Report No. 45945-AFR AFRICA Economic Partnership Agreements between Africa and the European Union: What to do Now? Summary Report October 2008 Poverty Reduction and Economic Management (PREM) Africa Region Document of the World Bank Currency Equivalents (October 1, 2008) US$1.00 = 0.710 1.00 = US$1.408 Abbreviations and Acronyms ACP African, Caribbean, and Pacific countries AGOA Africa Growth and Opportunity Act CARIFORUM Caribbean Forum of the African, Caribbean and Pacific Group of States CEMAC Communauté Economique et Monétaire de l'Afrique Centrale (Central African Economic and Monetary Community) CET Common External Tariff COMESA Common Market for Eastern and Southern Africa EAC East African Community EBA Everything-But-Arms EC European Commission ECOWAS Economic Community of West African States EDP European Development Fund EPA Economic Partnership Agreement ESA Eastern and Southern Africa EU European Union FDI Foreign Direct Investment FTA Free Trade Area GDP Gross Domestic Product GSP Generalized System of Preferences HS Harmonized System IPR International Property Right LDC Least Developed Country MFN Most Favored Nation NTB Non-tariff barrier REC Regional Economic Community RoO Rules of Origin RTA Regional Trade Agreement (or Area) SACU Southern African Customs Union SADC Southern African Development Community SSA Sub-Saharan Africa UEMOA Union Economique et Monétaire Ouest-africaine (West African Economic and Monetary Union) UNCTAD United Nations Conference on Trade and Development US United States WTO World Trade Organization Vice President : Obiageli K. Ezekwesili (AFRVP) Sector Director : Sudhir Shetty (AFTPM) Task Team Leader : Linda Van Gelder (AFTPM) ii Table of Contents ACKNOWLEDGEMENTS.......................................................................................................V EXECUTIVE SUMMARY......................................................................................................VI MAP..................................................................................................................................XXI 1. INTRODUCTION: A DEVELOPMENT PERSPECTIVE ON EPAS ..................................... 1 2. INTERIM EPAS: IMPLEMENTATION AND COMPLEMENTARY REFORMS ................... 3 A. The Outcome of EPA Negotiations to Date................................................................. 3 WTO Compatibility: The Driver of the EPA Process .................................................... 3 Revised Approach: A Flexible, Variable-Geometry Design .......................................... 4 Overview of Interim EPAs Initialed to Date.................................................................... 5 B. Access of African Exports to the EU Market ............................................................. 7 Tariff Preferences ............................................................................................................ 8 The EPAs' Rules of Origin.............................................................................................. 8 C. Liberalization of African Trade and Accompanying Reforms ............................... 13 EPA-Countries' Imports from the EU and Tariff Structures......................................... 14 Commitments to Liberalize Import................................................................................ 15 Potential Negative Effects ............................................................................................. 17 Parallel Reduction in Tariffs on All Imports ................................................................. 19 Potential Revenue Losses and Stronger Domestic Tax Systems................................... 21 D. Complementary Reforms to Generate A Strong Supply Response....................... 23 Supply-side Constraints and Competitiveness Problems............................................... 23 Types of Reforms Required to Spark a Strong Supply-Response ................................. 25 E. Supportive Development Assistance.......................................................................... 26 The EU's and its Member States' Aid Programs........................................................... 26 F. Risks and Mitigating Measures.................................................................................. 28 G. Conclusion: An Opportunity to Accelerate Trade-Related Reforms................... 30 3. FULL EPAS: REGIONAL INTEGRATION AND LIBERALIZING GOODS & SERVICES TRADE......................................................................................................................... 32 A. Regional Trade Integration and Open Trade Policies............................................. 32 Regional Participation in Interim and Framework EPAs .............................................. 32 Constraints to Expanding Intra-African Trade .............................................................. 34 Importance of Open Trade Policies ............................................................................... 36 A New Strategy for Regional Trade Integration............................................................ 39 Potential Role of Full EPAs in Advancing Regional Trade Integration........................ 41 B. Liberalizing Trade in Services and Foreign Direct Investment.............................. 44 Africa's Exports of Services to the EU Market ............................................................. 44 Liberalization of Imports of Services and Regulatory Reform...................................... 45 The Caribbean EPA: Provisions Relevant to Africa..................................................... 47 Enhancements and Complementary Competitiveness Reforms .................................... 49 iii Using EPAs to Improve Regulatory Environment for Services and Investments ......... 51 ANNEX A: THE CARIBBEAN EPA'S APPROACH TO THE LIBERALIZATION OF TRADE IN SERVICES ................................................................................................................... 53 REFERENCES..................................................................................................................... 56 Graphs Graph 1: Non-oil Imports from the African EPA-Countries as a Share of their Major Trading Partners'non-Oil Merchandise Imports, 1985-2006 ........................................................ 2 Graph 2: Destinations of the African EPA-Countries' non-Oil Merchandise Exports, 1985-2006 7 Graph 3: Graph 3: EU and US Imports of Clothing from African LDCs, 1995-2007 ................ 10 Graph 4: Sources of the African EPA-Countries' Merchandise Imports, 1985-2006.................. 14 Graph 5: EU Non-oil Imports from African LDCs, 1995-2007................................................... 23 Tables Table 1: Estimated Shares of Imports from the EU to be Liberalized.......................................... 16 Table 2: Estimated Revenue Losses from EPAs for Four COMESA Countries.......................... 21 Table 3: Comparative Investment Rates and Total Factor Productivity....................................... 24 Table 4: Comparison of Average Rankings of Major Developing Regions................................. 24 Table 5: Classification of EPA-Groups by Common External Tariffs......................................... 41 iv Acknowledgements T his study of Economic Partnership Agreements and related trade and regional integration issues was supported by grants from the Trade Window of the Bank- Netherlands Partnership Program trust fund and the Multi-donor Trust Fund for trade. The study was carried out by a team of consultants led by Larry Hinkle and composed of Sheela Ahluwalia, Luis de Azcarate, Elke Kreuzwieser, Jonathan Munemo, Zhen Kun Wang, and Fahrettin Yagci. The study team would like to thank Richard Newfarmer for his contributions to previous phases of the World Bank's analytical work on EPAs, Paul Brenton and Mombert Hoppe for their inputs on the EU's preference schemes and rules of origin, and Olivier Cattaneo for inputs on several legal issues. Soamiely Andriamananjara, Bernard Hoekman, and Richard Newfarmer were the peer reviewers for the study. Jean-Pierre Chauffour, Sumana Dhar, Philip English, Ian Gillson, Bernard Hoekman, Guggi Laryea, Richard Newfarmer, Salomon Samen, and Volker Treichel provided helpful written comments on drafts. Guggi Laryea's reporting on EPAs from Brussels has been valuable for keeping up with events while the reports were under preparation. v Executive Summary 1. In November-December 2007, the European Union and 18 African countries initialed interim Economic Partnership Agreements (EPAs) providing for reciprocal liberalization of merchandise trade with each other. In the first half of 2008, further negotiations were launched for full EPAs to address additional issues concerning regional trade integration, liberalization of trade in services, foreign direct investment, and trade- related regulations. This report analyzes from a development perspective the steps required for successful implementation of interim EPAs and the potential role of full EPAs in advancing regional trade integration in Africa and in addressing the liberalization of trade in services and related issues.1 Implementation of Interim EPAs and Complementary Reforms 2. After adoption of a flexible, variable-geometry approach to EPAs in the quarter of 2007, eighteen of the 46 African countries eligible for EPAs quickly reached agreement with the EU on interim EPAs. These established bilateral free trade arrangements for merchandise trade to replace the Cotonou trade-preferences when they expired in January 2008. The interim EPAs contain non-binding rendezvous clauses stating the parties' intentions to continue negotiations towards full EPAs in a period of 12-18 months. Despite their name, interim EPAs are in fact permanent in that they will govern trade between the eighteen African signatories and the EU unless and until replaced by another trade agreement. 3. Uneven Participation in Interim EPAs. There were sharp differences in participation in interim EPAs between LDCs and non-LDCs and between the eastern- southern and the western-central African regions. These variations reflect both the different market-access alternatives faced by LDCs and non-LDCs after the expiry of the Cotonou trade preferences and the diversity of regional and country trade policies and interests. Ten of Africa's 14 non-LDCs, which otherwise would have reverted to the EU's less favorable tariff preferences under its General System of Preferences (GSP) after expiration of the Cotonou trade-preferences, initialed interim EPAs. In contrast, only eight of Africa's 33 LDCs, which would continue to have tariff-free, quota-free access to the EU market under its EBA program, initialed interim EPAs. There were similar differences in participation rates between regions: fifteen of the 23 countries in eastern and southern Africa initialed interim EPAs, but only three of the 24 countries in the western and central Africa did so. 1The terms "development friendly" and "pro-development" as used here in assessing EPAs refer to policies that will accelerate global and regional trade integration in Africa and sustained long-term growth of real GDP per capita. The term "trade integration" is used to refer to increases in the ratio of total trade (that is, imports plus exports of goods and non-factor services) to GDP. vi 4. Establishment of a WTO-compliant Trade Regime. Interim EPAs bring the trade relations between the EU and the signatory countries into line with WTO rules and thus achieve one of the key objectives set out in the Cotonou Agreement. They establish reciprocal, asymmetric arrangements for preferential merchandise trade between the EU and the EPA-signatories. Under the interim EPAs, the EU has granted immediate tariff- free and quota-free access to its market for 100% of its imports from the EPA-signatories, with short transition periods only for rice and sugar and more liberal rules of origin for clothing exports. The EPA-signatories are required to reciprocate by gradually granting tariff free access to their own markets for at least 80% of their imports from the EU within a 15 year transition period. 5. Implementation Requirements. To realize the potential of interim EPAs for accelerating development, key actions will need to be taken on both the EU and the African sides during their implementation. From the EU, further liberalization of the EPA's current restrictive rules of origin will be the key step. From the African EPA- signatories, a series of tariff, tax, and other complementary reforms to improve competitiveness will be needed. These two sets of actions are discussed in order in the following sub-sections. Access to the EU Market and the EPA's Rules of Origin 6. The EU's Tariff Preferences. Although its relative importance has declined over the last two decades, the EU is still by far the largest single market for Africa's non-oil exports, absorbing 51% ($24 billion) of the total in 2006, a share that is six times the 8% share of Africa's second largest market (the US) for its non-oil exports. Now, the 18 EPA-signatories have tariff-free, quota-free access to the EU market under their interim EPAs; all African LDCs have similar tariff-free, quota-free market access under the EU's Everything-but-Arms (EBA) program; the three oil-exporting non-LDCs not participating in EPAs have tariff preferences under the EU's Generalized System of Preferences (GSP); and South Africa has its own separate free trade agreement with the EU. This combination of trade arrangements will maintain (or, in some cases improve) Africa's previous preferential access to the EU market under the Cotonou Agreement and has avoided significant disruption of current Africa-EU trade. 7. The EPAs' Restrictive Rules of Origin. By far the most important remaining market-access issue is the further liberalization of the EPAs' rules of origin, which seriously limit the benefits of the tariff-free, quota-free market access provided under interim EPAs. The improvements in market access under interim EPAs relative to the previous Cotonou tariff-preferences are limited, with the notable exception of the more liberal rules of origin for clothing exports. Hence, there is no compelling reason to expect that the export performance (other than of clothing) of the EPA-signatories will significantly improve under interim EPAs unless the EPA's rules of origin are further liberalized and supply constraints and competitiveness problems in the EPA-signatories are addressed. vii 8. Substantially liberalizing the EPA's current rules of origin will thus be important for the ultimate success of the interim EPAs in increasing and diversifying the signatories' exports to the EU. It is in the interest of Africa EPA-countries to aggressively pursue such liberalization in future negotiations. In particular, simpler, standardized rules of origin need to be established by replacing all of the current detailed product and process specific rules of origin with a uniform change of tariff heading at the HS 6-digit level or 10% value added rule. In addition, EPA-signatories will need to take steps to overcome supply constraints and address as discussed below. Liberalization of African Trade and Accompanying Reforms 9. Limited and Selective Market Opening under Interim EPAs. The opening of African economies under EPAs will be significantly smaller than many of the press reports might lead one to believe. The 18 interim EPAs initialed to date provide for the elimination of tariffs on 80-85% of imports from the European Union over a 10 to 15 year period as required for WTO-compatibility but will lead to only limited, selective market opening. The countries entering into interim EPAs appear to have all followed fairly similar approaches in choosing the 80-85% of imports from the EU on which to eliminate tariffs under EPAs. In most cases, the first tranche of the scheduled tariff "reductions" consists largely of binding current applied zero-rate MFN tariffs at a preferential zero-rate for imports from the EU. Tariffs eliminated on imports from the EU in subsequent tranches appear to have been chosen in most cases by starting with the products bearing the lowest non-zero MFN tariff rates and working upwards until 80% of imports from the EU are covered by zero-rate preferential tariffs. The timing of the import liberalization tranches under the interim EPAs has, in all cases, been set to eliminate tariffs on at least 80% of imports from the EU in 15 years or less. 10. Because of the simple structures of production in Africa's LDCs and its other small low income economies, the exclusion of 20% of imports from liberalization will leave high tariffs in place on most domestically produced products in most countries. In addition, because the existing structure of tariffs and trade with EU varied considerably from country to country, both the composition of the 20% of imports from the EU tariffs on which tariffs will remain and the tariffs on imports from non-EU sources will vary significantly from country to country despite the similarity of approaches in choosing the 80% of imports from the EU on which to eliminate tariffs. A problematic aspect of the import liberalization schedules under the interim EPAs is thus that they establish a minimalist norm of very slow and highly selective tariff reductions, leaving high levels of protection in place on virtually all import-competing domestic products. 11. Potential Negative Effects of Preferential Elimination of Tariffs on Imports only from the EU. Since many of the EPA-countries still have high and distorted MFN tariffs and the EU supplies of 34% of their merchandise imports, preferential elimination of tariffs on 80% of imports from the EU could have some significant negative effects. In one-half of the EPA-countries, the tariff rates on 45% of tariff lines are 15% or higher and the maximum tariff rate is 30% or higher, reaching 60% or more in some regions. As a result, effective protection rates of 60%-90% for value added in import-competing viii domestic industries are still quite common. Preferential elimination of tariffs on imports from the EU, in the presence of still high MFN tariffs, will further distort tariff structures and production incentives. Tariff preferences for the EU may also lead to some costly diversion of trade from low-cost non-EU suppliers to higher-cost EU suppliers and monopolistic pricing by EU exporters and implicit transfers of forgone tariff revenues from African governments to EU exporters. Over time, tariff preferences for the EU could also have a hub-and-spoke effect on the EPA-signatories' pattern of trade that could reinforce the current low level of trade complementarity among their economies.2 12. Necessary Parallel Reductions in MFN Tariffs on Imports from All Sources. To avoid the negative effects that could result from the preferential elimination of tariffs on imports from the EU under interim EPAs, the signatory countries will need to unilaterally (or regionally in the case of customs unions) reduce tariffs on imports from all countries in all sectors, including, in particular, the main domestic import-competing sectors, to lower the prices of goods imported from other sources and increase competition in African markets both with imports from EU-suppliers and with overly protected domestic producers. At a minimum, three reforms in MFN tariffs will be needed to reduce the inherent trade-diverting, distortionary effects of tariff preferences for the EU and avoid unduly favoring EU suppliers. First, any remaining quantitative restrictions, import licensing requirements, discriminatory domestic taxes on imports, and similar trade-distorting NTBs need to be eliminated. Second, phased reductions in MFN tariffs, initially focused on cutting down tariff peaks to lower the maximum MFN tariff rate to 15% or less, are needed to reduce the most excessive rates of protection, to avoid increasing effective protection for import-competing industries as tariffs are eliminated on inputs imported from the EU, and to increase competitiveness generally. Third, the MFN tariffs on products that will be imported duty-free from the European Union under interim EPAs will need to be lowered to 5% or less prior to the elimination of the tariffs on imports from the EU to avoid creating an excessive price advantage for EU suppliers and limit trade diversion. 13. The foregoing three reforms constitute a basic tariff reform program that all signatories of interim EPAs will need to implement. Depending upon a country's initial tariff structure and its choice of imports from the EU to liberalize, some countries could find that this basic package would be inadequate to eliminate all of the serious distortions in their tariff structure and, hence, would need to implement additional tariff reductions. In addition, more active reformers may want to undertake a faster and more fundamental liberalization and restructuring of their tariff systems. 2 In the case of free (preferential) trade agreements, like the interim EPAs, when a hub country or region (that is, the EU) enters into free trade agreements with various small countries like the EPA-signatories (the spokes) and the latter do not similarly liberalize trade among themselves, the hub country tends to benefit more because it has free access to all markets whereas the spokes only have free access to the hub market. This hub-and-spoke effect increases the incentive for hub-country exporters to invest in the hub country, rather than in the spokes, in order to serve all of the markets. The potential hub-and-spoke effect from interim EPAs can be mitigated by liberalization of trade among the EPA-signatories. ix 14. Potential Revenue Losses and the Strengthening of Domestic Tax Systems. Potential losses of tariff revenue as a result of the tariff reductions under interim EPAs are an important concern for some countries. The most credible of the publicly available estimates of likely revenue losses from eliminating tariffs on imports from the EU suggest for four countries in eastern and southern Africa that these losses may amount to only 1% to 4% of total tax revenues. However, because of low collection efficiency and narrow existing domestic tax bases, low income developing countries typically are more dependent on trade taxes than other countries and sometimes have serious difficulty in replacing revenues lost due to tariff reductions. 15. All countries integrating into the global trading system must sooner or later replace tariff revenues with increased domestic taxation. Irrespective of the precise magnitude of the eventual revenue losses, domestic tax systems will need to be strengthened and tax and customs administration improved to replace tariff revenues that will be lost because of tariff reductions under interim EPAs and the necessary parallel reductions in MFN tariffs. As the elimination of tariffs will be limited to 80% of imports from the EU and, in most cases, will take place 10 to 15 years from now, EPA-signatories will have a long period in which to strengthen their revenue systems to offset these losses. Complementary Measures to Generate a Strong Supply Response 16. Supply Constraints and Competitiveness Problems. Alone, improved market access is unlikely to be enough to substantially accelerate the growth and diversification of the EPA-signatories' exports to the EU as limited supply capacity and competitiveness problems are the binding constraints to expansion and diversification of trade in most African countries. A wide range of supply-side constraints in African LDCs has inhibited a strong supply-response to their tariff-free, quota-free access to the EU market since the introduction of the EBA program in 2001; and their share in total EU non-oil imports has continued to decline. Most of Africa's non-LDCs face the same types of supply constraints and competitiveness problems as its LDCs. Hence, unless these constraints are effectively addressed, the immediate response of both the LDCs' and the non-LDCs' exports to tariff-free, quota-free access to the EU market under interim EPAs is likely to be weak. 17. Both the level and the efficiency of investment in Africa will need to increase to achieve a strong supply response to take advantage of tariff-free, quota-free market access under interim EPAs. However, in most African countries, the prospects for increasing investment are limited because of serious weaknesses in the business and investment climate. Africa has the lowest rankings of any major developing region for both its competitiveness by the Global Competitiveness Report 2007-2008 and for its business climate by Doing Business 2008. Reported deficiencies in infrastructure, governance, institutional capacity, the macro-economic environment, and the regulatory framework increase the cost of doing business and lower productivity and competitiveness in Africa. x 18. Types of Reforms Needed. Substantial improvements in the business and investment climate will be needed over the next decade to benefit fully from improved market access and trade liberalization. In addition to implementing MFN tariff reductions and strengthening domestic tax systems to avoid negative outcomes from the preferential elimination of tariffs on imports only from the EU, EPA-signatories would need to use EPAs to help leverage a series of complementary trade, fiscal, regulatory, and institutional reforms to improve the business climate, increase investment in infrastructure, liberalize imports of services, and facilitate trade. Most of the reforms needed to provoke a strong supply response require unilateral domestic actions. Since the importance of different competitiveness problems, the opportunities for correcting these, and progress under existing reform programs vary from country to country, reform priorities will have to be determined on a country-specific basis rather than a regional one. Furthermore, some of these unilateral country-specific reforms may, in fact, be higher immediate priorities for the countries concerned than the reforms directly necessary to support EPA implementation. Each EPA-signatory will thus need to work out its own country-specific reform program. 19. Risks. Partial or non-implementation of necessary policy reforms is the primary risk faced by the 18 countries that have initialed interim EPAs. Some EPA-signatories may not be able to implement the required parallel MFN tariff reductions and restructuring of their revenue systems because of lack of political support or limited administrative capacity. Possible resulting negative outcomes are some trade diversion, monopolistic pricing effects, further distorted tariff structures, and revenue losses. The remedy for any such negative effects is implementation of the necessary accompanying MFN tariff reductions and tax reforms. However, even if the reforms needed to avoid possible negative effects are implemented, the supply response could be limited in countries that do not unilaterally implement investment-climate reforms and other supply-side measures. 20. Supporting Development Assistance. The opportunity to effectively coordinate trade and aid with a major trading partner and aid donor is a distinct advantage of the EPA-process relative both to the multilateral WTO trade negotiations and to most other bilateral trade negotiations, where the links between aid and trade have often been tenuous or non-existent. Development assistance can help with adjustment costs for firms and workers, temporarily make up for revenue losses from tariff reductions, support the implementation of trade and business climate reforms, and improve infrastructure to take advantage of increased export opportunities. Potentially, countries aggressively implementing interim EPAs and related reforms could receive significant amounts of additional development assistance and aid for trade from donors making performance based aid allocations. Because many EPA-signatories have yet to formulate complementary reform programs, they have also been slow to identify specific requirements for development assistance to support these and need to do so as they design their reform programs. xi Full EPAs: Regional Integration and Liberalizing Trade in Goods and Services 21. In addition to increasing global trade integration, EPAs are also intended to promote regional integration in Africa, which is an important political as well as economic objective for many EPA-countries, and provide an important opportunity to liberalize trade in service and related issues. This section reviews the implications of the EPAs initialed thus far for regional trade integration, the constraints to expanding intra- African trade, and the potential roles of EPAs and open trade policies in advancing regional trade integration and liberalization in services imports. Negotiations of Framework Regional EPAs 22. Eastern and Southern Africa. As noted earlier, there were pronounced regional differences between Eastern-Southern and Western-Central Africa in participation in EPAs. In Eastern and Southern Africa, the flexible variable-geometry EPA-design adopted in October 2007 led to quick agreement on regional EPAs with EAC, ESA, and SADC EPA-groups, as well as to the interim EPAs with 15 countries. These three EPA- groups initialed "framework" regional EPAs covering WTO-compatible reciprocal import liberalization, development cooperation, and plans for working towards full EPAs in 2008-9. Two different types of regional EPAs emerged from the regional negotiations: a customs-union based approach that was adopted by EAC and an approach based on free (preferential) trade areas that was adopted by the ESA and SADC EPA-groups. 23. Western and Central Africa. In contrast, Western and Central Africa, where only three countries initialed "interim" EPAs, showed little initial interest in even framework regional EPAs. No regional EPAs have yet been initialed in these two regions, and negotiations of these did not advance very far in 2007. 24. Implications for Regional Trade Integration. Alone, in their current form, most of the interim and framework EPAs will do little to advance regional trade integration in Africa. The one important exception is EAC: its member countries entered into an interim EPA as a custom union with one common schedule of imports from the EU to be liberalized that could facilitate the liberalization of intra-EAC trade in the same products. The other interim and framework EPAs have not yet reinforced regional trade integration efforts in Africa. Furthermore, as explained earlier, EPA-signatories will need to make parallel unilateral reductions in their MFN tariffs to limit the hub-and-spoke effects that could be caused by their implementing bilateral free trade arrangements with EU without simultaneously liberalizing intra-African trade. However, those EPA-signatories that do liberalize imports from the EU under interim EPAs and implement the necessary accompanying reductions in MFN tariffs will be better positioned to further liberalize their intra-African trade, as well as their global trade, than countries not doing so. xii Constraints to Expanding Intra-African Trade 25. In contrast to Africa's trade with the European Union, intra-African trade is quite low. Officially recorded intra-regional trade in the EPA-negotiating groups averaged only 6.9% of the groups' total imports plus exports in 2006. Additionally, the share of intra-regional trade in the EPA-groups' total trade has not changed significantly since 1990 except in EAC, where it has increased slightly. 26. Policy and Institutional Barriers to Intra-African Trade. Intra-African trade is low for two reasons. First, despite the proliferation of preferential trade agreements in Africa, substantial policy barriers to intra-regional trade still remain within both free trade areas and customs unions in all five of the regional EPA-groups. The current mixture of "free" trade agreements that have myriad exceptions, protectionist rules of origin, and trade-distorting NTBs and customs "unions" that maintain customs and other barriers to intra-union trade has done relatively little to eliminate the policy and institutional barriers to intra-African trade. Furthermore, some natural trading partners in Africa with substantial trade between themselves belong to different regional trade areas (RTAs) and EPA-groups. Trade between these countries thus does not benefit from measures designed to liberalize intra-regional trade within RTAs and EPA-groups. 27. Low Trade Complementarity. Second, medium-term prospects for expanding intra-African trade would be modest because of the similarity of the African exports. Because of the small size of the EPA-countries' economies and the similarities in their exports, production structures, and income levels, trade complementarity among their economies is low. These similarities limit the range of products that can be exchanged with regional partners. The products exported by African countries--mostly primary commodities--are not the main imports of African countries, which consist of manufactures and capital goods sourced mainly from the European Union and Africa's other major global trading partners. Other than South Africa, only a few African countries have even a limited capacity to supply manufactured products, in most cases simple consumer goods, to regional markets. Trade complementarity is determined by the structure of production and the composition of demand in the countries concerned To the extent that trade complementarity changes over time, it changes only slowly as economies develop. In the medium term, Africa's current imports of manufactures, machinery, and equipment cannot be domestically produced to expand intra-Africa trade because most African countries lack the capacity to do so and will need a long period of time to develop more diversified economies. 28. Consequently, even if the various policy, institutional, and infrastructure constraints to intra-African trade could be overcome, medium-term prospects for expanding intra-African trade would be modest because the similarity of the EPA- countries' exports limits the range of products that can be exchanged with regional partners. Intra-African trade is likely to continue to grow less rapidly in the medium term than Africa's international trade with countries in other that have more complementary economies, and the potential gains from regional trade integration are likely to be relatively small compared to those from global trade integration. Supporting the xiii expansion and diversification of international trade should, therefore, be the primary objective of trade integration efforts. Effective regional trade integration can, however, provide limited additional opportunities for some small African economies, particularly landlocked ones with few other options, to take advantage of scale economies and increase competition, efficiency, and competitiveness. A New Strategy for Trade Integration 29. Unsuccessful Common-Trade-Policy Strategies. For three decades or more, most African regional trade areas (RTAs) have sought to promote intra-regional trade by emphasizing common trade policies, particularly common external tariffs, to harmonize trade regimes across their member countries. However, differences among their members in economic conditions, trade interests and policies, and political willingness to liberalize intra-African trade have often led to paper agreements on common trade policies that could not be effectively implemented, as reflected in the numerous existing policy and institutional obstacles to intra-African trade, because of underlying divergences between countries in interests and policies. The current combination of "free" trade agreements that have numerous exceptions, protectionist rules of origin, and trade-distorting NTBs, and customs "unions" that maintain customs and other barriers to intra-union trade has done relatively little to eliminate the policy and institutional barriers to intra-African trade. The expansion of intra-African trade is thus still constrained by a long list of policy and institutional barriers: free trade arrangements and common external tariff, where they exist, are only partially implemented; numerous restrictive non-tariff barriers to intra-African trade are still maintained; customs administration is inefficient, and corruption is often a problem; and transit and transport services are poor and costly. Effectively promoting regional trade in Africa, whether through full EPAs or independently of them will thus be no simple matter. 30. Export-Oriented Trade Policies and Regional Trade Integration. If the EPA- countries wish to accelerate regional trade integration in parallel with the more important global trade integration, a radically different policy approach from the unsuccessful common-trade-policy strategies followed thus far will be needed. Export-push strategies have growth, efficiency, and distributional advantages relative to regional import- substituting ones in Africa. The low trade-complementarity of most African economies both makes it difficult to expand intra-African trade in the medium term and magnifies the importance of global trade integration and open trade policies for them. 31. Moreover, broadening the scope of intra-African free trade is difficult in countries where external (MFN) tariffs are still high. The current high tariffs on most domestically produced import-competing products result in high protection levels and rents for the firms that benefit from these. Elimination of tariffs on intra-African trade (whether through FTAs or customs unions) in import-competing products that are protected by high tariffs will cause substantial reductions in these rents. The loss of such rents is strongly opposed by the vested interests concerned. Hence, it is often politically difficult or impossible to effectively implement free trade in products bearing high MFN tariffs. In addition, preferential tariff-free importation from regional producers of a product otherwise subject to a high MFN tariff gives the regional producers a substantial xiv price advantage over external ones and can cause a significant revenue loss for the government as well as diversion of trade from more efficient suppliers outside of the region. Consequently, government commitment to implementing free intra-African trade in high tariff products is often weak. As a result, most high-tariff products that compete with domestic production are usually explicitly excluded from arrangements for free intra-African trade and, when not excluded, are constrained by protectionist rules of origin and trade-distorting NTBs. Open, low-tariff trade policies are, therefore, important for advancing regional trade integration in Africa. 32. Regional integration strategies thus need to be outward-oriented, and supporting the expansion and diversification of international trade should be a primary objective of regional trade integration efforts. The essential components of an open trade strategy are elimination of trade-distorting NTBs, reductions in MFN tariffs, particularly peak rates, and effective implementation of measures to ensure that exporters have ready access to competitively-priced duty-free inputs (such as export processing zones). 33. A Flexible Variable-Geometry Approach Focusing on Free Intra-African Trade. Because of the wide differences among the members of the various EPA-groups in initial conditions, trade interests and policies, and political readiness to liberalize, the variable-geometry approach adopted for EPAs in the last quarter of 2007 will remain essential. The priority for regional trade policy will also need to be on phased implementation of free intra-African trade in all products (or, for those products for which free trade is not initially possible, preferential trade) both within and between regional trade areas and the liberalization of the restrictive rules of origin governing free/preferential intra-African trade. An additional important component of a new regional integration strategy would be to focus collective efforts on improving the regional business climate and competitiveness, an objective which the inclusion of the liberalization of trade in services and investment in full EPAs could help advance as discussed below. Potential Role of Full EPAs in Advancing Regional Trade Integration 34. Full EPAs could, potentially help accelerate regional trade integration by (a) providing a dynamic stimulus and a coordinating mechanism for regions to undertake reforms to liberalize trade on an MFN basis and to improve the business climate and competitiveness; (b) facilitating liberalization of intra- and inter-regional policies governing intra-African trade by addressing such issues as elimination of trade-distorting NTBs, creation of more effective regional preference schemes, and liberalization of the restrictive rules of origin governing free/preferential intra-African trade; and by (c) enhancing the credibility of regional integration by locking-in reforms in an international treaty. Concluding full EPAs in a flexible variable-geometry format is an achievable objective from some African regions. However, significant progress in actually liberalizing intra-Africa trade under full EPAs will require an open trade strategy as discussed above. It will also have to be largely African driven as the incentives that the EU can offer in terms of new market access will be limited unless the EU can substantially liberalize the EPAs' rules of origin as recommended in paragraph 8. xv Liberalization of Imports of Services and Regulatory Reform 35. In addition to helping to catalyze progress on regional trade integration, full EPAs could also provide an important opportunity to address the liberalization of trade in services and related issues. Services are both an important input contributing to the competitiveness of goods production in the world market and growth sectors in their own right. Two thirds of the EPA-countries' services exports go to the EU, roughly half of their services imports come from the EU, and the EU is the source of more than 80% of their foreign direct investment. 36. Because of the current underdeveloped nature of the service sector in most of Africa, the major gains from the liberalization of trade in services are likely to come from the liberalization of imports of services rather than from increased African exports of services. In particular, liberalization of foreign direct investment in services (delivery of services by GATS Mode 3, commercial presence) offers the largest potential for generating efficiency gains that are needed for increasing the competitiveness of Africa's merchandise exports and its economies more generally. For many services ­ particularly, finance, telecommunications, transport, and energy -- the gains from increased investment, trade-induced improvements in efficiency and competitiveness, and more rapid growth will be magnified because these play such a major role in all types of production. The boost in competitiveness needed to enable African economies to reap the full benefits of merchandise trade liberalization under "interim" EPAs will not materialize without a parallel opening of the services sector. 37. Establishment of an Appropriate Regulatory Framework. Because of their characteristics, serious problems can result from the liberalization of some service sectors without a satisfactory regulatory framework. Sectors in which large fixed investments are necessary for the efficient provision of services are often imperfectly competitive even with free entry. Market failures caused by monopoly power, asymmetric information, or externalities in the provision of services create the need for appropriate regulation. In addition, some services, such as finance and telecommunication, require some regulation to prevent destabilizing or anticompetitive practices. Thus, to reap the full benefits of liberalizing imports of services and foreign direct investment in these service sectors, establishment of an appropriate regulatory framework will be indispensable. 38. Complementary MFN Liberalization of Service Imports from all Sources. As in the case of merchandise trade, the African EPA-countries would benefit more from reducing their barriers to service imports from all foreign sources rather than liberalizing on a preferential basis with just the EU. In cases where network and capital intensive services are not already open to foreign investors, preferential liberalization of these services for investors only from the EU could give EU service providers a permanent advantage even if the services concerned are eventually liberalized multilaterally. Because the EU is unlikely to be the most efficient provider of all network and capital intensive services, preferential liberalization of imports of less-competitive services from the EU could lead to sub-optimal development of these sectors in the EPA-countries xvi when the investors from the EU are not the most efficient available providers of the services concerned. Liberalization of services imports from all sources tends to increase competition among service providers and is likely to lead to greater efficiency gains and more rapid growth. The Caribbean EPA: Provisions Relevant to Africa 39. In December the EU signed a full EPA with the 16 CARIFORUM countries in the Caribbean that establishes a comprehensive framework for future actions to address liberalization of trade in services, foreign direct investment, and trade-related regulations. Many of the Caribbean EPA's provisions concerning the liberalization of services and foreign direct investment are equally relevant to Africa and are worth consideration in designing full EPAs. These provisions would, however, need to be both enhanced and accompanied by complementary unilateral business climate reforms in order to have significant positive effects on investment and competitiveness. 40. Provisions Relevant for African EPAs. As in the Caribbean, the investment provisions of African full EPAs should include the principles of free entry, national treatment, and non-discrimination applicable to both goods and services producing sectors. Because of the wide range of EPA-outcomes thus far in different regions and the large variations among different African countries in readiness to implement reforms, the flexible design and phased approach to the liberalization of services and investment in the Caribbean EPA with its built-in schedule for further progressive liberalization also appear well suited for full EPAs in Africa. With this type of flexible design, reform priorities in African EPAs could reflect the interests and constraints of individual countries. The Caribbean EPA's competition and regulatory provisions would also be appropriate for some full EPAs in Africa. Such competition and regulatory provisions could provide a catalyst and coordinating mechanism for setting up or strengthening national and regional competition and regulatory agencies. Inclusion of a regional preference clause in full EPAs similar to that in the Caribbean EPA could provide an additional stimulus to intra- and inter-RTA efforts to liberalize to trade in services and investment in Africa. 41. Required Enhancements. The Caribbean EPA has three shortcomings concerning the liberalization of imports of services and foreign direct investment that would, need to be corrected for a similar approach to have greater positive effects on investment and competitiveness in Africa. First, in order to promote as much competition as possible, to attract investment by the most efficient service providers, and benefit fully from economies of scale over the long term, services imports and investment should be liberalized from all countries on a MFN basis in African EPAs instead of just vis-à-vis the European Union and regionally as in the Caribbean EPA. Second, the Caribbean experience has shown that the liberalization of trade in services is, in fact, more technically and politically demanding than the liberalization of merchandise trade and that the results from the initial effort risk being minimal. In order to reap any immediate substantive benefits from a flexible framework and phased approach to liberalization participating African countries will need to identify priority service sectors with good development potential and formulate country-specific and regional proposals xvii for actually liberalizing these in the first phase rather than leaving all new liberalization effort to the future as in the Caribbean. Third, the benefit of including investment in production of goods (merchandise), as well as services, in full EPAs will be limited by the interim EPAs' current restrictive rules of origin. Liberalizing the rules of origin for merchandise trade as recommended in paragraph 8 above will be particularly important for attracting FDI for production of products that are currently unable to benefit from tariff-free, quota-free access to the EU market because of the interim EPAs' current rules of origin. In addition, because of their low levels of service sector development, many of the African EPA-countries will probably not be able to increase their exports of services to the EU substantially in the medium term; and liberalization of the EPAs' rules of origin for merchandise trade will be an important incentive for these countries to liberalize imports of services and investment from the EU. 42. Necessary Complementary Competitiveness Reforms. Unless additional unilateral steps are taken to further improve the business climate and competitiveness, the impact of guaranteed free entry and non-discrimination under full EPAs would be likely to be small even on FDI and would be insignificant on domestic investment. The effect of these reforms on investment and competitiveness in the African EPA-countries would be exclusively, or primarily, on foreign direct investment. In addition, all of the other legal, regulatory, governance, financial sector, and infrastructure problems that have led to low business climate and competitiveness ratings in the African EPA-countries would remain as disincentives to both foreign and domestic investment. Hence, additional unilateral reforms in the business climate and competitiveness are likely to be essential for raising investment rates, increasing productivity, and generating a strong supply-response in the EPA-participants as discussed in paragraph 17 above. Using EPAs to Improve the Regulatory Environment for Services and Investments 43. As in the case of regional trade integration, the liberalization of services imports and investment under full EPAs needs to be seen primarily as an opportunity for stimulating reforms that are necessary in the long term for increasing global and regional trade integration. Inclusion of trade in services and investment in full EPAs could provide an important opportunity for participating countries to liberalize their service sectors and investment regimes globally and regionally, while obtaining financial and technical assistance from the EU for building capacity in service sectors for which they make liberalization commitments. However, unless the EU is able to liberalize the EPAs' rules of origin for merchandise trade under full EPAs, African countries will have little incentive for liberalizing imports of services and investment under full EPAs other than the desire to accelerate the reforms involved. The reform process would, therefore, have to be African-driven. xviii Conclusion: An Opportunity to Accelerate Trade-Related Reforms 44. Interim EPAs. For those African countries and sub-regions that chose to enter into interim EPAs, the EPA-process could provide an opportunity to leverage action on needed trade and related supply-side issues and thus accelerate trade integration and growth. The requirement for EPA-signatories to eliminate tariffs on imports from the European Union offers an opportunity for them to accelerate MFN trade liberalization and the strengthening of domestic revenue systems and to lock in these reforms in a way that makes them credible to the rest of the world. The improvements in access to the EU market under EPAs and the associated technical and financial assistance from the European Union and its member countries could create favorable external conditions for undertaking related trade and investment-climate reforms. 45. Participation of Additional Countries in Interim EPAs. Should additional African LDCs and oil-exporting non-LDCs consider entering into EPAs that provide for WTO-compliant asymmetric, reciprocal liberalization of merchandise trade? Conditionally, yes. However, the primary motivation for doing so would have to be the desire to accelerate reforms needed for faster global and regional trade integration. Improvements in access to the EU market could also be a potentially significant benefit for (a) countries that can substantially increase clothing exports to the EU, (b) LDCs that grow fast enough in the medium term to graduate from LDC status and hence from eligibility for the EU's EBA program, and (c) oil-exporters that need to diversify their exports. On the other hand, for LDCs and oil-exporters that are not interested in or are not able to accelerate reforms, the EU's EBA and GSP tariff preferences, from which these countries currently benefit, provide satisfactory access to the EU market. 46. Full EPAs. Full EPAs offer a broader opportunity to tackle additional reforms needed to advance regional trade integration and the liberalization of trade services. This study's recommendations for advancing the reforms related to both interim and full EPAs are summarized in Box 1 below. Note that, other than the liberalization of the EPAs' rules of origin, the actions required for achieving pro-development outcomes from EPAs will need to be taken primarily by the EPA-signatories. 47. The ability of individual countries to use full EPAs to leverage a wide range of competitiveness reforms will be critical in determining how much each country will ultimately benefit. The relevance and usefulness of full EPAs for specific countries will depend both upon the priority for each country of the EPA-related reforms affecting services and foreign direct investment and of the necessary complementary reforms for improving competitiveness and upon the country's capacity to implement these reforms. All of the reforms involved in full EPAs could also be implemented unilaterally by active reformers, although the catalytic effects of the full EPAs, the visible political commitment to liberalization from entering into these, and the locking-in of open investment policies through a treaty would be forgone. xix Box 1: Summary of Recommendations Implementation of Interim EPAs and Complementary Measures 1. Revised Rules of Origin for EPAs (paragraph 2.30): a. Eliminate all the current complex product and process specific rules of origin. b. Eliminate the requirement for certificates of origin for products having preference margins of five percentage points or less. . c. Adopt a uniform change of tariff heading at the HS 6-digit level or ten percent value-added rule, with the choice between these two criteria left to the exporter concerned. The foregoing recommendation requires action by the EU. All of the subsequent recommendations concern reforms by the EPA-signatories. 2. Parallel MFN tariff reductions (paragraph 2.49): a. Eliminate any remaining quantitative restrictions, import licensing requirements, discriminatory domestic taxes on imports, and similar trade-distorting NTBs. b. Introduce phased reductions in MFN tariffs to lower the maximum MFN tariff rate to 15 percent or less; c. Lower the MFN tariffs on products that will be imported duty-free from the European Union to five percent or less prior to the elimination of the tariffs on imports from the EU. 3. Strengthening of domestic tax systems and improving customs administration to replace tariff revenue lost because of tariff reductions: Carry out detailed country-specific analyses of the trade diversion/monopolistic pricing effects and the revenue implications of the interim EPA's import liberalization schedules and prepare plans for reducing MFN tariffs and replacing foregone revenues (paragraphs 2.51 and 2.57). 4. Improvements in competitiveness and the business climate to spark a strong supply response: Prepare or, in cases where the equivalent already exists, review and revise as needed, country-specific plans for priority reforms to improve competitiveness and the business climate (paragraphs 2.63-2.64). 5. Preparation of proposals for funding financial and technical support for the above reforms (paragraphs 2.69-2.70). Full EPAs, Regional Trade Integration, and Liberalization of Trade in Services 6. Implementation of a new strategy for export-oriented trade integration based on: a. Open trade policy including elimination of all trade-distorting NTBs, reductions in MFN tariffs, and effective implementation of measures such as export processing zones to ensure that exporters have ready access to competitively-priced duty-free inputs (paragraph 3.22); and b. A flexible variable-geometry approach including (i) effective phased implementation of first preferential and the free intra-African trade in all products within and between African RTAs; (ii) the full liberalization of the restrictive rules of origin governing preferential/free intra-African trade; and (iii) inclusion of a regional preference clause in full EPAs (paragraphs 3.26-3.27 and 3.32). 7. Liberalization of services imports and investment on both a global (MFN) and a regional basis at the same time as liberalization vis-à-vis the EU is scheduled (paragraphs 3.42-3.44 and 3.51). 8. Where regulation of services is still deficient, adoption of a phased approach to liberalization focusing on building regulatory capacity and identification of priority sectors for liberalization and formulation of country-specific proposals for doing so (paragraphs 3.43-3.45 and 3.53), including: a. initially liberalizing only those sectors that require little regulation or where the existing regulatory framework is satisfactory; and b. adopting a clear timeframe for improving the (national and regional) regulatory framework and regulatory capacity for the liberalization of additional sectors. xx Map xxi ) :a ricfA oup(7 a,bii not Gr bers ent rne A )L( maN s s 6 EP a,ybLi me em m agre outhS queib EPA )L( )L( nd EPA DC acirfA DC e SA adrt EPAs SA eerf 5 amzo ing ibia ing Egypt, gola EPAs ohto uth (L), Other eas 5 DC Ar SA nedigS M gnistoN An CU edn anawstoB esL amN ilazawS SA Sig gnistoN So a golnA m. ngitisxe an aderT : : progra oup:rg A has icarfA 1) icarfA ) APE EB acrifA 2008. EU's rn (1 (5 A rn ES Regional eht hetrof outhS 6 g oup ni .ew nuaryaJfo as outheS Gr-A ) 4 outheS -A Group gni elbig eli abb is 3 ) miZ Existin nda EP )L( (L EP )L( ipatcirt C, noitautsi s PAsE to n usitir g ia )L( )L( nda s )L( (L) LD (L), n C a (L)id paton an iab hett on ertsaE ASE ESA osro )L(i )L( )L( MOC( scarag EPA )L(iw bia EA iana ecl ed moC da leselhc bweab hiop and as, au ref Ma M eyS imZ gninist outbijD eartirE an Et Mala Sud amZ ertsaE EPA ay ed C enK ndaawR Ug runuB anzT ersb maZ me but xxii m s,ell Relati ESA Sign No EA Sign ASE onstiai he in ) MOC ot Seyc neg (8 p )L( )L( hertO APE s,uitirua onsitacifissalc These Groupsytr :a ni 3 M, :E ry. gni ricfA Grou AP ionnU blicu )L( epicni (L)i unt Rep Pr Co awla NOT M Coun al As As and )L( ipatcirt pat ), DCA). C-E omst (L (T n nacir ognoC carifA o edpo EP of ral EP e vel car EPA Centr Cus oore Aflar )L( ineauGlia blic or nt ng moT ent De na CEMA CA EPAs at nosi)ASS M ed amC Afric CE Sign gningistoN entC hadC epuR bonaG quE CerehtO gniistoN Sao ongC.R.D Least a inyrtnuoc agasda M reemgA (L), oni o rat h 1 6) denotes")L(" 48t Cong Coope (1 2 he(t R. D. and : oup .puorG- aila up: ent moS gro opml icarfA Gr-A 4 )L( )L( EPA ). APE (L Deve (L) DC ste EP er ngo SA aderT W SA ATF 2 EPAs )L( )L( )L( )L( EPAs Faso ) EPAs ia Co.R eht het a a )L( eoneL )L( insietrnu al carifAt D. ni OWCE SA edreV ausisB-a EPAs ing aib ai air m arr Ivoid' EPAs ing ani )L( ing anit -- WO Ga Sie OA nin )L(li (Lreg neg Wesr Be Ma Ni Se gnis ur Ma coforeb ngit UE EC nedgiS hanG gnis epaC ineuG beriL geiN Not UEM nedgiS etoC gnis rkuB ineuG ogoT Not Othe Not muN and,li het 1 azwS paic parti htiw 1. INTRODUCTION: A DEVELOPMENT PERSPECTIVE ON EPAS 1.1 Need for Improved Trade Performance. Expanding trade is essential for sustaining and accelerating growth in sub-Saharan Africa.3 The European Union is Africa's largest trading partner providing a market for 51% of Africa's non-oil merchandise exports and supplying 34% of its merchandise imports in 2006. The trade relationship between Africa and the EU is thus important for the continent's development. However, despite the trade preferences accorded to African countries by the European Union under a series of Lomé Conventions and continued under the Cotonou Agreement in June 2000, the performance of African exports to the EU has been dismal. The African EPA-countries' exports have increased little in real terms; and their share in the EU's non-oil imports has fallen steadily from 1.7% in 1985 to 0.6% in 2006.4 This prolonged decline is largely a reflection of competitiveness problems and supply constraints in the African EPA-countries, although the limits on market access under the Lomé-Cotonou trade regime have also been a restraining factor. The performance of the EPA-countries' non-oil exports in most other major markets has been similarly weak, and these countries' exports currently account for less than a 1% share of non-oil imports in all of their major trading partners' markets (Graph 1). 1.2 Economic Partnership Agreements. EPAs are intended to reformulate the trade preferences accorded to the African, Caribbean, and Pacific (ACP) countries under the Lomé-Cotonou agreements to make them compatible with World Trade Organization (WTO) rules as well as more effective in promoting ACP-EU trade and more supportive of African regional integration and broader development goals. After five years of negotiations, in November-December 2007, the EU and 18 African countries initialed interim EPAs providing for WTO-compliant reciprocal liberalization of merchandise trade.5 In the first half of 2008, further negotiations were launched for full EPAs to address additional issues concerning regional trade integration, liberalization of trade in services, foreign direct investment, and the trade-related regulatory framework. 3 In this study, the terms Africa and sub-Saharan Africa are used interchangeably. When the context requires greater precision, the term sub-Saharan Africa is used. 4 The term "African EPA-countries" is used here to refer to the 45 sub-Saharan African countries that signed the Cotonou Agreement and are eligible to participate in EPAs. The other two sub-Saharan African countries are Somalia, which did not sign the Cotonou Agreement and is not eligible for an EPA, and South Africa, which did not sign the Cotonou Agreement but which the EU has offered to include in the EPA process if South Africa wishes to participate. Because South Africa's economy is so much larger and more advanced than those of the other African EPA-countries, data reported in this study for the African EPA- countries does not include South Africa unless specifically noted. Oil accounted for $126 billion (73%) of the EPA-countries' total exports of $174 billion in 2006. Eight of the EPA-countries are classified as oil- exporters by the World Bank, and oil represents an average of 80% of these eight countries' total exports. See the full report for a further discussion of the performance and structure of the EPA-countries trade. 5The term "EPA-signatories" is used here to refer to the 18 African countries that have initialed interim EPAs. See the map on page xxi and the table on page xxiii showing the EPA-signatories and non- signatories in Africa. 1 Graph 1: Non-oil Imports from the African EPA-Countries as a Share of their Major Trading Partners'non-Oil Merchandise Imports, 1985-2006 1.8 1.8 1.6 1.6 1.4 1.4 1.2 1.2 1 1 % % 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 198 198 198 198 198 1990 1991 1992 1993 1994 1995 1996 1997 1998 199 200 200 200 200 200 2005 2006 5 6 7 8 9 9 0 1 2 3 4 EU Japan US China Source: UN COMTRADE for SITC 2 data, accessed August in 2008. 1.3 Objective and Coverage of this Report. This report addresses the question raised in its title: now that 18 interim EPAs have been initialed and negotiations of full EPAs have been launched, what should African countries and regional EPA-groups do? Part II of the report analyzes the outcome of the EPA negotiations thus far, the interim EPAs' implications for the trade and related policies of participating African countries, and the reforms required for successful implementation of interim EPAs. Part III examines the potential role of full EPAs, in advancing regional trade integration, open trade policies, and the liberalization of trade in services and foreign direct investment in Africa. The intended audience for this report is primarily policy makers and their advisors in the African EPA-countries, but it may also be of interest to those in the broader development community concerned with Africa. 1.4 A Development Perspective. Both the European Union and the ACP countries have repeatedly emphasized their desire to use the EPAs as instruments of development. This study thus analyzes the EPAs from a development perspective. The terms "development friendly" and "pro-development" as used here in assessing EPAs refer to policies that will accelerate global and regional trade integration in Africa and sustained long-term growth of real GDP per capita. The term "trade integration" is used to refer to increases in the ratio of total trade (that is, imports plus exports of goods and non-factor services) to GDP. 2 2. INTERIM EPAS: IMPLEMENTATION AND COMPLEMENTARY REFORMS 2.1 This part of the report looks at how interim EPAs could contribute to trade integration and development in Africa. It begins with an overview of the outcome of EPA negotiations to date and the interim agreements that have been initialed by 18 African counties. It then discusses the implementation measures that will be necessary for interim EPAs to achieve favorable development outcomes through the improved access of Africa's exports to the EU market, the liberalization of the EPA-signatories' imports from the EU, and complementary tariff, tax, and competitiveness reforms. A. The Outcome of EPA Negotiations to Date WTO Compatibility: The Driver of the EPA Process 2.2 The primary driver of the interim EPAs has been the incompatibility of the Cotonou Agreement's unilateral trade preferences for the ACP countries with WTO rules. The Cotonou preferences, like those under the preceding Lomé Conventions, did not comply with the WTO's "enabling clause." This clause permits developed WTO member countries to give unilateral preferential treatment to imports from only two groups of developing countries: either all LDCs or all developing countries. The Cotonou trade preferences are inconsistent with the enabling clause because they, like the earlier Lomé preferences, were not extended to all developing countries and because some ACP countries were not LDCs.6 2.3 Hence, the European Union needed to obtain waivers from the WTO, first for the Lomé Convention in 1994 and then for the Cotonou Agreement at the Doha ministerial in 2001. The most recent waiver for the Cotonou trade preferences was scheduled to expire at the end of 2007, and a number of WTO members would have strongly opposed extending it. Expiration of the WTO waiver at the end of 2007 reinforced the Cotonou Agreement's target of completing the negotiations of the EPAs by the same date. It made the need for replacing the Cotonou trade preferences with a WTO-compliant reciprocal trade agreement the effective driver of the EPA process. 2.4 In order to bring trade relations between the European Union and the ACP countries into line with WTO rules and standard global practices, the Cotonou Agreement provided for replacing the existing relationship of unilateral trade preferences for the ACP countries in the EU market with EPAs creating reciprocal, but asymmetrical, free trade arrangements between these two groups of countries. WTO's rules require that developed and developing countries entering into reciprocal free trade agreements together liberalize "substantially all trade" within a "reasonable length of time." To meet this requirement under the planned EPA's asymmetrical trade liberalization, the 6 For the same reason, the U.S. has had to apply for a waiver for the trade preferences granted under its African Growth and Opportunity Act (AGOA). This waiver has not yet been granted because some non- African LDCs have demanded equivalent treatment. See the full report for a longer, more detailed discussion of various aspects of the WTO compatibility and waiver issues. 3 European Union proposed to provide immediate tariff-free, quota-free access to its markets for 100% of its imports from the EPA-countries, with short transition periods for only rice and sugar. The EPA-countries were expected to reciprocate by gradually providing back-loaded tariff-free access to their own markets for about 80-85% of EU exports over a twelve to fifteen year period.7 Revised Approach: A Flexible, Variable-Geometry Design 2.5 The Original EPA Design. In addition to establishing a WTO-compliant trade relationship between the EU and the ACP countries, the EPA-process has had two other central, inter-related objectives. The first objective is to accelerate the development of the ACP countries. The second objective is to promote regional integration among the ACP countries as a primary means of helping to accelerate their development. To help achieve these objectives, the EU's original EPA-design had three key features: a comprehensive reform agenda including liberalization of trade in services, foreign direct investment, competition, and trade-related regulatory issues; universal regional integration with, where possible, common regional external tariffs; and coordination of the EU's trade and aid programs. 2.6 In the autumn of 2006, after five years of discussions and negotiations, fundamental disagreements over the original EPA design persisted both between African countries and the EU and among African countries themselves. Underlying the numerous policy disputes about EPAs were four basic problems with the original design and the EU's initial approach to the negotiations. First, and most fundamentally, many African countries had little political interest in liberalizing imports or in using EPAs to accelerate implementation of trade-related reforms unless a politically attractive quid pro quo was available. Second, as discussed below, the incentives provided by EPAs for the 33 African LDCs and for its three oil-exporting non-LDCs to undertake the reforms required were inadequate unless the counties themselves had their own independent motivation for doing so. Third, even though membership in the regional EPA-groups was self- determined, the original four large and diverse regional groups turned out to have little coherence in their trade policies and interests and could not agree on common external tariffs. Fourth, the EU and many African countries had quite different concepts of the roles of trade and aid in the EPAs. While the EU viewed the EPAs primarily as trade agreements and instruments for accelerating reforms, many African countries hoped to use them to obtain additional development aid. In the end, lack of political interest in trade liberalization, inadequate incentives, the incoherence of the regional EPA-groups, and disagreements over development assistance necessitated a revised negotiating approach for achieving the WTO-compatibility objective by the December 2007 deadline. 2.7 The Revised EPA-Design. The EU's revised approach, formally adopted in October 2007, provided for greater flexibility in the geographic scope, policy coverage, and timing of EPAs. First, under a new variable-geometry design, interim EPAs would 7 As explained in the full report, this degree of liberalization of bilateral trade is needed to comply with WTO's Article XXIV governing regional trade agreements between developed and developing countries. 4 be negotiated with individual countries or groups of countries that wished to enter into these. Despite their name, these interim EPAs would in fact be permanent in that they would remain in effect indefinitely, unless and until replaced by a full EPA or another trade agreement. Second, policy coverage of interim EPAs would be flexible. Interim EPAs would include, at a minimum, reciprocal WTO-compatible market access arrangements for merchandise trade along with all other aspects of EPAs on which it was possible to conclude negotiations before the end of 2007. Negotiations would then continue on all other outstanding issues required to reach agreement on a full EPAs. And, third, the timing of entry into EPAs would also be flexible, with regional EPAs remaining open to any countries in a region that initially chose not to participate in an EPA. 2.8 Although the revised EPA-design is substantially more flexible than the original comprehensive approach, the scope for flexibility was necessarily constrained by WTO rules covering regional trade agreements. Thus, to comply with WTO rules, under the interim EPAs the European Union is providing immediate tariff-free, quota-free access to its markets for 100% of its imports from the EPA-signatories, with short transition periods for rice and sugar. The EPA-signatories are required to reciprocate by gradually providing tariff-free access to their own markets for 80-85% of EU exports in no more than fifteen years. The essential feature of an interim EPA is that it in fact satisfy the requirements for WTO-compliant reciprocal liberalization of merchandise trade under free trade agreements between developed and developing countries. Overview of Interim EPAs Initialed to Date 2.9 The European Union initialed interim EPAs liberalizing merchandise trade with 18 African countries just prior to the Cotonou Agreement's January 2008 deadline for putting these in place (see the map on page xxi and the table on page xxiii showing the EPA-signatories and non-signatories in Africa). Fifteen of the 18 country agreements are parties to regional framework-EPAs initialed by the EAC, ESA, and SADC EPA- groups. The remaining three interim EPAs are free-standing, country-specific EPAs signed by Cameroon, Côte d'Ivoire, and Ghana. These interim EPAs provide, at a minimum, for reciprocal, asymmetric WTO-compatible preferential market access arrangements for merchandise trade.8 As noted earlier, the interim EPAs are in fact permanent in that they will remain in effect indefinitely, unless and until replaced by another trade agreement. The EU also plans to continue efforts to negotiate "full" EPAs with more comprehensive regional and policy coverage. 2.10 There were sharp differences in participation rates in interim EPAs between LDCs and non-LDCs. Ten of Africa's 14 non-LDCs initialed interim EPAs, but only eight of its 33 LDCs did so. This outcome reflects the quite different market access alternatives faced by non-LDCs and LDCs post-Cotonou and the interaction of these alternatives with regional and country-specific factors. 8 The only aspect of the interim EPAs that has thus far turned out to be controversial at the WTO is their MFN clause, which requires the EPA-signatories to extend to the EU any improvements in market access that the EPA-signatories subsequently grant to other major trading countries. This issue is discussed further in the full report. 5 2.11 Non-LDCs: An Unattractive Alternative Led to a High Participation in Interim EPAs. Any of Africa's 14 non-LDCs, other than South Africa, that did not initial interim EPAs would have reverted to the EU's less favorable general system of preferences (GSP) in January 2008. Nine of these countries, all of which initialed interim EPAs, would have faced significant losses of market access if they switched to GSP status.9 Taking into account both avoidance of losses from reverting to GSP status and likely near term gains from improved market access under EPAs, these nine countries all had significant immediate market-access incentives to enter into interim EPAs and did so. Botswana, which would have faced a smaller immediate loss, also entered into an "initial" EPA. The remaining four non-participating non-LDCs are South Africa, which already had a satisfactory, separate, pre-existing free trade agreement with the EU, and three oil exporters (Nigeria, Gabon, and Congo (Rep)), which will lose virtually nothing from the switch from Cotonou to GSP preferences because their principal exports face zero or very low MFN or GSP tariffs in the EU. 2.12 LDCs: An Attractive Alternative Led to Low Participation in Interim EPAs. In contrast, LDCs had neither an immediate need nor a strong incentive to enter into interim EPAs. Even without EPAs, the 33 African LDCs would still have full tariff-free, quota- free access to the EU market under its Everything but Arms (EBA) program. Shifting to the EBA program might possibly cause some limited losses of markets for a few products in the longer term (because of the EBA program's slightly more restrictive rules of origin than the Cotonou Agreement's); but these were not a significant concern for most LDCs. With automatic eligibility for the EBA arrangement, many LDCs thus had little interest in liberalizing their imports from the EU to obtain similar market access under EPAs. As a result, the participation of LDCs in interim EPAs has been quite limited. Only eight of the 33 eligible African LDCs have signed interim EPAs. Moreover, four of these (Burundi, Rwanda, Tanzania, and Uganda) are members of EAC, which entered as a customs union into an interim EPA under the leadership of Kenya, a non-LDC with strong incentives to sign an interim EPA. 2.13 Continued Effort to Negotiate Full EPAs with All Sub-Saharan African Countries. The EU's long-term objective remains "full" EPAs that are both supportive of regional integration with the participation of all countries in each regional EPA-group and that cover all trade-related policy issues. The timing of entry into full EPAs is flexible. Countries can choose to revert temporarily to the EU's basic GSP or its EBA arrangement for an indefinite period. But regional EPAs would remain open to any country in a regional EPA-group that wishes to participate, and the EU has already initiated further negotiations of full EPAs with all of the African regional EPA-groups. 2.14 Although the greater flexibility of the revised EPA-design has permitted accommodating the diverse economic conditions, interests, and preferences in the various regional EPA-groups, a number of issues will require continued attention during the implementation of the interim EPAs. These include, on the export side, primarily the EPA's restrictive rules of origin (RoO) and, on the import side, the reciprocal elimination 9 See the full report for estimates of the magnitude of the potential losses. 6 of the EPA-countries' tariffs on 80-85% of their imports from the EU and the necessary accompanying policy reforms. B. Access of African Exports to the EU Market 2.15 Weak Performance of Africa's Non-oil Exports to the EU. The European Union is by far the largest single market for the non-oil exports of the African EPA-countries, accounting for 51% ($24 billion) of the total in 2006, a share that is six times the 8% share of the second largest market (the US) for their non-oil exports. However, despite the Lomé-Cotonou trade preferences, over the last decade the African EPA-countries' non-oil exports to the EU have grown at only 3.0% annually in nominal terms, less than one-half of the 6.3% growth rate of their non-oil exports to the rest of the world. The share of the EPA-countries' non-oil exports going to the European Union has thus fallen steadily from 67% in 1985 to 51% in 2006 (Graph 2). In addition, very little progress has been made in the diversification of Africa's exports to the EU market: sales of the EPA- countries' most significant labor-intensive manufactured export, clothing, also grew very much faster to the rest of the world (19% annually) than to the EU (2% annually). Graph 2: Destinations of the African EPA-Countries' non-Oil Merchandise Exports, 1985-2006 (as a % of their total non-oil merchandise exports) 80 80 70 70 60 60 50 50 % 40 40 % 30 30 20 20 10 10 0 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 EU Japan US China Source: UN COMTRADE for Africa SITC 2 mirror data, accessed in August 2008. 2.16 The Challenge for EPAs. A central challenge for EPAs is thus to accelerate export growth and diversification in Africa well beyond what has been taking place under the Lomé-Cotonou regime. The current combination of interim EPAs and the EU's EBA and GSP programs will maintain Africa's preferential access to the EU market and avoid significant disruption of current Africa-EU trade. However, the EPA-countries are faced with both market-access and supply-side constraints to export growth and diversification. If the EPAs are to help African countries quickly achieve better export performance than that experienced under the Lomé-Cotonou regime, both sets of constraints will need to be addressed ­ market-access constraints by the EU and supply-side and competitiveness constraints by the EPA-signatories. The rest of this section discusses the tariffs and other 7 factors affecting the access of the EPA-countries' exports to the EU market. Subsequent sections address trade policy and other supply-side constraints to increasing exports. Tariff Preferences 2.17 Countries Signing Interim EPAs. The 18 African countries that have signed interim EPAs now have full tariff-free, quota-free access for all exports to the EU (with transition periods for rice and sugar until 2010 and 2015, respectively) together with less restrictive rules of origin (RoO) for clothing and, in some cases, fish products. The preferential access to the EU market under interim EPAs is also permanent and more legally secure than under the Cotonou regime. 2.18 LDCs Not-signing Interim EPAs. The 25 African LDCs that did not sign interim EPAs still have full tariff-free, quota-free access to the EU market under its EBA program. The EBA program is also WTO-compliant and thus more legally secure than the Cotonou preferences. However, like the Cotonou preferences, the EBA preferences are subject to periodic renewal, and their RoO for clothing and fishery products are more restrictive than those under EPAs. 2.19 Non-LDCs Not-signing Interim EPAs. Of the four non-LDCs that did not sign interim EPAs, South Africa has experienced no change in its access to the EU market, which has been governed by its free trade agreement (the Trade, Development, and Cooperation Agreement or TDCA) both before and after the expiration of the Cotonou preferences.10 Congo (Rep.), Gabon, and Nigeria have reverted from the Cotonou preferences to the EU's GSP regime. But, as large oil and other primary product exporters, these three countries have not been significantly affected by this change. 2.20 The EPA process has, thus, successfully achieved the immediate objective of replacing the Cotonou tariff preferences with alternative, WTO-compliant arrangements that provide equivalent or improved market access for all African countries. The remaining market access issues concern primarily the EPAs' restrictive RoO.11 The EPAs' Rules of Origin 2.21 When tariffs and quotas are lowered or eliminated in a preferential trade arrangement, rules of origin are needed to ensure that only goods actually produced in the intended beneficiary countries are granted preferential access. Otherwise, products from third countries could be transshipped through the preference-receiving country to evade tariffs in the preference-granting country.12 It is usually straightforward to determine the 10The provisions and implications of the TDCA are discussed further in the full report. 11See the full report for a discussion of other non-tariff barriers to market access such as sanitary and phyto-sanitary standards, anti-dumping measures, and import-defense provisions. 12Such transshipment of products through a preference-receiving country is referred to as trade deflection. Avoiding trade deflection is generally in the interest of the preference-granting country so that it does not give up tariff revenues to support the exports of unintended beneficiaries in third countries. Avoiding trade deflection may, in some cases, also be in the interest of the preference-receiving country if its exports would otherwise be displaced or have their price advantage eroded by the deflected exports from third countries. 8 origin of goods that are produced entirely within a preference-receiving country. Problems in determining origin arise when goods are only partially produced (or substantially transformed) in a preference-receiving country using imported inputs. In these cases, substantial transformation is usually defined using one or more of three basic criteria: change in tariff heading, minimum domestic value-added required, and specific processing steps taking place in a preference receiving country.13 2.22 Although RoO are necessary and serve a legitimate purpose, in practice the rules of origin for trade preferences are usually determined unilaterally by the preference giving country and are often manipulated to achieve other objectives, mainly protecting domestic producers. When domestic interests are allowed to influence the scope and details of RoO, the outcome tends to be far more restrictive than necessary to prevent trade deflection. Product-specific RoO, in particular, tend to be systematically captured by producer interests in the dominant partner in a preferential trade agreement. Too often, RoO seriously limit the actual improvement in market access resulting from a preferential trade agreement; and the objective of promoting the exports of the preference receiving country is thus undermined. 2.23 The Cotonou Rules of Origin. For the past three decades, the ACP countries' preferential access to the EU market has been governed by rules of origin that are complex and protectionist for many items. These product-specific RoO under the Lomé and Cotonou Agreements ran more than 360 pages in length. They impose detailed, often complex, processing requirements that are very demanding for small low income economies like those in Africa to meet and, in fact, substantially limited the benefits of tariff-free, quota-free access to the EU market.14 2.24 RoO have effectively circumscribed the scope of the Cotonou preferences and constrained exports from both African LDCs and non-LDCs. The rules denied producers in the preference-receiving countries the freedom to choose the source of their inputs, and global sourcing of competitively priced inputs was not possible under the Cotonou RoO. As a result, export industries using imported inputs (such as those producing clothing from imported fabrics) that could have a substantial economic and development impact have been denied preferential access to the EU. In some cases, the result of these restrictions is that investment in such export industries does not take place. Brenton and Ozden (2007) further point out that restrictive rules discriminate most strongly against small developing countries where the possibilities of local sourcing are most limited. More importantly, the experience of the past 30 years has shown that restrictive RoO in preferential trade arrangements have done nothing to stimulate the development of integrated production structures in developing countries. Rather, in a globalized world, they constrain firms in developing countries from integrating into global and regional production networks. 13Many rules of origin also contain provisions permitting "accumulation" of processing in specified countries. Accumulation may potentially be important in expanding the scope of preferential trade and its impact on regional integration. It can enable a preference-receiving exporting country to count inputs produced in specific third countries in complying with origin requirements if these inputs are further processed in the exporting country claiming the preferences. 14See the full report for a further discussion of this point. 9 2.25 What evidence is there that, when rules of origin are less restrictive, export performance may improve? Because the Lomé-Cotonou RoO were so restrictive for so long, there is, as discussed earlier, plenty of evidence of little improvement in the growth or diversification of the ACP countries' export performance under the Lomé-Cotonou regime. However, there are so few cases of liberalization of RoO for preferential trade arrangements that it is hard to find empirical examples of the benefits from doing so. The most telling example, and the one that ultimately motivated the EC to act, is the successful expansion of clothing exports from the African EPA-countries under the provision of the U.S.'s Africa Growth and Opportunity Act's (AGOA) trade preferences. The AGOA rules of origin (as an exception to the U.S.'s standard RoO under free trade agreements) permit African clothing exporters to temporarily utilize fabrics imported from third countries outside Africa instead of having to source the fabrics from within Africa or the United States. A comparison of the performance of African countries in exporting clothing under AGOA with their performance under the EBA program clearly illustrates the restrictive effect of the EU's rules of origin for clothing. Exports of clothing from African LDCs to the European Union have stagnated despite the introduction of tariff-free, quota free market access under the EBA initiative in 2001, whereas clothing exports to the United States under similar AGOA tariff preferences have grown very strongly. In 2000, exports of clothing from African LDCs to the EU and the U.S. were about equal; but by 2004 clothing exports to the U.S. were almost four times the value of those to the EU (Graph 3). Graph 3: Graph 3: EU and US Imports of Clothing from African LDCs, 1995-2007 (in millions of US dollars) 900 900 800 800 700 700 600 600 sn 500 500 llioi m $,SU400 400 300 300 200 200 100 100 0 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 EU Imports US Imports Source: Adapted from Benton and Hoppe (2006) and updated with UN COMTRADE SITC 3 data, accessed in September 2008. 10 2.26 Revisions to the Cotonou Rules of Origin in EPAs. The Cotonou Agreement provided for its RoO to be reviewed as part of the EPA negotiations. Progress was initially slow on the EU-side in formalizing a new more liberal approach to RoO. The ACP group was also unable to effectively articulate a united view on the need for more liberal RoO for EPAs. The absence on either side of a clear, bold vision of more liberal rules of origin resulted in only limited ad hoc revisions of the Cotonou RoO in a few sector-specific areas. Apart from these sector-specific changes and some minor changes in cumulation provisions, the RoO under the interim EPAs remain the same as those under the Cotonou Agreement. 2.27 Under interim EPAs, the Cotonou Agreement's product-specific RoO have been relaxed somewhat for textiles and clothing, fish, and a few processed agricultural products. The changes are the most significant in the rules governing textile and clothing. The new rules switch from the Cotonou Agreement's "two stages processing" criterion to a "single-stage processing" criterion for EPAs ­ that is, the local conversion of non-originating yarn to fabric or fabric to clothing is now sufficient to confer origin. This change will enable clothing manufactures to source competitively-priced fabric worldwide as permitted under the RoO for AGOA. 2.28 The liberalization of the rule of origin for textiles and clothing is clearly a significant improvement that will hopefully have the same kind of impact on Africa's clothing exports to the EU that the similar AGOA rule has had on Africa's clothing exports to the US. The changes in the product-specific rules for fish and selected processed agricultural products will also provide some benefits in a few cases. However, except for the liberalized RoO for clothing exports, the improvements in the EPAs' rules of origin fall well short of the kind of dramatic liberalization that is needed to provide a strong market-access incentive for the reforms necessary to accelerate export growth and diversification in Africa's LDCs and other small low income economies. 2.29 Plans for Reviewing the EPAs' RoO. Interim EPAs include a provision for reviews of their RoO beginning no later than three years after signature. These reviews are to cover both technical amendments and possible substantive improvements. The EC is also in the process of carrying out a thorough review of the rules of origin under all of the EU's various preference arrangements. At present, it is unclear how much the EU may be willing to eventually liberalize its RoO for imports from EPA-countries. However, if a high (30-40%) value-added criterion, reportedly under consideration, is adopted, the potential for developing new export sectors under EPAs could be quite limited. 2.30 Steps for Further Liberalizing the EPAs' RoO. Much could be done over the medium and longer term to further liberalize the interim EPAs rules of origin. The following three reforms, in particular, would encourage the development of new export industries and the expansion and diversification of the EPA-countries' exports: a. Elimination of all the current complex product- and process-specific RoO. This reform would greatly increase transparency, simplify administration, and lower costs. 11 b. Elimination of the requirement for certificates of origin for products having preference margins of five percentage points or less. This reform would reduce the cost of compliance for obtaining small preferences. It would create only minimal scope for trade deflection, which in practice would be more costly than the value of the small tariff preferences that could be obtained through it. c. Adoption of a uniform change of tariff heading at the HS 6-digit level or ten percent value-added rule, with the choice between these two criteria left to the exporter concerned. This rule would be both far easier for exporters in EPA- signatories to comply with and far less restrictive than the current complex product- and process-specific RoO. Together, the above reforms would allow African exporters to become part of global production chains and source inputs at the most competitive prices globally. 2.31 The Cumulation Debate. Some authors argue further that a major short-comings of the RoO under interim EPAs is that the broad cumulation previously permitted among all ACP countries by the Cotonou Agreement is now limited to the smaller group of EPA-signatories.15 The objective of regional economic integration, these authors contend, will be very much undermined by the fragmented country coverage of interim EPAs and the limitation of cumulation to only EPA-signatories. In our view, however, the cumulation provisions of EPAs are a secondary concern in terms of their effects on both export development and regional integration. Cumulation among African countries under the Cotonou Agreement was quasi, if not totally, non-existent; and no one has cited even a single case of cumulation that will be disrupted by the modified cumulation provisions under the EPAs. Medium-term prospects for cumulation among African LDCs and the region's other small low-income economies are also quite limited because of the similarities in the structures of their exports. Extending the coverage of the EPA's cumulation provisions to all African countries (including South Africa), rather than limiting it just to EPA-signatories, could be helpful in the long run. However, the steps to liberalize the EPAs' rules of origin discussed above are a much higher priority for improving export performance and would also greatly reduce the need for improved cumulation provisions. Regional integration can also be much more effectively pursued through measures other than improved cumulation provisions as discussed at length in the section of this report on full EPAs. If, however, the EU were to eventually adopt a value- added criterion of 20% or more, permitting cumulation with all developing countries (including, specifically, all the major competitive exporters of inputs such as China, India, and Brazil) would be a helpful second-best measure.16 15 See, for example, Stevens and Bilal (2008) and Nauman (2008). 16 A much inferior nth-best alternative would be to try to identify revisions in specific processing requirements that constrain exports of new products that are potentially important for EPA-signatories. Doing so, however, would require an expensive time-consuming research effort that is likely to be largely speculative and inconclusive plus willingness on the part of the EU to accept ad hoc changes in its standard approach to RoO in preferential trade agreements with developing countries. See the full report for a further discussion of issues concerning rules of origin. 12 2.32 Importance of Fully Liberalizing the EPAs' RoO. Improving the EPA's current RoO will be critical to the ultimate success of the EPA process. There is substantial scope for further liberalization of the EPAs' rules of origin, and the ultimate effect of EPAs on the access of participating countries' exports to the EU market will depend greatly upon the decisions that the EU takes when revising its RoO. Liberalization of the EPAs' rules of origin would increase the effective market access of the EPA-countries and provide important benefits for African manufactured and processed agricultural exports.17 It would also be an important stimulus for foreign direct investment in export sectors where utilization of tariff-free, quota-free access to the EU market is currently blocked by the EPAs' RoO and is a necessary enhancement for the investment provisions of full EPAs as discussed in paragraph 3.55 below. Furthermore, the value of tariff-free, quota-free access of EPA-signatories to the EU market will gradually be reduced by preference erosion as tariffs on imports from other sources are cut. Additional liberalization of the RoO is the only instrument available for further improving the market access of EPA- signatories and providing a continuing strong incentive for the implementation of EPA- related reforms. In addition, the WTO's approach to special and differential treatment in the Doha round has put the RoO problem at center stage for future trade negotiations between LDCs and developed countries. EPAs offer the best opportunity that African countries will have for addressing this issue in a favorable forum and for setting a good precedent for subsequent negotiations with other developed countries, and it is in their interest to aggressively pursue such liberalization in future negotiations over the implementation of interim EPAs and the design of full EPAs. C. Liberalization of African Trade and Accompanying Reforms 2.33 The reciprocal liberalization of the imports by the ACP countries under EPAs has been the subject of much debate and resistance on by African governments based on both legitimate and imagined fears of increased import competition from more efficient European exporters. This section gives a brief overview of the EPA-countries' imports from the EU and the structure of their import tariffs. It then examines the reciprocal elimination of the EPA-countries' tariffs on 80-85% of their imports from the EU; the parallel unilateral reductions in tariffs on imports from other sources necessary to mitigate the negative effects of trade diversion and monopolistic pricing; and the measures needed to offset the loss of government revenues from reductions in tariffs on imports from the EU and other sources. The three subsequent sections then take up the complementary reforms that are needed to spark a strong supply-response to improved 17 If the experience with liberal rules of origin for clothing exports under AGOA (Graph 3) is any guide, the market access gains from a full liberalization of the EPAs' restrictive rules of origin are potentially substantial, although of uncertain magnitude and probably unquantifiable. The U.K.'s Commission for Africa made a similar recommendation -- a uniform 10% value-added rule -- to the above reforms in 2005 (Commission for Africa, 2005). Attempting to quantify the benefits from a liberalization of this type would require a time consuming, expensive research effort that is likely ultimately turn out to be inconclusive. The potential risks from liberalizing the EPAs' rules of origin are, however, virtually nil. The most common objection to simplified, non-restrictive rules of origin is that such liberal rules may lead to creation of superficial assembly production and employment with little domestic value added and few backward linkages (so called "screw-driver" industries). However, decades of experience have shown that in small low-income countries restrictive rules of origin have generally blocked potential exports without creating significant backward linkages. See the full report for a further discussion of these points. 13 access to the EU market and import liberalization, the role of development assistance in supporting the reform process, and the risks posed by interim EPAs. EPA-Countries' Imports from the EU and Tariff Structures 2.34 Imports from the EU. The European Union's dominance in African import markets has declined steadily since the 1980s, with its share of Africa's merchandise imports falling from 56% in 1985 to 47% in 1995 and 34% in 2006 as other suppliers, such as China and India, have become more competitive. However, the EU remains, by a substantial margin, Africa's largest supplier of imports. The 34% share of Africa's merchandise imports supplied by the EU is still nearly three times the 12% share of China, Africa's second largest source of imports (Graph 4). Graph 4: Sources of the African EPA-Countries' Merchandise Imports, 1985-2006 (as a % of their total merchandise imports) 70.0 70.0 60.0 60.0 50.0 50.0 40.0 40.0 % % 30.0 30.0 20.0 20.0 10.0 10.0 0.0 0.0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 EU China US South Africa Japan Source: UN COMTRADE for SITC 2 mirror data, accessed in August 2008. 2.35 The Structure of Tariffs in the EPA-Countries. Despite the tariff reforms of the past decade, current MFN tariffs are still relatively high in African countries. In the median EPA-country, the average MFN tariff is 14%, and tariff rates on 45% of tariff lines are classified as peak rates by international standards (that is, these rates are 15% or higher). For the countries with the highest tariffs in each region, the average MFN tariff is 20-30%. The maximum tariff rate is 30% or higher in one-half of the EPA-countries and still reaches 60% or more in some regions. Furthermore, African countries have a strong tendency to levy peak rates on all imports of products that are produced domestically. In addition, minimal or zero tariffs are typically imposed on the imported inputs used in producing import-competing goods; and domestic value-added is fairly low because only the final stages of the manufacturing process are usually carried out in African countries. As a result, effective protection rates of 60-90% for value added in import-competing industries are still quite common. 14 Commitments to Liberalize Import 2.36 Similarity of Approaches to Import Liberalization. The 18 interim EPAs initialed to date provide for the elimination of tariffs on 80-85% of imports from the European Union over a ten to fifteen year period as required for WTO-compatibility. All of the countries entering into interim EPAs appear to have followed fairly similar approaches in choosing the 80-85% of imports from the EU on which to eliminate tariffs under EPAs. In most cases, the first tranche of the scheduled tariff "reductions" consists largely of binding current applied zero-rate MFN tariffs at a preferential zero-rate for imports from the EU. Tariffs eliminated on imports from the EU in subsequent tranches appear to have been chosen in most cases by starting with the products bearing the lowest non-zero MFN tariff rates and working upwards until 80% of imports from the EU are covered by zero-rate preferential tariffs. The timing of the import liberalization tranches under the interim EPAs has, in all cases, been set to eliminate tariffs on at least 80% of imports from the EU in 15 years or less. However, the existing structure of tariffs and trade with the EU varied considerably from country to country. As a result, the composition of the exclusion lists (that is, the 20% of imports from the EU tariffs on which tariffs will remain) and the tariffs on non-EU imports of the products to be liberalized vary significantly from country to country despite the similarity of approaches in choosing the 80% of imports from the EU on which to eliminate tariffs. 2.37 Two special cases are worth noting. First, the SADC EPA-signatories have a shorter liberalization period than the signatories in other EPA-groups because its four SACU members needed to align their liberalization schedules with the de facto one for the SACU CET. (South Africa, the fifth SACU member, had previously established this liberalization schedule when negotiating its current bilateral free trade agreement (the TDCA) with the EU, which requires eliminating tariffs on 86% of imports from the EU by 2014). Second, the EAC EPA-group, like the other regional EPA-groups, is scheduled to eliminate tariffs on 80% of its imports from the EU in 15 years but has an additional 11 year period to eliminate tariffs on another 2% of its imports from the EU.18 2.38 Limited and Selective Market Opening. The additional opening of African economies under interim EPAs is likely to be significantly smaller than many of the press reports might lead one to believe. This outcome is illustrated in Table 1 for selected countries from the ESA and SADC EPA-groups. In all countries other than Botswana and Zimbabwe, over 90% of the imports on which tariffs will be eliminated currently carry MFN tariff rates of 10% or less. The 20% of imports from the EU bearing the highest MFN tariff rates, products with prohibitive MFN tariffs, and products not imported from the EU have been largely exempted from the import liberalization process. Because of the simple structures of production in the LDCs and Africa's other low- income economies, the retention of tariffs on 20% of imports from the EU appears in most cases to have been sufficient to exclude from the liberalization process most domestic products likely to have faced strong competition from tariff-free imports from the EU. 18See the full report for more detailed information about the import liberalization commitments of different countries and regional EPA-groups. 15 2.39 The selection of the 20% of imports on which to maintain tariffs was probably the most difficult economic decision required for designing an interim EPA, involving trade- offs among revenue, protection, and anti-export bias concerns. Because of the delays in the EPA negotiations, some of these decisions appear to have been made hurriedly in the rush to meet the December deadline. In the process, some inappropriate choices, from a public policy point of view, may have been made in the selection of products on which tariffs on imports from the EU will be maintained, with too much weight being given to protectionist interests and too little to public revenue mobilization. As the tariff rates on imports from the EU will be bound by their EPA treaties, EPA-signatories will not be able to correct these errors unilaterally. Hence, in the course of EPA implementation it would be desirable for the EU and the EPA-signatories concerned to jointly review and, if necessary, revise the lists of imports from the EU on which the EPA-countries will maintain tariffs (but with firm commitments not to extend the transitional period nor to reduce the percentage of trade on which tariffs will be eliminated). Table 1: Estimated Shares of Imports from the EU to be Liberalized Classified by MFN Tariff Rates (Data for selected ESA and SADC countries) Proportion of EU imports by MFN tariff (in %) Total EU of which % of Cumulative trade to be trade with import liberalized tariff rate 0% >0-5% >5-10% share with >10-20% >20% (%) 10% tariffs **** 10% (1) (2) (3) (4) = (5) (6) (7) * (8) = (4)/(7) (1)+(2)+(3) SACU Botswana 55.7 3.0 3.3 62.0 6.1 31.9 86.0 72 Swaziland 65.3 5.1 10.9 81.3 14.6 4.2 86.0 95 Lesotho 68.4 10.3 7.6 86.3 11.4 2.3 86.0 100 Namibia 71.3 14.3 7.5 93.1 5.0 2.0 86.0 108 Other SADC Mozambique 15.4 37.1 30.0 82.5 5.8 11.7 78.5 105 ESA Madagascar 8.3 5.1 61.3 74.7 25.3 0.0 80.7 93 Mauritius 90.3 1.1 1.4 92.8 3.4 3.8 95.6 97 Zimbabwe 15.2 26.1 14.4 55.7 20.5 23.9 80.0 70 ESA non-signatories** Malawi 36.1 24.7 11.7 72.5 13.7 13.9 **80.0 **91 Zambia *** 24.7 51.0 1.4 77.1 19.2 3.7 **80.0 **96 Notes: * Percentages are calculated from import liberalization schedules contained in initialed EPAs. **For non- signatory countries, a theoretical target share of 80% of imports was assumed in the case of future participation in an EPA liberalizing merchandise trade. ***Zambia initialed the ESA regional framework-EPA but did not enter into any import liberalization commitments. Data are for 2005 in most cases. **** Share estimated under the assumption that countries included in their import liberalization schedules those HS tariff lines with tariff rates of 10% or less. Source: Adapted from Rollo and Gillson (2008). Data sources: COMTRADE and UNCTAD TRAINS. 2.40 A Minimalist Approach. The most problematic aspect of the import liberalization schedules under interim EPAs is that, de facto, they establish a minimalist norm of very slow and highly selective tariff reductions combined with still high levels of protection on virtually all import-competing domestic products. Some of the interim EPAs, in fact, permit the re-imposition of tariffs on selected products on "infant industry" grounds, a 16 provision which may turn out to be an open invitation for protectionist interests to lobby for higher tariffs. For substantially accelerating growth and trade integration, a more decisive approach to trade liberalization is likely to be needed. The EPA precedent could, unfortunately, serve as a disincentive for adopting more proactive approaches. On the other hand, the import liberalization commitments under interim EPAs do at least provide a starting point for accelerating trade liberalization in those countries ready to do so. Potential Negative Effects 2.41 Over the next fifteen years, the 18 EPA-signatories will thus have to gradually eliminate the tariffs on at least 80% of their imports from the EU. Although many of the arguments against asymmetrical reciprocal trade liberalization under EPAs have been overstated or have little merit, there are still several reasons for legitimate concern about the preferential elimination of tariffs on 80% of the EPA-signatories' imports from the EU. This sub-section discusses three of the potential negative effects of the preferential elimination of tariffs on imports from the EU -distortions in the structure of tariffs in the EPA-signatories, trade diversion/monopolistic pricing, and a hub-and-spoke pattern of trade - and the reforms needed to minimize these. Potential revenue losses and offsetting measures are discussed in the following sub-section. An additional concern, considered in the section in Part III on regional trade integration, is whether the liberalization of imports from the EU under the interim EPAs is likely to lead to increased regional trade integration or regional fragmentation in Africa. 2.42 Distorted Tariff Structures. Partial preferential elimination of tariffs on imports from the European Union under EPAs could distort the structure of tariffs in the EPA- signatories and increase anti-export bias. Countries initialing interim EPAs will have to eliminate tariffs on 80% of their imports from the EU but will retain their current tariffs on the other 20%. Those sectors on which tariffs on imports from the EU are removed may experience trade diversion or monopolistic pricing by EU suppliers (or some of both), as discussed below. In addition, those sectors for which current high tariff rates are maintained will continue to have little incentive to increase their efficiency or to reduce monopolistic profits. In addition, removal of tariffs on inputs imported from the EU will raise effective protection rates for import-competing industries using these inputs and thus reduce the relative incentive to export, even to the EU. Hence, partial preferential elimination of tariffs could further increase existing anti-export bias by actually raising the already high effective protection rates for import-competing domestic industries. At the same time, it will eliminate revenues from tariffs on imports from the EU that do not compete with domestic production. In some cases, this increased anti- export bias may limit the supply-response to the EPAs' tariff-free, quota-free access to the EU market. 2.43 Trade Diversion and Monopolistic Pricing. As noted earlier, many of EPA- countries still have high and distorted MFN tariffs. Preferential elimination of tariffs on 80% of imports from the EU under interim EPAs could, in the presence of the high MFN tariffs in some EPA-signatories, lead to the following: (a) some costly diversion of trade from low-cost non-EU suppliers to higher-cost EU suppliers; (b) some monopolistic pricing by EU exporters and implicit transfers of forgone tariff revenues from African 17 governments to EU exporters; (c) further distortions in tariff structures and production incentives, as discussed above; and (d) some revenue losses, as discussed below. The actual results of preferential elimination of tariffs on imports from the EU will depend upon both the competitiveness of EU exporters relative to those elsewhere in the world and upon the extent of competition among EU exporters.19 The EU's declining aggregate import market share and declining competitiveness in some sectors relative to Asian suppliers increase the risk of trade diversion to less efficient EU suppliers when EPA- signatories eliminate the tariffs on imports from the EU alone. The exclusion of most higher tariff items from liberalization and the large share of liberalized imports with MFN tariff rates of 10% or less will, however, tend to moderate these effects. 2.44 The only publicly available estimates of trade diversion under interim EPAs that have been made since the liberalization commitments of the EPA-signatories have been determined are by de Melo and Carrère (2008).20 They estimate that, in the median EPA- signatory, 1.4% of imports will be diverted to the EU from the rest of the world by the preferential tariff reductions under EPAs. Estimates of trade diversion from interim EPAs range from minimal levels in SACU (which is already in the process of eliminating tariffs on 86% of its imports from the EU as a result of the EU-South Africa free trade agreement) to 2% - 4% of imports in seven countries (Burundi, Cameroon, Côte d'Ivoire, Ghana, Madagascar, Rwanda, and Zimbabwe). Because so many of the imports from the EU on which tariffs will be eliminated currently have MFN tariffs of 0% - 5%, estimated average trade diversion effects are not large in macro terms. However, trade diversion of products with tariffs exceeding 5% is likely to be significantly higher than the average. These are the products on which offsetting reductions in tariffs on imports from the rest of the world will need to focus. 2.45 Little or no empirical work has been done on the risk of monopolistic pricing by EU suppliers in Africa, and there are some differences of view about the importance of this concern. Advocates of EPAs have tended to assume that the EU economy is sufficiently large and competitive that inadequate competition and monopolistic pricing are not likely to be serious problems. Some other observers familiar with supply conditions in small African markets, which are costly to serve and are dominated by a few suppliers in some sectors, tend to be more skeptical about the existence of sufficient effective competition among EU exporters in all sectors. Inadequate competition could cause both an immediate static loss from the higher costs imposed on the affected sectors and a larger long-term dynamic loss through slower technological and structural change. 2.46 A Hub-and-Spoke Pattern of Trade. In the case of free (preferential) trade agreements, like the interim EPAs, when a hub country or region (that is, the EU) enters into FTAs with various small countries like the EPA-signatories (the spokes) and the latter do not similarly liberalize trade among themselves, the hub country tends to benefit more because it has free access to all markets whereas the spokes only have free access to 19See the full report for a discussion of this point. 20 A number of illustrative estimates of trade diversion were made prior to the initialing of interim EPAs. However, these overstated the likely magnitude of trade diversion as they assumed elimination of tariffs on 100% of imports from the EU, rather than just on the 80% of imports with the lowest tariff rates, and had various other methodological problems contributing to overestimates. 18 the hub market. This hub-and-spoke effect increases the incentive for hub-country exporters to invest in the hub country, rather than in the spokes, in order to serve all of the markets.21 The potential hub-and-spoke effect from interim EPAs can be mitigated by liberalization of trade among the EPA-signatories. Parallel Reduction in Tariffs on All Imports 2.47 Granting preferential free trade status to most imports from the European Union under EPAs without lowering the current high MFN tariffs in Africa could induce trade diversion, allow European suppliers to profit excessively from selling at high prices in Africa's protected markets, and increase anti-export bias. Because of the distortions that can result from preferential import liberalization under free trade agreements, proactive developing countries (such as Chile and Mexico) lowered their MFN tariffs to a uniform rate of around 10% on a comprehensive basis before eliminating tariffs on a preferential basis under their free trade agreements with the EU and the US. These proactive integrators followed up on their initial MFN cuts both with further reductions in their MFN tariffs to the 5-6% range and with negotiation of additional free trade agreements with other large trading partners.22 2.48 To avoid the negative effects that could result from partial preferential liberalization of imports from the European Union, EPA-signatories will need to unilaterally (or regionally in the case of customs unions) reduce tariffs on imports from all countries in all sectors, including, in particular, the main domestic import-competing sectors, to increase competition and lower the prices of tradable goods in Africa.23 Such parallel MFN and preferential liberalizations are beneficial because of (a) the reduction in trade diversion and monopoly pricing effects that preferential liberalizations alone can cause; (b) the standard static benefits from increases in trade resulting from tariff reductions; and (c) the much larger longer-term dynamic gains from a more open economy, greater competition, and the accelerated inflow of technology, know-how, and ideas through improved access to competitively priced imports. MFN trade liberalization would also facilitate expansion of intra-African trade generally since other African trading partners would benefit from it. MFN trade liberalization would also reduce the "hub-and-spoke" problems that could otherwise be caused by the EU's entering into separate free trade arrangements with the five regional EPA-groups.24 2.49 Minimum Tariff Reforms Needed. At a minimum, three reforms in MFN tariffs will be needed to reduce the inherent trade-diverting, distortionary effects of tariff preferences for the EU and avoid unduly favoring EU suppliers. First, any remaining 21 Any hub-and-spoke effects would moderately reinforce the lack of trade complementarity among African economies discussed in paragraphs 3.12 and 3.13 in the section below on regional trade integration. 22The EPA-signatories could also try to pursue trade liberalization by negotiating a series of bilateral free trade agreements with their major trading partners similar to the interim EPAs. However, this approach is probably not feasible at present as developed and large developing countries have shown little interest in it. In addition, most of the EPA-signatories, particularly the LDCs, do not have the negotiation and administrative capacity to implement a series of bilateral trade agreements. 23 MFN liberalization is also very important for imports of services as discussed in Part III below. 24 See Part III below for a further discussion of the gains from a more open, export-oriented trade policy. 19 quantitative restrictions, import licensing requirements, discriminatory domestic taxes levied at higher rates on imports than on similar domestically produced goods, and other trade-distorting non-tariff barriers (NTBs) should be eliminated.25 Second, phased reductions in MFN tariffs, initially focused on cutting down tariff peaks to lower the maximum MFN tariff rate to 15 percent or less are needed. These are necessary to reduce the most excessive rates of protection, to avoid increasing effective protection for import-competing industries as tariffs are eliminated on inputs imported from the EU, and to increase competitiveness generally. Third, the MFN tariffs on products that will be imported duty-free from the EU will need to be lowered to five percent or less prior to the elimination of the tariffs on imports from the EU in order to avoid creating an excessive price advantage for EU suppliers and limit trade diversion. This third reform may not be too difficult to implement since a large percentage of the imports from the EU on which tariffs are being eliminated currently bear MFN tariff rates of 10% or less. 2.50 The foregoing three reforms constitute a basic tariff reform program that all signatories of interim EPAs will need to implement. These steps will reduce the inherent trade-diverting, distortionary effects of tariff preferences for the EU, avoid favoring EU suppliers relative to regional ones, and facilitate the further liberalization of both global and intra-African trade. Depending upon their initial tariff structures and their choices of imports from the EU to liberalize, some countries could find that this basic package would be inadequate to eliminate all of the serious distortions in their tariff structure and, hence, would need to implement additional tariff reductions. In addition, more proactive reformers may want to undertake a faster and more fundamental liberalization and restructuring of their tariff systems. 2.51 Timing of Reforms. EPA-signatories will need to complete the above unilateral reductions in tariffs on imports from all countries before (or, at the latest, at the same time as) implementing the preferential reductions in tariffs on imports from the European Union under EPAs. This pre-emptive timing is necessary so that trade diversion is minimized and EU exporters do not have a chance to sell into highly protected domestic markets in Africa, and thereby obtain large implicit transfers of forgone tariff revenues, before MFN tariffs are reduced.26 Because of the heavily back-loaded schedules for eliminating tariffs on imports from the EU, in most EPA-signatories it will be a number of years before any non-zero tariff rates are actually cut. However, now that the negotiated liberalization schedules are available, it is an opportune time to carry out detailed country-specific analyses of trade diversion and monopolistic pricing effects and formulate plans for reducing MFN tariffs and replacing forgone tariff revenues as discussed below. 2.52 Implications for Interim EPAs. Because of their potential importance and the uncertainty surrounding their likely magnitude, trade diversion and monopolistic pricing 25 In cases where transitional arrangements are needed for the elimination of some NTBs, these should be replaced with tariffs (that is, "tariffized" in WTO terminology). These tariffs should then be phased down along with the country's other MFN tariffs. 26 Similarly, the MFN tariff reductions would need to precede (or take place at the same time as) the liberalization of intra-regional trade so that high-cost regional exporters are not given excessive protection in the markets of other members of the same regional trade area. 20 effects merit careful monitoring during the implementation of interim EPAs so that appropriate policy adjustments can be made as required. Avoiding monopolistic pricing is an area in which the EU could potentially be helpful by applying its own competition laws to the behavior of EU firms in African markets and by providing technical and financial assistance for strengthening African competition laws and institutions. Potential Revenue Losses and Stronger Domestic Tax Systems 2.53 Potential losses of tariff revenue under interim EPAs are an important concern for some countries. Import tariffs remain a substantial source of government revenue in many African countries. Since the European Union is the largest source of imports for most African coun-tries, some countries fear losing significant tariff revenues from eliminating all tariffs on 80-85% of imports from the EU. Various initial estimates of potential revenue losses suggested that the eventual losses might be significant for some countries (5-10% of total revenues) and substantial in a few extreme cases (10-15% of revenues). However, most of the analyses of revenue losses made prior to the initialing of interim had serious flaws and, as a consequence, significantly overstated the likely revenue losses to be expected from implementing interim EPA.27 Table 2: Estimated Revenue Losses from EPAs for Four COMESA Countries (assuming elimination of tariffs on all imports from EU) Country Estimated Revenue Loss in % Revenue structure in % as a share as a share as a as a share Tariff Trade Tariff of tariff of revenue share of of total revenue/ tax revenue/ revenue from trade total tax Gov Trade revenue/ Total tax taxes revenue revenue tax Total tax revenue incl grants revenue revenue EPA signatories Madagascar -29.9 -11.2 -4.1 -2.6 34.1 37.2 12.7 EPA non- signatories Ethiopia -17.7 -8.0 -3.4 -2.1 44.3 43.3 19.2 Malawi -6.5 -2.4 -0.8 -0.4 33.9 35.4 12.0 Zambia -17.7 -5.5 -1.7 -0.6 28.7 31.7 9.1 Notes: Percentage changes and ratios are calculated based on collected tariffs. Revenue data are for 2006, except for Ethiopia (2005). Trade taxes are defined as the sum of customs tariffs + VAT + excise duties + other miscellaneous duties on imports. Assumed cross price elasticity of substitution between imports from EU and from the rest of the world is -2. Source: Brenton, Hoppe, and von Uexkull (2007) 2.54 The most accurate of the publicly available estimates of likely revenue losses from eliminating tariffs on imports from the EU ­ 1% to 4% of total tax revenues -- are less than half of the initial estimates cited above. These more recent estimates were made by Brenton, Hoppe, and von Uexkull (2007) for four African LDCs. Detailed data were available at the HS-6 tariff-line level for imports coming from the EU and the rest of the 27See the full report for a discussion of the various methodological issues encountered in making accurate estimates of likely revenue losses. 21 world, applied (statutory) tariffs, exemptions, actual collected tariffs, and VAT and excise taxes levied on imports (Table 2). Their work highlights the importance of taking into account both (a) tariff exemptions, which often cause estimated losses in tariff revenues computed by applying statutory tariffs to import values to be substantially overstated, and (b) revenues from VAT and excise taxes on imports, which accounted for 56%­ 71% of revenues from trade taxes in the four-country sample and typically increase as imports rise when tariffs are cut. Brenton's and associates' estimates were completed in November 2007 before the initialing of the interim EPA agreements and assume elimination of tariffs on 100% of imports from the EU rather than just on the 80% of imports with the generally lowest tariffs. Hence, even their estimates overstate by at least 20% the likely revenue loss expected at the end of the transition period in 2023. Since the liberalization schedules and exclusion lists are now known, Brenton's and associates' estimates can be updated and their methodology applied to other countries in order to make more accurate estimates of likely revenue losses both for EPA-signatories and other countries considering entering into EPAs that would liberalize merchandise trade with the EU. 2.55 Because of low collection efficiency and narrow existing domestic tax bases, low income developing countries typically are more dependent on trade taxes than other countries and often have serious difficulty in replacing revenues lost due to tariff reductions. Baunsgaard and Keen (2005), using panel data for 111 countries over 25 years, find that high and middle income countries have been quite successful in recovering from other tax sources the revenues that they have lost because of trade liberalizations. In contrast, they find that "revenue recovery has been extremely weak in low-income countries (which are those most dependent on trade tax revenues): they have recovered, at best, no more than about 30 cents on each lost dollar. Nor is there much evidence that the presence of a value-added tax has in itself made it easier to cope with the revenue effects of trade liberalization." 2.56 However, all countries desiring to integrate into the global trading system must sooner or later replace tariff revenues with increased domestic taxation. Irrespective of the precise magnitude of the revenue losses, domestic tax systems will need to be strengthened and tax and customs administration improved to replace the tariff revenues that will be lost because of tariff reductions under interim EPAs and the necessary parallel reductions in MFN tariffs. Because the elimination of tariffs will be limited to 80% of imports from the EU and, in most cases, will take place 10 to 15 years from now, EPA-signatories will have a long period in which to strengthen their revenue systems in anticipation of these losses, which in most cases should be manageable. 2.57 Timing of Reforms. Since improving tax administration requires capacity and institution building that take time, the EPA-signatories need to make an early start on the reforms necessary for replacing forgone tariff revenues. Now that the actual schedules for removal of tariffs on imports from the EU are known, detailed country-specific analysis of their revenue implications can be made and accompanying reforms in MFN tariffs and domestic tax systems designed. 22 D. Complementary Reforms to Generate A Strong Supply Response 2.58 Improved market access alone is unlikely to be enough to substantially accelerate the growth and diversification of the EPA-signatories' exports to the EU. A wide range of supply-side constraints in African LDCs has inhibited a strong supply response to their tariff-free, quota-free access to the EU market under its EBA program. Since the implementation of the EBA program in March 2001, the nominal value of the EU's non- oil imports from African LDCs has increased only from E7.3 billion in 2001 to E7.8 billion in 2007. The share of the EU's non-oil imports coming from African LDCs has thus continued its steady long-term decline (Graph 5). Most of Africa's non-LDCs face many of the same types of supply constraints and competitiveness problems as its LDCs. Thus, unless these constraints are effectively addressed, the immediate impact of tariff- free, quota-free market access under interim EPAs on most of the non-LDCs' exports to the EU is not likely to be much greater than the EBA program's impact on the LDCs' exports. Substantial improvements in the business and investment climate will also be needed, as discussed below. Graph 5: EU Non-oil Imports from African LDCs, 1995-2007 (as a % of total EU Imports) 1.4 1.4 1.2 1.2 1 1 0.8 0.8 % % 0.6 0.6 0.4 0.4 0.2 0.2 0 0 5 6 7 8 9 0 1 2 3 4 5 6 7 199 199 199 199 199 200 200 200 200 200 200 200 200 Total Non-Oil Manufactured Goods Agricultural Materials Other Primary Products Source: UN COMTRADE SITC 2 data., accessed in September 2008. Supply-side Constraints and Competitiveness Problems 2.59 Low Level and Efficiency of Investment. Both the level and the efficiency of investment in Africa will need to increase to achieve a strong supply response to improvements in market access. Investment levels in Africa have historically been lower than in other developing regions. In 2004-2006, Africa had the lowest weighted average investment rate of any major developing region, although the average annual growth rate of investment in Africa has been close to the median for all regions since 1990 (Table 3). 23 At about 20% of GDP, average gross capital formation in Africa is too low to sustain the 7% or higher growth rates of GDP needed to achieve the millennium development goals.28 To accelerate the growth rate of real GDP in Africa to 7% or higher, the ratio of gross capital formation to GDP needs to rise to a range of 25%-30% of GDP, with the required level depending upon how successful African countries are in accelerating their slow growth in factor productivity. Table 3: Comparative Investment Rates and Total Factor Productivity in Major Developing Regions Gross Capital Formation Total Factor Productivity 2005* 1990-2005 2005 1990-2005 growth % of GDP growth rates relative to U.S. (U.S.=100) rates East Asia & Pacific 38.3 9.4 8.4 5.1 South Asia 30.7 7.6 5.8 2.3 Middle East & North Africa 26.5 4.1 13.3 0.5 Eastern Europe & Central Asia 23.0 -1.7 21.7 2.2 Latin America & Caribbean 20.3 3.8 19.3 0.2 Sub-Saharan Africa 19.7 4.1 5.6 0.2 * 2004-2006 averages, where available Sources: GCF data from World Development Indicators, accessed on 5/9/2008 from Development Data Platform. TFP data from Poncet (2006), published in Global Economic Prospects 2008. Table 4: Comparison of Average Rankings of Major Developing Regions on Business Climate and Competitiveness Indicators, 2007-2008 Doing Business Global Ranking Competitiveness Ranking East Asia & Pacific 77 50 Eastern Europe & Central Asia 76 73 Latin America & Caribbean 87 83 Middle East & North Africa 96 53 South Asia 107 90 Sub-Saharan Africa 136 106 Total number of countries in sample 178 131 Average Ranking 89 66 Number of African countries in sample 46 23 Sources: Doing Business 2008 and the Global Competitiveness Report 2007-2008 Note: Because the Doing Business 2008 data cover 178 countries whereas the Global Competitiveness Index covers only 131 countries, the average country ranking under the Doing Business Indicators (89) is thus 23 points lower than the average country ranking under the Global Competitiveness index (66). A second difference is that Doing Business 2008 reports data for 46 sub-Saharan African countries, whereas the Global Competitiveness Report 2007-2008 includes only 23 sub-Saharan African countries. 28World Bank staff estimate. See the full report for a further discussion of target investment rates. 24 2.60 Low FDI except in Extractive Industries. Although foreign direct investment (FDI) accounts for 13% of total investment in Africa, most FDI currently goes into petroleum and other extractive industries. Production of competitive exports depends upon access to foreign technology, much of which often comes through foreign direct investment. Typically, the farther behind an economy is technologically, the more it requires access to foreign technology. Greater FDI flows, of which more than 80% came from the EU in 2004-2006, will need to be attracted into sectors other than petroleum and mineral exports, particularly labor-intensive and agro-processing exports, infrastructure, and services. 2.61 A Problematic Business Climate and Weak Competitiveness. In most African countries, the prospects for increasing investment and long-term growth, and taking advantage of tariff-free, quota-free market access under EPAs, are constrained by serious weaknesses in the business and investment climate. Africa has the lowest rankings of any major developing region for both its competitiveness by the Global Competitiveness Report 2007-2008 and for its business climate by Doing Business 2008 (Table 4 above). Deficiencies in infrastructure, governance, institutional capacity, the macro-economic environment, and the regulatory framework increase the cost of doing business and lower productivity and competitiveness in Africa.29 Types of Reforms Required to Spark a Strong Supply-Response 2.62 In recent years, a number of EPA-countries have undertaken programs to create an environment more favorable to investment and growth. But, as the above data on competitiveness and the business climate show, in order to take full advantage of improved market access under the EPAs, many African countries will need to step up their reform efforts. 2.63 In addition to implementing MFN tariff reductions and strengthening domestic tax systems to avoid negative outcomes from the preferential elimination of tariffs on imports from only the EU, EPA-signatories could use EPAs to help leverage a series of complementary trade, fiscal, business climate, institutional, and regional integration reforms over the next decade to prompt a strong supply response and benefit fully from increased trade integration. Examples of the general types of complementary reforms needed are: a. Improvements in the business climate through country-specific reforms to create a supportive legal, regulatory, financial-sector, and governance environment. b. Investments in infrastructure to support a strong supply response from the private sector, including, where appropriate, carefully designed and managed export processing zones. c. Liberalization of trade in services and investment in key trade-related services and industries on a phased MFN basis, as appropriate legal and regulatory environments are established. 29See the full report for a longer discussion of competitiveness issues. 25 d. Trade facilitation measures such as improving customs administration and transportation facilities and services, including ports. In addition to the above broad agenda of trade and investment-climate related reforms, EPAs will need to be accompanied by good macro-economic management and a competitive exchange rate policy that encourages the expansion and diversification of exports. Other country-specific unilateral reforms may also be needed in other key areas (such as labor training or technological support) in order to achieve a substantial supply response. 2.64 Reform Priorities. Most of the reforms needed to improve the business climate and boost the supply response require unilateral domestic actions. In some cases, some of these unilateral reforms may be higher immediate priorities than reforms directly needed to support EPA implementation. Hence, the determination of priorities will have to be based on individual country circumstances rather than on generic approaches; and each EPA-signatory will need to work out its own country-specific reform program. However, in selected cases, there may also be scope for some collective regional actions, particularly in Eastern and Southern Africa, where many countries have initialed interim EPAs. E. Supportive Development Assistance 2.65 Adequate supporting development assistance is necessary for reforms under EPAs to have strong development benefits. Development assistance will be needed to help with adjustment costs for firms and workers, temporarily make up for revenue losses from tariff reductions, support the implementation of trade and business climate reforms, and improve infrastructure to take advantage of increased export opportunities. 2.66 One of the objectives of the EPA-process is to encourage additional trade-related development assistance and to ensure effective coordination of trade and aid. The European Union is Africa's largest single aid donor and is active in a wide range of sectors. In principle, the EU's financial and technical cooperation programs could provide substantial assistance for overcoming problems and taking advantage of the opportunities created by the improved market access and trade liberalization under EPAs. In addition, the European Commission (EC) has an extensive network of contacts with other large aid donors, particularly its member governments. Hence, the EU has considerable scope for effectively coordinating aid with the trade and investment climate reform programs under EPAs and for mobilizing the financial and technical support required for their success. The opportunity to coordinate trade and aid with a major trading partner and aid donor is a distinct advantage of the EPA-process relative both to the multilateral WTO trade negotiations and to most other bilateral trade negotiations, where the links between aid and trade have been tenuous or non-existent. The EU's and its Member States' Aid Programs 2.67 The EU has consistently treated the question of trade-related development assistance under EPAs as an exercise in reprogramming already committed aid flows and 26 in coordinating support with other donors.30 The EU increased the resources of the European Development Fund (EDF) by E6 billion from E16 billion for the 9th EDF (2003-2007) to E22 billion ($28+ billion) for the 10th EDF (2008-12), an increase in nominal terms of 35% but significantly smaller in real per capita terms. The EU has clearly indicated that the 10th EDF's resources will not be further increased and that additional aid to support EPAs would have to come from the EU's member states or other sources. However, as part of the G-8's effort to mobilize increased funding for development in Africa, EU member states have pledged substantial additional resources for Africa that would raise their aid from the current level of E20 billion per year to E30 billion per year by 2010 and to E40 billion per year by 2015. EU member states are to be encouraged to direct a substantial part of this increase to countries implementing EPAs.31 The EC is also promoting the establishment of regionally managed funding mechanisms as financial conduits for aid flows supporting the EPA process, and regional funds are mentioned in some EPAs.32 2.68 Support for Interim EPAs. The interim EPAs include general EU commitment to provide adequate and timely development assistance to EPA-signatories but are vague as to specifics. The additional amounts of assistance that have been identified thus far are also quite small in both absolute and relative terms. Hence, eventual aid availability will depend largely on the aid decisions that the EU and its member countries make during the implementation of the EPAs. Furthermore, improved coordination of the EU's aid and trade programs will not be easy. The general directorates for trade and aid at the EC, which have different constituencies and different perspectives, have operated largely independently in the EPA process thus far. EPAs are more likely to achieve pro- development outcomes, the more that the EU can do (a) to strengthen the coordination and administration of its aid and trade programs and (b) to create positive incentives for EPA-countries to implement politically challenging trade and related reforms. The availability of adequate development assistance will be a particularly important incentive to sign and implement EPAs for LDCs, which will gain relatively little in new market access from shifting from the EBA program to EPAs. 2.69 In order to facilitate the timely implementation of preferential and MFN tariff reductions (and also the liberalization of intra-African trade), proactive financial and technical support will be needed for the strengthening of domestic tax system and offsetting any temporary large revenue losses. One particularly useful confidence- 30The amount, additionality, and modalities of the development assistance that would accompany the EPAs has been perhaps the most divisive issue in the long running negotiations. This controversy is discussed further in the full report. 31 In addition, the EC currently provides about E1 billion in trade-related assistance to developing countries. In the fall of 2007, the EC and EU member states indicated that their total trade-related aid for all developing countries would be increased from the current level of E1 billion per year to E2 billion per year by 2010, with nearly all of the increase coming from the EU's member states. A substantial part-- 40% to 50% -- of this additional E1 billion a year in trade-related aid pledged for 2010 is to go to ACP countries entering into EPAs as additions to their aid allocations from the European Development Fund. According to EC estimates, EU member states currently allocate just under 40% of their trade-related assistance to the ACP countries and intend to raise this figure by E100 million to 50%. 32 The ESA EPA-group has also recently prepared a specific action plan for the use of regional funding. 27 building measure could be to provide contingent back-up grant-financing for tariff losses and related revenue shortfalls while tax reform progresses, tax administration is improved, and foregone tariff revenues are gradually replaced. Such contingent grant- financing could be particularly helpful for LDCs, which as discussed earlier, often have serious difficulties in replacing revenues foregone through tariff reforms. However, such a temporary financing facility would need to be linked to observed progress in tax reform and designed and implemented in a way that it does not become permanent. 2.70 Because many EPA-signatories have yet to formulate reform programs, they have also been slow to identify specific requirements for development assistance to support these. They need to do so now that their EPAs are in place. Potentially, countries aggressively implementing interim EPAs and related reforms could receive significant amounts of additional development assistance and aid for trade from donors that make performance based aid allocations. Those countries that take a proactive approach to implementing EPAs and complementary reforms will also be the best positioned to benefit from any general increases in aid availability. However, the actual allocation of available aid resources between trade-related assistance and other development priorities is primarily a decision that will be made by recipient countries, which will need to allocate adequate funding for financing tax reforms and infrastructure development. Moreover, if EPA-signatories really wish to accelerate growth and trade integration, they will need to proactively implement the necessary reforms and cannot afford to wait for donors to finance each of them. F. Risks and Mitigating Measures 2.71 Possible Negative Effects. The one significant downside risk of interim EPAs is that they could create some new economic difficulties if, because of lack of political support or limited administrative capacity, the signatories are not able to satisfactorily implement the required parallel MFN tariff reductions and the restructuring of their revenue systems. In this case, the signatories could experience some trade diversion, monopolistic pricing, further distorted tariff structures, and revenue losses. To mitigate this risk, EPA-signatories need to assess now on a country-by-country basis the possible magnitude of negative effects and formulate plans for dealing with these, as recommended in paragraphs 2.51 and 2.57 above. The remedy for these negative effects is to implement the necessary accompanying MFN tariff reductions and tax reforms.33 2.72 Non-Implementation of Complementary Policy Reforms. A second risk ­ this one to the magnitude of the upside gains -- is that signatory countries may not realize the potential development benefits of EPAs because of non-implementation of 33Alternatively, LDCs have the option of abandoning their interim EPAs and utilizing the EBA program for tariff-free, quota free access to the EU-market. Conceivably, a few non-LDCs might also decide for political reasons that they prefer GSP access to the EU market to liberalizing imports as required by their interim EPAs and thus withdraw from these. Another potential complication is that both LDCs and non- LDCs that are members of a customs union (such as EAC) participating in interim EPAs might need to renegotiate some provisions of their membership in the customs union if they withdraw from its EPA. 28 complementary policy reforms.34 Even if the tariff and reforms required to avoid negative effects are carried out, the supply response to trade liberalization and improved access to the EU market could be limited in countries that do not unilaterally implement investment-climate reforms and other supply-side measures or that receive inadequate development assistance. Since some countries and regions have shown little interest in using EPAs to leverage action on domestic and regional reforms, non- or partial implementation of the complementary reforms necessary to prompt a strong-supply response is a possibility. In addition, because of the ready availability of the soft option of simply switching to the EPA program for tariff-free, quota-free access to the EU market, the market-access incentives provided by EPAs for LDCs to undertake EPA- related reforms may be minimal. The risk of non-implementation will be heightened if (a) the EPAs' rules of origin are not improved to make them more substantially more favorable than those of the EBA program, (b) financial incentives for continued EPA implementation are weak, and (c) effective enforcement and dispute settlement mechanisms are not established. 2.73 Dispute Settlement and Enforcement. Strengthening of two aspects of EPAs would help to reduce the risks of non-implementation of policy reforms. First, if EPAs are to be credible instruments for leveraging and locking in reforms, the inevitable trade and investment disputes will need to be equitably settled and decisions enforced. Explicit dispute settlement and enforcement provisions are, however, not included in most of the interim EPAs initialed to date. The Caribbean EPA, in contrast, sets out the complete, generally accepted spectrum of recourse for dispute settlement -- mandatory consultation, mediation, and arbitration -- and addresses compensation, remedies, and settlement procedures. Each step in the process is subject to time limits, and arbitration follows international best practice. The addition of similar approaches to dispute settlement and enforcement to African EPAs could strengthen the credibility of their EPA-related reforms. In addition, the policy incentives for the EPA-signatories to adhere to interim EPAs as currently formulated are weak, and ways to strengthen these incentives need to be found as discussed above. 2.74 Monitoring of Implementation. Second, systematic monitoring and evaluation of EPA implementation would facilitate policy adjustments if problems arise. During the implementation of EPAs, the following concerns will need to be addressed and appropriate policy adjustments made when required: (a) further liberalization of the EPAs' rules of origin, (b) possible negative effects of the EPA-signatories' elimination of tariffs on imports from the EU, (c) the implementation of necessary accompanying reforms, (d) the delivery of EPA-related development assistance, and (e) other unforeseen issues that may arise later. However, the African interim EPAs appear to contain no provisions concerning their monitoring, evaluation, and revision. At this point, it is unclear exactly what arrangements will be made for doing so; but the CARIFORUM EPA again provides some ideas about how such arrangements might work.35 34Another small risk is possible implementation difficulties on the EU-side, which are discussed in the full report. 35See the full report for a further discussion of possible arrangements for monitoring and evaluation. 29 G. Conclusion: An Opportunity to Accelerate Trade-Related Reforms 2.75 Interim EPAs offer two principal types of potential benefits: some -- modest in the case of LDCs -- improvement in access to the EU market and an opportunity to accelerate trade- related reforms that are needed in the long term.36 2.76 Improved Access to the EU Market. For LDCs (which already have tariff-free, quota-free access to the EU market under the EBA program), the indefinite duration and greater legal security of preferential access to the EU market obtained under an EPA- treaty compared to those under the EU's unilateral preference scheme are worth having but not decisive gains. In contrast, for non-LDCs, tariff-free, quota-free market access under an EPA is a significant improvement over the more limited GSP tariff preferences. For both groups of countries, more liberal rules of origin for exports of textiles and clothing are also a potentially significant improvement in market access under EPAs. Any further gains in market access from EPAs would come largely from future liberalization of the EPAs' current RoO. If the EPAs' current RoO can be replaced with a uniform change of tariff heading or 10% value added rule as recommended above, substantial new market access could be obtained. 2.77 Acceleration of Reforms. More importantly, for those African countries and sub- regions that chose to enter into interim EPAs, the EPA-process could provide a unique opportunity to leverage action on needed trade and related supply-side issues and thus accelerate trade integration and growth. EPAs could provide a framework for interested countries to benefit earlier from reforms that they will otherwise eventually need to implement unilaterally such as liberalization of international trade in goods and services, trade facilitation measures, improvements in investment and competition policies, and liberalization of intra-African trade. The requirement for EPA-signatories to eliminate tariffs on imports from the EU provides an opportunity for them to accelerate MFN trade liberalization and the strengthening of domestic revenue systems under potentially favorable conditions, to lock in these reforms in a globally credible way, and to increase regional trade integration. The improvements in access to the EU market under EPAs and the associated technical and financial assistance from the European Union and its member countries could create uniquely favorable external conditions for undertaking trade and investment-climate reforms. 2.78 Participation of Additional Countries in EPAs. Should additional African LDCs and oil-exporting non-LDCs consider entering into EPAs that provide for WTO- compliant asymmetric, reciprocal liberalization of merchandise trade? Conditionally, yes. However, the primary motivation for doing so would have to be the desire to accelerate reforms needed for faster global and regional trade integration. Improvements in access to the EU market could also be a potentially significant benefit for (a) countries that can substantially increase clothing exports to the EU, (b) LDCs that grow fast 36Increased development assistance is a third potential benefit, although its likely magnitude is quite uncertain. 30 enough in the medium term to graduate from LDC status and hence from eligibility for the EU's EBA program, and (c) oil-exporters that need to diversify their exports. On the other hand, for LDCs and oil-exporters that are not interested in or are not able to accelerate reforms, the EU's EBA and GSP tariff preferences, from which these countries currently benefit, provide satisfactory access to the EU market. 31 3. FULL EPAS: REGIONAL INTEGRATION AND LIBERALIZING GOODS & SERVICES TRADE 3.1 This point marks a change of focus. The report's attention shifts here from the implementation of the interim EPAs that have been initialed to the design of possible full EPAs. The following sections first consider the potential role of EPAs and open trade policy in advancing regional trade integration in Africa and then examine how full EPAs might facilitate the liberalization of trade in services and investment. A. Regional Trade Integration and Open Trade Policies 3.2 In addition to restructuring the trade relationship between the EU and the ACP countries, EPAs are also intended to promote regional integration, which is an important political as well as economic objective for many African countries. Despite much rhetoric about the importance of regional integration in the EPA context, substantive discussions of it made little progress until October-November 2007 because of (a) disputes over the geographic configuration of EPA-groups in eastern and southern Africa, (b) conflicting trade interests and views between LDCs and non-LDCs and between oil exporters and non-oil exporters concerning the merits of opening up to EU imports on a reciprocal basis, and (c) similarly divided views over including trade in services and related issues in the negotiations. Serious difficulties were also encountered with the EPA's original approach to regional trade integration. However, the more flexible variable-geometry EPA-design that the EU adopted in October 2007 was a marked improvement over the original design and its adoption unblocked the stalemated negotiations in Eastern and Southern Africa. The following sub-sections analyze the diverse regional outcomes of the EPA negotiations to date, the constraints to expanding intra-African trade, open trade policy and the strategy for regional trade integration, and the potential contribution of full EPAs to advancing regional trade integration.37 Regional Participation in Interim and Framework EPAs 3.3 Eastern and Southern Africa. As discussed earlier, there were marked differences between eastern-southern and western-central Africa in participation in interim EPAs. With the exception of the conflict-torn Horn of Africa area, a large majority of the countries in the Eastern and Southern Africa regions have initialed interim EPAs. Taking the ESA, EAC, and SADC EPA-groups together, 15 out of 23 countries in 37Regional cooperation and integration can take many forms other than regional trade agreements: joint defense and peacekeeping arrangements; monetary unions; river basin development authorities; shared educational and research facilities; multi-country public health programs; and power pools, gas pipelines, and transport infrastructure, to name a few. All of these can be found in one or more of Africa's sub- regions, with various types of regional economic groupings and institutional arrangements dealing with different combinations of them. Each of these forms of regional cooperation is important, both in its own right and for its contribution to cooperation and policy coordination among countries. The focus in this section, like that of EPAs, is primarily on one form of integration ­ increased intra-African trade -- and related policy measures affecting the trade performance and investment climate in the EPA-countries. 32 eastern and southern Africa initialed interim EPAs (see the map on page xxi and the table on page xxiii). The coverage of the triangle bounded by Kenya on the north, Angola on the west, and South Africa on the south is even more complete. Of the 19 countries in this triangle, 15 signed interim EPAs. Of the four countries in eastern and southern Africa that have not signed interim EPAs, South Africa already has an existing free trade agreement with the EU; and Zambia signed the ESA framework EPA, although it opted out of its trade provisions and uses the EBA preferences instead. Only Angola and Malawi are relying solely on the EBA program. 3.4 The flexible variable-geometry EPA-design adopted in October 2007 also permitted quick agreement on "framework" regional EPAs with the EAC, ESA, and SADC groups. Two different types of regional EPAs emerged from the regional negotiations: a customs-union based approach that was adopted by EAC and an approach based on a free (preferential) trade area that was adopted by ESA and SADC. These three EPA-groups' framework EPAs cover WTO-compatible reciprocal import liberalization, development cooperation, and plans for working towards a full EPA in 2008-9. 3.5 Western and Central Africa. In contrast to the relatively positive response to the revised EPA design in eastern and southern Africa, there was much less immediate interest in western and central Africa. Only three countries out of 24 initialed interim EPAs. Both of these regions are composed largely of LDCs and oil-exporting non-LDCs that had little to gain in the near term from entering into interim EPAs. The three EPAs initialed to date in western and central Africa are free-standing country-specific ones with Cameroon, Côte d'Ivoire, and Ghana, which could have lost significant trade preferences by reverting from Cotonou to GSP status in January 2008. Negotiations of regional EPAs with the ECOWAS and CEMAC EPA-groups also had made little progress by the end of 2007 because of divergences in their trade structures, policies, and interests; and no regional EPAs yet have been signed in these two regions. 3.6 Implications for Regional Trade Integration. Alone, in their current form, most of the interim EPAs will do little to advance regional trade integration in Africa. The one important exception is EAC. Its member countries entered into an interim EPA as a custom union. They also agreed on one common schedule of imports from the EU to be liberalized that could facilitate the liberalization of intra-EAC trade in the same products. The other interim and framework EPAs have not yet reinforced regional trade integration efforts in Africa. Furthermore, as explained earlier, EPA-signatories will need to make parallel unilateral reductions in their MFN tariffs to limit the hub-and-spoke effects that could be caused by their implementing bilateral free trade arrangements with EU without simultaneously liberalizing intra-African trade. In some cases where EPA-signatories have chosen to eliminate tariffs on different lists of imports from the EU, regional negotiations may also be desirable to harmonize regional liberalization plans. However, those EPA-signatories that do liberalize imports from the EU under interim EPAs and implement the necessary accompanying reductions in MFN tariffs will be better 33 positioned than those not doing so to further liberalize their intra-African countries as well as their global trade.38 3.7 Because of the haste with which decisions had to be taken between the adoption of the revised EPA-design in late October 2007 and the December 2007 deadline for putting a WTO-compliant trade regime in place and the absence of sufficient preparatory work, most substantive issues concerning regional trade integration were left for future negotiations in the interim and framework EPAs. These agreements include only general provisions about further negotiation of full EPAs to agree on concrete actions to advance regional trade integration, but specific proposals for doing so have yet to be formulated in any detail in most of the regional EPA-groups. Constraints to Expanding Intra-African Trade 3.8 In contrast to trade with the European Union, intra-African trade, which EPAs are also intended to promote, is low. Officially recorded intra-regional trade in the EPA- negotiating groups averaged only 7% of the groups' total imports plus exports in 2006, ranging from a high of 14% of total imports and exports in the EAC EPA-group, all of whose members belong to the EAC customs union, to a low of 2% in the CEMAC EPA- group.39 The share of intra-regional trade in the EPA-groups' total trade has not changed significantly since 1990 except in EAC, where it has increased by 1.4 percentage points. Intra-African and transit trade is relatively more important for the fifteen landlocked African countries, for a handful of coastal countries that have some light manufacturing capacity, and for many border communities. 3.9 Intra-African trade is low for two reasons. First, significant policy and non-policy barriers to trade remain within both free trade areas and customs unions. Second, because of the small size of the EPA-countries' economies and the similarities in their production structures and income levels, trade complementarity among their economies is low. 3.10 Policy and Institutional Barriers to Intra-African Trade. Despite the proliferation of preferential trade agreements in Africa, significant barriers to intra- regional trade still remain within free trade areas, and even within customs unions, in all 38 Stevens, Bilal, and associates (2008) argue that one result of countries in the same regional trade arrangement liberalizing different lists of imports from the EU on different schedules will be further divergences in national trade policies and greater obstacles to regional trade integration. However, further analysis of the 18 interim EPAs shows that only the three independent country-specific interim EPAs for Côte d'Ivoire, Ghana, and Cameroon may pose problems for regional trade integration efforts in their two EPA-groups (ECOWAS and CEMAC)depending upon the approach to trade integration that is eventually implemented by these two EPA-groups. See the full report for a further discussion of the implications of interim EPAs for regional trade integration versus fragmentation. 39 The share of intra-regional in total trade is also at the higher end of the range in the UEMOA customs union, 10.8%, but was very low in the CEMAC customs union, 1.9%. Comparable figures are not available for SACU, which probably has relatively high intra-regional trade because of the size and diversity of the South African economy. The figures cited here are the shares of intra-EPA group imports and exports in the EPA-group's total imports and exports. Data are from IMF Direction of Trade Statistics. See the full report for a discussion of the various statistical problems involved in accurately measuring regional trade in Africa. 34 five of the regional EPA-groups. None of the three "free" trade areas involved in the EPA negotiations (ECOWAS, COMESA, and SADC) has really free internal trade among its members.40 Intraregional trade is only marginally freer within Africa's four customs unions. With the exception of SACU, all of these are, in fact, partial customs unions in the sense that, although they have official common external tariffs (CETs), actual implementation of these "common" external tariffs varies among member countries, there is no pooling of customs revenues, member countries maintain customs barriers at the custom unions' internal borders, and trade within the customs union is subject to restrictive rules of origin.41 Intraregional free trade, where effectively implemented, is generally limited to a small group of selected products. In addition, non-tariff barriers such as cumbersome and costly transit arrangements, inefficiencies and corruption in customs administration, both official and unauthorized road blocks and "tolls" on key transit routes, and the discriminatory application of higher domestic taxes to imports than to similar domestic products are additional major obstacles. Other internal barriers result from the lack of common product standards and systems of certification and from the absence of mechanisms for ensuring compliance with regional trade agreements and for resolving disputes. 3.11 Furthermore, some natural trading partners in Africa with substantial trade between themselves belong to different regional trade areas and EPA-groups.42 Thus, in some cases the liberalization of inter-regional trade is as important as the liberalization of intra-regional trade. However, actual and potential inter-regional trade between countries belonging to different RTAs and EPA-groups will not benefit from measures designed to liberalize intra-regional trade within these groups.43 3.12 Low Trade Complementarity. Second, even if the above policy, institutional, and infrastructure constraints could be overcome, medium-term prospects for expanding intra-African trade would be modest because of the similarity of the African exports. These similarities limit the range of products that can be exchanged with regional partners. The products exported by African countries--mostly primary commodities-- are not the main imports of African countries, which consist of manufactures and capital goods sourced mainly from the European Union and Africa's other major global trading 40All of these free trade areas would be more accurately described as "preferential" rather than as "free" trade areas. Within all of the free (preferential) trade areas, intra-regional trade is subject to restrictive rules of origin. Thirteen of COMESA's nineteen members are now implementing its free trade agreement, but they still maintain trade-distorting non-tariff barriers on trade with each other. Intra-regional trade is even more restricted in SADC, except within SACU. The ECOWAS free trade agreement has not yet been effectively implemented. Thus serious tariff and non-tariff barriers still constrain both intra and inter- regional trade in Africa. 41 Such nominal "customs union" in which member countries do not fully implement the "common" external tariff or maintain barriers to internal trade has is, in reality, hardly a customs union in any meaningful sense and has only limited regional trade integration effects. 42 For example, Nigeria belongs to ECOWAS, whereas Cameroon on its southern border belongs to CEMAC; and Congo (Dem. Rep.) belongs to the CEMAC EPA-group, whereas Burundi and Rwanda belong to EAC. The cross-border trade between these countries thus does not benefit from measures to liberalize intra-regional trade within ECOWAS, CEMAC, and EAC. 43 Inter-regional trade could, in fact, be disrupted by the more universal implementation of regional common external tariffs on trade with countries in other regional groupings. 35 partners. Other than South Africa, only a few African countries have even a limited capacity to supply manufactured products, in most cases simple consumer goods, to regional markets. Furthermore, integrated cross-border production networks and corresponding trade flows have yet to develop because of low levels of development, in general, and of manufacturing, in particular, as well as because of the various policy, institutional, and infrastructure constraints to intra-African trade noted above.44 Thus, the potential gains from regional trade integration are likely to be relatively small compared to those from global trade integration. 3.13 Trade complementarity will not increase quickly; and, as a result, intra-African trade is likely to continue to grow less rapidly in the medium term than Africa's world trade. Trade complementarity is determined by the structure of production and the composition of demand in the countries concerned. To the extent that trade complementarity changes over time, it changes only slowly as economies develop. In the medium term, Africa's current imports of manufactures, machinery, and equipment cannot be domestically produced to expand intra-Africa trade because most African countries lack the capacity to do so and will need a long period of time to develop more diversified economies. Consequently, the share of intra-African trade in the EPA- countries total trade is likely to continue to increase little, if at all, in the medium term. Supporting the expansion and diversification of international trade should, therefore, be the primary objective of trade integration efforts. Effective regional trade integration can, however, provide limited additional opportunities for some small African economies, particularly landlocked ones with few other options, to take advantage of scale economies and increase competition, efficiency, and competitiveness. Importance of Open Trade Policies 3.14 Although regional integration is an important objective of the African countries, low trade complementarity will remain an enduring constraint to the rapid growth of intra-African trade, and prospects for expanding intra-African trade will be limited in the medium term. International trade is, and will remain, economically much more important for African countries than intra-African trade. Its expansion and diversification is thus likely to have a much greater impact on overall growth and poverty reduction than is the growth of regional trade. Intra-African trade will necessarily play a secondary role to global trade in accelerating growth, and the non-complementarity of most African economies magnifies the importance of global trade integration and open trade policies for them. 3.15 Advantages of Export-Push Strategies. Export push strategies have growth, efficiency, and distributional advantages for Africa relative to regional import- substituting strategies. In small economies, export-oriented industries may grow faster and to a much larger eventual size than import substitution permits. European Union and other foreign markets are very much larger than those of the regional EPA groups. This large market size permits rapid expansion of output and employment in each export industry that masters the required technological and managerial skills and becomes 44 See the full report for a discussion of empirical measures of trade complementarity in Africa. 36 competitive enough to serve them. Even one or two such industries can thus yield a dramatic economic acceleration in a relatively short period. In contrast, to have much impact on the growth of output and employment, a general import substitution strategy is far more demanding: it requires mastering new technologies and attaining reasonable levels of efficiency in a wide range of industries because the scope for expansion of each is relatively small. An immediate export-push strategy also necessitates a focus on efficiency and competitiveness right from the start. To be competitive, export industries require access to low cost inputs of both goods and services, and tariff-free imports are important for obtaining these. 3.16 The distributional effects of export-led growth also tend to be more attractive than those of import-substitution. A comparative advantage in exporting labor intensive manufactured or agricultural products and an expansion of these industries can result in rapid growth of employment and a wide sharing of the income gains from growth. Import-substitution in more capital intensive sectors, on the other hand, tends to generate rents for owners but only a few niche jobs for privileged workers. In addition, import- substituting industries typically produce consumption products for the low-income mass market (because they would not competitive in producing for the small high-income market) or simple widely used inputs. The protection granted to these industries is thus usually passed on in the form of higher prices (resulting from inefficiency or high profits) to low-income consumers or to other domestic producers. 3.17 Problems with Regional Import Substitution Strategies. In contrast, even a general regional import-substituting strategy is likely both to slow growth and to have negative distributional effects between countries as well as within them. In light of the poor track record of import substitution policies in the small African economies, many proponents of protectionist policies are now arguing that these policies should be extended from the national level to the larger economies of Africa's regional trade areas. They argue that a period of inward-oriented regional integration--that is, a "regional integration first" strategy--is needed to build up domestic industries before liberalizing the external trade of Africa's regional trade areas and facing global competition. However, even region-wide markets in the African EPA-groups are too small to offer much scope for economies of scale and meaningful competition among multiple producers. The growth and employment creating prospects for most, if not all, import- substituting industries in these markets are necessarily quite limited. Hence, a general regional import substitution strategy, like the country import substitution strategies that have preceded it, is likely to maintain a range of inefficient industries that will be a continuing burden on the member countries' economies, and an obstacle to further trade liberalization, rather than the basis for a future export expansion. If costly trade- diversion and monopoly pricing effects are to be avoided in the regional integration process, liberalization of external trade needs to take place ahead of, or at least simultaneously with, the liberalization of intra-regional trade, rather than follow it. Moreover, the transfer of technology and know-how and the rapid implementation of new business opportunities through ready access to competitively priced imports are key elements in the dynamics of the growth process but can be seriously retarded by import substitution strategies. 37 3.18 In addition, regional import substitution tends to be politically divisive. A high regional tariff on an import-substitute benefits primarily its producers in the country (or countries) where they are located, at the expense of users of the product in all the members of the RTA. Since manufacturing activity tends to be heavily concentrated in one or two countries in each of the African EPA-groups (Cameroon in CEMAC, Kenya in EAC, Nigeria in ECOWAS, and South Africa in SADC), the producers benefiting from a regional imports-substitution strategy will tend to be similarly concentrated, with resulting divergence of interests and political tensions among countries. 3.19 Open Trade Policies and Regional Trade Integration. Moreover, trade openness is likely to be the driver of regional trade integration rather than vice versa as envisaged in the "regional integration first" strategy. In particular, open trade policies are far more important than common external tariffs for advancing trade integration in Africa. Broadening the scope of intra-African free trade is difficult in countries where external tariffs are still high. In such cases, preferential duty free importation of a product otherwise subject to a high tariff can give regional suppliers a substantial price advantage over external ones and cause a significant revenue loss for the government as well as possible diversion of trade from more efficient suppliers outside of the region. Even in "customs unions," a high common external tariff creates strong incentives for protected domestic producers to seek restrictive rules of origin to prevent competing imports from entering from other members of the customs union as well as to try to maintain or increase high common external tariffs. 3.20 In an era of increasing global trade integration, countries with open trade policies are justifiably unwilling to adopt the policies of their more protectionist regional neighbors. In most cases, a country's lack of political interest in effectively implementing open trade policies will unfortunately also imply a similar lack of progress on regional trade integration. And, if the protectionist countries are not ready to liberalize, neither agreement on low common external tariffs nor free intra-African trade is likely to be possible with these countries. Outward-oriented countries should thus continue opening their trade policies as rapidly as possible rather than seeking, or accepting, inward-looking common external trade policies that will hold them back as well as the regions and countries sponsoring these policies. 3.21 The readiness of countries to pursue open trade policies, both globally and regionally, will thus be a key determinant of progress on intra-African trade integration. In addition to open trade policies, other complementary actions -- such as trade facilitation, investment in required infrastructure, and establishment of a favorable investment and business climate and increased trade-related investment -- are, of course, equally necessary. But these actions too are more likely in an open outward-looking economic environment than in a closed inward-looking one. Conversely, in most cases, lack of political interest in effectively implementing open trade policies will unfortunately also imply lack of progress on other measures to expand trade and promote regional trade integration. In practice, global and regional trade integration are likely to turn out to be complements rather than substitutes; and, the faster that global integration proceeds, the faster will be the pace of regional trade integration of intra-African trade. It is, therefore, essential that regional integration policies be outward-oriented. 38 A New Strategy for Regional Trade Integration 3.22 Given the limited trade complementarity of African economies and the numerous policy, institutional, and physical barriers to trade among them, effectively promoting intra-African trade, whether through full EPAs or independently of them, will be no simple matter. A new strategy will be necessary for doing so. The key feature of such a strategy will be a more open trade policy environment that is more favorable to the liberalization of intra-African trade. The essential components of an open trade strategy are elimination of trade-distorting NTBs, reductions in MFN tariffs, particularly peak rates, and effective implementation of measures (such as export processing zones) to ensure that exporters have ready access to competitively priced tariff-free imported inputs. The parallel unilateral reductions in MFN tariffs that need to accompany the elimination of tariffs on imports from the EU under interim EPAs would be a helpful first step to liberalize the EPA-countries' intra-African trade as well as their global trade and to improve the environment for further liberalization. 3.23 Unsuccessful Common Trade Policy Strategies. As discussed above, the current combination of "free" trade agreements that have numerous exceptions, protectionist rules of origin, and trade-distorting NTBs, and customs "unions" that maintain customs and other barriers to intra-union trade has done relatively little to eliminate the policy and institutional barriers to intra-African trade. The expansion of intra-African trade is thus still constrained by a long list of policy and institutional barriers: free trade arrangements and common external tariff, where they exist, are only partially implemented; numerous restrictive non-tariff barriers to intra-African trade are still maintained; customs administration is inefficient, and corruption is often a problem; and transit and transport services are poor and costly. Unlike low trade complementarity, these policy and institutional obstacles could be overcome, or at least substantially eased, in the medium term. 3.24 For three decades or more, most African RTAs have sought to promote intra- regional trade by emphasizing common trade policies, particularly common external tariffs, to harmonize trade regimes across their member countries. However, differences among their members in economic conditions, trade interests and policies, and political willingness to liberalize intra-African trade have often led to paper agreements on common trade policies that could not be effectively implemented, as reflected in the numerous existing policy and institutional obstacles to intra-African trade, because of underlying divergences between countries in interests and policies. In many cases, common trade policy strategies have tended to restrain the liberalization of intra-African trade to the pace of trade policy reform in an RTA's most reluctant member countries. 3.25 An important reason for the failure of the common-trade-policies approach is that MFN tariff rates are still high. Most, if not all, imports of products that compete with domestically produced ones are subject to the countries' maximum MFN tariff rates, which are still 20-30% or higher depending upon the country. Broadening the scope of intra-African free trade is difficult in countries where external (MFN) tariffs are still high. The current high tariffs on most domestically produced import-competing products result in high protection levels and rents for the firms that benefit from these. Elimination of 39 tariffs on intra-African trade (whether through FTAs or customs unions) in import- competing products that are protected by high tariffs will cause substantial reductions in these rents. The loss of such rents is strongly opposed by the vested interests concerned. Hence, it is often politically difficult or impossible to effectively implement free trade in products bearing high MFN tariffs. In addition, preferential tariff-free importation from regional producers of a product otherwise subject to a high MFN tariff gives the regional producers a substantial price advantage over external ones and can cause a significant revenue loss for the government as well as diversion of trade from more efficient suppliers outside of the region. Consequently, government commitment to implementing free intra-African trade in high tariff products is often weak. As a result, most high-tariff products that compete with domestic production are usually explicitly excluded from arrangements for free intra-African trade and, when not excluded, are constrained by protectionist rules of origin and trade-distorting NTBs. Open, low-tariff trade policies are, therefore, important for advancing regional trade integration in Africa. 3.26 A Flexible Variable-Geometry Approach Focusing on Free Intra-African Trade. If EPA-countries wish to accelerate regional trade integration in parallel with the more important global trade integration, a radically different policy approach will be needed. Rather than continuing often futile efforts to "harmonize" divergent policies through paper agreements on "common" trade policies, a new more pragmatic strategy for promoting intra-African trade is needed that emphasizes implementation of open global and regional trade policies as rapidly as individual countries are ready to do so. Because of the wide differences among the members of the various EPA-groups in initial conditions, trade interests and policies, and political readiness to liberalize, the variable- geometry approach adopted for EPAs in 2007 will remain essential. Continuation of the variable-geometry approach will help to accommodate the current mixture of free (preferential) trade areas, customs unions, and countries with independent trade policies wishing neither to implement the existing trade arrangements nor to liberalize imports from the EU. The larger and more diverse a regional EPA-group is, the more important will be free trade areas and variable-geometry arrangements to allow for diversity. Rather than using up scarce technical/managerial resources and political capital in pushing hard for adoption of unachievable or counterproductive "common" external tariffs, the priority for regional trade policy should, instead, be effective implementation of free intra-African trade in all products (or, in those products for which free trade is not initially possible, preferential trade) both within and between regional trade areas and the liberalization of the restrictive rules of origin governing free/preferential intra-African trade.45 An additional important component of a new regional integration strategy would be to focus collective efforts on improving the regional competitiveness, an objective which the inclusion of the liberalization of trade in services and investment in full EPAs could help advance as discussed in section B.4 below. 45As long as peak MFN tariffs still exceed 15%, free trade in high tariff products may be too strongly opposed by the protected industries concerned and tariff preferences for intra-African trade in these products may need to be considered as a temporary second-best option. 40 3.27 Free Intra-African Trade and Convergence Towards Low Common External Tariffs. Continuing with a flexible variable geometry approach would not mean that the African RTAs concerned would need to give up any ambition of eventually becoming effective full customs unions. Rather, for most African RTAs, the immediate priority for advancing regional trade integration is effectively implementing free/preferential intra- African trade for most products. Freeing up intra-African trade requires, in addition to further reducing MFN tariffs and eliminating tariffs on intra-African trade, removing all remaining trade-distorting NTBs, liberalizing rules of origin for intra-African trade, improving customs administration and trade facilitation, and investing in the necessary supporting infrastructure. Pursuit of more open trade policies should over time reduce the current policy disparities among countries and lead to convergence at lower MFN tariff levels, reducing protectionist influences and facilitating eventual negotiation of efficient common external tariffs. At the same time as low MFN tariffs and free intra- African trade in most products are implemented, the capacity of regional institutions to manage a full customs union can be developed so that their member countries can delegate the necessary political authority these institutions when they are ready to do so politically. Potential Role of Full EPAs in Advancing Regional Trade Integration 3.28 Full EPAs could help accelerate regional trade integration by (a) providing a dynamic stimulus and a coordinating mechanism for regions to undertake unilateral MFN reforms to liberalize trade and to improve the business environment and competitiveness; (b) facilitating liberalization of intra- and inter-regional policies governing intra-African trade by addressing such issues as elimination of trade distorting NTBs, creation of more effective regional preference schemes, and liberalization of the restrictive rules of origin governing intra-African trade; and by (c) enhancing the credibility of regional integration by locking-in reforms in an international treaty. Concluding full EPAs in a flexible variable-geometry format appears to be an achievable objective from some African regions. Table 5: Classification of EPA-Groups by Common External Tariffs and Common List of Imports from the EU to be Liberalized Common Import Liberalization List Yes No No CARIFORUM ESA SADC Common External Tariff Cote D'Ivoire (UEMOA) Yes EAC SACU Cameroon (CEMAC) Ghana (ECOWAS) 3.29 Regional EPAs in Eastern and Southern Africa. The different approaches to regional coordination that have been adopted to date by the various regional EPA-groups are classified in Table 5 by whether or not a group adopted common external tariffs and by whether or not it agreed to liberalize a common lists of imports from the EU.46 EAC 46 The Caribbean EPA, which is also included in Table 5, adopted a different approach to regional trade integration from any of those adopted thus far in Africa. The Caribbean countries all agreed to eliminate 41 is the only African EPA-group that has successfully concluded negotiations for liberalizing imports from the EU for the EPA-group as a whole and has been able to enter into an interim EPA as an entire customs union. It also agreed to eliminate tariffs on a common list of imports from the EU, a step which should facilitate moving towards free intra-EAC trade in the same products. 3.30 In the other two EPA-groups that have concluded framework agreements, ESA and SADC, a flexible variable geometry approach was needed and has worked well. In both cases, the framework EPAs include both regional provisions that apply to all members in the group and country or sub-group specific provisions that apply only to specific countries or sub-groups. By thus permitting varying degrees of commitment at the regional, sub-regional, and country levels that reflect the diversity of conditions and national interests, a variable geometry approach can encourage more countries to join in the negotiations for full EPAs -- particularly if opting out of the liberalization of imports from the EU is permitted, as in the case of Zambia's participation in ESA's framework EPA. Because of the diversity of conditions and trade interests among their members, both the ESA and SADC groups will need to continue negotiations of full EPAs on a variable geometry basis. 3.31 Regional EPAs in Western and Central Africa. Similarly, in western and central Africa, the best option for negotiating full EPAs is to adopt a variable geometry approach similar to the one used in the ESA and the SADC groups. Because all members of the ECOWAS and CEMAC EPA-groups have not implemented the ECOWAS/UEMOA or CEMAC common external tariffs and because those that have implemented these CETs maintain customs posts at internal borders, do not pool customs revenues, and apply rules of origin on imports from other members of the customs union, these two groups' trade regimes are effectively free/preferential trade areas with partially harmonized external trade policies. There is little possibility in the medium term of all members of the ECOWAS or CEMAC EPA-groups' effectively implementing common external tariffs and then successfully negotiating a common list of imports from the EU to be liberalized under a full EPA. However, the two groups could easily participate in full EPAs as free/preferential trade areas in the same manner as ESA and SADC with separate country-specific lists of imports from the EU to be liberalized.47 In addition, in the ECOWAS EPA-group it is particularly important to maintain the current low four band common external tariff with its maximum rate of 20% for the 14 countries that have already adopted this CET and to avoid the addition of a higher fifth tariff band, or any other increase in tariffs, to accommodate Nigeria's more protectionist trade policies. tariffs on the same common list of imports from the EU and included a regional preference clause in their EPA extending the same tariff concessions to each other as to the EU. These steps will, in effect, create free intra-Caribbean trade in the products concerned as the tariffs on imports of these from the EU are eliminated. This approach to intra-Caribbean free trade was facilitated by the fact that most of the Caribbean countries concerned were non-LDCs, which had to agree to reciprocal import liberalization in order to preserve preferential access to the EU market for their exports. 47Such an approach could, for example, accommodate the different import liberalization lists negotiated by Côte d'Ivoire and Ghana for their interim EPAs discussed in footnote 32 above. See the full report for a further discussion of alternative variable-geometry arrangements with and without common external tariffs and with and without common lists of imports from the EU to be liberalized. 42 3.32 Measures to Reduce Policy Barriers to Intra-African Trade. Appropriately designed and effectively implemented full EPAs could play a useful role in establishing free intra-African trade. EPA-countries could use these to help accelerate action on long- standing policy and institutional barriers to expanding regional trade. In addition to eliminating tariffs on intra-African trade within and between RTAs, a package of four broad measures would remove (or substantially reduce) most of the other policy barriers to intra-African trade: a. Removal of Trade-Distorting NTBs. All trade-distorting NTBs need to be eliminated as soon as possible.48 Continuing or, worse, increasing the use of trade-distorting NTBs would make preferential and MFN tariff reductions meaningless and could raise, rather than reduce, the policy barriers to intra- African trade. b. Liberalization of Rules of Origin for Intra-African Trade. Simple, standardized rules of origin need to be established by taking the same steps as recommended above for improving access to the EU market -- that is, replacing all of the current product and process specific rules of origin with a uniform change of tariff heading at the HS 6-digit level or 10% value added rule. c. Regional Preference Clauses. Each full EPA should include a regional preference clause (similar to that in the CARIFORUM EPA) granting to other EPA-countries whatever tariff and other concessions are granted to the EU in the full EPA. Such regional preference clauses would encourage intra-African trade without imposing significant new adjustment costs on the EPA-signatories since tariffs on the goods concerned would be eliminated on imports from the EU in any case. These preferences could subsequently be extended to the members of the other African EPA-groups. d. Comprehensive but Low Tariff Preferences for Intra-African Trade. For those products for which it is not possible to move immediately to free intra-African trade, a comprehensive but low tariff preference margin for all participating African countries of a 5 to 10 percentage point reduction from the applicable MFN tariff for all imports should be established. Such a preference scheme would encourage trade between the members of different RTAs that does not benefit from steps to free up intra-regional trade.49 48 The term "trade-distorting NTBs" is used here to refer to measures such as quantitative restrictions, discriminatory domestic taxes, and import licensing that are used for protectionist purposes in order to distinguish these from WTO-compatible measures such as legitimate sanitary and phyto-sanitary standards and safety and public health regulations. 49A 5 to 10 percentage point preference margin would cause smaller revenue losses and be less threatening to protected industries than efforts to completely eliminate the tariffs on products currently bearing high tariffs, which are likely to be more strongly opposed or lead to re-imposition of NTBs. An additional reason for holding the regional preference margin to no more that 5 to 10 percentage points is to limit the risk of trade diversion. As MFN tariffs are eventually reduced over time, the preference margin for intra- African trade would cause the actual applied tariff rates on intra-African trade to fall toward zero. Such a preference scheme, like the regional preference clauses in the preceding subparagraph, would be compatible with WTO rules since it would involve only developing countries and hence would be permitted under WTO's enabling clause. See the full report for further discussion of the foregoing points. 43 Full EPAs could also help accelerate regional trade integration by supporting trade facilitation and capacity building for making and implementing open trade policies and by financing the development of regional infrastructure. 3.33 Reform Incentives. The negotiating schedules and deadlines for the full EPAs could provide, as those for the interim EPAs have, a useful dynamic impetus for further global and regional integration in a situation where progress might otherwise be halting. However, significant progress in actually liberalizing intra-Africa trade under full EPAs will require an open trade strategy as discussed above. The liberalization of intra-African trade will also have to be largely African driven as the incentives that the EU can offer in terms of new market access will be relatively limited unless it is willing to substantially liberalize the EPAs' restrictive rules of origin. Hence, what can be achieved to advance regional trade integration under full EPAs could largely depend upon the willingness of the governments concerned to use these to accelerate action on difficult longstanding policy and institutional barriers to expanding intra-African trade. B. Liberalizing Trade in Services and Foreign Direct Investment 3.34 The services sector is currently the largest and fastest growing economic activity in most of the developed and developing world, and its share in world trade and investment is rising. Services are both an important input contributing to the competitiveness of goods production in the world market and growth sectors in their own right. On average, services accounted for 28% of the EPA-countries' total imports of goods and services in 2003-2004 and 15% of their total exports of goods and services. Two thirds of the EPA-countries' services exports go to the EU, roughly half of their services imports come from the EU, and the EU is the source of more than 80% of their foreign direct investment. 3.35 In addition to helping to catalyze progress on regional trade integration, full EPAs could also provide an important opportunity to address issues concerning trade in services and investment. This section examines the possibilities for increasing the access of the African EPA-countries' service exports to the EU market, the measures needed to maximize the positive effects of liberalizing the EPA-countries imports of services and investment from the EU, the relevance for Africa of the Caribbean region's experience in liberalizing services and investment under its full EPA, and the complementary reforms that will be needed. Africa's Exports of Services to the EU Market 3.36 In the average African country, exports of services are about 1/5 the level of its merchandise exports and account for 5% of GDP. The EU is actually a somewhat more important market for Africa's service exports than for its merchandise exports, absorbing 66% of Africa's $15 billion in service exports in comparison to 54% of its non-oil merchandise exports. Travel and transportation services account for the bulk, 70%, of Africa's service exports to the EU. 44 3.37 African services exports that could potentially benefit under EPAs are tourism, call centers and data processing, and provision of services through the temporary presence of natural persons (GATS service delivery Mode 4). However, the gains in the near term from improvements in market access for these services, although worth pursuing, are likely to be limited because the EU has minimal restrictions on the access of African exporters of these services to its market other than on the temporary presence of natural persons (Mode 4). The binding constraints to exports of services are limited capacity and supply bottlenecks rather than market access. Development assistance including technical assistance and cooperation with the EU provided in support of full EPAs could, however, be helpful in addressing supply side issues.50 3.38 Tourism. Tourism, as Africa's most important service export, would seem, prima facie, to be a possible beneficiary of full EPAs and is probably the least controversial in Africa for the binding liberalization of EU direct investment as most countries would like to increase their foreign exchange earnings from the sector. However, the expansion of EU tourism to Africa may, itself, face few policy constraints that could be addressed in EPAs. But the tourism sector could probably benefit from improvements in the regulation of related services such as scheduled and charter air services. In view of the importance of tourism for some African countries, it is worth investigating in more detail to determine whether there are any tourism-related policy constraints that EPAs could alleviate. Liberalization of Imports of Services and Regulatory Reform 3.39 Services accounted for 28% of Africa's total imports of goods and services in 2003-2004. The residual category "other commercial services" accounted for the largest share of Africa's services imports with 44% of the total, followed by transportation services with 35%, travel services with 17%, and government services with 5%.51 The EU, with a 48% market share, is the largest supplier of services imports to Africa. 3.40 Importance of Investment in Services. Because of the current underdeveloped nature of the service sector in most of Africa, major gains from the liberalization of trade in services are likely to come from the liberalization of imports of services rather than from increased African exports of services. In particular, liberalization of foreign direct investment in services (delivery of services by GATS Mode 3, commercial presence) offers the largest potential for generating efficiency gains that are needed for increasing the competitiveness of Africa's merchandise exports and its economies more generally. Barriers to entry of investment in services in some African countries have led to 50Full EPAs also offer an opportunity to facilitate the temporary presence of natural persons for business purposes, including business services sellers, contractual service suppliers, and independent professionals (delivery of services by GATS Mode 4). In the Caribbean EPA, the EU's liberalization commitments include six categories of natural persons, 29 sub-sectors for contractual service suppliers, and 11 sub- sectors for independent professionals. Similar provisions to those in the Caribbean EPA might be made for expanding employment of African temporary service providers in the European Union. See the full report for a more detailed discussion of Mode 4 issues. 51The only available aggregate statistics on imports of services breakdown commercial (non-government) services into transport, travel, and a catch-all residual category, "other commercial services." 45 inefficient provision of important services such as finance, telecommunications, and transportation.52 For many services, the gains from increased investment, trade-induced improvements in efficiency and competitiveness, and their resulting more rapid growth will be magnified because these play such a major role in all types of production: financial services help turn savings into investment; telecommunications facilitate the diffusion of knowledge and the coordination of production; transport systems make the movement of goods and people more efficient; and education builds the stock of human capital. The largest gains from liberalization of imports of services in Africa are likely to come from increased foreign investment and commercial presence (Mode 3) in three sectors that provide important inputs for both Africa's merchandise exports and for production for its domestic markets: finance, telecommunications, and transport. 3.41 Need for an Appropriate Regulatory Framework. To reap the full benefits of liberalizing imports of services and foreign direct investment in these, establishment of an appropriate regulatory framework will be indispensable for some of the service sectors concerned. Because of the characteristics of some service sectors, some serious problems can result from their liberalization without an adequate regulatory framework. Market failures caused by monopoly power, asymmetric information, or externalities in the provision of services create the need for appropriate regulation. Sectors in which large fixed investments are necessary for the efficient provision of services are often imperfectly competitive even with free entry. In addition, some services, such as finance and telecommunication, require some regulation to prevent destabilizing or anticompetitive practices. 3.42 In fact, the timing of the liberalization of different service sectors should be determined by the capacity of African countries to implement the required accompanying regulatory reforms. Establishment of an appropriate regulatory framework should precede the opening up of the sector concerned to set the rules of the game for new investors. But improving the competitiveness of the services sector is too important for accelerating economic growth to be put on indefinite hold because of the lack of regulatory capacity. In cases where inadequate regulatory capacity is a constraint to liberalization, clear plans and a definite timeframe need to be established for upgrading this capacity. 3.43 Complementary MFN Liberalization of Services Imports and Investment from All Sources. As in the case of merchandise trade, the African EPA-countries would benefit more from reducing their barriers to service imports from all foreign sources rather than liberalizing on a preferential basis with just the EU. Most barriers to imports of services usually distinguish between foreign and domestic firms but not between firms from different foreign countries. Thus, removal of the barriers for all foreign firms, rather than just for EU firms, is the natural policy response for opening up services sectors to imports. Liberalization of services imports from all countries tends to increase competition among service providers and is likely to lead to greater efficiency gains and more rapid growth. 52See the full report for a further discussion of this point. 46 3.44 In addition, in cases where network and capital intensive services are not already open to all foreign investors, preferential liberalization of these services for investors from the EU could give EU service providers a permanent advantage even if the services concerned are eventually liberalized multilaterally. Large fixed investments imply that the first entrant, or "first mover," is likely to have a permanent advantage because the marginal costs of the first mover are likely to be much lower than average costs of new entrants. The irreversibility of investment decisions and the first-mover advantage make it wasteful not to invest in the most efficient service at the start and make it hard to replace an inefficient network or capital intensive service provider once in place. Because the EU is unlikely to be the most efficient provider of all network and capital intensive services, preferential liberalization of imports of these services from the EU could lead to sub-optimal development of these sectors in the EPA-countries when the first movers from the EU are not the most efficient available providers of the services concerned. 3.45 In order to attract investment by the most efficient service providers and benefit fully from economies of scale over the long term, services imports and investment should thus be liberalized on both a global (MFN) and an intraregional basis at the same time as they are liberalized vis-à-vis the European Union. The necessary MFN liberalization of services imports and investment could be accomplished either by including such provisions in a full EPA or through parallel unilateral reforms. The Caribbean EPA: Provisions Relevant to Africa 3.46 In December 2007 the EU signed a full EPA with the 16 CARIFORUM countries in the Caribbean. CARIFORUM went farther than any of the African EPA-groups. Its full EPA establishes a comprehensive framework for future actions to address liberalization of trade in services, foreign direct investment, and trade-related regulations. The Caribbean EPA is of particular interest to Africa because it gives a clearer idea of what the EU envisages for full EPAs. Its provisions concerning the liberalization of trade in services, foreign direct investment, and trade-related regulatory issues are summarized in Annex A. Many of these provisions are relevant to Africa and are worth consideration in designing full EPAs. 3.47 Provisions Relevant for African EPAs. The main features of the Caribbean that appear particularly relevant for the design of the liberalization of trade in services, foreign direct investment provisions, and the trade-related regulatory framework for full EPAs are: a. Coverage of all four GATS modes of service delivery and foreign direct investment in production of goods as well as services; b. Progressive, reciprocal, asymmetric bilateral liberalization of trade in services and investment; 47 c. Objectives of free-entry, non-discriminatory (national treatment), and MFN treatment of service providers and investors from the signatory countries;53 d. A flexible, phased approach to implementation with separate country-by-country lists of commitments and initial commitments to liberalization only in the sectors where this can be readily implemented; e. Schedules for reviewing progress in implementation and initiating negotiations of subsequent phases of liberalization; and f. A regional preference clause extending the same treatment of services and investment granted to the EU to other members of the EPA-groups.54 3.48 Because of the wide range of EPA-outcomes thus far in different regions and the large variations among African countries in readiness to implement reforms, the flexible design and phased approach to the liberalization of services and investment in the Caribbean EPA with its built-in schedule for further progressive liberalization also appear well suited for full EPAs in Africa. With this type of flexible design, reform priorities in African EPAs could reflect the interests and constraints of individual countries. In Africa, particular emphasis will need to be given to capacity building for improving the regulatory framework for services liberalization. The phased approach could allow countries to focus on capacity building and initially liberalize only those sectors that require little regulation or where the existing regulatory framework is satisfactory, leaving remaining sectors for the future phases with a commitment to continuing negotiations over further progressive liberalization. 3.49 The Caribbean EPA's competition and regulatory provisions also seem appropriate for full EPAs in Africa. As discussed in Annex A, these provisions emphasize the importance of establishing policies and institutions both for preventing practices that substantially lessen competition and for regulating services appropriately when necessary. Such competition and regulatory provisions could provide a catalyst and coordinating mechanism for setting up or strengthening national and regional competition agencies in Africa. The EU also has consider-able experience in competition and regulation and is potentially an important source of technical assistance in this area. 3.50 Negotiating Incentives. The incentives facing the CARIFORUM countries when they negotiated their full EPA were quite different from those now facing the African EPA-countries. Most of the participants in the negotiations for the Caribbean EPA were non-LDCs with a relatively strong incentive to enter into an EPA to preserve their 53It is also advisable that full EPAs cover other investor protections in order to establish a fully investment friendly environment in the signatory countries. Investment protections are often found in bilateral investment treaties rather than in free trade agreements like EPAs. It is therefore desirable to review any bilateral investment agreements that African countries may have signed with the EU (and the US) to help develop a position for EPA negotiations. 54Such a regional preference clause could provide an additional stimulus to intra and inter-REC efforts for liberali-zation of trade in services and investment. Although a regional preference clause could become redundant if all members of an African EPA-group liberalize services and investment on a MFN basis, it would be helpful if some do not and could also have a constructive symbolic value as a visible sign of political support for regional integration efforts. 48 preferential access to the EU market for their merchandise exports, and the liberalization of imports of services and investment was part of a package that included tariff-free, quota-free access to the EU market for merchandise exports. It is unclear how much new liberalization of service imports the Caribbean EPA will actually stimulate over time unless the EU is able to offer further improvements in the CARIFORUM countries access to its own market. 3.51 In contrast, nearly all of the African EPA-countries now have tariff-free, quota- free access to the EU market for their merchandise exports under either interim EPAs or the EBA program. The market-access incentives for liberalizing imports of services and investment under full EPAs, in fact, appear weaker than those faced by African LDCs when considering interim EPAs, a situation which led to only 8 of these 33 LDCs' entering into interim EPAs. Moreover, in the case of full EPAs, the weak market-access incentives for liberalizing services and investment apply to non-LDCs as well as to LDCs. The one important improvement in market access that the EU could offer to countries liberalizing services and investment under full EPAs is the full liberalization of the EPA's rules of origin for merchandise trade, as recommended in paragraph 2.31 above. It is, however, uncertain whether or not the EU would eventually agree to fully liberalize the rules of origin for full EPAs. 3.52 If the EU turns out in the end to be unwilling to fully liberalize the rules of origin for merchandise trade under full EPAs, African countries may have little incentive for liberalizing imports of services and investment under full EPAs other than the desire to accelerate the reforms involved. In this case, the liberalization process, as in the case of regional trade integration, would have to be largely African driven; and only genuine reformers may be interested in participating in full EPAs with meaningful substantive provisions for liberalizing services and investment. Enhancements and Complementary Competitiveness Reforms 3.53 Necessary Enhancements. The services and investment provision in the Caribbean EPA would need to be both enhanced and accompanied by complementary unilateral business climate reforms in order to have greater positive effects on investment and competitiveness in Africa. The Caribbean EPA has three shortcomings concerning the liberalization of imports of services and foreign direct investment that would need to be corrected for in full EPAs for Africa. First, in order to promote as much competition as possible, attract as much investment as possible from the most efficient service providers, benefit fully from economies of scale over the long term, imports of services and investment should be liberalized from all countries on a MFN and regional basis at the same time as they are liberalized vis-à-vis the European Union. Unless African EPA- countries provide identical treatment to foreign investors and service providers from all countries, investment flows from other countries would not be likely to increase and might even be deterred if nationals of third countries perceive the investment climate in some sectors as biased in favor of the EU by an EPA. 3.54 Second, to reap any immediate substantive benefits from a flexible framework phased approach for liberalization like that in the Caribbean EPA participating African 49 countries will need to identify priority service sectors with good development potential and formulate country-specific (and, where feasible, regional) proposals for actually liberalizing these in the first phase. The Caribbean experience to date has shown that the liberalization of trade in services is, in fact, more technically and politically demanding than the liberalization of merchandise trade and that the results of the initial effort risk being minimal. Hence, an important step to prepare for negotiations of full EPAs will be for the African EPA-countries to undertake detailed reviews at the country and regional level to identify priority sectors for liberalizing trade in services and investment. Some countries, particularly LDCs, may have little capacity to carry out the necessary analyses and require additional time and assistance to prepare adequately for negotiations. 3.55 Third, the effect of including investment in production of goods (merchandise), as well as services, in full EPAs will be limited by the interim EPAs' current restrictive rules of origin. Full EPAs need to be enhanced by fully liberalizing the rules of origin for merchandise trade as recommended in paragraph 2.31. This enhancement would be particularly important for attracting FDI for export production of products that are currently unable to benefit from tariff-free, quota-free access to the EU market under interim EPAs because of their rules of origin. 3.56 Likely Effects of MFN Application of Enhanced CARIFORUM Type Reforms. Enhanced MFN liberalization of imports of services and investment and the trade-related regulatory provisions of the type included in the Caribbean EPA would have two broad effects on foreign direct investment in the sectors to which they are applied. First provision of services and investment in designated sectors would be permanently opened to all foreign service- providers and investors through the provisions on free entry, national treatment, and non-discrimination; and the permanency of this opening would be locked-in by treaty. Second, a competitive environment would be established for all foreign service-providers and investors that has the type of minimum regulatory framework expected by trading partners in open, globalizing developing economies: institutional arrangements to maintain competition, prohibition of corrupt practices, protection of IPRs, core labor standards, and application of global environmental regulations. 3.57 In the first instance, the likely impact of the above measures on investment and competitiveness in the African EPA-countries would be exclusively, or primarily, on foreign direct investment. Its ultimate magnitude would depend upon (a) the number and the sizes of sectors with potentially attractive profit opportunities that are opened up for foreign investment by participating countries and (b) the presence or absence of other important obstacles to increased investment. Thus, even if broadly and effectively implemented, guarantees of free entry and non-discrimination for foreign investment could be useful but only the first steps for easing current constraints to increased private investment. 3.58 Complementary Competitiveness Reforms. General improvements in the business climate and competitiveness to help raise both foreign and domestic investment rates are likely to be necessary for generating a strong supply-response to improved market access and trade policies in the EPA-signatories as discussed earlier in section 50 I.D. Except in cases where barriers to entry are the only important constraint to increased private investment, even enhanced CARIFORUM type reforms would not address all of the other legal, regulatory, governance, financial sector, and infrastructure problems that have led to low business climate and competitiveness ratings in the African EPA- countries, which would remain as disincentives to both foreign and domestic investment. Hence, unless unilateral steps are taken to further improve business climate and competitiveness, the impact of free entry and non-discrimination would be likely to be small even on FDI and would be insignificant on domestic investment. 3.59 Additional unilateral reforms in the business climate and competitiveness are thus likely to be essential for raising investment rates and increasing productivity. The readiness and ability of individual countries to use full EPAs to leverage a wide range of such reforms will be critical in determining how much a particular country will ultimately benefit. Although proactive governments could equally well carry out the necessary reforms unilaterally, the locking-in of open investment policies and the visible political commitment of entering into full EPAs would be foregone. Using EPAs to Improve Regulatory Environment for Services and Investments 3.60 Given the low levels of domestic investment, access to technology, and managerial skills in most African countries, an important part of the necessary increase in both the level and the efficiency of investment will need to come from greater inflows of foreign direct investment. Moreover, the boost in competitiveness necessary for African economies to reap the full benefits of merchandise trade liberalization under interim EPAs will not materialize without a parallel opening of the services sector. Opening up the services sector would increase its direct contribution to the growth of exports, GDP, and employment in African countries. The indirect contribution of service liberalization would be equally large through improving the quality and reducing the costs of the core infrastructure and business services and thereby helping to raise productivity and lower cost in other sectors. Full EPAs could provide an opportunity for the African countries to improve their investment climate in order to attract FDI particularly in services and export oriented sectors. 3.61 An Instrument for Accelerating Reforms. As in the case of merchandise trade, the liberalization of services imports and investment under EPAs needs to be seen primarily as an instrument for stimulating reforms that are necessary in the long term for increasing global and regional trade integration. All of the reforms involved in full EPAs could also be implemented unilaterally by any country desiring to do so at a time of its own choosing. However, inclusion of trade in services and investment in full EPAs would provide an important opportunity for African EPA-countries to liberalize their service sectors and investment regimes globally and regionally, while obtaining financial and technical assistance from the EU and other donors for building capacity in service sectors for which they make liberalization commitments. In addition, if the EU agrees to fully liberalize the rules of origin for merchandise trade under full EPAs, African countries would also have an attractive market-access incentive for liberalizing services and investment under full EPAs. Because individual countries could act largely independently, it may also be easier to make significant progress in liberalizing imports 51 of services and investment under full EPAs than to advance regional trade integration, which requires coordinated actions by multiple countries. 3.62 Reform Priorities. The relevance and usefulness of full EPAs for specific countries will depend upon the priority for each of the EPA reforms affecting services and foreign direct investment and of the necessary complementary reforms for improving the general business climate as well as upon the country's capacity to implement these reforms. On the one hand, for countries for which catalyzing trade-related reforms for services and investment is a priority, full EPAs could provide a significant opportunity to accelerate, extend, deepen, and lock-in useful reforms with coordinated technical and financial assistance from the EU. On the other hand, for countries for which such reforms are not a priority at present or which are unable to implement these reforms, reliance on the EBA program, an interim EPA (which will remain in effect indefinitely), or GSP for access to the EU market is a reasonable alternative for the time being and participation in full EPAs could be considered at a future date. 52 Annex A: The Caribbean EPA's Approach to the Liberalization of Trade in Services 55 A.1. Liberalization Strategy. The liberalization of services under the Caribbean EPA covers all four GATS modes of service delivery. The liberalization is to be progressive, reciprocal, and asymmetric. The EU commits to eventually liberalize 94% of service sectors in the official WTO sectoral classification list while the target shares for the LDC and non-LDCs in CARIFORUM are 65% and 75%, respectively. The Caribbean EPA also includes a regional preference clause extending to the other CARIFORUM participants in a phased manner whatever trade and related concessions are granted to the EU by any CARIFORUM participant. A.2. Liberalization Objectives. The EU's approach to investment in bilateral treaties has been to generalize to all economic sectors its provisions that were originally designed to govern the commercial presence of service sector firms. The Caribbean EPA adopts this approach. Its main provisions are intended to ensure non-discrimination between domestic and EU investors through commitments on: a. Freedom of entry: Signatory countries are to refrain from maintaining or adopting measures to limit the number of commercial firms (through numerical quotas or other requirements such as economic needs tests) or the total quantity of output. Limitations on the participation of foreign capital in terms of maximum percentage and measures requiring specific types of establishment, such as joint ventures, are also prohibited. b. National treatment: Signatory countries grant to investors from other signatories treatment no less favorable than that they accord to their own investors.56 c. Most-favored-nation treatment: Signatory countries also grant to investors from the other signatories treatment no less favorable than the most favorable that they accord to similar investors from any third country. These broad provisions apply in principle to all goods producing and service activities, including the delivery of services by suppliers from abroad and consumption abroad (GATS Modes 1 and 2 of service delivery) as well as through commercial presence (Mode 3).57 55For a further discussion of the CARIFORUM EPA's provisions concerning the liberalization of trade in services and related issues, see the full report and also German Federal Ministry of Cooperation and Development (2008) 56 Public enterprises are, however, exempted from this non-discrimination clause if their existing regulations are inconsistent with it. 57With the exception of the standard GATS exclusions, that is: audio-visual services, national maritime cabotage, and national and international air transport services. To the GATS exceptions, the EU adds the mining, manufacturing, and processing of nuclear materials and the production of or trade in arms, munitions, and war material. 53 A Flexible Design and a Phased Approach to Implementation A.3. Minimal Initial Commitments. The Caribbean EPA adopts a flexible, phased approach to implementation, with separate country-by-country lists of commitments and initial commitments to liberalization only in the sectors where this can be readily implemented. The initial implementation under the Caribbean EPA of the above broad principles governing foreign direct investment is limited. The sectors to which the agreed principles will be applied are determined by a positive list approach, that is: the sub- sectors to be liberalized must be explicitly listed in the agreement or its annexes, and the EPA's investment provisions do not apply to any sectors that are not listed. (In contrast, in a negative list approach, which, despite its name, is generally the more liberal option, investment in all sectors is automatically liberalized except for those sectors which are explicitly excluded.) The initial liberalization, the specific commitments of which are presented in an annex, reflects largely, if not entirely, the binding of the countries' current legislation and pre-existing commitments with little, if any, new liberalization. New liberalization in the treatment of foreign direct investment will depend upon future rounds of negotiations with the EU. A.4. The Caribbean EPA's main text also contains detailed provisions governing the commercial presence of services providers in six sectors: computer services, courier services, telecommunications, financial services, international maritime transport, and tourism services. Other sub-sectors to be liberalized for commercial presence and foreign investment under the Caribbean EPA are identified in its annexes, which contain detailed country-by-country lists of the sectors to which its general principles, and various limitations on these, apply. Again, the CARIFORUM countries' initial commitments appear to be largely a re-statement of the legislative provisions that are currently in effect. A.5. Plans for Further Negotiations and Progressive Liberalization. Schedules for reviewing progress in implementation and initiating negotiations of subsequent phases of liberalization are integral parts of the Caribbean EPA. The legal framework for investment, the investment environment, and actual flows of investment between the parties are to be reviewed before the end of the third year and at regular intervals thereafter. There are no references to sanctions for non-compliance with agreement, and the monitoring of its implementation is to be part of the general review of the legal framework and environment for investment and the assessment of actual investment flows. The EPA includes a rendezvous clause committing the parties to entering into further negotiations on additional sectors within five years to gradually reach the agreed targets for market access. The Regulatory Framework A.6. Competition. The Caribbean EPA also includes a broad range of provisions discussing competition, the regulation of services, and most of the other more important trade-related regulatory issues. It stresses the importance of free and undistorted competition in trade relations and states that practices that prevent or substantially lessen 54 competition and abuse of market power are incompatible with the proper functioning of the EPA. The CARIFORUM countries acknowledge the importance of their regulatory bodies' being independent from commercial interests and indicate their intention of putting in place within five years the necessary laws and institutions to address anti- competitive practices within their jurisdictions. However, the EPA does not contain any binding substantive commitments to new policy actions. A.7. Regulation of Services. The Caribbean EPA places heavy emphasis on effective and transparent regulation, including independent regulatory bodies, especially for network services like telecommunications. The Caribbean EPA's regulatory provisions address some sector specific issues such as the obligation of universal service, regulation of licensing and inter-connection for telecommunications services, observance of "internationally agreed standards for regulation and supervision in the financial sector," prohibition of "cargo-sharing arrangements in bilateral agreements with third countries concerning maritime transport" to foster competition and avoid cartelization, and sustainable development for tourism. Some sections also contain a sector-specific paragraph on development cooperation and technical assistance. A.8. Trade-Related Regulatory Provisions. In addition, the Caribbean EPA has a number of provisions covering most of the more important trade-related regulatory issues including customs administration and trade facilitation, intellectual property rights, anti-corruption, public procurement, core labor standards, and environmental safeguards. All of the provisions impose minimal demands on the CARIFORUM countries, mostly requiring only adherence to previous international commitments or existing legislation, and appear to be reasonable starting points for LDCs and other low income countries in addressing these issues in bilateral trade agreements with developed countries. 55 References Baunsgaard, Thomas, and Michael Keen. 2005. "Tax Revenue and (or?) Trade Liberalization," IMF Working Paper WP/05/112, International Monetary Fund, Washington, D.C., June. Brenton, Paul, and Mombert Hoppe. 2006. "The Africa Growth and Opportunity Act: Exports and Development in Sub-Saharan Africa," Policy Research Working Paper #3996, World Bank, Washington, D.C. Brenton, Paul, and Caglar Ozden. 2007. "The Effectiveness of the EU and US Trade Prefer- ences for LDCs: Rules of Origin in the Clothing Sector," in European Union Trade Politics and Development: Everything but Arms Unraveled, G. Faber and J. Orbie, editors, Routledge: Oxon. Brenton, Paul, Mombert Hoppe, and Erik von Uexkull. 2007. "Evaluating the Revenue Effects of Trade Policy Options for COMESA countries: the Impact of a Customs Union and an EPA with the European Union". Paper prepared as part of the World Bank ­ COMESA Joint Work Program for Regional Integration and EPA negotiations. Washington, DC. Carrère, Céline, and Jaime de Melo, 2008. "Meeting the Challenges for a Successful Conclusion of Economic Partnership Agreements." Mimeo prepared for the World Bank, CNRS, CERDI, University of Geneva and CEPR. Commission for Africa, 2005. Our Common Interest: Report of the Commission for Africa, Chapter 8, "More Trade and Fairer Trade". www.commissionforafrica. org, London, March. German Federal Ministry for Economic Cooperation and Development. 2008. "Cariforum EPA and beyond: Recommendations for negotiation on Services and Trade related Issues in EPAs." Trade Programme at Gesellschaft für Technicnische Zusammenarbeit (GTZ). Eschborn, Germany. July. Naumann, E. 2008. "Rule of Origin and EPAs: What has been Agreed? What does it Mean? What Next?" Trade Law Centre for Southern Africa, March. Rollo, Valentina and Ian Gillson. 2008. "The Characteristics of Economic Partnership Agreements in Southern Africa: A Brief Analysis of the Merchandise Trade Provisions." Mimeo, the World Bank, Washington, D.C., February. Stevens, Christopher, Sansoussi Bilal, and associates, 2008. "The New EPAs: Comparative Analysis of their Content and Challenges for 2008." Overseas Development Institute, London, and European Centre for Development Policy Management, Brussels, March. 56