Trade Regionalism in Europe: Towards an Integrated Approach*

Transcription

1 Journal of Common Market Studies March 2000 Vol. 38, No. 1 pp Trade Regionalism in Europe: Towards an Integrated Approach* Université Libre de Bruxelles and EC 1. Introduction The political and economic maps of Europe have changed beyond recognition since the European Community (EC) was established 40 years ago. On the political side, although the Iron Curtain has fallen and the foundations of an ever closer union among the peoples of Europe invoked by the EC founding fathers have been laid, Europe remains divided. Currently, Europe defined here as the set of states belonging to the Council of Europe consists of 40 countries, which fall into three groups: 15 states belonging to the European Union (EU); 12 EU candidates; 1 and 13 other countries. 2,3 Back in 1960, when the main European divide was between East and West, the current 40 countries formed only 33 *An earlier version of the article was presented at the CEPR/Yrjö Jahnsson Foundation/ZEI Conference on Regionalism in Europe: Geometries and Strategies After 2000 held in Bonn on 6 7 November 1998 and at the WIIW 25 Year Anniversary Conference on Shaping the New Europe held in Vienna on November I am grateful to several conference participants for helpful comments, particularly Jorge Braga de Macedo, Carl Hamilton, Jim Rollo, Jürgen von Hagen, Helen Wallace and Alan Winters. I have also benefited from suggestions by Iain Begg and an anonymous reviewer. Views expressed here are mine alone and should not be attributed to the European Commission. 1 The ten countries with Europe Agreements (Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic and Slovenia), plus Cyprus and Turkey. 2 Albania, Andorra, Croatia, the former Yugoslav republic of Macedonia (FYROM), Iceland, Liechtenstein, Malta, Moldova, Norway, Russia, San Marino, Switzerland and Ukraine. 3 As of December 1998, three European countries did not belong to the Council of Europe and are therefore largely ignored in this article: Belarus, Bosnia and the former republic of Yugoslavia. Georgia, which became the 41st member of the Council of Europe in April 1999, is also ignored., 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA

2 152 separate entities: 24 in the West (the same number as today); and 9 in the East (compared to 16 today). 4 The redrawing of Europe s political map has been accompanied by an explosion in the number of regional trade agreements (RTAs) inside the continent. In 1960, there were essentially only three RTAs among European countries: two in Western Europe, namely the EC and the European Free Trade Association (EFTA); one in Eastern Europe, i.e. the Council for Mutual Economic Assistance (CMEA); and none between Eastern and Western European countries. Today, there are more than 90 RTAs inside Europe: about 15 in Western Europe; about 20 in Eastern Europe; and roughly 55 between Eastern and Western European countries. The purpose of this article is to analyse the spread of RTAs inside Europe, the problems it generates and possible solutions to these problems. The article starts by providing a detailed description of the complex web of regional trade agreements currently in force inside Europe and of their evolution since the early 1960s. It then examines the problems associated with the existence of nearly 100 RTAs inside Europe and offers some possible solutions. A central element in all these solutions is the prospect for further EU enlargement. II. The Network of European RTAs in 1998 and its Expansion since 1960 More than ever, the current pan-european trading architecture is a hub-andspoke system of bilateral regional trade agreements, a point already noted by Richard Baldwin in his insightful study Towards an Integrated Europe published in The EC is the focal point of the system, being party to bilateral RTAs with 20 of the 25 other European countries (belonging to the Council of Europe), which are often themselves connected to each other by bilateral or plurilateral RTAs. This architecture is the direct descendant of the trading system established in 1973 in Western Europe, after the first EC enlargement, but it also owes much to recent developments in Eastern Europe. The Situation in 1998 Table 1 attempts to describe the full complexity of the system of regional trade agreements in force inside Europe at the end of It shows that: 4 Of the nine old entities, five have remained untouched (Albania, Bulgaria, Hungary, Poland and Romania), three (Czechoslovakia, Soviet Union and Yugoslavia) have split up and one (East Germany) has merged with another entity. The term Eastern Europe is used throughout to refer to all the 16 current Member States of the Council of Europe that are (partly or wholly) the successors of these entities.

3 TRADE REGIONALISM IN EUROPE 153 All but two of the 40 European states (members of the Council of Europe) are parties to at least one RTA with another European country (the two exceptions are Albania and Russia). Most states are parties to several RTAs. Altogether there are 90 RTAs currently in force among European states concerning trade in goods. 5 All but four of these 90 RTAs are bilateral agreements. The only plurilateral RTAs are: the European Community (EC), with 15 members; the European Free Trade Association (EFTA), with four members (Iceland, Liechtenstein, Norway and Switzerland); the Central European Free Trade Area (CEFTA), with seven members (Bulgaria, Czech Republic, Hungary, Poland, Romania, Slovak Republic and Slovenia); and the Baltic Free Trade Area (BFTA), with three members (Estonia, Latvia and Lithuania). Three of these plurilateral RTAs are free-trade areas (FTAs): EFTA, CEFTA and BFTA. The EC is a customs union (CU). All but seven of the 86 bilateral RTAs between European countries are free-trade areas. The only bilateral customs unions are: between the EC and five partners (Andorra, Cyprus, Malta, San Marino and Turkey); between the Czech and Slovak republics; and between Liechtenstein and Switzerland. The EC is party to bilateral RTAs with 20 of the 25 states which, besides EC members, belong to the Council of Europe; Liechtenstein and Switzerland are parties to 13 bilateral RTAs; Iceland, Norway and Slovenia to 12; Lithuania to 11; and the Czech Republic, the Slovak Republic and Turkey to 10. The other countries are parties to fewer than ten bilateral RTAs. All but eight of the 79 bilateral FTAs involve Central and East European countries (CEECs). The eight exceptions are: the 4 FTAs between the EC and the individual EFTA members; and the four FTAs between Turkey and the individual EFTA members. All of the 71 bilateral FTAs involving CEECs date from after the collapse of the Soviet Union and the demise of the CMEA in These 71 FTAs divide into 56 East West FTAs with the EC and its long-standing Western partners (EFTA and Turkey), and 15 East East FTAs between former CMEA partners. All but three of the 15 bilateral East East FTAs involve countries that are both parties to a bilateral FTA with the EC. The three exceptions are: the Moldova Romania FTA; the Croatia Slovenia FTA; and the Lithuania Ukraine FTA. Croatia, Moldova and Ukraine are not parties to FTAs with the EC. It is apparent that Europe is riddled with a crisscross of almost exclusively bilateral regional trade agreements. It is also clear that, although the European 5 There are also nine RTAs covering trade in services.

4 Table 1: Regional Trade Agreements Among the 40 Member States of the Council of Europe Notified to the GATT/WTO and in Force in December 1998 Regional Trade Agreements of Members of the Council of Europe Members of the (excluding EC Members) Council of Europe With the European With Other Council of Europe Members (excluding EC Communities Members) (GATT XXIV) GATS V) (Agreements Notified under GATT Article XXIV) Albania Andorra CU Bulgaria (Bul) FTA EI CEFTA member; FTAs with EFTA members Croatia (Cro) FTA with Slo Cyprus CU Czech Republic (Cze) FTA EI CU with Slk; CEFTA member; FTAs with EFTA members, Est, FYROM*, Lat and Lit Estonia (Est) FTA BFTA* member; FTAs with EFTA members, Cze, Slk and Slo FYRO Macedonia FTA* FTA with Cze and Slo Hungary (Hun) FTA EI CEFTA member; FTAs with EFTA members and Tky Iceland FTA* EI (EEA) EFTA member; FTAs with Bul, Cze, Est, Hun, Lat, Lit, Pol, Rom, Slk, Slo and Tky Latvia (Lat) FTA BFTA* member; FTAs with EFTA members, Cze, Slk and Slo Liechtenstein (Lie) FTA EI (EEA) CU with Swi; EFTA member; FTAs with Bul, Cze, Est, Hun, Lat, Lit, Pol, Rom, Slk, Slo and Tky Lithuania (Lit) FTA BFTA* member; FTAs with EFTA members, Cze, Pol, Slk, Slo, Tky and Ukr* Malta CU Moldova (Mol) FTAs with Rom and Tky 154

5 Table 1: (Contd) Regional Trade Agreements of Members of the Council of Europe Members of the (excluding EC Members) Council of Europe With the European With Other Council of Europe Members (excluding EC Communities Members) (GATT XXIV) GATS V) (Agreements Notified under GATT Article XXIV) Norway FTA* EI (EEA) EFTA member; FTAs with Bul, Cze, Est, Hun, Lat, Lit, Pol, Rom, Slk, Slo and Tky Poland (Pol) FTA EI CEFTA member; FTAs with EFTA members and Lit Romania (Rom) FTA EI CEFTA member; FTAs with EFTA members, Mol and Tky Russian Federation San Marino CU* Slovak Republic (Slk) FTA EI CU with Cze; CEFTA member; FTAs with EFTA members, Bul, Est, Lat and Lit Slovenia (Slo) FTA CEFTA member; FTAs with EFTA members, Bul, Cro, Est, FYROM, Lat, Lit and Tky* Switzerland (Swi) FTA CU with Lie; EFTA member; FTAs with Bul, Cze, Est, Hun, Lat, Lit, Pol, Rom, Slk, Slo and Tky Turkey (Tky) CU FTAs with EFTA members, Hun, Lit, Mol, Rom and Slo* Ukraine (Ukr) FTA with Lit* TRADE REGIONALISM IN EUROPE Note: *Not notified to GATT/WTO. Abbreviations: BFTA (Baltic Free Trade Area); CEFTA (Central European Free Trade Area); CU (customs union); EEA (European Economic Area); EFTA (European Free Trade Association); EI (economic integration); FTA (free trade area). 155

6 156 network of RTAs has been in place since about 1960, its scope has vastly increased during the past ten years, mostly as result of political changes in Eastern Europe. Moreover, it is manifest that the EC is the nucleus of the system. The Evolution since 1960 The network of regional trade agreements existing at the moment in Europe is the result of two separate, but complementary, evolutions. One pertains to Western Europe, the other concerns Eastern Europe. In 1960, the 24 countries of Western Europe fell into three groups: the six members of the EC (Belgium, France, Germany, Italy, Luxembourg and the Netherlands); the seven members of EFTA (Austria, Denmark, Norway, Portugal, Sweden, Switzerland and the United Kingdom); and countries belonging to neither the EC nor EFTA (Finland, which became a de facto member of EFTA in 1961, Greece, Ireland, Spain, Turkey and six other countries which are ignored here). 6 Two events in the history of the EC are particularly noteworthy for understanding the evolution of the network of RTAs in Western Europe. One relates to the widening of the EC, the other to its deepening. The first event is the first EC enlargement, which took place in The membership of Denmark, Ireland and the United Kingdom immediately prompted the establishment of free-trade areas between the EC and each of the remaining EFTA members. 7 Since this time, the architecture of RTAs in Western Europe consists of one plurilateral customs union (the EC), one plurilateral freetrade area (EFTA), and a series of bilateral FTAs between the EC and individual EFTA members. Further enlargements took place in 1981 (Greece) and 1986 (Portugal and Spain). The second event is the launching of the single market programme in the mid- 1980s. Krugman (1988) was the first to note that EC deepening might harm EFTA-based firms by lowering the cost of their EC-based competitors. Baldwin (1994, 1995, 1997) also held that the single market programme posed a special threat to EFTA countries since, on average, 60 per cent of their exports went to the EC market. According to Baldwin, this threat triggered a domino effect in Europe as EFTA countries rushed one after the other to apply for EC membership: Austria in 1989; Sweden in 1991; and Finland, Norway and Switzerland in In 1994, Austria, Sweden, Finland and Norway (together with Iceland and Liechtenstein) established the European Economic Area (EEA), a single markettype arrangement, with the EC. One year later, the first three countries joined the EC. 6 Andorra, Cyprus, Iceland, Liechtenstein, Malta and San Marino. 7 The existing FTAs between the EC and the four current EFTA members (Iceland, Liechtenstein, Norway and Switzerland) are their direct successors.

7 TRADE REGIONALISM IN EUROPE 157 The domino hypothesis is confirmed by Sapir (1997), which estimates a series of annual gravity equations on trade flows among Western European countries for the period This study finds that, starting in 1989, both intra-efta trade and trade between the EC and EFTA turned out to be, other things equal, significantly inferior to intra-ec flows. It estimates that during the period , intra-efta trade was, ceteris paribus, times smaller than intra-ec trade, while EC EFTA trade was times smaller than intra- EC trade. By contrast, the evolution of the network of RTAs in Eastern Europe and between Eastern and Western Europe was originally largely tinted by the evolution of the political situation on the continent. In 1960, the nine countries of Eastern Europe (Albania, Bulgaria, Czechoslovakia, East Germany, Hungary, Poland, Romania, Soviet Union and Yugoslavia) 8 were all participants in a regional trade agreement under the aegis of CMEA. At the same time, none of these countries maintained any RTA with the EC or any other Western European trade entity. 9 This situation remained essentially unchanged until With the German unification in 1990 and the demise of the CMEA along with the collapse of the Soviet Union in 1991, the countries of Eastern Europe were able radically to modify their political and economic alliances. The 16 countries of Eastern Europe belonging to the Council of Europe form at the moment two distinct groups. The first group comprises ten EU candidates (the countries with Europe Agreements listed in footnote 1), of which five (Czech Republic, Estonia, Hungary, Poland and Slovenia) have already started membership negotiations. Each of these 10 CEECs is a participant in bilateral FTAs with the EC and with the four EFTA members. In addition, each country belongs to a plurilateral FTA (CEFTA or BFTA) and is party to several bilateral FTAs with other members of the group. Baldwin (1994 and 1997) argues that these new RTAs owe their existence not only to political factors (above all the desire to join the EU), but also to domino effects. He claims that, since nearly all (industrial) trade in Western Europe was duty free, and since Western Europe was paramount for the CEECs, FTAs with the EC and EFTA were essential for levelling the playing field for CEEC exporters. The second group consists of six countries that are not currently EU candidates (Albania, Croatia, FYRO Macedonia, Moldova, Russia and Ukraine). None of these countries is a participant to a bilateral FTA with either the EC 10 8 Albania did not participate in CMEA after 1961, and Yugoslavia only held associate status in CMEA. 9 However, trade between East and West Germany benefited from a special status under the EC Protocol on German Internal Trade and Connected Problems. 10 There is one exception, Macedonia, which is party to a bilateral FTA (not notified to GATT/WTO) with the EC.

8 158 or an EFTA member, but a few are parties to bilateral FTAs with CEFTA members (Croatia, FYROM and Moldova) or BFTA members (Ukraine). Albania and Russia do not belong to any RTA with other Member States of the Council of Europe. No doubt, domino effects will also impact on these countries. The desire to level the playing field for their exporters is bound to prompt these countries to seek an expansion of their European RTA network in the near future. In conclusion, Europe is crisscrossed by a complex network of regional trade agreements that has developed throughout the continent and is focused on the EC. The spreading of regionalism inside Europe seems inexorable as long as access to important markets above all the EC remains relatively unfavourable for exporters located in some countries. 11 III. Problems with the Pan-European Trading System and Possible Solutions The current pan-european trading system described in the previous section suffers from four major problems. The first and most important problem is discrimination. The rule inside Europe is for each of the 26 (the EC plus 25) nations to treat its trading partners differently. Non-discrimination is the exception. Discrimination arises from two factors. Firstly, not all countries have regional trade agreements with all their trading partners. Only one-third of all European country pairs belongs to a freetrade area or a customs union. 12 Secondly, most free-trade areas and customs unions have idiosyncratic features. Discrimination is bad economics because it distorts the allocation of resources in favour of firms and nations with better market access. It is also bad politics because it breeds resentment. The second problem derives partly from the first one. The multiplicity of RTAs creates important administrative trade costs. In the case of FTAs (the prevailing type of RTA in Europe), these costs are further inflated by the existence of rules of origin, which are necessary in order to prevent trade deflection related to different tariff rates among the members. 13 Krueger (1995) points out that, in addition to administrative costs, rules of origin often generate efficiency costs associated with increased external protection. The possibility of 11 Obviously the same logic applies equally to countries outside Europe that also depend heavily on the European and especially EC market for their exports. 12 With 26 countries, 325 (i.e. 26x25 divided by 2) RTAs would be required to ensure that all countries are parties to bilateral RTAs. Given the existence of EFTA, CEFTA and BFTA (which cover 19 pairs of countries), the required number falls to 306. In reality, as indicated in Section II, there are only 86 bilateral RTAs inside Europe. 13 In an FTA, each member sets its own tariff on outside imports. In the absence of rules of origin, imports would come into the FTA through the low-tariff country and be trans-shipped duty-free into high-tariff countries. Rules of origin allow discrimination between goods produced inside the FTA, which qualify for duty-free treatment, and goods produced outside the FTA, which are subject to tariffs.

9 TRADE REGIONALISM IN EUROPE 159 raising external protection via the presence of origin rules also tends to provoke lobbying activity. The third problem is the investment-deterrent effect associated with hub-andspoke systems. As pointed out by Baldwin (1994), the hub-and-spoke FTAs in Europe tend to marginalize the many spokes in favour of the EC hub. This problem is partly alleviated by the creation of a pan-european system of diagonal origin cumulation covering the EC, EFTA and the ten CEEC EU candidates. The last problem relates to EU enlargement and to the status of RTAs between EU candidates and other non-eu countries. The issue derives from the fact that the EC is a customs union and, therefore, that its members cannot participate in bilateral RTAs. Two situations may cause this problem. First, each of the 12 EU candidates is party to bilateral and/or plurilateral RTAs with other EU candidates that may join the EU later than itself. Potential situations of this type include: the customs union between the Czech and Slovak republics; the Baltic Free Trade Area among Estonia, Latvia and Lithuania; and the FTA between Bulgaria and Slovenia. Second, some of the 12 EU candidates are parties to bilateral RTAs with countries that are not (at the moment) EU candidates. Examples of this situation include: the FTA between Croatia and Slovenia; the FTA between Moldova and Turkey; and the FTA between Lithuania and the Ukraine. In both situations, a country acceding to the EU would have to withdraw from such RTAs, which could disrupt existing trade flows. Whether or not the other parties to these RTAs would be entitled to compensation under GATT Art. XXIV:6 is another matter. In essence, these problems arise because Europe is neither a mere collection of 40 nations trading with each other according to the sacrosanct principle of nondiscrimination, nor a continental union with 40 Member States. Instead, it is a hub-and-spoke trading system, with the EU as focal point and future EU members as spoke countries, waiting to become a European Union of nearly 30 countries. It seems, therefore, that an obvious solution to all the above problems would be to incorporate as soon as possible all 12 candidate countries into the Union. However appealing it may be, this solution carries two major drawbacks. The first is that it may easily take a decade or so for all the candidates to become EU members. The second drawback lies in the fact that, even after incorporating all the 12 candidates, the EU would still not be isomorphic with the whole of Europe. Several solutions to the problems displayed by the pan-european trading system can be envisaged. The first solution, proposed by Baldwin (1994), would be to set up an Association of Association Agreements (AAA), consisting of all EU members and all the CEECs with Europe Agreements. The AAA would maintain bilateral

10 160 FTAs with individual EFTA members, thereby guaranteeing duty-free trade among the 29 AAA and EFTA members. The second, more radical, solution would be to introduce a regional most favoured nation (MFN) clause in regional trade agreements among European trading nations. In keeping with the language of GATT Art. I, the clause would require that: Any advantage, favour, privilege or immunity granted by any European trading nation to any product originating in or destined for any other European trading nation shall be accorded immediately and unconditionally to the like product originating in or destined for the territory of all other European trading nations. In parallel with Art. I, which applies only to GATT/WTO members, the regional MFN clause should be limited to the sub-set of European trading nations that belongs to the WTO. The introduction of the regional MFN clause would transform the current trading system between the EC hub and 20 spoke countries into a Pan-European Free Trade Area (PEFTA). The third and last solution would go one step further and create a Pan- European Customs Union (PECU), i.e. PEFTA with the EU s common external tariff. 14 The main advantage of this solution over the FTA option would be to remove the need for origin rules inside the area, thereby saving on important costs. At the same time, as the EU Turkey customs union plainly demonstrates, a customs union is not a stable equilibrium unless it involves EU membership. No sizeable country can accept the de facto surrender of its commercial policy to the EU without the serious prospect of rapid accession to the Union. An important issue for all three arrangements (AAA, PEFTA and PECU), is the range of products and trade barriers covered. One option would be to restrict free trade to industrial goods, and to suppress only tariffs and quantitative restrictions. This relatively narrow approach, excluding ranges of products (agricultural goods and services) and trade barriers, would be in line with many of the existing European RTAs. But obviously, this option may be neither feasible nor desirable. In the end, it is clear that there is an inevitable trade-off between the size and the depth of the proposed trade arrangement. The larger its size, the more shallow the extent of integration. For the sake of non-discrimination, the new pan- European trade arrangement based on regional MFN status should include as many countries as possible, including all EU candidates. At the same time, these countries should not be held back in their effort toward EU membership. 14 Jim Rollo has pointed out to me that the establishment of a continental European customs union has already been envisaged by Aghion et al. (1992). However, their purpose was to ensure the permanent removal of the Cold War divisions of Europe and an irreversible commitment to market development by Central and Eastern European countries, rather than to solve the maze constituted by the current pan-european trading architecture.

11 TRADE REGIONALISM IN EUROPE 161 What does this imply for the pan-european architecture? A possible model, incorporating ideas put forth by Pisani-Ferry (1998) and Emerson (1998), would run as follows. All the nations of Europe should belong to PEFTA, if and when they are WTO members. All those fulfilling general criteria and willing to adhere to its aim, would become members of the European Union. The latter would comprise three, instead of the current two, integration levels: (1) the customs union, which would comprise the current EU members and the present candidates meeting the required conditions; (2) the single market, which would include the current EU members and some of the present candidates able and willing to conform to the relevant Community legislation; and (3) the monetary union, in which membership would remain subject to meeting convergence criteria. This scheme offers two main advantages. First, PEFTA would ensure the end of discrimination although this may apply only to trade in industrial products. Second, a three-tier EU would allow rapid enlargement of its membership, thereby removing political and economic uncertainty for the current candidates, while at the same time preserving the dynamics towards greater integration. These two features derive from a basic principle: the need to adapt or create pan- European institutions reflecting the diversity of situations. At the same time, the proposed scheme also raises a number difficult questions. What of nations such as Russia, the Ukraine or the FR of Yugoslavia which may not meet the conditions for WTO membership for quite a while and would therefore remain excluded from PEFTA for several years? What would be the implication of PEFTA for the transatlantic relationship if Russia were to join it? What would be the implication of PEFTA for other EU regional trading partners and for the multilateral trading system? What of Turkey which already has a customs union with the EU but may not yet fulfil some of the conditions for membership of the Union? These questions are clearly difficult but, to my mind, they do not preclude a PEFTA solution. The issues raised in this article have become all the more urgent with the end of the war in Yugoslavia and the necessity to find solutions for the development and integration of Southeast Europe. They require a frank and open debate. Correspondence: André Sapir ECARE, Université Libre de Bruxelles 39 Avenue Roosevelt 1050 Brussels, Belgium

12 162 References Aghion, P., Burgess, R., Fitoussi, J-P. and Messerlin, P.A. (1992) Towards the Establishment of a Continental European Customs Unions. In Flemming, J. and Rollo, J. (eds) Trade Payments and Adjustment in Central and Eastern Europe (London: Royal Institute of International Affairs). Baldwin, R.E. (1994) Towards an Integrated Europe (London: Centre for Economic Policy Research). Baldwin, R.E. (1995) A Domino Theory of Regionalism. In Baldwin, R.E., Haapranta, P. and Kiander, J. (eds) Expanding European Regionalism: The EU s New Members (Cambridge: Cambridge University Press). Baldwin, R.E. (1997) The Causes of Regionalism. The World Economy, Vol. 20, pp Emerson, M. (1998) Redrawing the Map of Europe (Basingstoke: Macmillan). Krueger, A.O. (1995) Free Trade Agreements Versus Customs Unions. NBER Working Paper No (Cambridge, MA: National Bureau of Economic Research). Krugman, P.R. (1988) EFTA and EFTA Occasional Paper No. 23 (Geneva: European Free Trade Association). Pisani-Ferry, J. (1998) Dealing with Diversity. In Zysman, J. and Schwartz, A. (eds) Enlarging Europe: The Industrial Foundations of a New Political Reality (Berkeley: University of California Press). Sapir, A. (1997) Domino Effects in Western European Trade, CEPR Discussion Paper No (London: Centre for Economic Policy Research).