I am delighted to be with you tonight and honoured to share some of my thinking with all of you and with Dean Anderson and Professor Merit Janow. As you know Professor Janow is a member of the WTO Appellate Body and as Director-General, I must remain neutral and I am not supposed to try to influence her opinion in any way. Tonight I hope I will!

When I was first invited to speak at Columbia on multilateral versus bilateral free trade agreements I immediately thought of my good friend Professor Bhagwati and was reluctant to embark on a topic on which he has already said so much. However, with the current suspension of the WTO negotiations under the Doha Agenda and the flurry of bilateral activity around the world, I decided to accept tonight's challenge and dig deeper into what motivates this activity.

Regional trade agreements have always existed. Indeed, one of the first international legal instruments to leave its trace in history was the commercial treaty between the Egyptian Pharaoh Amenophis IV and the king of Alasia during the XIV century BC. This treaty exempted Cypriot traders from customs duties in exchange for the importation of a certain quantity of copper and wood.

Nothing has fundamentally changed since: at the beginning of the XXI century countries continue to negotiate preferential trade agreements. However, the impact of these agreements into the world trading system has led us to agree that they be notified to the WTO where they are to be checked for consistency with WTO rules.

  

The upsides of preferential trade agreements

Since the creation of the GATT and its successor the WTO, 362 regional trade agreements have been notified to the WTO, of which 211 are currently in force. But if we take into account agreements that are in force but have not been notified, those signed but not yet in force, those currently being negotiated, and those in the proposal stage, we arrive at a figure of close to 400 agreements that could be implemented by 2010.

It is also worth noting that there is a growing trend away from the traditional concept of integration among natural trading partners in geographically contiguous countries, towards cross-regional trade agreements among countries across regions and hemispheres. Examples abound, among them EFTA-Chile, US-Australia or EC-South Africa to name but a few.

Regional trade agreements among developing countries are also increasing. Of those notified to the WTO and currently in force, 15 were notified during the 50 years of the GATT, while the other 23 have been notified in the 11 years since the establishment of the WTO. In addition, the majority of these agreements, currently in force but not notified to the WTO, are between developing countries. Key developing countries — China, India or Brazil — appear almost to be competing among each other to establish themselves as regional trade agreement hubs. The result of this race is an increasing number of South-South trade agreements under negotiation.

The numbers have increased but more importantly their coverage has broadened. Agreements with services provisions are increasingly common. And an increasing number of agreements contain trade provisions in areas not regulated multilaterally.

A recent World Bank study, Global Economic Prospects, estimated that the percentage of trade taking place under preferences, may vary from 40% to 15% according to the caveats used. This is quite a blow to the “Most Favoured Nation” principle which is one of the cornerstones of the WTO. These numbers are impressive but the question we must ask ourselves is why so many countries are accepting rules and disciplines at the bilateral level that they are not prepared to accept at the multilateral level and whether and how these agreements can be beneficial.

Because of similarities in interests and often more common values, bilateral trade agreements can go into new areas such as investment, competition, labour standards or environment, where there is no consensus among WTO Members. Fewer parties means that preferential trade agreements can be concluded within a shorter period of time. Short-term political or geo-strategic consideration may often lead countries to enter into regional trade agreements with a limited number of partners, discriminating against all other WTO Members.

Particularly with regard to services, preferential access can create a so-called “first-mover advantage”, due to the fact that location-specific sunk costs of production are important in many services sectors. Accordingly, preferential access in such sectors as telecommunications or financial services can result in a long-term advantage for an inferior supplier, even if preferences are subsequently extended to other countries.

For developing countries negotiating with more powerful developed countries, there is usually the expectation of exclusive preferential benefits, as well as expectations of development assistance and other non-trade rewards. They are also viewed as an instrument to get 'brownie points' and gain an advantage over other WTO Members.

Bilateral trade agreements are also useful for negotiators to learn how to negotiate and become comfortable with the difficult terminology of trade, thus contributing to building and reinforcing a country's trade institutions. Many regional trade agreements have been the bedrock for peace and greater political stability. Finally, they may be used as instruments for domestic reform in areas where the multilateral system offers a weaker leverage.

But regional trade agreements have inherent limitations. If mismanaged or not adequately integrated in, and subordinated to the multilateral trade system, they can nullify their positive contribution to our world economy.

Distortions, downsides, risks and limits of regional trade agreements

By definition, preferential trade agreements create a discriminatory environment for non-parties. These may see their exports fall as products are increasingly sourced from parties within the preferential trade agreement and they might experience a loss of FDI as investment is channelled into the members that have preferential access to a larger market. But, preferential trade agreements can also lead to the diversion of trade among their partners if imports from an economically inefficient regional trade agreement partner displace more competitive imports produced elsewhere. Only a few preferential trade agreements actually result in totally free trade among the parties. Often restrictions on sensitive products remain in place and studies have shown that restrictive rules of origin can result in exporters using the multilateral rates rather than preferential rates, especially if the preferential margin is small.

The conclusion of regional trade agreements can also create an incentive for even further discrimination, when countries outside the agreement quickly try to conclude agreements with others that are within it so they are not excluded. This has been called the “domino” or “bandwagon effect” and is the reason behind much of the regional trade agreement activity seen in Asia recently. There is also the phenomenon of “addictive regionalism” whereby countries attempt to create preferential relationships with all their major trading partners. Under this scenario, some countries, particularly the developing countries with small domestic markets, risk being left on the sidelines and further marginalized.

The proliferation of regional trade agreements can greatly complicate the trading environment, creating a web of incoherent rules, and intricate rules of origin. An increasing number of WTO Members are party to ten or more regional trade agreements, most of which for a given Member, contain agreement-specific rules of origin. This complicates the production processes of suppliers who may be obliged to tailor their products for different preferential markets in order to satisfy rules of origin. It also complicates life for customs officials who are obliged to assess the same product differently depending on its origin, thus compromising the transparency of the trading regime. Borrowing the expression used by Professor Bagwhati — this is where we begin to have a real “spaghetti bowl” of twisted rules of origin.

Regional trade agreements can divert negotiating energy and resources from multilateral fora – which is particularly serious for developing countries with limited capacities; we are already seeing evidence of this where given the suspension of the WTO negotiations, staff in many of our Members is being reassigned to their regional trade agreements departments.

In addition to their distortive impacts, preferential trade agreements have real limitations: some issues can simply not be handled at all in regional trade agreements. Take for instance, negotiations to eliminate or reduce trade distorting agricultural subsidies, or fisheries subsidies. There is no such a thing as a “bilateral” farmer or a “bilateral” poultry and a “multilateral” farmer (or poultry?). Subsidies are given to farmers for all their poultry production. The same is true for rules on anti-dumping.

Bilateral trade agreements in services tend to exclude or only partially cover those areas where the larger partners are confronted with internal resistance to foreign competition, let alone genuine reform: in other words, such agreements tend to do little to liberalize trade in sensitive sectors. This is not surprising insofar as the coordinating Ministries of Trade are confronted with some heavyweights — Finance, Telecommunications, Labour, Transport, Health or Education — for which 'multilateralism' does not rank high on their priority list, in particular if compared to 'stakeholder interests' which may be more conveniently pursued in a bilateral context.

Bilateral dispute settlement arrangements cannot replace the multilateral WTO dispute settlement system, the benefits of which derive from the general application and implications of the WTO Dispute Settlement Body rulings.

Preferential trade agreements may also lead to the creation of political alliances, where the price that must be paid by a developing country for signing a regional trade agreement with a developed country is that it must support the broader policy stance of that country in the WTO or elsewhere. In this sense, bilateral agreements do not rectify power imbalances between partners. It is all very well if your name is Brazil, China, EU, India or US, but for Ghana, Cambodia or Peru, the bilateral leverage is much smaller than the multilateral one.

Another limitation to regional trade agreements resides in the self-erosion of offered preferences: a market access preference given to a first partner in a regional trade agreement loses its value when the same market access is offered to another partner for another regional trade agreement. At the same time, those countries that are excluded suffer increasing discrimination as the number of preferential partnerships continue to multiply. This problem is particularly pronounced in Hubs and Spokes situations where the spokes are vulnerable to future trade deals by the hub. Take the example of the Economic Partnership Agreement between the EU and the ACP countries. The ACP's strong reliance on preferential market access in the EU may be affected with the EU signalling intentions to expand its preferential partners.

Finally, regional trade agreements often act as disincentives to multilateral negotiations given the costs involved with the phasing out of preferences. Some countries conduct 90% or more of their trade with preferential partners and thus run the risk of developing strong domestic lobbies resistant to any liberalization of tariffs at the multilateral level which will undermine their preferences. In the area of services, it is now clear that a number of Members have entered into bilateral trade agreements that contain significant improvements over their existing multilateral commitments in both the scope and the depth of market liberalization. These often go well beyond GATS offers on the table in the on-going round of negotiations. For instance, through a number of its bilateral agreements, the United States has gained significant market access on services, compared to what the US 's bilateral partners have committed, or are willing to offer, at the WTO.

One wonders whether the ongoing regional trade agreements hyperactivity has not incited some Members to make minimal offers so as to have further negotiating chips to offer in various regional trade agreement negotiations. Under the concept of competitive liberalization, aren't bilateral agreements supposed to transform participating countries into multilateral free traders? Those taking part in bilateral agreements should be prepared to lead the way and show support for multilateralism by narrowing the gap between regional trade agreement commitments and GATS offers. In any event, the benefits that developing countries can get in return for their services commitments, is much greater in the WTO than in bilateral agreements.

While regional trade agreements can be used as laboratories for new issues such as labour rights, investment, competition or open-skies agreements, it is also in these agreements that countries may lose rights and privileges negotiated collectively in the WTO: access to medicines, data collection prescriptions or special and differential provisions, to name but a few.

From the above it is clear that while regional trade agreements have the capacity to complement the multilateral trading system, they can also be a source of trade diversion and difficulty. Mindful of these strengths and shortcomings, the old GATT fathers and, more recently the WTO Members, negotiated disciplines relating to the way bilateral or regional trade accords should be designed with a view to them not being trade distortive, that they should complement the WTO system and bring about their full positive potential.

  

The position of the WTO vis-à-vis regional trade agreements

The GATT and now the WTO recognize the conditional right of Members to form regional trade agreements and to the extent necessary, to set aside some WTO obligations. The WTO also recognizes that regional trade agreements are exceptions to the multilateral principle of the WTO and therefore, when challenged, it is for the regional trade agreement Member to justify its WTO compatibility.

The WTO imposes three types of substantive conditions for regional agreements to be WTO consistent. First with respect to the overall impact of the regional trade agreements vis-à-vis other Members. In free-trade areas and customs unions, there is the obligation not to raise barriers to trade with third parties. This is quantifiable in terms of tariffs, but less easy to measure in terms of other trade regulations such as standards or rules of origin. For instance a restrictive rule of origin, which requires a high percentage of regional components, could penalise suppliers outside the regional trade agreement.

Second with reference to what we call the “external requirement”. A free trade agreement cannot lead to higher imports duties for its members while a customs union must harmonise the external trade policies of its members and compensate affected non-members accordingly.

Thirdly, there are conditions dealing with the “internal dimension” of regional trade agreements: both tariffs and other restrictive regulations of commerce must be phased out on “substantially” all trade. Again the tariff component can be quantified, but it is harder to determine in the case of other restrictive trade regulations as there is no agreed definition of the term.

As a general principle we can say that WTO authorizes regional trade agreements the operation of which should not lead to situations where the non-regional trade agreement members would “pay the price” of internal preferences.

In order to ensure coherence between the WTO and regional trade agreements, regional trade agreements are to be “promptly” notified to the WTO and reviewed by its peers before the regional trade agreement is implemented. In practice, regional trade agreements are usually notified after the application of preferential treatment, often with a considerable delay. Furthermore, the consensus rule on WTO decision-making has resulted in the review process of the WTO Committee on Regional Trade Agreements being blocked. Only one regional trade agreement has been explicitly considered by the committee as being WTO consistent. So in case of disagreement the only available route is the WTO dispute settlement mechanism. But are WTO panels and the Appellate Body the best forum to assess the legal, economic and political value of any specific regional trade agreement?

Finally, the so-called Enabling Clause allows for regional trade agreements among developing countries to be subject to much less stringent disciplines than those imposed by the GATT/WTO. The exact parameters of those disciplines and the lack of notification and transparency of such agreements is becoming a source of concern as more important players plunge into this route.

  

Future avenues

Regional trade agreements are now a growing feature of the landscape which is here to stay. Thus, bearing in mind that the WTO allows them, the challenge we face today is how to ensure that they contribute to the health of the world trading system by minimising the risks that they constrain global welfare and limit economies of scale. We must ensure regional trade agreements which are complementary — and not a substitute — to the multilateral trading system. We need to work on operationalising the GATT/WTO on regional agreements to ensure effective legal and economic surveillance. It is with this objective in mind that WTO Members decided to include this issue in the agenda of the on-going negotiations under the Doha Agenda.

What can we do to improve the cohabitation of bilateral and multilateral trade agreements? I can think of positive regional efforts such as the APEC Guidelines on regional trade agreements. The APEC's best practices include comprehensiveness (i.e. providing for liberalization in all sectors), transparency, measures to improve trade facilitation, simple rules of origin which facilitate trade, and the option of accession for third parties. They complement the WTO rules and encourage consistency with WTO disciplines.

I believe we also have to deal with the spaghetti bowl of rules of origin. Harmonization of rules of origin that are simple, easy to apply and non-restrictive across different regional trade agreements would simplify trading conditions and contribute greatly to transparency. Alternatively, while aiming towards harmonization, can we think of some form of hierarchy of rules along the pattern of the Anti-Dumping or the Customs Valuation Agreements, which provide for alternative methods for calculating prices when the preferred and usually simplest method is not relevant or effective. Can we think of presumptions, or equivalences, or mutual recognition principles for rules of origin which could be rebutted through burden of proof reversal?

WTO Members have already reached an agreement – which still needs to be formally adopted — on a new transparency mechanism that calls for the notification of new regional trade agreements before the application of preferential treatment. This new transparency mechanism envisages an enhanced role for the WTO Secretariat whereby the Secretariat, on its own responsibility and in full consultation with the parties, shall prepare a factual presentation of all regional trade agreements notified to the WTO. At the moment the process is voluntary. The factual presentation is a means to enhance the transparency of regional trade agreements by providing a systematic view of regional trade agreements' trade liberalization and regulatory aspects. This is a first step in the right direction to ensure that bilateral and multilateral agreements remain friends and never become foes.

Now that the WTO negotiations have been suspended, it becomes essential that bilateral agreements do not become the easy way out. Letting down the WTO is not a solution, given the inherent economic, political and systemic value of the multilateral trading system, in which bilateral agreements can find a positive expression. If the multilateral system dies away, so does the positive potential of regional trade agreements. If you would allow me an analogy with French cuisine, regional trade agreements are the pepper in a good sauce which is the multilateral agreements. Pepper adds taste and can improve a sauce but pepper alone is not tasty, and good pepper in a poor sauce, will not do the trick! Use the wrong recipe and it will be a disastrous dinner.

A strong and modern multilateral trading system coupled with regional trade agreements which amplify rather than undermine its benefits is the mature answer to this generation of new regional trade agreements.

Thank you for your attention.