The last decade witnessed a huge upsurge in the number of regional trade agreements (RTAs). In the five years since the establishment of the World Trade Organization (WTO) in January 1995, it has been notified of 69 new RTAs. A total of 113 were in force as of December 1999, including 91 RTAs in goods notified under Article XXIV, the WTO rule that allows countries to form trading blocs as long as they do not become more protectionist towards outsiders.

The European Union (EU) is by far the most active WTO member in terms of RTA participation. It is party to 28 of the 91 RTAs in goods notified under Article XXIV. The EU, being at its heart a customs union, is itself a RTA under Article XXIV. In addition, it has 27 bilateral RTAs with third countries. This is a significant increase on the position of ten years ago when the EU was party to 18 RTAs in goods. It is also party to eight of the eleven RTAs in services. In contrast, the United States is only party to three RTAs in total while Japan belongs to none at all.

At the beginning of the 1990s, the EU’s bilateral RTAs in goods fell into two distinct categories: reciprocal RTAs – customs unions or free trade areas providing reciprocal free access to both parties; and non-reciprocal RTAs – cooperation agreements providing duty-free access only for the non-EU partner. At the time, the implicit EU policy was that reciprocal customs unions and free trade areas were essentially reserved for potential EU members while non-reciprocal cooperation agreements were generally for countries outside Europe.

By the end of the decade, the shape and content of EU regionalism had been radically transformed. Not only had the number of RTAs increased substantially but the substance of the agreements had also changed dramatically. First, the fall of the Iron Curtain and the redrawing of Europe’s political map resulted in the mushrooming of bilateral RTAs with the countries of Central and Eastern Europe. The second change was the decision gradually to eliminate non-reciprocal RTAs and set up reciprocal RTAs with countries outside Europe, which are highly unlikely to become EU members. In 1990, the EU had just one reciprocal RTA with a non-European country (Israel). By the end of 2000, five such RTAs will be in force (with Israel, Morocco, the Palestinian Authority, Tunisia and South Africa) and several others are already in the pipeline.

In the meantime, the redrawing of Europe’s political map has led to a veritable outbreak of intra-European RTAs. Today, there are dozens of RTAs in Europe. The European ‘spaghetti bowl’ could be characterized as containing three layers. The nucleus of the system would be the EU and its 20 or so bilateral European RTAs. The second layer would comprise the European Free Trade Area (EFTA) and its dozen bilateral intra-European RTAs. The last stratum would consist of the nearly 30 RTAs among European countries that belong to neither the EU nor EFTA.

The present pan-European trading system could be seen as suffering from three major weaknesses. The first and foremost is the potential for widespread discrimination. A country like the Czech Republic, for example, faces numerous different rules depending on whether it is trading with an EU member, an EFTA member, the Slovak Republic, its partners in the Central European Free Trade Area or various other countries and trade groupings.

The second problem is the investment-deterrent effect associated with ‘hub-and-spoke’ systems, where the ‘hub’ country has free access to all ‘spokes’ but each ‘spoke’ country has only free access to the ‘hub’.

The last problem relates to EU enlargement and the status of RTAs between EU candidates and certain other non-EU countries. This issue derives from the fact that the EU is a customs union, which implies that its members cannot be parties to bilateral RTAs. Two situations may cause problems. One is that nearly all the 13 EU accession candidates participate in bilateral RTAs with other candidates that may join later than they do. The other is that some of the EU candidates are parties to bilateral RTAs with countries that are not (at least, at the moment) EU candidates.

A solution that I have proposed to these problems would be the creation of a Pan-European free trade area (PEFTA) incorporating all the countries of Europe that belong to the WTO. But obviously, given the central role of the EU in the system, such a solution cannot be envisaged unless, at the same time, the participants describe a clear vision of the pan-European political architecture. A possible model for the latter could run as follows. All the nations of Europe should belong to PEFTA if and when they become WTO members. And all the members of PEFTA fulfilling general criteria and willing and able to adhere to its aims would become members of the EU.

In this scheme, the EU would comprise three, instead of the current two, levels of integration. First, there would be the customs union, which would contain the current EU members plus the candidates meeting the required conditions. Second, there would be the single market, which would include the current EU members plus some of the present candidates able and willing to conform to the relevant EU legislation. And third, there would be the monetary union, in which membership would remain subject to meeting the convergence criteria.

Even without such a scheme, EU regionalism has clearly taken a new direction in recent years with its full-blooded RTAs with countries outside the European continent. And as a result of parallel initiatives by the United States, the world trading system now finds itself on virgin territory.

The EU and the United States have both implemented RTAs with neighbouring developing countries designed to ‘lock in’ their economic reforms and foster regional stability. Both have also taken important steps towards preferential trading agreements with countries outside their immediate vicinity that offer important market potential. At the same time, the EU is also considering establishing a series of RTAs with its traditional African, Caribbean and Pacific (ACP) partners with a view to fostering economic reforms and reinforcing traditional ties with potentially significant markets.

If the Free Trade Area of the Americas (FTAA) initiative and the idea of EU/ACP free trade areas are implemented, we could witness the emergence of two major ‘hegemon-centred’ trading blocs by the end of the first decade of the twenty-first century. One would be focused on an EU of 25-plus members, encompassing most non-EU European countries (including, perhaps at some stage, countries like Russia and Ukraine), African countries from the North to the South of the continent, and some countries in the Near East. The other, centred on the United States, would comprise the entire American continent.

The EU-centred bloc is, and would continue to be, a hub-and-spoke system of bilateral agreements; in contrast, the US-centred bloc is, under NAFTA, and would continue to be under the FTAA, a single agreement. The crucial point is that both blocs are, and would remain, free trade areas rather than customs unions. Free trade areas tend to require more costly rules of origin than customs unions.

It need not happen but it is certainly possible that the two emerging trade blocs could become closed to one another or even antagonistic. The best way to lay aside existing worries would be to undertake an ambitious agenda of ‘global free trade in our time’. In the meantime, it is of the utmost importance that the EU and the United States, the two ‘hegemons’ of the system, agree to strengthen WTO rules on RTAs in order to minimize the discrimination they allow and maximize their potential for trade liberalization.

André Sapir