Critical mass as an alternative framework for multilateral trade negotiations.

Gallagher, Peter ; Stoler, Andrew

 The article posits that, over time, the sense of the World Trade
 Organization's so-called Single Undertaking has been perverted, and
 that the current interpretation requiring every WTO member to be
 obligated by all new Doha Round agreements is a major problem in the
 stalled negotiations. The authors' preliminary research supports the
 idea of conducting international trade negotiations in agriculture on
 the basis of a critical mass framework, where only those WTO members
 accounting for some nominated major percentage of trade would take on
 new obligations. The article recounts how this approach has worked
 before in the General Agreement on Tariffs and Trade (GATT) and the
 WTO, and suggests areas of further research in order to test the
 proposition with respect to agricultural trade. KEYWORDS:
 international negotiations, trade. World Trade Organization,
 agriculture, political economy.

**********

The outcome of the Uruguay Round of multilateral trade negotiations contributed significantly to strengthening and unifying the international trading system. The World Trade Organization (WTO) and the so-called Single Undertaking of its members, wherein all are party to all WTO agreements, participate equally in decisionmaking, and safeguard their rights under a strong central dispute settlement system, is a unique and invaluable institution that is essential to the effective governance of the global economy. The continuing success of the system depends on the WTO being able to show that it can effectively discharge its three principal roles of overseeing the implementation of the trading rules (via multilateral surveillance of the committees and councils overseeing the agreements, the various transparency obligations embedded in those agreements, and the Trade Policy Review Mechanism, or TPRM), settling disputes among members, and serving as a forum for negotiation of new and improved rules and increased trade liberalization.

At the time of this writing, the Doha Round of multilateral trade negotiations is in serious trouble. (1) The negotiators are years late in achieving the targets set out for themselves in the November 2001 declaration. In our view, a large part of the problem relates to the process and to the fact that a decision made in November 2001 to treat ''the conduct, conclusion and entry into force of the outcome of the negotiations as parts of a single undertaking" (2) has been accepted by participants as meaning that all WTO members must be involved in and obligated by the outcomes in all areas of the negotiations.

Given the economic diversity of the WTO's membership and the history of the organization, it makes little sense to expect the Single Undertaking approach to negotiations as it is currently understood to produce good outcomes. Even more important, such an approach does not need to be pursued in order to achieve an outcome with meaningful benefits for the global economy. In fact, insistence on the Single Undertaking approach will likely produce an inferior result in the Doha Round--if a result can ever be achieved on this basis.

The problems with the current approach to the negotiations are particularly acute in the case of trade in agriculture--the sector universally regarded as the most significant element of the Doha Round. In this article, we outline a hypothesis that an alternative framework for international trade negotiations for agricultural products may be likely to produce better outcomes for agriculture and for future trade negotiations more generally.

The argument of this article is as follows:

1. The WTO's Single Undertaking is valuable to the extent that it improves the adherence of developing and (especially) developed country members to all of the agreements.

2. But the Single Undertaking did not eliminate the contradictory treatment of reciprocal trade liberalization by the General Agreement on Tariffs and Trade (GATT). It encouraged members to embed GATT's principle of nonreciprocity for developing countries (including some of the largest trading economies) in a more coherent legal framework ensuring slower, more difficult progress.

3. Agreements that open markets on a reciprocal basis among members are more readily negotiated and are more likely to be robust over time.

4. The Single Undertaking makes it almost impossible to negotiate effective agreements because it gives parties nonreciprocal rights over the content and management of an agreement.

5. Critical mass agreements provide a familiar and practical framework for non-Single Undertaking agreements that would remain consistent with all of the other principles of the WTO. They retain an element of nonreciprocity by offering most-favored-nation (MFN) treatment for nonmembers, but they quarantine its debilitating effects.

A Contradiction in the GATT/WTO System

The GATT/WTO system has struggled, almost from its beginnings in 1947, with an apparent contradiction between its mandate to liberalize markets through negotiation of trade agreements and its objective of universal membership including rich and poor economies. In keeping with its universal vocation, the GATT/WTO system has objectives that sound, at least, redistributive. "Recognizing further that there is need for positive efforts designed to ensure that developing countries, and especially the least developed among them, secure a share in the growth in international trade commensurate with the needs of their economic development." (3) But in a redistributive framework, you cannot ask the poor for cuts in trade barriers if you define them as a payment or a concession made to other economies, including the rich.

From one perspective, this contradiction looks like it is only semantic; a problem due to the mercantilist language that the political economy of trade enforces on the GATT/WTO system. Roughly two centuries of economic consensus that mercantilism fails to capture the national interest has had no impact on language and, most important, on the policy practice, of governments that are still bound in most circumstances to adopt a mercantilist stance in order to secure support for more open markets at home, let alone abroad. Similarly, governments often describe the support that trade brings to the development of developing countries in "redistributive" terms, even when they know well that the gains from trade are not a zero-sum distribution of income, but accrue to both sides of an exchange, even between unequals.

However, the WTO's problem is much deeper than confusion over language. It permeates the activities of the organization, at every turn, creating a stultifying tension between its high ambitions, characterized in the Doha ministerial declaration as both liberalizing and distributive, and the reality that no developing country member is obliged to take

an action that would secure the benefits of liberalization for themselves or their trading partners.

 Recalling the Preamble to the Marrakesh Agreement, we shall continue
 to make positive efforts designed to ensure that developing
 countries, and especially the least-developed among them, secure a
 share in the growth of world trade commensurate with the needs of
 their economic development. In this context, enhanced market access,
 balanced rules, and well targeted, sustainably financed technical
 assistance and capacity-building programmes have important roles to
 play. (4)

Schooled by GATT's five decades of managing this contradiction, the Doha ministerial declaration links its distributive objectives to the market liberalizing mechanism by means of elisions (enhanced access to whose market?), ambiguities ("balanced" how?), and a promise of some aid for trade capacity that has, for once, proved substantial.

Shorn of the drafting tricks, however, the qualifications attached to the Doha work program suggest that no developing country will be obliged to enhance access to its own market in order to secure these benefits for itself or for its trading partners, even on the basis of reciprocity. This is spelled out clearly toward the end of the Doha declaration.

 The negotiations and the other aspects of the Work Programme shall
 take fully into account the principle of special and differential
 treatment for developing and least-developed countries embodied in:
 Part IV of the GATT 1994; the Decision of 28 November 1979 on
 Differential and More Favourable Treatment, Reciprocity and Fuller
 Participation of Developing Countries; the Uruguay Round Decision
 on Measures in Favour of Least-Developed Countries; and all other
 relevant WTO provisions. (5)

The history of GATT echoes through this paragraph, especially in the reference to the 1979 Framework Agreements, adopted around the time that developing countries assumed majority membership of GATT. The framework contains the so-called Enabling Clause (formerly embedded in a pair of 1971 waivers from Article I of GATT) that regularizes unilateral preferences for developing country trade. It broadens developing country access to balance of payments restrictions and to infant industry protection in Article XVIII and confirms a broad-ranging principle of nonreciprocity for developing countries in GATT negotiations.

 (5.) The developed countries do not expect reciprocity for
 commitments made by them in trade negotiations to reduce or remove
 tariffs and other barriers to the trade of developing countries,
 i.e., the developed countries do not expect the developing
 countries, in the course of trade negotiations, to make contributions
 which are inconsistent with their individual development, financial
 and trade needs. Developed contracting parties shall therefore not
 seek, neither shall less-developed contracting parties be required to
 make, concessions that are inconsistent with the latter's
 development, financial and trade needs. (6)

The Framework Agreements mark the first--but it turns out, not the last point in GATT/WTO's history at which the mercantilist idea that trade could be a resource transfer from rich to poor seemed to trump the more orthodox theory that trade's contribution to development consists in rewarding a better allocation of domestic resources.

The vision of trade as transfer explains the framework's repudiation of reciprocal obligations on the part of developing country members. From this point on, agreements that invoke the Framework Agreements (including as does the Doha declaration) should include one-way concessions to developing countries that offer them an apparent terms-of-trade gain for free. (7)

Politically, this project proved to be as naive as it seems: developed countries responded to the framework stipulations of nonreciprocity on the part of developing countries by making no concessions that really mattered on their own part. (8) Economically, too, it was an illusion; the tax on developing country imports that the framework preserves under the cover of nonreciprocity is also a tax on their exports. (9) Developing countries' economic gains from trade can come only from the exploitation of their specialization in production. A reallocation of resources at home and abroad, driven by price signals from imports as well as exports, "automatically" directs the reallocation. So the reciprocal trade agreements framework--albeit adopted by GATT's founders for foreign policy reasons--cooperates closely with trade's engine of economic development.

The Mark I and Mark II Versions of the Single Undertaking

The Framework Agreements' attack on reciprocal trade agreements was far from a complete triumph for two reasons. First, not all developing country governments were duped by the mercantilist prescription for economic policies. From around the time that the Framework Agreements were adopted, and for most of the next two decades, developing countries (and smaller developed countries, including Australia and New Zealand) showed that they understood the benefits of greater openness to trade as compared to the autonomous industrialization policies of the 1950s and 1960s. Until at least the end of the Uruguay Round (or until the financial institutions crisis in East Asian economies in the late 1990s), many developing countries adhered to trade policies described in the Washington Consensus. (10) Even now, despite the post-Washington Consensus rediscovery that there are many "recipes," economic integration still figures prominently in successful recipes. And there has been no evidence of a resurgence of autonomous growth theory.

Second, the developing countries' insistence on nonreciprocity was a strategic failure. It demanded that developed countries adopt a pragmatic attitude toward developing country participation in GATT. The developed countries concluded that nonreciprocity by economies that, in the early 1980s, had small markets was an acceptable price to pay for their (nominal) adherence to the GATT. In fact, the developed countries extended this pragmatic approach to their own obligations, which they tended to treat on an "a la carte" basis. (11) The United States, especially, clung to its provisional application of GATT, resisting pressure (12) to ratify the permanent treaty negotiated in 1955 (the Organization for Trade Cooperation) because its "provisional" application of GATT allowed it to deny an injury test to importers under its countervailing duty laws and to grandfather other GATT-inconsistent measures. But the United States was not alone: the early development of the European Communities saw the introduction of many measures that were doubtfully compliant with GATT provisions, including the system of variable levies on agricultural imports that Uruguay directly challenged in its wide-ranging 1961 Article XXIII dispute and the growing use of export subsidies that were uncontrolled by the provisions of Article XVI. Robert E. Hudec (13) points to the absence of any waiver for a developed country using balance-of-payments restrictions (although developing countries sought many). The greatest of the "pragmatic" solutions, however, was the development after 1961 of the short-term and long-term agreements on cotton leading to the 1974 Multifibre Arrangement that, effectively and in a discriminatory manner, by using quotas otherwise prohibited in GATT, cartelized large segments of the global textile and garment market in order to safeguard developed country industries.

The high-water mark of this a la carte approach to GATT agreements was--it now seems surprising to say--the original Single Undertaking principle adopted at the start of the Uruguay Round in 1986 (the Mark I version of the Single Undertaking).

Today, there is a widespread belief that the Uruguay Round's Single Undertaking always meant the same as it does now: a process that automatically obligates all WTO members to be bound by all of the WTO agreements. In fact, it did not. When the contracting parties (CPs) of GATT decided in 1986 that "the launching, the conduct and the implementation of the outcome of the negotiations shall be treated as parts of a single undertaking," (14) they actually had quite the opposite intention.

The Punta del Este ministerial meeting that launched the Uruguay Round took place after difficult and protracted debates in Geneva, where many GATT CPs vigorously opposed negotiations in GATT on trade in services. At the Punta del Este meeting, the key developing countries--led by Brazil and India--continued to insist that negotiations on trade in services were inappropriate as a topic to be dealt with in GATT. This position led to a division of the ministerial declaration into Part I "Negotiations on Trade in Goods" and Part II "Negotiations on Trade in Services." In Part I, it was explicitly stated that the goods negotiations were to be "within the framework and under the aegis of the General Agreement on Tariffs and Trade." Technically, the services negotiations were to be conducted outside the GATT framework--albeit with the support of the GATT secretariat and with the application of GATT procedures and practices. How the "respective results" (a term clearly indicating a degree of separateness) would be internationally implemented was to be decided after the round at a special session of the CPs.

The Single Undertaking grouped the parts of the negotiation for procedural purposes (including timing) as an eventual decision on international implementation would be taken at a single special session. But the Single Undertaking as it was expressed in 1986 in no way was interpreted as implying that all participants in the negotiations would need to take on all of the resulting obligations, especially, those resulting from the services negotiations. This was also true for the negotiations foreseen on the Tokyo Round Codes (which most Uruguay Round participants were not party to in 1986). (15) Under the terms of the Single Undertaking (Mark I), it would have been possible to envisage a final decision at the special session in which some GATT CPs joined the services trade agreement, for example, and some did not.

It was only much later in the Uruguay Round, around 1992 to 1993, that the anticipated agreement to create a new Multilateral Trade Organization began to be seen by the Quad countries (16) as offering an opportunity not to be missed to rid the new system of free riders and to ensure that all future agreements would be binding on all member countries. This was an opportunity to restore the reciprocal basis of obligations that the Framework Agreements had weakened. By deciding to leave behind GATT's less-than-reciprocal, somewhat a la carte, MFN framework and start over with an entirely new structure, the Quad members were able to force Uruguay Round participants into accepting obligations under all of the new system's agreements with the exception of the International Dairy Agreement, the International Bovine Meat Agreement, and the Government Procurement and Civil Aircraft Codes. (17) Members of WTO had no a la carte options at all: it was a case of "accept everything or remain outside the multilateral system."

The Mark II version of the Single Undertaking required all WTO members to sign on to every agreement, including the renegotiated Tokyo Round Codes; to engage on services as well as goods; to submit to a dispute mechanism that automatically determined disputes and facilitated retaliation in the case of noncompliance; (18) to tariffy and bind all agricultural duties; to meet the terms of the formulas for agricultural and non-agricultural market access reforms; and to accept a reformulation of special and differential treatment in terms of thresholds and implementation time frames rather than as status exemptions or preferences.

It seemed only just that, in the same step, the developed countries bound themselves more closely to the letter of their obligations. The developed countries lost most of the privileges that the looser constitutional framework of GATT allowed them. The rules could no longer be provisionally applied, which put an end to the waivers. The processes of the dispute system could no longer be avoided. Although retaliation was now an automatic option when remedies were delayed, unilateral retaliatory legislation was now outlawed, and arbitrary safeguards became impossible. And devices such as variable levies; minimum import prices; quotas on garments, farm products, and footwear; and agricultural export subsidies that shored up special and differential treatment of developing countries for developed country industries were gone and could not be replaced.

The Quad could get away with changing the Mark I version of the Single Undertaking into the Mark II version only because of the creation of the new organization. But Mark II did not reach back into history. GATT and its procedures, some associated provisions including the Framework Agreements, and even some of its dispute decisions were imported into the new organization, thereby laying up some (almost certainly unanticipated) problems for the future.

Doha Round Difficulties

The Mark II version of the Single Undertaking appeared to swing the pendulum a long way from the 1979 Framework Agreements toward legally binding reciprocal participation by developing countries. On at least one reading, the swing was over the top, binding developing countries to reciprocate more than they received on account of the huge transfers that, in principle, they were debited by their new TRIPS obligations. (19) In any case, a reversion began immediately with the implementation debates that slowed WTO's continuing negotiations (20) procedure to a halt and caused the collapse of the 1999 Seattle ministerial conference in recriminations and confusion. (21) By the time of the Doha ministerial conference in 2001, the language of the pragmatic GATT standoff had returned, as paragraph 50 of the declaration indicates.

 (50.) The negotiations and the other aspects of the Work Programme
 shall take fully into account the principle of special and
 differential treatment for developing and least-developed countries
 embodied in: Part IV of the GATT 1994; the Decision of 28 Novembe
 1979 on Differential and More Favourable Treatment, Reciprocity and
 Fuller Participation of Developing Countries; the Uruguay Round
 Decision on Measures in Favour of Least-Developed Countries; and
 all other relevant WTO provisions.

Although the developed countries' old laissez-faire GATT-style pragmatism could not be maintained in the WTO's more coherent legal framework, the privileges for developing countries under the 1979 Framework Agreements actually gathered new force thanks to the Single Undertaking: they became more effective than they ever had been in the GATT. Until 1994, the agreement on the less-than-complete reciprocity of developing countries affected only the balance of market access concessions. The impact might, at most, be found in the modalities of an agreed tariff-cutting formula. Because, under GATT, developing countries were not obliged to join every agreement (e.g., the Tokyo Round Codes), they had no need to secure the inclusion of framework-type language in every aspect of the legal framework. But under the WTO's Single Undertaking, the terms of the developing country exceptions have assumed a broader sweep. Every agreement must now be capable of attracting every member's participation, which means that the Framework Agreements, that formerly qualified only the expectations of members about the outcome of negotiations, have become an integral "modality" of the negotiations in the WTO.

It is not difficult to see where this has led: to agreements that define non-reciprocity in fine procedural detail. In the May 2008 modalities text proposed by the chairman of the negotiating group on agriculture, (22) developing countries have lower thresholds of achievement that would ensure their less-than-reciprocal obligations. But they also have a broad range of status exceptions and product carve-outs or exemptions that will ensure that more than half of them--not counting the Least Developed Countries who have no market access obligations at all--have still lesser market access obligations. In nonagricultural market access products, the proposed modalities suggest larger absolute cuts in tariffs by developing countries leading, however, to significantly higher final tariff rates. The developing countries will also access broad flexibilities to distribute a range of product exemptions and to employ status exceptions.

As we write this article, the outcome of the negotiations remains clouded by disputes over the headline numbers in agriculture, the formulas for NAMA, and members' intentions on services market access. We cannot define the impact of the lowered ambitions, but do have a guide in the estimates made for the World Bank by Kym Anderson, Will Martin, and Dominique van der Mensbrugghe. (23) A pessimistic model scenario projects that low- and middle-income countries would derive zero welfare gains from an agricultural reform scenario that cuts domestic support, phases out export subsidies, and reduces the bound-rate overhang of agriculture tariffs with exceptions for special and sensitive products and less-than-reciprocal cuts for developing countries. But that low-ball modeling scenario is no more pessimistic than the modalities currently proposed for agriculture in the Doha Round. According to Anderson, Martin, and van der Mensbrugghe, if developing countries were to drop their demand for less-than-reciprocal market access obligations in both NAMA and agriculture, low- and middle-income countries would share in (at least) a $40 billion (2001 dollars) boost to annual income out of an overall global increase of $120 billion by 2015. (24)

An Alternative Approach--Critical Mass

The Mark II version of the Single Undertaking delivered a paradoxical outcome: a World Trade Organization in which nonreciprocity for developing countries is more pervasive than ever and the resulting disaffection among members stronger than it ever was under GATT. There has to be a better way to reach global trade agreements; but what?

Successful negotiating outcomes in the GATT/WTO system have been based not on a Single Undertaking approach, but rather on what has come to be called the critical mass approach. It is important to realize that, over the years, the critical mass approach applied not only to the negotiation of trade liberalization, but also to the negotiation of the GATT system's rules.

Fifty-three years ago, a review session of the GATT CPs reexamined the GATT approach to disciplining subsidies. The 1955 review session eventually led to the incorporation into GATT Article XVI of the rules in paragraphs 2 to 5 (or Section B--"Additional Provisions on Export Subsidies"). New paragraphs 2 and 3 dealt with export subsidies to agriculture, basing new rules on the effects of subsidies in world markets. A different approach was applied to nonprimary (or industrial) product export subsidies in new Article XVI:4, which attempted to prohibit export subsidies on industrial products as from I January 1958. Early experience with compliance with this new rule was unsatisfactory and, in 1960, a working party on the provisions of Article XVI:4 produced a "Declaration Giving Effect to the Provisions of Article XVI:4" that entered into effect for its signatories in November 1962. At a time when forty-two governments were CPs to GATT, only seventeen (25) signed the declaration. The new obligations applied to the seventeen signatories, but rights under Article XVI: 4 accrued to all forty-two CPs. Clearly, this apparent lack of reciprocity did not stop the seventeen from signing on because they must have considered that they collectively constituted a critical mass of CPs likely to engage in meaningful export subsidies on industrial products. (26) The fact that the agreement among these seventeen CPs was a successful path to disciplining export subsidies was evidenced later--first by the fact that the provisions served as the legal basis for the Domestic International Sales Corporation (DISC) case brought by the European Community against the United States and, subsequently, by its plurilateral extension through the Tokyo Round Subsidies Code and, eventually, to all WTO members at the end of the Uruguay Round.

Another example of early critical mass in rule-making negotiations was the Kennedy Round's 1967 Agreement on Implementation of Article VI--the first Antidumping Code--which was limited in participation to those CPs and members of the European Economic Community (EEC) that accepted it. Nevertheless, it entered into force for those governments on 1 July 1968 and remained in force as supplemental obligations for these countries until it was replaced by the Tokyo Round Code in 1979.

As the Warwick Commission report notes, the Tokyo Round nontariff measures (Codes) are also good examples of how differing numbers of GATT CPs decided to join different types of rules agreements and where they obviously considered that a sufficient critical mass of membership was achieved to make undertaking the agreements' obligations worthwhile. The Codes were an exercise in critical mass with an element of variable geometry. An aspect to the Tokyo Round Codes that distinguished certain of them from other critical mass exercises is that they were implemented on a non-MFN or conditional MFN basis.

These examples from the past make clear that a critical mass approach is not limited to a market access framework. Of course, it is true that, in effect, all multilateral trade negotiations on trade liberalization and market access are largely critical mass negotiations. That is because, once a certain degree of success is achieved by way of participation and commitments, participants have repeatedly demonstrated that they are not willing to jeopardize an outcome by insisting that every player in the system has to be a part of the deal. The Information Technology Agreement (ITA) as well as the results of the 1997 financial services and basic telecommunications negotiations are frequently cited as WTO-era agreements reached through a critical mass approach. In the case of the ITA, critical mass was set at a level of 90 percent of world trade in the covered products, and this level of participation was reached in April 1997 with just forty WTO members. Fifty-six offers, representing seventy countries, were submitted by the deadline of 12 December 1997 in the financial services negotiations. A similar threshold for critical mass is evident in the basic telecommunications services negotiations, where sixty-nine governments made commitments covering markets that accounted for more than 91 percent of global telecommunications revenues. The ITA, unlike some of the Tokyo Round Codes, is a multilateral agreement because the benefits of the ITA access improvements are available to all WTO members on an MFN basis. The same is true of the Financial Services Agreement and the Agreement on Basic Telecommunications Services. But there clearly is no reason to believe that the outcome of these negotiations would have been more successful if the framework adopted for the negotiations had insisted on participation by all WTO members at the time.

From this brief history, it is evident that the critical mass approach to negotiations has worked well in producing successful outcomes in trade negotiations both for rules and for market access. A key question that needs to be seriously addressed for the future is whether a critical mass approach would be a viable framework for reaching agreement on multilateral agricultural liberalization in the WTO.

Can Critical Mass Work for Agriculture?

In this article, the working definition of a critical mass framework for trade negotiations assumes that an effective and meaningful result can be obtained when it is crafted and agreed to by those WTO members whose combined share of world trade in the products covered by the negotiation is great enough that these members consider that nonparticipation by others is not a significant impediment to the effectiveness of their agreement. In addition, for any agreement negotiated in the WTO context, the participants in the agreement must be willing to implement their obligations under the critical mass agreement on an MFN basis.

The Theory Behind Critical Mass in Agriculture

Why are the current negotiations on trade in agriculture so complex? Why has it been necessary for the chairman to produce draft modalities running to nearly eighty pages, with numerous provisions designed to give differential treatment in the results to particular products or groups of countries? Among the important reasons for this complexity is the number of countries involved in the talks that play only a small role in international trade in agriculture. These countries may have an interest in efficient world markets for agriculture (e.g., because agriculture continues to provide a high proportion of employment in their economies), but they may not have the capacity or the resources to participate in agriculture trade negotiations and do not, in any case, have agricultural trade objectives that call for reciprocal participation. Examples include many African states, like Botswana or Cote d'Ivoire. Looked at from an export perspective, even India and Indonesia could fall into this category. Insistence on special treatment as a condition for these countries' participation in the talks and to an eventual consensus outcome only adds to the complexity of the talks and lowers the overall ambition of the agreement. Thus, the first premise of this article is that these countries need not be party to the negotiation or the implementation of an agriculture agreement in order for the agreement to be an effective part of the WTO framework. Neither is their participation in the negotiations necessary from their own perspective if the agreement is implemented on an MFN basis.

Who Needs to Be Part of a Critical Mass Deal?

In the approach that we set out to test, we will apply the same basic parameters as those adopted in the past negotiations on information technology products, basic telecommunications services, and financial services. We will assume that a working hypothesis for a critical mass negotiation in agriculture requires the participation of WTO members accounting for 90 percent of global trade in agricultural products. (27)

Does an effective critical mass deal on agriculture need to cover trade in all product categories traded internationally? If the answer is yes, is it necessary to cover all products with the same approach or could the approach to the negotiations be differentiated by product? If it is not necessary to cover all products, which products need to be covered for a deal to be effective?

Although a critical mass agreement might be conceptualized as covering trade in only a single commodity group (e.g., grains), the global interests of agricultural exporting and importing countries, and the relationship of developments in agricultural negotiations to trade negotiations in other products and services, probably requires a multisectoral approach at a minimum. Still, we should be able to imagine a collection of critical mass agreements across a range of commodities that account for most global trade in agriculture and a more bilateralized request-and-offer approach to less significant products in international trade.

The basic telecommunications, financial services, and ITA agreements all used a critical mass approach successfully in what were essentially market access agreements. As one of the three pillars of the Agreement on Agriculture, market access is clearly indispensable to any negotiated result, and there is no reason a priori to think that agricultural market access issues should not be amenable to treatment in a critical mass negotiating framework. But there are two other pillars that we also need to address: export competition and domestic supports. How might these topics be addressed in this critical mass scenario?

Export competition. Notwithstanding decades of publicity surrounding the negative effects of export subsidies on agricultural trade, the export competition pillar of the Doha Round negotiations has turned out to be the topic where it has proven relatively easy to reach consensus among WTO members. Agreement to prohibit export subsidies and measures with equivalent effect seems to be at hand. Thus, we would hope that the question of disciplines for these measures might be put "on the shelf since they apparently accord strongly with the overall agricultural policy directions of the European Community and the United States that, in the past, have accounted for almost all export subsidies. If that is not the case, and the Doha Round either fails or leaves export subsidies in place, then we suggest an approach that would require agreement on the disciplines to be applied to export subsidies by countries accounting for 90 percent of global exports and imports of an agricultural commodity on a commodity-by-commodity basis. (28) An alternative might be to require agreement on export subsidy disciplines by only the WTO members accounting for 90 percent of global exports in the commodity concerned because the importing countries may have little practical interest in limiting the use of export subsidies.

Domestic supports. Domestic supports can operate both to distort trade through overproduction of agricultural commodities in exporting countries and to nullify or impair the effect of market access concessions in importing markets. Therefore, this critical mass negotiation should require agreement on domestic support disciplines to the extent necessary to avoid circumvention of market access undertakings by all of the exporting and importing countries that account for 90 percent of global trade in an agricultural commodity on a commodity-by-commodity basis.

For the purposes of this preliminary examination of how critical mass might operate in agriculture, we have taken the top thirty commodities in international trade in agriculture (by value) and identified those countries that, when their share of exports are ranked in descending order, cumulatively account for more than 90 percent of global exports in these commodities. We have done the same thing for imports. (29)

What this initial analysis shows is that--on this basis--only a relatively small number of WTO members are significant participants in global agricultural trade. For wheat (HS 100190), only seven countries (30) account for 91 percent plus of exports and thirty-six countries account for 90 percent of imports. For other commodities, the numbers are even smaller. In the case of maize (HS 100590), the 90 percent threshold is met with four exporters and twenty-six importers. For the two most highly traded categories of bovine meat (HS 020130 and 020230), only nine exporters and twenty importers account for 90 percent of trade. On a single commodity basis, only in the case of refined sugar (HS 170199) trade, does the number of countries add up to more than fifty--with forty-seven importers and fifteen exporters. And, in the case of sugar, ten of the importers and two of the exporters are not currently WTO members.

In total, and for the commodities examined in this initial analysis, just eighty (31) WTO members make the chart. Twenty-one of these are significantly involved in trade in only one commodity. In addition, for many of the importing countries included in the 90 percent critical mass groupings, their share of world imports of the commodity in question is less than 1 percent of the total. In the case of wheat, for example, thirteen of the thirty-six importers individually import less than 1 percent of wheat imports.

When we group global trade in wheat, durum wheat, soybeans, maize, and barley, we find that fifty-three countries (counting the EU-27 as one) account for more than 90 percent of all imports and exports of these five commodities. Taking the two most highly traded milk powder categories (HS 040210 and 040221) together, sixty-five countries, many of them not WTO members (including Algeria, Iraq, Syria, Libya, Yemen, Tajikistan, and Sudan), account for 90 percent of trade. Of the importers of milk powder, more than one-third account individually for less than 1 percent of global imports of the product.

Suppose that the threshold level for a critical mass agreement is reduced to 80 percent of global trade in covered commodities. In wheat, two exporters and thirteen importing countries are lost; in maize, two exporters and eleven importing countries are lost. To reach an 80 percent threshold for soybeans, only two exporting countries and five importers are needed. An 80 percent threshold may not be far-fetched given that many of the importing countries have little alternative except to import the commodities in question and those left out of the critical mass sample might elect to join the agreement anyway.

Problems with the Critical Mass Model for Agriculture

Even if the statistics seem to show us that only a relatively small subset of WTO members would need to participate in a negotiation covering a critical mass of 90 percent of imports and exports of key internationally traded agricultural commodities, there are several other factors that may affect the success of a critical mass approach to negotiations. (32) Two of the most important are discussed below.

Large differences in assistance regimes. Even where a large proportion of trade is due to a small number of countries, these countries may nevertheless impose a wide range of protective measures on the products in question. Some participants may have high tariffs and tariff rate quotas protecting administered prices or farmer supports while others have a liberal import regime. Critical mass would not make it any easier for Japan or Korea to liberalize rice. On the other hand, it is clear that many countries that import food products do so because they have an excess demand that can be met only from imports.

For high-protection and high-support countries, the question is whether they would be willing to reduce protection on all supply, including domestic production. There is no a priori answer: governments must take into account both economic and political factors that vary from product to product and market to market. But there are several reasons for some confidence that they will look positively at the opportunities.

* Successful liberalization of a large proportion of the global market for an agricultural commodity under a critical mass agreement is likely to lead to higher and more stable world prices for that commodity. The implied consumer penalty can be ameliorated by reducing import barriers, and the implied bonus of higher prices for domestic producers should reduce their need for protection.

* Recent developments in food prices in world markets driven by, among other things, demographic factors that seem to represent a secular change suggest an adverse terms-of-trade development for net importers. Food importers cannot reverse this trend, but they can cut their losses from food price inflation by, among other things, cutting their border barriers.

* Importers who perceive that a critical mass agreement may go ahead without their participation face no risk of trade discrimination if the critical mass is implemented under MFN provisions of the GATT. But nonmember-ship means no right to contribute to the management of the agreement; governments of protected markets are likely to think twice about being left out of the information flows that come from participation in a global regime. India found itself in this situation at the time that the ITA was being concluded.

* Governments that are especially concerned about the security of import supply may recognize that one consequence of being outside the agreement, and of maintaining high import barriers in a globally liberalized market, will be reduced interaction between their industries and the global industry. This will affect the commercial (contractual) underpinnings of supply security and increase reliance on expensive government measures to secure supply (e.g., by maintaining public stocks).

Large differences in export regimes. It is not clear that we need to be concerned any longer with the problem of export subsidies; however, some of the exporting countries will continue to be large users of domestic supports and these will need to be disciplined in any deal to prevent circumvention of market access concessions. Given the relatively small number of countries that time and again show up as the major exporters of agricultural commodities, negotiating domestic support disciplines might be an easier task than it has been to date in the WTO.

Conclusion

Moving to a critical mass approach to agricultural trade negotiations clearly will not solve all of the problems that have been manifest in the Doha Round negotiations. It also is unclear whether a critical mass approach would prove to be politically acceptable to all of the countries that would need to participate. But the challenges of a critical mass approach to the liberalization of agricultural markets are no wider in scope (access, supports, subsidies) and significantly better focused than the challenges faced by the current approach to agricultural trade liberalization. Critical mass agreements promise a relatively simple framework for reciprocal agreements that are consistent with GATT nondiscrimination provisions. They suggest a means for crafting effective disciplines, to secure mutual advantage in a relatively brief negotiation among those countries most concerned, and could reinvigorate the WTO's role in global trade reform.

In this article, we have looked at critical mass from the standpoint of a negotiation within the framework of the WTO. We might also examine the possibilities for reaching a critical mass on agricultural trade in a regional context. So far, preferential (bilateral or regional) trade agreements have largely been unable to deal effectively with the products that are hard-core protected in certain markets. Subsidies have also not been easy to address in free trade agreements (FTAs). But suppose the FTA was a large FTA; for example, the mooted ASEAN+6, Comprehensive Economic Partnership in East Asia (CEPEA) arrangement. Could two-thirds of the world's population reach a good deal on agriculture among themselves even if the United States, European Community, and Latin America were on the outside of the deal?

The research reported in this article is only the beginning of an important research project that is designed to take an in-depth look at these and other questions in an effort to reach some more well-documented conclusions on the political economy of critical mass negotiations and other alternative approaches for trade liberalization in agriculture. Over the next eighteen months, we hope to be in a better position to answer these questions.

Notes

Peter Gallagher is a trade and public policy analyst who provides advice and analysis on international trade, public policy, and business communications. His clients include Australian food companies, industry organizations, government agencies, and international institutions. Andrew Stoler is the executive director of the Institute for International Trade at the University of Adelaide. Previously, he served as deputy director-general of the World Trade Organization from 1999 to 2002, following a long career with the Office of the US Trade Representative.

(1.) This article is a contribution to the research project "Viability of Alternative Frameworks for Agricultural Trade Negotiations," cosponsored by the Australian Rural Industries Research and Development Corporation (RIRDC) and the Australian Centre for International Agricultural Research (ACLAR). Elements of this article were originally produced for the April 2008 Warwick Commission seminar held at the University of Melbourne.

(2.) Ministerial Declaration of 14 November 2001, para. 47.

(3.) Agreement Establishing the World Trade Organization (Marrakesh Agreement), Preamble.

(4.) Doha Ministerial Declaration, para. 2.

(5.) Ibid., para. 50.

(6.) "Differential and More Favourable Treatment, Reciprocity and Fuller Participation of Developing Countries," GATT Council Decision of 28 November 1979, L/4903.

(7.) Here is an intuitive demonstration of the intended mechanism. Consider a two-product developing country economy where prices are determined by the ratio of importables to exportables (i/x). Now, hold protection in this economy steady so that the price of importables (i) is unchanged while reducing protection in developed country export markets. The price of exportables rises in the developing country, owing to the greater demand for exports (to the developed market) so that the price ratio of importables to exportables (i/x) falls: a terms of trade gain to the developing country.

(8.) The nonreciprocity of developing countries in the market access negotiations of the Dillon, Kennedy, and Tokyo Rounds is documented in the GATT study "The Tokyo Round of Multilateral Trade Negotiations, Report by the Director-General of GATT" (Geneva: GATT, 1979); and in William R. Cline et al., Trade Negotiations in the Tokyo Round--A Quantitative Assessment (Washington, DC: Brookings Institution, 1978). Fourteen developed countries applied the Tokyo Round tariff formula: another twenty, mostly developing countries, participated to some extent in the tariff negotiations. The GATT study estimated that the final reduction on industrialized products was about 38 percent and that the cut on products of interest to developing countries that were included in the final results was approximately 35 percent (simple averages). The reluctance of developing countries (and Australia) to participate in the Tokyo Round market access negotiations reflected not only on their trade policies but also on the imbalanced objectives of the negotiations--carried forward from the Kennedy Round--that concerned, chiefly, a subset of nonagricultural products. Although developed countries offered cuts in "tropical products," no tariff formula was applied to cuts in temperate farm products.

(9.) The Lemer equivalence: as the demand for exportables increases, output in that sector rises, wages and other specialized factor prices rise, and the price of the exportable rises too. At a higher price relative to imports, demand in the export market (where the importable goods are produced) falls.

(10.) J. Williamson, "The Washington Consensus as Policy Prescription for Development," paper presented at the Practitioners in Development Lecture, World Bank, Washington, DC, January 2004.

(11.) Robert E. Hudec, Developing Countries in the GATT Legal System (London: Gower, for the Trade Policy Research Center, 1987).

(12.) From Australian trade minister John McEwen, among others, see GATT Summary Record of Eleventh Session, 25 October 1956, SR. 11/7, available at www.wto.org/gatt_docs/English/SULPDF/90270417.pdf (accessed 1 July 2008).

(13.) Hudec, Developing Countries, p. 36.

(14.) Ministerial Declaration on the Uruguay Round (20 September 1986), para. I.B.(ii)

(15.) The Codes, to which few developing countries adhered, were renegotiated as agreements or "understandings" in the Uruguay Round.

(16.) The Quad was composed of the European Communities, the United States, Canada, and Japan. The Quad members met regularly on an informal basis to exchange information and, where possible, to coordinate their positions in the Uruguay Round negotiations.

(17.) These four agreements were seen to be of limited interest and application to most WTO members (only a few countries produce large civil aircraft or are major exporters of dairy products), and it was agreed that it was not necessary or desirable to make participation in the agreements a requirement of WTO membership.

(18.) Hudec, Developing Countries, p. 103.

(19.) J. M. Finger, The Doha Agenda and Development: A View from the Uruguay Round (Manila: Asian Development Bank, 2002).

(20.) The commitment embodied in, for example, Article 20 of the Agreement on Agriculture to continue negotiations on the "reform process" as part of the normal work of the WTO without relaunching a comprehensive round of negotiations.

(21.) Peter Gallagher, The First Ten Years of the WTO (Cambridge: Cambridge University Press, 2005).

(22.) Revised Draft Modalities for Agriculture, TN/AG/W/4/Rev.2, 19 May 2008.

(23.) Kym Anderson, Will Martin, and Dominique van der Mensbrugghe, "Doha Merchandise Trade Reform: What's at Stake for Developing Countries?" World Bank Economic Review 20, no. 2 (2006): 169-195.

(24.) Anderson, Martin, and van der Mensbrugghe, "Doha Merchandise Trade Reform," are at pains to point out that these projections are based on static estimates that experience suggests are likely to significantly undershoot actual outcomes, including the dynamic ("second round") gains from liberalization as firms and households adjust to new opportunities.

(25.) Austria, Belgium, Canada, Denmark, France, the Federal Republic of Germany, Italy, Japan, Luxembourg, the Netherlands, New Zealand, Norway, Sweden, Switzerland, the United Kingdom, the United States, and Zimbabwe.

(26.) Article XVI:4 was unbalanced, as far as most agricultural exporters were concerned. As Hudec, Developing Countries, notes, no "reciprocal" offer from them would have secured more equitable product coverage at the time (p. 73). The improved discipline on industrial subsidies worked, however, without the participation of the majority of GATT CPs because it had been accepted by a critical mass of countries accounting for most industrial product exports.

(27.) For the purposes of this initial article in the project, we are basing the analysts on trade in several product groups that are widely recognized as significant in both export and import trade, including grains, dairy products, and meat products. Later in the RIRDC-ACIAR research project, of which this article is a contribution, the categories of products will be expanded importantly.

(28.) WTO members not participating in the critical mass negotiations might be requested to operate any export subsidy programs they maintain in ways that do not undercut the effectiveness of the implementation of the critical mass agreement.

(29.) For now, we have based this analysis on the 2006 ITC TRADEMAP data at six-digit HS (harmonized system of tariff nomenclature) level.

(30.) The EC-27 are counted as one for purposes of this analysis.

(31.) This is 106 of the WTO's 152 members if the EC-27 are counted separately.

(32.) Among those that we do not examine here is the impact of critical mass agreements on what is sometimes called the "geometry" of WTO membership.