Publicaties van de faculteit Rechtsgeleerdheid van de Universiteit Utrecht. Deze publicatie is beschikbaar via de publicatiewebsite van het G.J. Wiarda Instituut, Utrechts Instituut voor Rechtswetenschappelijk onderzoek.

Economic Partnership Agreements and Regional Integration among ACP Countries

ECSA Seventh Biennial International Conference, May 31-June 2, 2001, Madison, Wisconsin future EU-ACP relations future EU-ACP relations future EU-ACP relations future EU-ACP relations 1future EU-ACP relations future EU-ACP relations future EU-ACP relations future EU-ACP relations

Gerrit Faber

Economic Institute, Utrecht University, Utrecht, The Netherlands, E-mail: g.faber@econ.uu.nl

The author thanks Hein Roelfsema for his critical comments.

1. Introduction

What would be the conclusions of an unbiased reader, going through the second title - Economic and trade cooperation - of the Cotonou Agreement? The following conclusions would probably be drawn. First, there is a strong emphasis on WTO conformity. Apparently, the wish not to deviate from WTO rules is so strong that WTO conformity has been upgraded to an objective of this co-operation (art. 34). One would expect it to be a principle at best. Articles 36 and 37 reiterate WTO conformity. Second, the transition period between 2000 and 2007 will be used to negotiate economic partnership agreements (EPAs) between (groups of) ACP countries and the EU that ' shall aim notably at establishing the timetable for the progressive removal of barriers to trade between the Parties in accordance with the relevant WTO rules' (art. 37.7). This hints at the reciprocity of the trade preferences under the EPAs which will require the ACP countries to give EC exports free entry on their markets. Third, there is strong emphasis on regional integration among ACP countries: there will be support for regional integration in the preparatory phase (art. 37.3) and the negotiations on the EPAs will take into account regional integration processes within the ACP (art. 37.5). It might be concluded that the EU and the ACP have a clear idea of the organisation of their future relationship: WTO compatible, reciprocal free trade agreements (FTAs) preferably built on regional integration among ACP countries. However, careful reading reveals that other outcomes are possible: non-least developed ACP countries 'might decide that they are not in a position to enter into EPAs' (art. 37.6). Alternatives will be examined for these countries. For least-developed ACP countries, 'duty free access for essentially all products' will be introduced (art. 37.9). It is striking that the non-reciprocal solutions get comparatively little attention; by going into much more detail the Agreement as such seems to prefer the reciprocal trade arrangements for the post-2007 period.

The option of reciprocity was not an idea that the ACP group enthusiastically accepted. In November 1996, the European Commission published a green paper containing options for the future shape of the EU-ACP trade co-operation (European Commission, 1996). Two main directions could be discerned [1]: coverage of ACP exports by the EU's Generalized System of Preferences (GSP) or the introduction of uniform or differentiated reciprocity. The first option may be attractive for the least-developed ACP, as their preferential position remains the same, for as long as they remain least developed. However, the GSP is unilateral and the preferential margins fall as the general and sectoral competitiveness of the beneficiary countries increases. The ACP group was not in favour of the GSP solution. The group first aimed at maintaining the status quo, and, realising that this position would be unproductive, subsequently accepted the principle of reciprocity, embodied in FTAs (ECDPM, 1998).. Introduction. Introduction. Introduction. Introduction. Introduction. Introduction. Introduction. Introduction

The EPA approach to the renewal of the trade regime of the Lomé Convention presents a number of questions. The first is, whether FTAs between the ACP countries and the EU are likely to produce better results than the non-reciprocal Lomé preferences. Second, even if promising on theoretical grounds, is it likely that (groups of) ACP countries will be able to conclude viable FTAs (or EPAs) with the EU? Which conditions will have to met in order to realise the potential benefits? This paper will address these questions. It provides an economic analysis of the FTA approach for the future EU-ACP relations.

The composition of the paper is as follows. In the second section the background for the new approach is given. The third section presents the economic theory of integration in order to explain the effectiveness of the present and the new approach. The fourth section formulates the conditions that have to be met to make FTAs successful. The fifth section addresses the question whether the new approach will work in practice. Conclusions are drawn in section six.

2. The economic theory of FTAs

Economic theory of regional integration can shed light on the question whether the EPA approach of the Cotonou agreement has the potential to contribute to ACP development. The potential effects of FTAs and customs unions (CU) are best shown by customs union theory. Viner (1950) shows that a CU affects trade in two ways. First, countries gain from preferences because of the trade creation effect. This takes place if a tariff reduction in favour of the partner country allows high-cost domestic production to be replaced by more efficient production imported from the partner country. Second, countries lose from a CU because of trade diversion, which arises if trade shifts from sources that have a comparative advantage to sources that have preferential access to markets. Two extensions of Viner's analysis are relevant. Lipsey (1960) has developed the concept of positive consumption effects that arise if a CU leads to lower prices for consumers. Another extension has been made to account for FTAs in particular. The absence of a common trade policy in an FTA can give rise to indirect trade deflection. This occurs if a member country characterised by relatively high trade barriers, shifts its imports from the world market to the partner, while the latter country - having relatively low trade barriers - enlarges its imports from the world market (Pelkmans, 1997). In practice, this means a shift of tariff income from the first to the latter country.

A country gains from participating in an FTA if the trade creation and the consumption effects outweigh trade diversion and indirect trade deflection. For this to happen three conditions are important. First, FTAs must result in a substantial fall of intra-FTA trade barriers. In that case, the price reduction resulting from preferential access has large trade creation effects. Second, if trade shifts to sources that are efficient (nearly as efficient as the most efficient producers on the world market), trade diversion and indirect trade deflection do not hurt very much. Third, indirect trade deflection will be absent if the high tariff partner simultaneously lowers its tariffs (and other trade barriers) to the level of the low tariff partner. Therefore, countries that join an FTA that consists of efficient producers are likely to gain from preferential trade, particularly if they lower their trade barriers on imports from third countries at the same time. [2]

Venables (2000) analyses the distribution of the gains from regional integration for different groups of countries, particularly poor countries and industrialised countries. He uses Ricardian and Heckscher-Ohlin-Armington models to extend CU theory to countries that are characterised by their factor endowment. Assume that two - poor - countries integrate that are unskilled labour abundant compared to the global average. One of the two is more extreme unskilled labour abundant than the other. After the FTA has been introduced, the extremely endowed member state will suffer trade diversion as its partner will have a comparative advantage in goods that use skilled labour in production. The non-extremely endowed country will gain from trade creation, as is substitutes its production of unskilled labour intensive goods for imports from its partner. The result is a divergence of real income levels. This tendency has been observed in the former East African Community and in the Economic Community of West Africa. Integration among industrialised countries, by contrast, is likely to give rise to convergence. In that case, the extremely endowed country is most skilled labour abundant and thus the richest of the two. Regional free trade will bring trade diversion to the country with the high endowment of skilled labour, which is the richest. Gains from trade creation will mainly flow to the poorer, less extremely endowed partner. The conclusion of this analysis is, that regional integration among poor countries tends to lead to income divergence, which encroaches upon the viability of these FTAs and CUs. The prospect for North/South FTAs is much better. Trade creation will be concentrated in the poor member state and trade diversion in the rich partner.

The trade creation, trade diversion, trade deflection and the consumption effects are static welfare effects of integration. Are these static welfare effects generally considered as being small, in the long run dynamic welfare effects are more important. First, trade may create positive production externalities in developing countries. If companies in these countries can produce more, they can create economies of scale and `learning by doing' effects. Second, preferences will change the competitive environment in which firms operate. For instance, this will reduce X-inefficiencies because of foreign competitive pressures and it will reduce domestic market power of firms.

Regional integration can have effects beyond the static and dynamic welfare effects. These non-traditional effects have in common that they increase the credibility of the countries' policies in ways that cannot be realised through unilateral or multilateral liberalisation (Francois, 1997). Two main non-traditional effects are distinguished: the effects of overcoming time inconsistency and signalling. Time inconsistency occurs if a policy is first best for the country in the long run - it improves welfare - and if there are strong incentives for the government to deviate from this first best policy in the short run. A well-known example in the area of trade policy is a change of the terms of trade. In a free trade situation, the economy will adapt and the allocation of factors of production will change in such a way that a maximum welfare will be attained. This occurs through falling factor rewards in, e.g. the import competing sector, and rising factor rewards in the export sector. However, a government will get strong incentives to intervene, e.g. to use trade barriers to compensate the falling wage level, to `protect the employment in the sector' according to the political rhetoric. Structural Adjustment Programmes are another case in point. [3] An FTA can help to overcome time inconsistency by introducing reciprocity in reforms that makes withdrawal costly as it will be followed by loss of market access in partner countries. A reciprocal regional trade arrangement, by making the cost of even a small deviation from an agreed trade liberalisation large, either by forcing the country to exit from the agreement or by having members punish the deviating country, makes it easier to overcome small temptations that culminate in a greatly distorted economy overall. It can be concluded that an FTA will contribute to overcoming time inconsistency of policy reforms

if the first best policy in a country is time inconsistent and

if the FTA binds the government to the time inconsistent policy and if a policy reversal - in extreme cases an exit of the FTA - has a cost that is higher than the benefits of the policy reversal.

The second non-traditional effect is signalling. By entering into an FTA a government may convey a message to foreign actors, e.g. investors, about the economic policy preferences of the government (make it clear that there is now a liberal government), about the condition of the economy (e.g. the export sector is competitive) or the future relationships of the country with its neighbours (conflicts will be solved by peaceful means). This effect will occur if two conditions are fulfilled:

there has to be a significant information asymmetry: e.g. there is substantial doubt about the government's commitment to reform;

there has to be a significant cost to enter the FTA which cannot be recovered at leaving, e.g. the negotiating cost, adaptation of laws and regulations, restructuring of the economy.

Ethier (1996) shows that FTAs that aim at locking in the less developed partner's economic reforms stimulate foreign direct investment. He argues that preferences and regional integration may have a dynamic effect on economic development. If reforms are successful and attract an inflow of FDI, countries gain competitiveness. However, if more countries reform simultaneously, they compete for FDIs. Trade preferences by an industrialised country to a particular developing country ensures that the FDI flows to that partner. Therefore, reciprocal preferences ensure the success of reforms and offer the incentive for other developing countries to follow.

Economic theory of CUs and FTAs contributes to the explanation for the lower than expected results from the non-reciprocal Lomé preferences. First, the positive static and dynamic effects of regional integration derive from falling prices and more competition in the importing country. In a situation of non-reciprocal trade preferences, the welfare improving trade creation and consumption effects will not occur in the preference-receiving countries as these countries do not reduce trade barriers, as a result of which prices of imported products would fall. On the contrary, developing countries have relatively high trade barriers and the ACP group in particular, as will be shown below. The convergence that could be the result of North/South integration (Venables 2000), was not allowed as far as the effect has to come from trade creation in the ACP countries. The main beneficial effects from the non-reciprocal preferences were to come from trade creation and diversion in the EU, which would raise import demand for the products from the ACP. [4] This is in fact an adapted version of the Infant Industry Argument. Like that argument, the reasoning for non-reciprocal preferences may backfire as well. If countries create industries in which they do not have a comparative advantage but which are only based on preferential market access, preferences have to be kept in place indefinitely and create vested interests to keep them in place.

Second, in the case of non-reciprocal preferences the dynamic effects in preference receiving countries are limited. As competition is not increased by lower trade barriers, economies of scale in export industries are the only source of potential dynamic effects. So on theoretical grounds, it was not realistic to expect very large effects on economic development, unless there were other strong factors, such as large preferential margins. [5] Furthermore, time-inconsistency and lack of credibility of economic policies cannot be solved by non-reciprocal preferences, as sanctions on deviations are completely lacking and no sunk cost are incurred to obtain the preferences.

The conclusion of this overview of economic theory is, that there are potential gains from regional integration between the EU and ACP countries. It is also clear that these gains cannot be taken for granted. In order to reap these gains certain conditions have to be met, be it in terms of initial conditions or in terms of policies that are pursued after the FTA has been introduced. Furthermore, integration theory mostly assumes that there is a high degree of information among partners, that free trade is immediately introduced for all goods and that policy switches (e.g. between tariff revenue and tax income) are without costs.

3. The sequencing of EPAs

The text of the Cotonou agreement shows that the parties have a clear sequence in their minds: groups of ACP countries integrate among themselves and these RIAs negotiate with the EU on EPAs. Without a proviso art. 35.2 of the agreement says: 'Économic and trade co-operation shall build on regional integration initiatives of ACP States, bearing in mind that regional integration is a key instrument for the integration of ACP countries into the world economy'. [6] This sounds attractive for three reasons. First, EPAs with groups of ACP countries lowers the number of these agreements which makes them more efficient instruments of co-operation for the EU, as every EPA has its negotiations, institutions, financial and technical support. Second, the ACP countries are in a better negotiating position if they operate in groups. This is not only a matter of power. Given the limited negotiating capacities of individual ACP countries, pooling their resources may increase the effectiveness of their proposals and the way these are handled. Third, the EU runs less risk of being considered as a neo-colonialist power if it negotiates with groups of ACP states instead of with individual countries, that by definition are very small compared to the EU.

However, the proposed sequence of local trading blocs that conclude EPAs with the EU also has (potential) shortcomings. The first shortcoming is, that the economic benefits of integration among (very) poor countries are often very small and, perhaps even more serious, may be unevenly distributed. In the last section the probability of a divergence in economic development among the members of a South/South RIA was reviewed. It is precisely on such RIAs that the EU wants to construct the EPAs. Table 1 gives some basic data on the RIAs that might conclude EPAs. With the exception of CARICOM, all RIAs mentioned have extremely low per capita levels of income. Of the 77 ACP countries, 37 belong to the least developed. They fit very well into Venables' model, which predicts that integration among these countries concentrates the trade diversion effects in the extremely endowed countries (i.e. the country with the highest share of low skilled labour) and the benefits in the less extremely endowed member. In addition, the economic effects of integration among countries in Sub-Saharan Africa (SSA) will most probably be very small. Trade among SSA countries is extremely low, and this also applies for intra-trade inside SSA RIAs. The share of intra-trade as a percentage of total exports was 8.6 for ECOWAS, 7.0 for COMESA, 5.1 for SADC and 10.4 for UEMOA in 1993 (Yeats, 1999: p. 28). In the foreseeable future, the prospects for SSA economic integration are not bright, the reason being that 'SSA countries have the export capacity to meet only a very small share of regional import needs' (Yeats, 1999: p. 67). Their exports are highly concentrated in a few products that are of very limited importance as imports of other SSA countries.

It might be argued that the economic effects of integration are of secondary importance, as the political objectives are the primary arguments for integration among developing countries. This might be the case, but it is equally true that if integration produces divergent economic development or small net benefits in the presence of concentrated and visible restructuring costs, support for integration will dwindle. This may bring the integration process to a standstill or even give rise to military conflicts. [7] To expect that these RIAs will become the foundations for FTAs with the EC may prove a delusion for many RIAs, particularly in SSA.

A second shortcoming - not isolated from the first one - is that there is a lot of overlap in membership of RIAs among developing countries. This applies for Africa in particular. Annex tables 2 and 3 give the membership of SSA RIAs. ECOWAS, UEMOA and CEMAC have together a membership of 22 countries. Eight of these countries are member of two RIAs. In East and South Africa, the overlapping membership is even more extreme. COMESA, EAC, IOC, IOR-ARC, SADC and SADC have a combined membership of 25 countries (not counting non-ACP countries). Of these, 8 countries have double membership, four have a triple membership and three are member of four RIAs. [8] Overlapping membership lowers the credibility of RIAs (Matambalya, 1999). By being member of more than one RIA, the countries concerned show that they want to keep open alternative options, which enable them to trade-off opportunities for regional schemes as they seem fit. However, regional integration is a reciprocal process by nature; it can only be successful if there is a firm commitment of the participating countries. Costs and benefits of regional integration are likely to be unevenly distributed over time. If a country has a difficult time as a result of restructuring of economic activities, a high degree of political commitment is required to defend the RIA domestically. Solidarity from the other member countries is helpful. If governments can easily put their stakes in a different RIA, the credibility of a particular RIA is undermined, which makes it a less attractive negotiating partner for an EPA with the EC. A related problem is, that most RIAs in SSA (CEMAC, COMESA, ECOWAS, SADC, SACU, UEMOA) have a CU or higher form of integration as final objective. A CU has a common external trade policy which excludes overlapping membership. [9] The fact that there is so much overlapping membership casts further doubts on the commitment of many countries to the objectives of the RIAs to which they belong.

A third shortcoming of ACP RIAs is that many of them have a history of lacking implementation of agreed liberalisation measures or policy reversals by member countries without sanctions. The complete collapse of the EAC in 1977 is a well-known example of non-implementation. The three members have picked up their integration plans again in the second half of the nineties. SADC members had concluded a Trade protocol in 1996 to establish a FTA in 2004. At the turn of the century, the protocol had not been ratified by a sufficient amount of countries. The COMESA FTA was launched in October 2000, many years behind schedule. The ECOWAS treaty envisaged a common market in 1989. However, this RIA "has failed to achieve free trade in goods, let alone in factor services" (World Bank, 2000) In many instances, countries have reinstated trade barriers on intra-RIA trade that had been removed earlier. Faber and Roelfsema mention such cases for Kenya, Tanzania, Zambia and Zimbabwe in the second half of the nineties (Faber and Roelfsema, 2000). As a result of this history of non-compliance with agreed objectives and measures, RIAs in SSA do not constitute lock-in mechanisms that improve the credibility of governments with respect to economic reforms.

In the Pacific region, the eight EU partners do not have a formal regional agreement or something that resembles such an arrangement. They are members of the South Pacific Forum Insular Countries (FICs) which is part of the South Pacific Forum which is mainly political in nature and supports members in negotiations with the EU. Papua New Guinea is member of the Asian-Pacific Economic Cooperation (APEC), which strives after regional free trade on a non-discriminatory basis. A viable RIA among Pacific ACP countries does not exist at the start of the transition period of the Cotonou Agreement. Waiting for such a RIA to merge might take a long time.

In the Caribbean region, regional integration has been established in CARICOM. Although the economic relations between its members (intra trade) are relatively small, there are working institutions such as a Caribbean Regional Negotiating Machinery (CRNM) which was established to formulate and implement coherent negotiating positions for multilateral, plurilateral and bilateral negotiations (Nicholls c.s. 2000, p. 1173). CARICOM has working non-reciprocal agreements with Canada, Colombia, Mexico and Venezuela.

The upshot of this section is, that it is unlikely that the sequence of, first, regional and then interregional integration, that underlies the trade section of the Cotonou Agreement, will prove realistic for most ACP countries. The likelihood of a divergent economic development among very poor partners in a RIA, the weak economic effects of integration in SSA and the lack of commitment to RIAs and the weak or absent lock-in effects in SSA in particular, do not bode well for the development of reasonably strong RIAs that cover most ACP countries, able to negotiate with the EU. This conclusion is underlined by the absence of a RIA among the eight Pacific ACP countries.

4. FTAs between the EU and the ACP?

Does this mean that the form of trade co-operation that is preferred in the Cotonou Agreement is bound to fail? This conclusion is too quick. Given the EU's strong preference for EPAs, a different strategy should or will be followed: the sequence will have to be reversed. Instead of waiting for viable RIAs to emerge as negotiating partners, EPAs should be used to reinforce regional integration among ACP countries. The EU will have to create its negotiating partners, with the exception of CARICOM. The question to be addressed in this section is, in which way and under which conditions EPAs will stimulate integration in ACP regions.

It can be argued that regional integration among groups of ACP countries will be stimulated by the simultaneous establishment of FTAs with the EC. The first argument is, that a North-South FTA has a tendency towards concentrating trade creation in the southern partners, which will produce a tendency of convergence with the EU, according to Venables' approach. Given the right economic policies - this assumption will be discussed below - viable economic activities will be stimulated. Second, regional integration is not costless, as economic theory assumes. In practice, poor countries refrain from integration because short term cost precede the long term benefits. Furthermore it is the lack of funds for redistribution that often hampers RIAs among poor countries. In an EPA, the EC will provide funds for the ACP RIAs that can be used to sustain social safety nets, to improve regional infrastructures and to increase institutional capacities to participate in regional integration and to change tax systems away from border taxes. Third, an ACP RIA that concludes an FTA with the EC will be a credible commitment mechanism. ACP countries that have opted for membership of this RIA cum FTA will have to surrender their trade policy powers to the RIA that negotiates with the EC. Overlapping membership will be impossible. Reversals will be sanctioned as the process is reciprocal in nature. Thus, a national government can much more easily stick to its economic reforms if these are embedded in such an EU/ACP FTA. These arguments will be further discussed below.

The first argument depends on the realisation sufficient beneficial economic effects. It is difficult to give an estimate for such a diverse groups of countries, where reliable statistical data are scarce. For some groups, trade with the EU is very small and is likely to remain so. For these countries the effects of EPAs will be relatively small. This applies in particular for the Pacific ACP countries. [10] A number of studies have been ordered by the European Commission in order to get an idea of the size of the static effects of EPAs. These studies were reviewed by McQueen (1999). The studies use different methodologies, which further complicates their comparability. Taking this into account, it seems fair to conclude that for most potential EPAs the static effects are small for all EPAs and small and negative for an EPA with SADC and EAC. For the Pacific and Caribbean region, the discontinuation of the commodity protocols for sugar, bananas and rum are much more important in quantitative terms than the static welfare effects of EPAs. The static effects do not only depend on the reciprocal liberalisation of trade between the ACP and the EU. Much will depend on the lowering of trade barriers that hamper imports from third countries into the ACP countries. Many of these countries have relatively high trade barriers. E.g., the average tariff of many Sub-Saharan African countries was higher than 20 per cent, while the incidence of non-tariff measures stood at 46 per cent at the end of the 1980s (Rodrik, 1998). Tables 4 and 5 give more detailed country information on tariffs and other charges per SSA country at the end of the nineties. CARICOM is implementing a CET reduction programme which will bring the number of tariff bands down to three while reducing the height of the tariffs. Still, the maximum rate will still be 20 per cent (Nicholls c.s. 2000). So there is sufficient scope for lowering tariffs on imports from the rest of the world which will diminish trade diversion, thus contributing to a better net effect. This means that ACP countries entering into EPAs should continue to participate in the WTO based multilateral liberalisation process.

Most economists agree that dynamic effects from integration will be more substantial than the static effects. As far as economic integration increases competition and the size of markets, production will get more efficient. It will rationalise (state) monopolies and force domestic firms to produce more efficiently. Small developing countries like the ACP countries in particular would benefit from these effects. This effect can only be realised if the trade liberalisation is combined with improved domestic market institutions that monitor competition and ensure that regulation does not segment markets. The EU can play an important role in this issue by supplementing trade liberalisation with deeper economic and political integration. Experience with FTAs in Central and Eastern Europe shows that FTAs in many cases include the `export of institutions and regulations'. Vogel (1995) shows that in a climate of trade liberalisation, countries will harmonise their standards towards the level of the countries with the highest standards. [11] If the EU takes an active stance not only in economic, but in legal (competition) and technical (standards) co-operation as well, the benefits from FTAs will be increased considerably.

The second argument mentioned at the start of this section is, that the EU will be involved in ACP integration through the financial support for economic restructuring, policy reform and capacity building. The gains from integration can only be reaped after a shrinking of inefficient lines of production. This process will be painful, the more so as employment and welfare levels are low in most ACP countries. It can reasonably expected that production will decline in the heavily protected manufacturing and service sectors. The restructuring is less painful the quicker production and employment expand in other sectors. In theory, this could happen in those sectors where the ACP have comparative advantages and the capacity to expand production quickly: labour intensive manufactures and agricultural products. An improvement of market access to the EU for these product categories would assist short term results. Apart from this, funds will be necessary for shifting government income from tariffs to taxes, for social safety nets, for capacity building in order to improve the implementation of public functions that are required if markets are to function for the benefit of the citizens at large and for retraining. The EU does not stand empty-handed. The financial protocol to the Cotonou Agreement supplies the 9th EDF with Euro 15.2 billion. Part of this is to be used for macro-economic and structural reforms, for sectoral policies and reforms and institutional development and capacity building, according to art. 60 of the Agreement. Under the Lomé Conventions, the EU was also involved in structural adjustment operations. Furthermore, the EU has supported economic integration among developing countries with financial and technical aid through the Cross Border Initiative (CBI) and Programme d'Assistance Régionale Intégré (PARI). By combining financial support and EPAs, one of the reasons why economic integration among developing countries has been so difficult to realise, may be taken away, viz. the absence of funds to bridge periods of painful adjustment in production and policies, to finance capacity building in the integration institution and its member states.

The third argument in favour of EPAs argued that an FTA increases the credibility of governments in their structural reform policies. As a result of long term reciprocity, liberalisation of trade in goods and services, of capital flows and the reform of institutions and regulation cannot be rescinded as easily as was the case in Structural Adjustment Programmes (Collier and Gunning, 1995) This anchor function is considered to be an important advantage of FTAs over non-reciprocal preferences (Francois, 1997). For an individual ACP country, a linking up with the EU makes a RIA credible: its institutions gain authority - they are negotiating with the EU - and the EU may sanction infringements on obligations. By linking ACP RIAs to EPAs, the RIA becomes much more attractive in terms of economic benefits and credibility. The question is, of course, whether the EU is willing to support sustained adjusters among the ACP and to sanction policy reversals by individual ACP countries or their regional institution? The answer is not evident. The EU has not a tradition of supporting ACP countries that seriously adjust their economy. Faber and Roelfsema (1997) concluded that ' Lomé instruments are not biased in favour of countries that are sustained adjusters'. This conclusion draws on research with respect to the distribution of benefits from aid and trade co-operation over ACP countries in the first half of the nineties. In her analysis of the way the EU stimulates ACP countries to meet their obligations with respect to human rights, good governance and democracy, Arts (2000) finds a number of cases where aid under Lomé was suspended. In the nineties, there is a growing willingness to resort to this kind of action, underpinned by more explicit articles in the Conventions. Trade preferences have not been used as a sanction up to now, although this was clearly possible under Lomé IV-bis. [12] The Cotonou Agreement opens the same possibility in art. 96. Thus, there is a growing preparedness to use sanctions against ACP countries not meeting their obligations in the areas of human rights, democracy and good governance. Under an EPA, the reciprocal nature of the agreement would make sanctions in the area of economic policies and trade a logical outcome. However, in practice this may turn out to be less clear. The obligations under an FTA agreement may be annulled by regulatory barriers or hidden subsidies. Deep integration may solve this problem. A different question is, whether the EU will be willing to play the role of hegemon in the proposed EPAs. This role implies not only the preparedness to apply negative sanctions in case an ACP country or ACP RIA does not meet its obligations under the EPA. What will the EU do in case external shocks destabilise an ACP economy to such an extent that it jeopardises its economic reforms introduced under the aegis of the EPA? Will the EU supply the ACP state with massive and immediate support, as the USA did with Mexico after the latter country ran into the peso crisis in 1994 just after NAFTA had been launched? As a number of ACP countries are tied to the Euro, and as most ACP economies are relatively small, the EU could do this. However, decision making in these issues has to be very quick. It is not certain, to say the least, that the EU could act effectively in such a situation. Furthermore, the USA has a higher stake in Mexico's economic performance than the EU has in ACP countries economic development. The conclusion is, that the function of EPAs as a commitment mechanism strongly depends on the preparedness and capacity of the EU to use sanctions to enforce the obligations under the EPAs and to protect the ACP countries against external shocks that undermine their capacity to meet their obligations.

5. Conclusions

New initiatives are needed to increase the effectiveness of the trade policy instruments the EU uses in its policy of development cooperation with the ACP countries. Merely giving non-reciprocal preferences has not worked very well. The main reason for this failure is that non-reciprocal preferences have not improved supply conditions in developing countries. FTAs in the shape of EPAs can give a new impetus to production and trade. For this strategy to succeed, the ACP and EU should avoid trade diversion and deflection and create as much trade as possible. This requires MFN liberalisation together with regional integration. This paper shows that EPAs could have important benefits for ACP economies by creating competitive market conditions, leading to dynamic welfare effects and by reinforcing economic reforms.

The success or failure of the new approach depends on a number of factors that have been analysed in this paper. This paper argues that the expectation underlying the Cotonou Agreement that effective negotiating partners in the shape of ACP RIAs will develop in the short term will not come true, with the exception of the Caribbean region. In the Pacific region, no ACP RIA is in the making. In SSA, regional integration suffers from overlapping membership, lack of commitment and small or negative economic effects. As a result, the EU will have to use EPAs as an instrument to stimulate viable RIAs among ACP countries in SSA. This puts a number of responsibilities on the EU in order to create successful EPAs. For the full benefits of regional free trade to arise, deep integration is necessary. Harmonisation of competition policy, of regulations with respect to product norms and standards is called for. Second, substantial funds will have to be provided by the EU for social safety nets, for capacity building in order to improve the implementation of public functions that are required if markets are to function for the benefit of the citizens at large and for retraining. In order to make EPAs successful, the market access for sensitive ACP export products should be improved. Third, the EU will have to be willing to act as the hegemon in EPAs, sanctioning infringements on obligations and supporting ACP countries that have run into externally caused difficulties threatening their participation in EPAs. If EPAs are to work as commitment mechanisms that effectively lock in ACP policy reforms, ACP countries should run the risk of losing substantial benefits if they do not meet their obligations, and feel secure that they will be supported in case of external shocks. As the negotiations on EPAs have not started at the moment of writing, there is no certainty about the willingness of the EU to take up this role. However, given the trends in sanctioning human rights violations, in supporting regional integration among developing countries, the relative size of the EDF and given the high stake the EU has set on EPAs, the new role that the EU will have to play to make EPAs successful, is not completely unrealistic.

References

Amjadi, A. and Yeats, A.J. (1995), Nontariff Barriers Africa Faces: What Did the Uruguay Round Accomplish and What Remains to be Done? Washington D.C.: The World Bank PRW Paper No. 1439

Aryeetey, E. and A.D. Oduro (1996), Regional Integration Efforts in Africa: An Overview, in: J.J. Teunissen (Ed.), Regionalism and the Global Economy. The Case of Africa. The Hague: Fondad, pp. 11-50

Babarinde, O.A. (1994), The Lomé Convention and Development. Avebury: Aldershot

Blokker, N., G. Faber and K. Hellingman, Trade Policies and the Developing Countries in a Changing World Economy, in: G. Faber (ed.), Trade Policy and Development. The Role of Europe in North-South trade; a multidisciplinary approach. The Hague: Rotterdam University Press, pp. 279-298

Collier, P. and J.W. Gunning (1995), Trade Policy and Regional Integration: Implications for the Relations between Europe and Africa, in: The World Economy. Vol. 18, no. 3, pp. 387-410

Credit (1998), Study on the Economic Impact of Introducing Reciprocity into the Trade Relations between the EU and EAC Countries. Nottingham: School of Economics

Drabek, Z. and S. Laird (1997), The New Liberalism: Trade Policy Developments in Emerging Markets. Geneva: World Trade Organization

ECDPM (1998), Comparing the ACP and EU Negotiating Mandates. Lomé Negotiating Brief No. 3. Maastricht: ECDPM

Idem (1999), The EC's Impact Studies on Regional Economic Partnership Agreements. Lomé Negotiating Brief No. 5. Maastricht: ECDPM

European Commission (1996), Green Paper on relations between the European Union and the ACP countries on the eve of the 21st century. Brussels: European Commission

Idem (1997), Communication to the Council and the European Parliament. Guidelines for the negotiation of new cooperation agreements with the African, Caribbean and Pacific (ACP) countries. Brussels: European Commission

Ethier, W.J. (1996), Regionalism in a Multilateral World, mimeo

Faber, G. (2000), Towards a Pan-European-Mediterranean Free Trade Area? in: P. van Dijck and G. Faber (Eds), The External Economic Dimension of the European Union. The Hague: Kluwer Law International

Faber, G. and H.J. Roelfsema (1997), The Lomé Convention and Development in Sub-Saharan Africa, in: European Foreign Affairs Review, Vol 2, pp. 376-386

Faber, G. and H.J. Roelfsema (2000), Evaluation of the Cross-Border Initiative, Volume 3: Research Report (DRN-ECO-IDD-Synergie, Antwerp, 2000)

Francois, J.F. (1997), External Bindings and the Credibility of Reform, in: A. Galal and B. Hoekman (Eds), Regional Partners in Global Markets: Limits and Possibilities of the Euro-Med Agreements. Cairo: ECES and London: CEPR, pp. 35 -49

Grossman, G. and E. Helpman (1994), Protection for Sale, in: American Economic Review. Vol. 84, pp. 667-690

Grossman, G. and E. Helpman(1995), The Politics of Free-Trade Agreements, in: American Economic Review. Vol. 85, 833-850

International Monetary Fund (1996), Liberal Trade and Regional Integration: A Prescription for Africa, in: IMF Survey. Vol. 25, nr 11, pp. 184-186

IDS (1998), Study on the Economic Impact of Introducing Reciprocity into the Trade Relations between the EU and CARICOM/Dominican Republic. Sussex: Institute of Development Studies

Imani Development LTD, Study on the Impact of Introducing Reciprocity into the Trade Relations between the EU and the SADC Region. Quatre Bornes: Imani Development Ltd

Johnson, O.E.G. (1995), Regional Integration in Sub-Saharan Africa, in: Journal of European Integration/Revue d'Integration Européenne, Vol. 18, no. 2/3, pp. 201-234

Lipsey, R.G. (1960), The theory of Customs Unions: a General Survey, in: Economic Journal, Vol. 70, pp. 496-513

Netherlands Economic Institute (1998), Introducing reciprocity into the trade relation between the EU and the Pacific ACP countries. Rotterdam: Netherlands Economic Institute

Matambalya, F.A.S.T. (1999), Post-Lomé IV EU-ACP Relations and Prospects for Regionalised and Multi-tier EU-ACP Co-operation Arrangements (Friedrich Ebert Stiftung, Dar es Salaam, 1999, mimeo)

Ng, F. and A.J. Yeats (1996), Open Economies Work Better: Did Africa's Protectionist Policies Cause Its Marginalization in World Trade? Washington D.C.: World Bank PRW Paper No. 1636

Pelkmans, J. (1997), European Integration. Methods and Economic Analysis. Harlow: Addison Wesley Longman

Planistat Belgique (1998), Etude de l'impact économique sur l'UDEAC/CEMAC de l'introduction de la réciprocité dan les relations commerciales UE-ACP. Brussels: Planistat Belgique

Rodrik, D. (1998), Trade Policy and Economic Performance in Sub-Saharan Africa. Cambridge Mas.: National Bureau of Economic Research

Stevens, Chr. (1999), EU Economic Policy towards Developing Countries after Lomé, in: P. van Dijck and G. Faber (Eds), The External Economic Dimension of the European Union. The Hague, London and Boston: Kluwer Law International

Thomas, R.H. (1997), The WTO and Trade Cooperation Between the ACP and the EU: Assessing the Options. ECDPM Working Paper No. 16, Maastricht: ECDPM

Viner, J. (1950), The Customs Union Issue. New York: Carnegie Endowment for International Peace; London: Stevens & Sons

Vogel, D. (1995), Trading Up. Consumer and Environmental Regulation in a Global Economy, Cambridge Ma.: Harvard University Press

Worldbank (2000), Trade Blocs Policy Research Report, at www.worldbank.org/wbiep/trade/RI_map.html

Yeats, A.J. (1994), What are OECD Preferences worth to Sub-Saharan Africa? Washington D.C.: The World Bank.

Yeats, A.J. (1999), What can be expected from African Regional Trade Arrangements? Some Empirical Evidence. Washington D.C.: The World Bank

Annex tables

Table 1 Selected economic indicators of ACP countries

Africa	GNP p.c. at PPP	HD Index
	Dollars 1999
Angola*	632	0405
Benin*	886	0411
Botswana	6032	0593
Burkina Faso*	898	0303
Burundi	553	0321
Cameroon	1444	0528
Cape Verde*	3497	0688
Central African Rep.	1131	0371
Chad*	816	0367
Comoros*	1360	0510
Congo	n.a.	0507
Congo (RDC)*	879	0430
Ivory Coast	1546	0420
Djibouti*	n.a.	0447
Equatorial Guinea*	n.a.	0555
Eritrea*	1012	0408
Ethiopia*	599	0309
Gabon	5325	0592
The Gambia*	1492	0396
Ghana	1793	0556
Guinea*	1761	0394
Guinea Bissau*	595	0331
Kenya	975	0508
Lesotho*	2058	0569
Liberia*	n.a.	n.a.
Madagascar*	766	0483
Malawi*	581	0385
Mali*	693	0380
Mauritania*	1522	0451
Mauritius	8652	0761
Mozambique*	797	0341
Namibia	5369	0632
Rwanda*	n.a.	0382
Sao Tomé & Principe*	1335	0547
Senegal	1341	0416
Seychelles	10381	0786
Sierra Leone*	414	0252
Somalia*	n.a.	n.a.
South Africa	8318	0697
Sudan*	1298	0477
Swaziland	4200	0655
Tanzania*	478	0415
Togo*	1346	0471
Uganda*	1136	0409
Zambia*	686	0420
Zimbabwe	2470	0555
Caribbean
Antigua and Barbuda	n.a.	0833
Bahamas	n.a.	0844
Barbados	n.a.	0858
Belize	4492	0777
Dominica	4825	0793
Dominican Republic	4653	0729
Grenada	5847	0785
Guyana	3242	0709
Haiti*	1407	0440
Jamaica	3276	0735
St Kitts and Nevis	9801	0798
St Lucia	5022	0728
St Vincent and the Grenadines	4667	0738
Suriname	n.a.	0766
Trinidad and Tobago	7262	0793
Pacific
Cook Islands	n.a.	n.a.
Fiji	4536	0769
Kiribati*	3186	n.a.
Marshall Islands	n.a.	n.a.
Micronesia	n.a.	n.a.
Nauru	n.a.	n.a.
Niue	n.a.	n.a.
Palau	n.a.	n.a.
Papua New Guinea	2263	0542
Solomon Islands*	1949	0614
Tonga	4281	n.a.
Tuvalu*	n.a.	n.a.
Vanuatu*	2771	0623
Samoa*	n.a.	0771

*= belonging to the groups of least developed countries

Table 2 Membership of RIAs in West and Central Africa

	ECOWAS	UEMOA	CEMAC	Overlap
Benin	*	*		2
Burkina Faso	*	*		2
Cameroon			*
Cape Verde	*
Central African R.			*
Chad			*
Congo			*
Côte d'Ivoire	*	*		2
Equatorial Guinea			*
Gabon			*
Ghana	*
Guinea	*
Guinea-Bissau	*	*		2
Liberia	*
Mali	*	*		2
Mauritania	*	*		2
Niger	*	*		2
Nigeria	*
Senegal	*	*		2
Sierra Leone	*
The Gambia	*
Togo	*

Tabel 3 Membership of RIAs in South and East Africa

	SADC	SACU	COMESA	EAC	IOC	IOR-ARC	Overlap

Angola*	*		*				2
Botswana*	*	*					2
Burundi			*
Comoros
DR Congo*			*
Djibouti			*
Egypt			*
Eritrea			*
Ethiopia			*
Kenya			*	*			2
Lesotho*	*	*					2
Madagascar			*		*	*	3
Malawi*	*		*				2
Mauritius	*		*		*	*	4
Mozambique*	*					*	2
Namibia	*	*	*				3
Rwanda			*
Seychelles	*		*		*	*	4
South Africa	*	*				*	3
Sudan			*
Swaziland	*	*	*				3
Tanzania*	*		*	*		*	4
Uganda			*	*			2
Zambia*	*		*				2
Zimbabwe	*

Table 4 Average applied tariffs and other duties and charges in West and Central Africa, in percentage terms

	Average applied Tariffs	Other duties & charges
Benin	6.7
Burkina Faso	23.8
Cameroon	15.5	26.7
Cape Verde
Central African R.	12.6	22.6
Chad	15.5	24.3
Congo	12.9	25.1
Côte d'Ivoire	15.9
Equatorial Guinea	18.2	18.5
Gabon	7.6	10.5
Ghana	n.a.
Guinea
Guinea-Bissau	n.a.
Liberia
Mali	16.7
Mauritania
Niger	18.1
Nigeria
Senegal	28.7
Sierra Leone
The Gambia
Togo	19.2

Tabel 5 Average applied tariffs and other duties and charges in South and East Africa, in percentage terms

	Average applied Tariffs	Other duties & charges
Angola	21.1
Botswana	12.3
Burundi	12.3
Comoros	30.0'
DR Congo
Djibouti
Egypt
Eritrea
Ethiopia
Kenya	14.0'	16.0
Lesotho
Madagascar	16.3
Malawi	15.5
Mauritius	21.5
Mozambique	13.1	5.0
Namibia	12.3
Rwanda	11.3	18.0
Seychelles	28.0'
South Africa
Sudan
Swaziland	12.3
Tanzania*	25.6	0
Uganda	9.0'	10.0
Zambia*	16.0
Zimbabwe	24.1
' = simple average