on the sourcing of inputs, forcing producers to use higher cost fabrics and materials, but also the costs and difficulties in proving conformity with these rules compared to the more liberal rules where fabrics can be globally sourced. 21 7. The Value of Preferences Previously we have looked at the importance of products subject to preferences in each country's total exports to the United States. The current economic value of these preferences will be determined by the amount of preferential trade and the margin of preference that is granted. Figure 3 shows the estimated value of preferences requested in 2002 as a proportion of total exports to the United States for each country grouped according to the preferential treatment that they receive. This shows the tariff payment that is foregone by the U.S. due to the preferences that each country is able to request.15 In what follows we implicitly assume that this amount of tariff revenue avoided is captured by the exporter in the African country, although in practice these rents may accrue to importers in the United States. Indeed, Ozden and Olarreaga (2003), suggest that two-thirds of the transfer under AGOA may actually go to U.S. importers. The left-hand section of Figure 3 shows that for AGOA beneficiaries that are not eligible for preferences on clothing the value of U.S. preferences under current trade structures is very small. For none of the countries in this group does the value of preferences exceed 1 percent of the value of exports to the United States. For 13 of these countries the value of preferences is less than 0.25 percent of the value of exports to the United States. Thus, the preferences available to these countries under AGOA and the GSP are having little apparent economic impact at present. The value of preferences are more substantial and more widespread for countries in the second group of AGOA beneficiaries; those which are eligible for duty-free access for clothing products. In this group there are 10 countries for which the value of current preferences in the United States exceeds 1 percent of the value of exports to the U.S. 15This is derived by multiplying the value of exports to the US requesting preferences by the preferential margin, the MFN tariff for each product, at the tariff line level. For agricultural products which are subject to tariff quotas we apply the out-of-quota duty rate. However, for certain products subject to what are called technical duties (comprising a duty related to the unreported content of the product, such as the sugar content) the value of preferences is obtained using duties calculated by USITC for AGOA countries in total (or when not available for all exporters) to derive an ad valorem equivalent tax. This means that for products subject to tariff quotas the average within and out of quota duty is used as the margin of preference. In practice these technical duties affected only 8 tariff codes of products currently exported by AGOA countries and do not have a significant impact on the results presented here. 22 However, for most of these countries the share of the United States in total exports is relatively small (see Appendix Table 1). This means that the current aggregate impact of these preferences on the economies of the beneficiaries will be slight. For example, the value of U.S. preferences for Malawi is estimated to be equal to 10 percent of the value of exports to the U.S., yet this is equivalent to just 1.6 percent of the total value of Malawi's exports and around 0.44 percent of GDP. The clear exception is Lesotho for whom the value of U.S. preferences amounts to 19 percent of the value of exports and, given the importance of the U.S. market for Lesotho, to more than 17 percent of the value of total exports and to approximately 6 percent of GDP in 2002. Figure 4 takes this analysis a little further by showing the absolute value in millions of dollars of the implicit transfers due to U.S. preferences. We start by noting that for the first group of AGOA beneficiaries there are substantial transfers only to the oil exporters, Gabon and Nigeria. Most of the benefits go to one country, Nigeria, or to those in the United States importing from Nigeria. In 2002 the preferences accorded to Nigeria led to a potential transfer of over $20 million. This transfer for Nigeria amounted to almost 10 percent of the total value of preferences for all countries under AGOA. Figure 4 shows that in absolute terms Lesotho receives the biggest benefits under AGOA. The estimated transfer in 2002 being $61 million, which represents 28 percent of the total calculated value of all AGOA preferences. The only other substantial transfers under AGOA at present go to Malawi, Madagascar, Swaziland, Mauritius, Kenya and South Africa. All of these countries are eligible for preferences on clothing products, although Mauritius and South Africa have not been granted use of the liberal rule of origin. These countries account for 86 percent of the total benefits derived by AGOA countries in 2002. For Lesotho, Madagascar, Swaziland, Mauritius and Kenya, 99 percent or more of the benefits of preferential access to the United States come from clothing. Thus, for these countries at present the ability to export clothing to the U.S. under preferences is the key to deriving benefits under AGOA. This will be a main factor affecting the ability of other countries to derive benefits from AGOA in the future. 23 So for non-oil exporting countries that are not eligible for preferences on clothing products the transfer under AGOA never exceeds $1 million and the benefits amount to less than 1 percent of the value of exports to the US. Given the small share of the United States in the overall exports of most of these countries, typically less than 10 percent, the broader macroeconomic impact of preferences will be negligible. Crudely assuming exports to be 100 percent of GDP would lead to such preferences being equivalent to 0.1 percent of GDP. The key issue for many of these countries is the strengthening of supply capacities and diversification to facilitate the growth of total exports. Preferences can play a role in encouraging these changes but they should be treated as part of a strategy toward creating a broad base for export expansion. The Integrated Framework (IF) for Trade Related Technical Assistance, when incorporated into poverty reduction strategies, provides a vehicle for addressing these issues, defining appropriate policy responses and mobilizing relevant resources. In this case it is necessary for both the United States and the international institutions to reconsider carefully eligibility for AGOA and the IF. There are a number of United Nations defined LDCs that are eligible for the IF but are not included under AGOA. There are also a number of non-LDCs that are eligible for AGOA but not the IF who may well benefit from an IF process. A key factor driving the growth of clothing exports, the key source of benefits under AGOA, has been the liberal rules of origin. It is currently proposed that these will be removed during 2004 and replaced by much more restrictive rules requiring the use of regionally sourced or U.S. fabrics. This will reduce the value of the available preferences as producers in Africa will have to shift to higher-cost sources of fabrics. Hence, it is most likely that the value of preferences calculated here will decline once the more restrictive rules of origin are enforced and the incentives to other countries to diversify into clothing will diminish. The quantity constraints on other key suppliers of clothing under the Agreement on Textiles and Clothing are also likely to have encouraged some of the investment in clothing in Africa. As we will argue below, making the rules of origin more restrictive will make it much more difficult for African suppliers to compete in the 24 U.S. market next year after the quotas have been removed, as is mandated by the agreement. 8. The Issue of Rules of Origin The granting of clothing benefits is dependent on the country operating an effective visa system, which is a "government-industry process that demonstrates that the goods for which benefits are claimed were in fact produced in a Sub-Saharan Africa country".16 Each shipment is accompanied by an original visa stamped on the original invoice. In addition, US Customs is permitted to visit factories, producers and exporters for verification purposes and AGOA countries must provide monthly export data to allow for reconciliation with US import statistics. This highlights that it is these mechanisms which are present to prevent shipments of clothing products from other countries illegally claiming reduced duties under AGOA through false declaration. Contrary to some beliefs it is not the role of the rules of origin to prevent illegal activity but rather to prevent legal transhipment whereby in the absence of rules of origin products produced in other countries are shipped to beneficiaries where they are subject to only minimal processing and then exported on to the preference granting market. In the case of AGOA the issue is whether assembly of imported fabrics into finished clothing, as is allowed for by the rules of origin currently applied to most beneficiaries of AGOA clothing provisions, is a sufficient process to enable African countries to derive substantial benefits from preferences. This rule entails that clothing producers in Africa can freely choose where to source their fabrics. The converse issue is whether more restrictive rules of origin would increase or decrease the benefits of AGOA to African countries. This is highly relevant since after October 2004 clothing from African countries will only receive preferences if the fabric is regionally produced or bought from the US. This requires that a much higher degree of processing must be carried out in the combined AGOA-US area but will reduce the 16Africa Growth and Opportunity Act Implementation Guide, October 2000, USTR. 25 competitiveness of African producers in the US market since they will have to source fabrics from a subset of higher-cost suppliers. Strict rules of origin are viewed by some as a mechanism for encouraging the development of integrated production structures within developing countries to maximize the impact on employment and to ensure that it is not just low value-added activities that are undertaken in the developing countries. However, there is no evidence that strict rules of origin over the past 30 years have done anything to stimulate the development of integrated production structures in developing countries. In fact such arguments have become redundant in the light of technological changes and global trade liberalization that have led to the fragmentation of production processes and the development of global networks of sourcing. Strict rules of origin act to constrain the ability of firms to integrate into these global and regional production networks and in effect act to dampen the location of any value-added activities. In the modern world economy, flexibility in the sourcing of inputs is a key element in international competitiveness. In the clothing industry, for example, modern analyses show that the key to moving up the value chain is to shift from simple assembly toward design and ultimately production of own label products.17 Limitations on the sourcing of materials will be a constraint rather than a stimulus to higher value-added activities. Thus, it is more than likely that the imposition of restrictive rules of origin rather than stimulating economic development raises costs of production by constraining access to cheap inputs and undermines the ability of firms to compete in overseas markets. It is also no coincidence that restrictive rules of origin act to protect U.S. clothing producers and stimulate demand for U.S. made fabrics. It is perhaps ironic that whilst technical assistance and advice to many of the countries in the region is stressing the importance of access to low cost imported inputs and the need to remove logistical and bureaucratic barriers for competitiveness and growth of exports, the rules of origin that will be 17See Mattila (2003), for example. 26 imposed under AGOA on clothing products in 2004 will act in exactly the opposite way. They will constrain access to low-cost imported inputs and will raise bureaucratic barriers to exports. A far better approach in stimulating the textile industry in Sub-Saharan Africa would be for the US to offer genuine preferences on all textile exports from these countries. The low utilization of clothing preferences by Mauritius and South Africa, the two countries which are currently subject to the more restrictive rules of origin is suggestive of the dampening effect of these rules on trade. Stevens and Kennan (2003) argue that whilst South Africa has the most developed textile industry in the region `its clothing sector cannot afford to use "originating cloth" without making its products uncompetitive'. Hence, the planned imposition of restrictive rules of origin on all AGOA beneficiaries in August 2004 is likely to have substantial negative impact on the flow of clothing products from Africa to the US. This is particularly important when bearing in mind that the removal of quotas from Asian exporters at the end of 2004 will lead to a much more competitive global market for textiles and clothing products. Forcing African suppliers to source fabrics from more expensive regional or US sources will handicap firms in these countries as they attempt to compete in the US market with products from China and other currently restricted suppliers in Asia. 9. Conclusions and Recommendations AGOA has yet to have a significant economic impact on one constituent group of countries ­ LDCs not eligible for clothing benefits. In 2002 no exports from these countries entered the United States under AGOA. For these countries AGOA will be of value if in the future it provides for the export of a greater range of products. A key issue is access to preferences on clothing. It is clear that AGOA preferences on clothing are the main source of gains under AGOA and have been a stimulus to export diversification for a small but increasing group of countries. Export diversification is a major factor underlying the ability of trade to contribute to growth and poverty reduction in these countries. 27 The special rule of origin that governs exports of these products from most beneficiaries is also important in this regard and its planned removal in 2004 will reduce the benefits currently being reaped by some countries and undermine the prospects for trade diversification by those countries that are currently, or will be in the future, seeking preferences on clothing. With these conclusions in mind we derive the following recommendations: 1. To entrench and enhance its benefits, AGOA should be extended over a much longer period18, if not made permanent. For many sectors it is likely that the current temporary and relatively short statutory period of AGOA constrains a significant investment response. 2. It is crucial that the liberal rules of origin for clothing products are not replaced, as planned, by restrictive rules in 2004. The liberal rules of origin have stimulated and facilitated trade in clothing; their removal will substantially reduce the value of AGOA preferences for many of the beneficiaries. Further, there is a range of countries that have only recently been granted access to clothing preferences and the application of strict rules of origin requiring sourcing of fabrics from the region or from the United States will severely hinder the ability of clothing producers in these countries to penetrate the U.S. market and will reduce the value of AGOA to these countries. There is also a group of countries that in the future could be granted clothing preferences, the attractiveness and benefits of which will be undermined with restrictive rules of origin. It is important that after the removal of quotas on textile and clothing products from Asian countries at the end of 2004 that African producers are not handcuffed in the new competitive conditions by restrictive rules of origin. 3. Countries not currently eligible should act to be granted access to preferences on clothing products. Preferences on clothing products, with liberal rules of origin, are a 18The Commission on Capital Flows to Africa (2003) has recommended that AGOA be extended to 2018 and that the liberal rules of origin also be extended until this date. 28 key source of benefits under AGOA. Those countries that have not yet been granted preferences on clothing should seek to upgrade customs procedures so as to satisfy AGOA requirements. Assistance from international institutions maybe of relevance. It is interesting to note that while exports of clothing to the United States from AGOA beneficiaries eligible for clothing benefits under the liberal rules of origin increased by 176 percent between 1999 and 2002, exports of clothing from AGOA beneficiaries not eligible for clothing preferences fell by almost 30 percent. Exports from countries with clothing preferences but more restrictive rules of origin increased by just over 30 percent. Thus, it is possible that the differential market access conditions under AGOA, including the nature of the rules of origin, may have contributed to the negative trend in exports from certain AGOA beneficiaries. This is perhaps an issue that warrants additional more detailed research. 4. AGOA should be enhanced by the removal of duties and quotas on all products. Extending preferences to excluded, high-duty, agricultural products and to industrial sectors such as textiles would offer greater scope for export diversification in beneficiary countries. Products that are subject to tariff quotas should be fully liberalized. While tariff quotas and the threat of safeguards remain, there is unlikely to be any significant investment in these sectors in AGOA beneficiaries and the value of the preferences on in- quota quantities will remain negligible. 5. Beneficiaries must address domestic constraints on investment and trade and incorporate actions relating to AGOA into a broad strategy for multilateral export diversification and expansion. While the changes proposed above would considerably enhance the opportunities offered by AGOA, they would not guarantee increases in exports or significant economic benefits. For example, there are some countries that are eligible for clothing benefits but do not yet export any clothing products to the United States. A conducive environment for investment and exports is crucial. Foreign direct investment can play an important role, at least in the initial stages of the development response, since foreign companies tend to have established marketing links to the U.S. market and bring strong technical and managerial capacities. 29 The analysis in this paper tends to reinforce the view of those who advocate the importance of multilateral trade liberalization through the WTO. For most African countries, while U.S. preferences can provide a stimulus to an initial export expansion of specific sectors, the overall benefits of preferential access are rather small and it is likely that for most countries the gains from multilateral liberalization will exceed any losses from preference erosion. Nevertheless, a small number of countries may be particularly affected by the erosion of existing preferences. However, this is not a reason to derail the multilateral process and any such difficulties are best addressed in the context of the overall development programs of the countries concerned with appropriate assistance from the multilateral institutions. Thus, identifying more carefully those who will be significantly affected by preference erosion is an important task for further research. In this light, AGOA preferences are not a panacea for success but rather should be seen as one element in the development of a strategy for export led growth and development. It is likely that in many AGOA beneficiaries the additional costs of operating relative to other locations compromise the margins of preference under AGOA. Only when customs clearance procedures are improved, costs of transport and other trade related services are reduced and corruption and other disincentives toward investment are removed will the full potential of preferential access to the United States under AGOA be realized. Addressing these issues now will allow for export growth to be broadly based and will provide a climate to ensure that as preferences decline with multilateral liberalization, an economic structure for continued export expansion is in place. 30 References Commission on Capital Flows to Africa (2003) `A Ten Year Strategy for Increasing Capital Flows to Africa'. Ianchovichina, E, A. Mattoo and M. Olarreaga (2002) `Unrestricted Market Access for Sub-Saharan Africa: How Much is it Worth and Who Pays?, Journal of African Economies, 10, 410-432 Lederman, D and C. Ozden (2003) `US Trade Preferences: All Are Not Created Equal', working paper, World Bank. Mattila, H (2003) `A Case Study of the Clothing Sector in Moldova', paper prepared for Moldova DTIS, mimeo, World Bank Mattoo, A, D. Roy and A. Subramanian (2003) `The Africa Growth and Opportunity Act and its Rules of Origin: Generosity Undermined?, The World Economy, 26, 829-51 Mwega, F and K. Muga (1999) `Africa and the World Trading System: A Case Study of Kenya', report prepared for AERC Project on Africa and the World Trading System, mimeo, World Bank Ozden, C and M. Olarreaga (2003) `AGOA and Apparel: Who Captures the Tariff Rent in the Presence of Preferential Market Access?', Mimeo, World Bank Stevens, C and J. Kennan (2003) `G8 Preferences for Africa: Country Studies', working paper, IDS. USTR (2003) `2003 Comprehensive Report on US Trade and Investment Policy Toward Sub-Saharan Africa and Implementation of the African Growth and Opportunity Act', USTR, Washington. 31 Appendix Table 1: The Importance of the U.S. as an Export Destination and Recent Evolution of Exports for AGOA Beneficiaries and other LDCs Total Exports Exports to Share of US in Growth of Growth of ($ Mill) 2002 US ($ Mill) Exports 2002 Total Exports Exports to US 2002 1999-2002 1999-2002 AGOA without Clothing Benin 237.3 0.7 0.3% -1.9% -96.2% Central African Republic 199.2 2.0 1.0% -18.8% -30.8% Chad 77.2 5.7 7.4% -28.2% -17.5% Congo, Dem. Rep. 1553.7 189.7 12.2% 29.3% -18.2% Congo, Republic of 2446.5 223.8 9.1% 32.3% -45.5% Cote d'Ivoire 4941.0 381.9 7.7% 5.5% 11.2% Djibouti 184.6 1.9 1.0% 11.7% 1638.8% Eritrea .. 0.4 ... ... -23.1% Gabon 3280.3 1622.0 49.4% -6.0% 7.2% Gambia, The 31.8 0.6 1.8% -67.3% 202.4% Guinea 856.6 71.6 8.4% 2.5% -38.0% Guinea-Bissau 128.5 0.0 0.0% 104.1% -51.5% Mali 181.7 2.6 1.4% -32.9% -70.9% Mauritania 583.5 0.9 0.2% 2.2% 23.2% Niger 152.7 0.9 0.6% -58.3% -81.5% Nigeria 19170.8 5819.6 30.4% 32.6% 39.5% Rwanda 147.2 3.1 2.1% 171.1% -16.3% Sao Tome and Principe 10.3 0.4 3.8% -25.0% -85.5% Seychelles 326.1 26.3 8.1% 117.0% 405.8% Sierra Leone 105.2 3.8 3.6% 1.6% -62.9% Total Group 34614.3 8357.9 24.1% 19.5% 22.0% Total Non-Oil 12163.2 916.3 7.5% 10.4% -21.4% AGOA plus Clothing with Liberal Rules of Origin Botswana 1749.9 29.7 1.7% 195.5% 75.5% Cameroon 2209.4 172.1 7.8% 5.6% 124.7% Cape Verde 23.0 1.8 7.9% 8.8% 2295.2% Ethiopia 538.6 25.7 4.8% 14.1% -15.1% Ghana 1665.5 115.6 6.9% -8.0% -44.8% Kenya 2467.3 189.2 7.7% 12.8% 78.2% Lesotho 357.3 321.5 90.0% 162.0% 190.1% Madagascar 895.5 215.9 24.1% 16.1% 169.2% Malawi 437.1 68.1 15.6% -17.3% 16.2% Mozambique 869.4 8.2 0.9% 225.2% -20.7% Namibia 801.2 57.4 7.2% 13.0% 91.3% Senegal 985.1 3.8 0.4% 20.0% -78.3% Swaziland 458.0 114.5 25.0% 54.7% 202.4% Tanzania 832.4 25.3 3.0% 19.7% -26.5% Uganda 373.4 15.2 4.1% -11.7% -25.0% Zambia 699.2 7.8 1.1% -21.6% -79.4% Total Group 15362.3 1371.7 8.9% 20.8% 56.4% 32 Appendix Table 1: The Importance of the US as an Export Destination and Recent Evolution of Exports for AGOA Beneficiaries and other LDCs (Cont.) Total Exports ($ Exports to US Share of US in Growth of Growth of Mill) ($ Mill) 2002 Exports 2002 Total Exports Exports to US 1999-2002 1999-2002 AGOA with Clothing but Restrictive Rules of Origin Mauritius 1850.7 280.4 15.2% -1.0% 8.6% South Africa 33018.9 4236.0 12.8% 11.8% 32.7% Total Group 34869.7 4516.4 13.0% 11.1% 30.9% US GSP LDC Angola 7949.4 3231.3 40.6% 64.5% 37.6% Bangladesh 6333.3 2133.7 33.7% 22.7% 11.0% Bhutan .. 0.8 ... ... 94.5% Burkina Faso 156.5 2.9 1.9% -7.9% 5.2% Burundi 33.9 0.7 2.0% -51.3% -90.2% Cambodia 1918.1 1070.7 55.8% 50.0% 80.8% Comoros 30.3 5.3 17.5% 147.2% 158.8% Equatorial Guinea 1959.1 572.6 29.2% 184.6% 1310.4% Haiti 313.6 254.6 81.2% -9.9% -15.4% Kiribati 37.3 1.1 3.0% 272.4% -21.9% Nepal 587.1 152.4 26.0% 0.8% -14.2% Samoa 73.4 6.4 8.7% 9.3% 20.1% Somalia 102.8 0.3 0.3% -19.2% 81.3% Togo 354.7 2.7 0.7% 10.3% -16.1% Vanuatu 90.1 2.8 3.1% 21.2% 40.0% Yemen, Republic of 2974.5 226.7 7.6% 31.8% 1449.4% Total Group 22914.2 7665.0 33.5% 43.2% 41.4% Total Non-Oil 14964.8 4433.7 29.6% 34.0% 44.3% UN LDC Afghanistan, Islamic Republic of 113.7 13.8 12.2% -15.5% 49.1% Lao P. D. Rep. 384.4 2.5 0.6% -10.6% -80.6% Liberia 1145.1 45.8 4.0% 88.1% 50.1% Maldives 235.7 114.0 48.4% 60.6% 107.6% Myanmar 2657.0 355.9 13.4% 99.9% 53.6% Solomon Islands 91.9 0.5 0.6% -45.6% -36.8% Sudan 2077.5 1.4 0.1% 168.1% 2258.8% Total Group 6705.4 533.8 8.0% 86.6% 57.0% Note: 1/ IMF Direction of Trade Mirror data 2/ USITC dataweb Mirror data * Country is grouped by 2002 AGOA status. Thus, some countries may be different from their current AGOA status. 33 34 35 36 37