THE BALI AGREEMENT, AT LAST AN ASSESSMENT FROM THE PERSPECTIVE OF DEVELOPING COUNTRIES WORKING PAPER | December 2014 EUGENIO DÍAZ-BONILLA AND DAVID LABORDE Name and Introduction On December 7, 2013, after several days of work and the usual posturing and drama, Members of the WTO closed the Ninth Ministerial Conference with an agreement on the organization’s first comprehensive multilateral trade package. Until that point, the trade agreements completed since the WTO’s creation in 1995 had been mainly regional and plurilateral ones, including some but not all WTO members. In many cases, these agreements were negotiated outside of the WTO altogether. The implementation of the Bali agreement should have taken place during 2014 but reached an impasse by the end of June; the reasons why will be discussed in detail below, but it was basically due to differences in opinion about the WTO’s treatment of public food stocks in developing countries.1 Only on November 27, 2014, almost a year after the original Bali Ministerial, did WTO members managed to patch up their differences. 2 This recent agreement allows the implementation of the assorted policy decisions that were supposed to have been settled at Bali but were held up by the dispute on public food stocks to finally proceed, and puts back on track the post-Bali work program that should now be defined by mid-2015. This paper discusses the results of the Bali Ministerial Conference of December 2013 (sometimes called the “Bali Pack- age”), the problems encountered during 2014, and how were they solved in November 2014, as well as the potential implications for the post-Bali work program which remains critical to unlocking the Doha Round. The perspective presented here is that of developing countries (although recognizing that this is a heterogeneous group). Our discussion is mainly aimed at policymakers, development practitioners, and the general public in developing countries who are interested in trade issues but who do not necessarily follow these negotiations closely. The paper tries to provide some historical and conceptual background on the topics covered, while at the same time including enough legal detail regarding the texts and specific trade discussions to serve as a basic reference. Since agricultural topics, particu- larly food security, have been key issues in the negotiations, they receive a more detailed treatment here. We recognize, however, that discussing agricultural issues in isolation will not provide an adequate picture of the Bali negotiations, during which countries tried to obtain a balance of rights and obligations (sometimes called “offensive” and “defensive” objec- tives). The rest of the paper is organized as follows. First, we provide a short historical narrative of the negotiations leading to the Bali Ministerial in December 2013, and of the issues that delayed the implementation process until the agreements reached in November 2014. Second, we try to place the negotiations and activities of the Bali Ministerial in the context of the different categories of work conducted in the WTO. A third section describes the legal aspects and implications of the 1 WTO Director General Azevêdo “We are very close to overcoming impasse on Bali implementation” http://www.wto.org/english/news_e/news14_e/dgra_26nov14_e.htm 2 WTO Director General Azevêdo ““The WTO has truly delivered”. http://www.wto.org/english/news_e/spra_e/spra16_e.htm 2 SUMMARY | APRIL 2010 Ministerial Decisions approved in Bali, and the modifications of November 2014. A fourth section summarizes the rele- vance of the Bali Package, particularly for developing countries. The final section concludes with some reflections about the thorny issues that may affect the future WTO program. A Short Historical Narrative of the Bali Package and the November 2014 Agreements The Bali agreement was a continuation of the process initiated by the Doha Round, which was launched in 2001 during the 4th WTO Ministerial Conference3 in an attempt to negotiate a multilateral deal building on the unfinished work of the previous Uruguay Round. From the point of view of many developing countries, the Doha negotiations were needed to better balance what they considered the excessive concessions that industrialized countries received in the Uruguay Round. These concessions related not only to new topics, such as intellectual property rights and services, but also to the legal room given to protect and subsidize industrialized countries’ agriculture and to the use of export subsidies for agricultural products (such subsidies are prohibited for industrial products under the GATT and WTO frameworks). The negotiating package discussed in Doha was even called the Doha Development Agenda in recognition of these complaints voiced by many developing countries.4 The Doha Ministerial Declaration (WTO, 2001) laid out an ambitious negotiating program with specific instructions for different topics, including agriculture, services, electronic commerce, and market access for non-agricultural products; intellectual property, investment, and competition policy; and government procure- ment, trade facilitation, WTO rules for regional trade agreements, the Dispute Settlement Understanding, and environmen- tal issues. Until Bali, however, the Doha negotiations had been languishing since the last sustained push to complete them col- lapsed in 2008.5 That collapse happened, in large part, because of disagreements between developed and developing countries regarding an adequate balance between the agricultural and non-agricultural commitments being negotiated, market access requests by developed countries, and differing opinions about “special and differential treatment” (SDT)6 for developing countries. There were two key factors in particular that derailed the 2008 negotiations.7 One was related to developing coun- tries’ food security concerns, focused at the time on the operation of a special safeguard mechanism to increase protection for selected staple products under specific conditions. The other factor was the cotton controversy between the US and several cotton-producing African countries (see more details below in the section related to cotton issues). Given the difficulties in pushing through negotiations on all of the Doha Round topics, WTO members decided to refo- cus their talks on a more limited number of issues, ones that seemed to have a better chance of leading to a general 3 The WTO Ministerial started after the formal creation of the WTO in 1995. The list from the latest to the earliest WTO Ministerial Conferences is as follows: Bali, 3-6 December 2013 (9th), Geneva, 15-17 December 2011 (8th), Geneva, 30 November - 2 December 2009 (7th), Hong Kong, 13-18 December 2005 (6th), Cancún, 10-14 September 2003 (5th), Doha, 9-13 November 2001 (4th), Seattle, November 30 – December 3, 1999 (3rd), Geneva, 18-20 May 1998 (2nd), and Singapore, 9-13 December 1996 (1st). As explained in more detail in the text, the Ministerial Conference is the supreme authority of the WTO. 4 The Doha Ministerial also addressed some of the problems created by the original agreement on Trade Related Intellectual Property Rights (TRIPS) with another Ministerial Declaration (Declaration on the TRIPS agreement and public health. WT/MIN(01)/DEC/2. 20 November 2001). That Declara- tion, recognizing “the gravity of the public health problems afflicting many developing and least-developed countries, especially those resulting from HIV/AIDS, tuberculosis, malaria and other epidemics,” made some of the requirements in the original text more flexible for developing countries. 5 When the negotiations collapsed, very sophisticated draft modalities texts were achieved, but not agreed. They are still used as a reference by many negotiators and observers and are called “Rev.4 Draft Modalities” in agriculture or “2008 Modalities” (TN/AG/W/4/Rev.4). 6 SDT refers to the notion that developing countries may need special treatment under trade rules, differentiated from the legal obligations of developed countries (more on this later). 7 Recently, Baghwati (2013) has argued that the failure was due to pressure from US business lobbies on the Obama Administration to try to get more concessions (mostly but not limited to market access) from large developing economies such as India for agriculture and Brazil for manufacturing. As is the case in the opposite narrative, in which the intransigence of large developing countries (particularly India for agriculture) was the main culprit, any verdict of guilt must recognize that in unsuccessful negotiations, it is usually the collective behavior of all actors which defines the results (although in the WTO consensual process, there is always the possibility of single actors blocking advances). 3 SUMMARY | APRIL 2010 agreement. From those efforts emerged the components of the Bali Package that was finally approved on December 7, 2013. The Bali Ministerial Declaration instructed WTO members to define a post-Bali work program during 2014. Crucially, a specific component of the Bali package-- the Agreement on Trade Facilitation (ATF)-- required a final review of the wording and the completion of some legal steps (explained below) to be sent to WTO members for final approval by their legisla- tures.8 Those wording and legal issues were scheduled to be resolved during a meeting of the WTO General Council in July 2014. That meeting broke down in disarray when India basically made the implementation of those final legal steps contingent on clarifications regarding another key component of the Bali package: the Ministerial Decision on Public Stockholding for Food Security Purposes. The topic had already brought negotiations to the verge of a breakdown at the Bali Ministerial and led to a one-day extension of discussions. Even after that, there were still places where the wording had alternative interpretations (see the discussion in Díaz-Bonilla, 2014), and India decided to hold up negotiations of the ATF until the text was clarified. From July to November, there was a flurry of negotiations to try to work out a compromise, until a final agreement on that wording was achieved in the WTO General Council meeting of November 27, 2014. The context and content of the Bali Package and of the November 2014 agreement are discussed in the next sections. CONTEXT OF THE BALI PACKAGE To provide some context, it is useful to note that the Ministerial Conference is the supreme authority of the WTO, com- prised of the top domestic authorities on trade issues designated by each Member. The Ministerial usually meets every two years and can make decisions on any WTO subject. The Ministerial is a meeting of governments; the WTO Secretariat (i.e. the WTO staff proper) do not serve as representatives of specific countries but rather as neutral support staff for the work done by the governmental delegations of diplomats accredited to the WTO. Under the Ministerial Conference is the WTO General Council (GC), which manages day-to-day issues at the WTO Headquarters in Geneva; and under the GC is a series of other councils and bodies, carrying out the different functions and activities of the WTO. These functions and activities can be organized around three main areas of work. The first is the negotiation of new agreements. While outside attention typically focuses on this area (i.e. trade negotiations), the two other areas of work are at least as important. The second group of activities is the work of the different WTO committees, where Members monitor and maintain a continuous dialogue on current trade practices and operations in the context of WTO legal commitments. The third area is the Dispute Settlement Mechanism (DSM) that Members utilize when they believe that their rights under specific WTO agreements have been negatively affected by other Members. Although the three areas include separate functions, they are linked in important ways: consultations under regular committees (the second area of work) may end up as cases in the DSM (the third area of work); both the activity of the committees and the DSM may lead to the identification of topics that may require further trade negotiations (the first area of work); and, obviously, the agreements negotiated under the first area of work provide the legal and operational issues that constitute the substance of both the work of the committees and the work of the DSM. Not all of the WTO’s work in any of the three areas necessarily reaches the level of decision-making by the ministers in the Ministerial Conferences. Some trade agreements are not multilateral; they may have links to the WTO legal framework but may be administered by a separate institutional architecture. One such category of agreements includes Plurilateral Trade Agreements (PTAs), which involve a subset of WTO members who agree on certain trade regulations for a specific sector and which are governed by the provisions of those agreements (Article IX (5) of the Agreement Establishing the WTO Organization).9 There is also a long list of Regional Trade Agreements (RTAs) involving two or more countries; as of July 2013, the WTO acknowledged 575 notifications of RTAs (counting goods and services), of which 379 were effective. 8 As discussed later, the Bali package had been agreed by representatives of the Executive Branches of the WTO Members, and the ATF, being a new WTO agreement required, in many cases, approval by the Legislative Branches. 9 The existing PTAs under the WTO umbrella are Trade in Civil Aircraft and Government Procurement. Currently there is an ongoing negotiation on Services, which, given that not all WTO Members are involved, would end up being a PTA if agreed. 4 SUMMARY | APRIL 2010 Currently, there are several RTAs that have been called “Mega Regional Trade Agreements” and are receiving particular attention: the Trans-Pacific Partnership (TPP) (including so far Australia, Brunei, Chile, Canada, Japan, Malaysia, Mexico, New Zealand, Peru, Singapore, the United States, and Vietnam), and the Transatlantic Trade and Investment Partnership (TTIP) negotiated between the United States and the European Union. The Bali Ministerial Conference must be placed within this broader context of WTO and non-WTO trade negotiations and trade issues taking place in parallel. In fact, one of the underlying themes in the Bali negotiations was the possibility that, if Ministers failed to reach an agreement, the WTO’s multilateral aspects would lose relevance given the advance of PTAs and RTAs. More generally, there is an ongoing debate about whether PTAs and RTAs are helping or undermining global trade and the multilateral trade system built around the WTO; what the legal and institutional links between PTAs and RTAs and the WTO framework should be; and the potential marginalization of developing countries that are not participating in those agreements, as well as the asymmetrical bargaining power of those that do participate (see Bouet and Laborde, 2009, where they quantify the implication of trade agreements between industrialized countries that exclude developing countries). The work done during the Bali Ministerial can be divided into three areas of concentration. The Bali Ministerial Decla- ration (WT/MIN(13)/DEC/W/1/Rev.1, 7 December 2013) is the basic document identifying all the decisions made at the Ministerial Conference. Table 1 includes a list of all relevant topics and documents. The main group includes the agree- ments and decisions related to the subset of components of the Doha Round (sometimes called the “Bali Package”) that were ready for Ministers to decide upon at the Bali Ministerial. This work is related to the “trade negotiating” activities (Section I in Table 1). Then there are decisions about the WTO work program post-Bali, focusing on future trade negotiating activities (Section II). Finally, there is another group of decisions in which the Ministerial Conference took note of the regular work of the General Council (mostly related to the second and third areas of work mentioned previously), issued further Ministerial decisions, and gave instructions for future regular work (section III). The adoption of the Decision on the Accession of the Republic of Yemen (WT/MIN(13)/24- WT/L/905), a Least Developed Country (LDC),10 by which Yemen became the 160th Member of the WTO, also belongs in this third group. 10 The category of LDCs is a specific denomination by the United Nations, based on development indicators. There are currently 48 LDCs determined by the United Nations, 35 of which have become WTO Members, counting the accession of Yemen during the Bali Ministerial (see next footnote). They include: Angola, Bangladesh, Benin, Burkina Faso, Burundi, Cambodia, Central African Republic, Chad, Democratic Republic of the Congo, Djibouti, Gambia, Guinea, Guinea-Bissau, Haiti, Lao People’s Democratic Republic, Lesotho, Madagascar, Malawi, Mali, Mauritania, Mozambique, Myanmar, Nepal, Niger, Rwanda, Samoa, Senegal, Sierra Leone, Solomon Islands, Tanzania, Togo, Uganda, Vanuatu, Zambia. There are other 8 LDCs that are WTO observers (Afghanistan, Bhutan, Comoros, Equatorial Guinea, Ethiopia, Liberia, Sao Tomé and Principe, and Sudan) and another 6 that are neither a Member nor an observer (Eritrea, Kiribati, Somalia, South Sudan, Timor-Leste, and Tuvalu). 5 SUMMARY | APRIL 2010 TABLE 1. DECISIONS AND DECLARATIONS OF THE BALI MINISTERIAL I. DOHA DEVELOPMENT AGENDA Trade Facilitation Agreement on Trade Facilitation (WT/MIN(13)/W/8) Agriculture and Cotton  General Services (WT/MIN(13)/W/9)  Public Stockholding for Food Security Purposes (WT/MIN(13)/W/10)  Understanding on Tariff Rate Quota Administration Provisions of Agricultural Products (WT/MIN(13)/W/11)  Export Competition (WT/MIN(13)/W/12)  Cotton (WT/MIN(13)/W/13) Development and LDC issues  Preferential Rules of Origin for Least-Developed Countries (WT/MIN(13)/W/14)  Operationalization of the Waiver Concerning Preferential Treatment to Services and Service Suppliers of Least-Developed Countries (WT/MIN(13)/W/15)  Duty-Free and Quota-Free (DFQF) Market Access for Least-Developed Countries (WT/MIN(13)/W/16)  Monitoring Mechanism on Special and Differential Treatment (WT/MIN(13)/W/17) II. POST-BALI WORK The relevant section of the Bali Ministerial Declaration (WT/MIN(13)/DEC/W/1/Rev.1) III. REGULAR WORK UNDER THE GENERAL COUNCIL  TRIPS Non-violation and Situation Complaints (WT/MIN(13)/W/2)  Work Program on Electronic Commerce (WT/MIN(13)/W/3)  Work Program on Small Economies (WT/MIN(13)/W/4)  Aid for Trade (WT/MIN(13)/W/5)  Trade and Transfer of Technology (WT/MIN(13)/W/6)  Decision on the Accession of the Republic of Yemen (WT/MIN(13)/24- WT/L/905) SOURCE: WT/MIN(13)/DEC/W/1/REV.1, 7 DECEMBER 2013. In its meeting on November 27, 2014, the WTO General Council approved the Decision on Public Stockholding for Food Security Purposes (WT/L/939), the Protocol of Amendment to insert the Trade Facilitation Agreement into Annex 1A of the WTO Agreement (WT/L/940) and open the Protocol for acceptance, and Decision on Post-Bali Work (WT/L/941). The General Council Chairman noted that “in adopting the three Decisions on Public Stockholding for Food Security Purposes, on the Protocol of Amendment for Trade Facilitation, and on Post-Bali work simultaneously, we are reaffirming the entirety of the Bali Ministerial mandates, including the priorities that Ministers identified at Bali” (http://www.wto.org/english/news_e/news14_e/gc_rpt_27nov14_e.htm). In what follows, we will focus on topics in Sections I and II of Table 1 (what was negotiated in the “Bali Package” as part of the Doha Development Agenda and the related future work on trade negotiation issues) and the Decisions of the November 2014 General Council. Anyone interested in the regular work under the General Council (Section III) can consult the main topics and related documents in the Bali Ministerial Declaration listed in Table 1.11 The Bali Package and Future Work As shown in Table 1, the Bali Package can be divided into three topics: a) trade facilitation, b) agriculture and cotton, and c) specific issues related to the special treatment of developing countries and LDCs under WTO rules. The modifications that emerged from the November 2014 agreements (the Decision on Public Stockholding for Food Security Purposes, WT/L/939, 11 The Ministerial Conference also took note of two decisions of the General Council taken in Geneva in response to the relevant mandates from the Eighth Ministerial: a) Decision adopted by the TRIPS Council concerning the extension of the transition period under Article 66.1 for Least-Developed Country Members in document IP/C/64, and b) Decision adopted by the General Council in July 2012 on the Accession of LDCs in document WT/L/508/Add.1. 6 SUMMARY | APRIL 2010 and the Protocol for the Trade Facilitation Agreement, WT/L/940) will also be highlighted where they apply. The Bali Ministerial Declaration did briefly define the parameters and procedures for post-Bali work; but, as previously mentioned, this work program was derailed by the controversy over public food stocks, and it was therefore updated by the Decision on Post-Bali Work (WT/L/941). The discussion of the post-Bali work program closes this section. TRADE FACILITATION Background As previously mentioned, trade facilitation was part of the negotiating program defined at the Doha Ministerial Declaration (WT/MIN(01)/DEC/1, 20 November 2001), focusing on customs procedures and related institutional and organizational issues. It was originally part of the four so-called “Singapore issues” that were discussed at the WTO’s First Ministerial Conference in 1996. The other three issues were trade and competition, trade and investment, and transparency in government procurement. All of these issues were included in the Doha negotiating agenda, but in 2003, several LDCs and developing countries asked for the exclusion of the three topics other than trade facilitation. Trade facilitation was formally accepted in 2004 by a core group of developing countries as part of the negotiating program in an effort to restart the Doha talks after the failure of the WTO Ministerial in Cancún in 2003 (ICTSD, 2013). The negotiating mandate on Trade Facilitation was included in Annex D of the so-called “July Package” (WT/L/579. 2 August 2004). Paragraph 1 stated that the objective was “to clarify and improve relevant aspects” of three articles of the WTO’s General Agreement on Tariffs and Trade (GATT) 1994: Article V (Freedom of Transit), Article VIII (Fees and Formali- ties connected with Importation and Exportation), and Article X (Publication and Administration of Trade Regulations). The final objective was “to further expediting the movement, release and clearance of goods, including goods in transit.” The same paragraph indicated that “negotiations shall also aim at enhancing technical assistance and support for capacity building in this area” and provide for “effective cooperation between customs or any other appropriate authorities on trade facilitation and customs compliance issues.” Paragraph 2 introduced a new concept by accepting that Special or Differential Treatment (SDT) “should extend be- yond the granting of traditional transition periods for implementing commitments” and that “the extent and the timing of entering into commitments shall be related to the implementation capacities of developing and least-developed Mem- bers.” To reinforce this point, it noted that “it is further agreed that those Members would not be obliged to undertake investments in infrastructure projects beyond their means.” This was a very relevant departure from previous approaches to SDT, suggesting that commitments be related to Members’ implementation capacity and the provision of effective technical support and foreign aid. It also directly ad- dressed developing countries’ complaints that the Uruguay Round committed them to WTO obligations without considera- tion of the costs involved. The novel idea that the commitments be related to the implementation capacity and the provision of effective technical support and foreign aid was incorporated in the Agreement on Trade Facilitation approved at Bali. Negotiations about the legal text of the July Package were complex. Some developing countries believed that the agreement would lead to import surges and create balance-of-payment problems. This contradicted the idea that several developing countries had about trade facilitation as a means to help them export more rather than become faster or more efficient importers. Developed countries, on the other hand, were concerned that linking commitments to the provision of foreign aid would significantly water down the WTO obligations envisaged in the ATF. There were also other technical issues, such as the treatment of pipelines as “trade in transit,” and complications aris- ing from customs cooperation stemming from confidentiality issues. As late as November 26, 2013, less than a week before the start of the Bali Ministerial, the text still had a significant number of brackets indicating areas in which the language had not been agreed upon (ICTSD, 2013). 7 SUMMARY | APRIL 2010 However, by the time the Bali Ministerial began the following week, all of these issues had either been resolved or dropped from the agreements (as in the case of the section on pipelines). Legal Aspects The ATF (document WT/MIN(13)/W/8) is the only new agreement negotiated in Bali. In order for this Agreement to enter into effect (i.e. become a binding commitment for the participating countries), it has to be ratified by two-thirds of the WTO members. As noted before, the heads of government and the ministers who approved the decisions in Bali represent their countries’ executive branches; in many countries, the concurrence of legislative bodies is also required by their respective constitutions to ratify international treaties. Before sending the final text for ratification, however, there were some steps that needed to come first, including a “cleaning up” of the text adopted in Bali. This clean up checks for word- ing and legal problems without changing the substance of what was agreed upon. That final text was to be reviewed and approved by the WTO General Council by July 31, 2014. After that, the ratification process in the different WTO members’ governments could start; the original Agreement indicated that countries had until July 31, 2015 to ratify the ATF. However, as noted before, the General Council of July 2014 was unable to reach an agreement because India condi- tioned its support to clarifications in the Bali Ministerial Decision on public food stocks. Negotiations continued until the dispute was resolved at the November 2014 meeting, as will be discussed further below. Regarding trade facilitation, the wording of the original agreement (WT/MIN(13)/W/8) was cleaned up and included in the document WT/L/940 approved in November 2014. This document opens with the decision of the WTO General Council approving the Protocol determining that the ATF will be included among the WTO trade agreements as Annex 1A of the Marrakesh Agreement (which established the WTO). It is followed by the actual text of the Protocol, establishing that the Protocol is open for acceptance by the WTO member countries, that it must be accepted in its entirety (i.e. without reservations by any country, unless all of the other WTO members agree otherwise), and that the ATF will enter into force and placed in Annex 1A to the WTO Agreement after the Agreement on Safeguards after two-thirds of the WTO members have ratified it (paragraph 3 of Article X of the WTO Agreement). An annex to the Protocol includes the final revised text of the ATF. The main changes in the final text were to correct some mistakes in the numbering in paragraphs, and to improve the grammar of some sections. The structure of the ATF was also changed. The original document (WT/MIN(13)/W/8) had institutional/implementation aspects in between specific commitments/disciplines, and SDT issues, while the new configuration in WT/L/940 shows three Sections: Section I pre- sents the specific commitments/disciplines; Section II covers SDT issues; and Section III includes institution- al/implementation aspects and final provisions. There is also an annex with a template for the notification of technical assistance needs under the SDT provisions. Section I includes 12 articles on the general obligations under the ATF regarding the following subjects: transparency and provision of information (Article 1); consultations before measures enter into force (Article 2); the possibility of advance rulings (Article 3); the appeal and review procedures when trade operators believe that custom operations are unfairly hindering their activities (Article 4); other measures to enhance impartiality, non-discrimination, and transparency (Articles 5); disciplines on fees and charges (Article 6); procedures for releasing and clearing goods (Article 7); cooperation between border agencies (Article 8); other aspects related to movement of goods intended for import under customs control (Article 9); procedures to simplify the movement of goods for export, import, and in transit (Article 10); specific disciplines to facilitate goods in transit (Article 11); and disciplines related to customs cooperation, including issues such as reciprocity, confidentiality, and so on (Article 12). Section II of the ATF includes the special and differential treatment for developing and LDCs. The original articles in this section (as presented in the Bali document WT/MIN(13)/W/8) were numbered differently than those in Section I; but, as part of the “cleaning up” of the text, the final document (WT/L/940) has the articles in Section II numbered 13 to 22 to sequentially follow the articles in Section I. In line with Annex D of the 2004 “July Package,” Section II includes the novel 8 SUMMARY | APRIL 2010 SDT provision already mentioned: the possibility that at least part of the potential commitments will become legal obliga- tions only if the country has received the technical or financial support it claims it needs to implement those commitments. While for developed countries, all obligations become binding when the ATF enters into force (i.e. after it is properly ratified by WTO members), the text allows developing countries and LDCs to divide the commitments under this Agree- ment into three categories (A, B, and C) and to phase them in under different schedules and conditions (Article 14 para- graph 1). Importantly, the designation of the obligations into the different categories is made directly by each developing country (Article 14, paragraph 2). There are three categories of provisions (Article 14, paragraph 1). Category A includes those obligations of Section I that a non-LDC developing country Member or an LDC Member has designated for early implementation, meaning imme- diately upon entry into force of the ATF in the case of non-LDC developing countries, and within one year of the entry into force for LDCs (Article 15). Category B contains provisions that can be implemented “on a date after a transitional period of time following the entry into force of this Agreement.” Developing countries must notify the Committee of Special and Differential Treatment of definitive dates for implementation of obligations under Category B no later than one year after the Agreement’s entry into force (Article 16 paragraph 1. b); LDCs must do the same by no later than two years (Article 16, paragraph 2b). Either group may ask for an extension of the period if necessary (Article 17). Category C includes those obligations of Section I which, according to the judgment of the Member in question, will require time and “the acquisition of implementation capacity through the provision of assistance and support for capacity building” in order to comply (Article 14, paragraph 1. c). In a footnote, it is clarified that such "assistance and support for capacity building” may “take the form of technical, financial, or any other mutually agreed form of assistance provided.” Developing countries must notify the Committee of their Category C obligations and the "assistance and support for capacity building” upon entry into force of the ATF (14.1. c). Within one year, they must inform the Committee of the “arrangements maintained or entered into that are necessary to provide assistance and support for capacity building to enable implementation of Category C” (14.1.d); and within 18 months, they “shall inform the Committee on progress in the provision of assistance and support” and at the same time “notify its list of definitive dates for implementation” (14.1. e). In the case of LDCs, the period to inform about needed capacity assistance is two years after entry in force of the ATF (14.2 d); after four years, LDCs must notify the Committee about their arrangements for capacity-building and indicative dates for implementation (14.2.e). Finally, after a total of five and a half years since ratification of the ATF, LDCs will notify the Committee of their definitive dates. Both non-LDC developing countries and LDCs also have the option of requesting extensions to the deadlines men- tioned for Categories A and B (14.1 b and 14.2. b), while for Category C they can invoke “the lack of donor support or lack of progress in the provision of assistance and support.” If the extension does not exceed 18 months (for a developing country) or 3 years (for an LDC), “the requesting Member is entitled to such additional time without any further action” (17.2). Article 17 sets out the procedures for requesting additional time. Article 19 allows some shifting of obligations between the B and C categories. If a developing country or an LDC believes that, after all other options have been exhausted, it still cannot implement a Category C commitment at all, it must notify the Trade Facilitation Committee (18.1), which will establish an Expert Group within 60 days; those experts must issue a recommendation to the Committee within 120 days of the group’s creation (18.2). During the process of study (up to 24 months), there cannot be cases under the Dispute Settlement Mechanism (18.5). In summary, the time periods for notification and implementation for developing countries and LDCs may last several years and depend on the provision of funds for technical assistance and capacity-building. The possibility of consultations (under Article XXII of GATT), claims of nullification, or impairment of concessions (under Article XXIII of GATT) are post- poned for two years for developing countries (20.1) and six years for LDCs (20.2) after the entry into force of the ATF for Category A commitments. The Ministerial Decision omits the case of developing countries’ Category B and C obligations 9 SUMMARY | APRIL 2010 and jumps to the case of those commitments by LDCs, which have protection from cases for eight years of after the implementation of the provision—rather than entry into force of the agreement, a noteworthy difference--under Articles XXII and XXIII of GATT 1994 (20.3). Additionally, claimants against an LDC must give particular consideration to their situation and exercise due restraint (20.4). Article 21 opens with donor Members (i.e. mostly developed countries) agreeing “to facilitate the provision of assis- tance and support for capacity building to developing country and least developed country Members, on mutually agreed terms and either bilaterally or through the appropriate international organizations.” Articles 21 and 22 and Annex 1 to the ATF then describe the information needed and the procedures for identifying and requesting such support. The wording of the obligations for donor Member countries implies less than a full obligation to provide technical assistance or funding. In return, non-LDC developing countries and LDCs may condition adherence to the implementation periods notified on receiving the expected support. Finally, Section III, Article 23 defines the institutional arrangements in place to administer the ATF within the WTO (creating a Committee on Trade Facilitation and defining the Committee’s functions and operations) and within the Member Countries (imposing on each the obligation to establish a National Committee on Trade Facilitation or to “desig- nate an existing mechanism to facilitate both domestic coordination and implementation of provisions of this Agree- ment.”). The ATF closes with Article 23 on Final Provisions, which indicates that all the provisions of this Agreement are binding on all WTO members, and that no reservations can be entered to those provisions without the consent of all Members. The Final Provision also indicates the starting date for implementation, the possibility that Members of a customs union or a regional economic arrangement could use regional approaches to assist in the implementation of their obligations, and the ATF’s links to other rights and obligations under GATT 1994, the Agreement on Technical Barriers to Trade, the Agreement on the Application of Sanitary and Phytosanitary Measures, and other topics, including the Dispute Settlement Mechanism. Finally, the Final Provision indicates that the commitments of developing countries and LDCs under Categories A, B, and C will be annexed to the ATF as an integral part of the Agreement. Significance of the ATF This agreement was mostly sought by developed countries-- they were the “demandeurs,” in trade diplomacy jargon-- but it can be argued that there are potential benefits for all categories of developing countries in the reduction of red tape and trade costs. The ATF may also limit the possibility of corruption and arbitrariness in customs operations and could, over time, improve the collection of customs revenues otherwise lost due to inefficiency and corruption. The ATF may also help small and medium-sized enterprises that currently find customs procedures incomprehensible or too costly. In a world of complex value chains, transparent, stable, and simplified customs procedures may allow developing countries and SMEs to become part of extended production arrangements. The simplified provisions for in-transit customs procedures would also help land-locked economies that trade through neighbors (Articles 10 and 11). For agricultural products, which are usually more perishable than non-agricultural goods, the reduction of time spent in customs processing may also improve trading conditions (Article 7 paragraph 9). Overall, if customs costs are calculated as an import tariff equivalent, some estimates have suggested that the Agree- ment, by reducing such tariffs, may lead to global economic benefits of up to one trillion USD (Hufbauer and Schott, 2013), although these figures should be considered as an upper, extremely optimistic bound.12 Using a more robust methodology in general equilibrium, but still without considering cost of implementation of the ATF, Decreux and Fontagne (2011) calculated a global gain of $72 billion USD. The OECD (Moïsé and Sorescu, 2013) calculates that improvements in trade- related information, the simplification and harmonization of documents, the streamlining of procedures, and the use of automated processes can lead to a reduction in trade costs of almost 14.5 percent for low-income countries, 15.5 percent for lower middle-income countries, and 13.2 percent for upper middle-income countries. Because the trade protection generated by ad-hoc customs procedures may differ significantly across products, economic agents, and even transactions, 12 The estimates for large effects have not been peer-reviewed. 10 SUMMARY | APRIL 2010 more transparent and uniform approaches to customs administration may reduce variability in trade protection even if the level of implicit trade protection does not change. Such a reduction is typically believed to increase welfare, although its benefits are more difficult to quantify. The problem for developing countries and LDCs is that building customs machinery takes money and effort, and these countries often face more pressing needs. Whatever the benefits of a better customs operation, there is an opportunity cost for the funds utilized that must be considered as well. Although the Agreement provides for external support in terms of money and technical assistance to build modernized customs system, it remains to be seen how effective those provi- sions are. At least for Category C commitments, a lack of effective external support for capacity-building can be invoked to evade the legal obligation under the agreement. As noted, this is a novel feature of the ATF. On the other hand, if the implementation of the ATF reduces corruption and leakages, it may lead to a net increase in government revenues enabling countries to better fund other priorities, particularly in the cases of low-income countries that depend more on trade taxes as public revenue. It must be also noted that developing countries, as explained previously, can self-define their commitments under the three categories with different time frames for implementation and urgency of commitments. This may dilute the impact of trade facilitation, as well as making the “one trillion dollar” evaluation of the deal even less realistic. Finally, developed countries will also face costs in implementing the ATF, since its impact will be less about removing pure inefficiencies but redistributing rents that are created by current laws and regulations in this area. The “Cuba Clarification” A final important point to note regarding trade facilitation is that the Bali Ministerial Declaration (WT/MIN(13)/DEC/W/1/Rev.1. 7 December 2013) includes the following language under the paragraph referring to the ATF: “In this regard, we reaffirm that the non-discrimination principle of Article V of GATT 1994 remains valid.” This refers to a complaint by Cuba, supported by some other Latin American countries, that its trade operations are being hurt by the US trade embargo; as a result, those countries decided near the end of the negotiations to withhold their support for the Bali Package. Given the consensual WTO decision-making system, if that group of countries had maintained their reserva- tions, the Bali Package would not have been approved. The language mentioned above was included to deal with those reservations; after this language was added, Cuba and its supporters accepted the whole agreement.13 AGRICULTURE AND COTTON A significant part of the language agreed on at Bali was based on the document “Revised Draft Modalities for Agriculture” (TN/AG/W/4/Rev.4, December 6, 2008), which in what follows will be referred to as the 2008 Modalities. Agriculture was of course part of the Doha Development Agenda, and this 2008 document was the last attempt to reach an agreement on agriculture before the general Doha talks collapsed in 2008. The text of this document, therefore, was never agreed upon; in the language of trade negotiations, it thus cannot be considered to be “stabilized.” In the negotiations leading up to the 9th Ministerial Conference and in the Bali Package, however, the link to the 2008 Modalities is clear (although with differ- ences discussed below). Background: General Services14 To better understand the implications of the text, it is useful to recall that Annex 2 of the Agreement on Agriculture (AoA) - -the so-called “Green Box,” although that name does not appear in any legal texts-- covers agricultural support programs 13 It should be noted that the debate about the Cuba embargo is not specific to the Doha Round and the Bali process. Cuba is raising this issue frequently at the WTO General Council. Indeed, the US sanctions started in 1962 and were extended by the Cuban Liberty and Democratic Solidarity Act (the Helms-Burton Act) of 1996 that applies restriction to third countries companies operating in Cuba. Interestingly, for this last reason, the issue was brought to the Dispute Settlement process of the WTO in 1996 by the European Union but the authority of the panel lapsed in 1998 after the EU suspended its request in 1997 (see Spanogle Jr. 1997 for a discussion). 14 The official Bali document is WT/MIN(13)/W/9. 11 SUMMARY | APRIL 2010 that meet “the fundamental requirement that they have no, or at most minimal, trade-distorting effects or effects on production” and “conform to the following basic criteria: (a) the support in question shall be provided through a publicly- funded government program (including government revenue foregone) not involving transfers from consumers; and, (b) the support in question shall not have the effect of providing price support” (Annex 2, paragraph 1). These Green Box measures do not have to be counted as distortionary domestic support (i.e. they are “exempted measures”). Annex 2 also includes an enumeration of Green Box domestic support measures, starting with a list in Paragraph 2 of “General Services” (GS) such as agricultural R&D, extension services, marketing information, and so on. These GS must “involve expenditures (or revenue foregone) in relation to programs which provide services or benefits to agriculture or the rural community. They shall not involve direct payments to producers or processors. Such programs, which include but are not restricted to the following list [emphasis added], shall meet the general criteria in paragraph 1” as well as other condi- tions that appear later in the second paragraph. Several developing countries claimed that such a list was biased toward GS that exist in rich countries and therefore may restrict other GS common in developing countries. These countries requested that the list be expanded to include programs such as land rehabilitation, soil conservation and resource management, drought management and flood control, rural employment programs, the issuing of land ownership titles, and settlement programs. The 2008 Modalities reflected that request; the list in Annex 2, paragraph 2, which originally included seven GS, was expanded in the proposed text of the 2008 Modalities to include a list of additional GS for developing countries. These additional GS were supposed “to promote rural development and poverty alleviation.” A group of developing countries asked that the language of the 2008 Modalities on General Services be approved at Bali. However, the final language in the Ministerial Decision on General Services (WT/MIN(13)/W/9) clearly departed from the 2008 Modalities, as will be discussed in the next section. Legal Aspects The Ministerial Decision approved in Bali refers to Annex 2, paragraph 2 of the AoA. “Members recognize the contribution that General Services programs can make to rural development, food security and poverty alleviation, particularly in developing countries,” including those related to “land reform and rural livelihood security that a number of developing countries have highlighted as particularly important in advancing these objectives.” The Bali Ministerial Decision goes on to say that “Members note that, subject to Annex 2 of the Agreement on Agricul- ture, the types of programs listed below could be considered as falling within the scope of the non-exhaustive list of general services programs in Annex 2, paragraph 2 of the AoA.” These would be “General Services programs related to land reform and rural livelihood security, such as: i. land rehabilitation; ii. soil conservation and resource management; iii. drought management and flood control; iv. rural employment programs; v. issuance of property titles; and vi. farmer settlement program, in order to promote rural development and poverty alleviation.” We highlight two points in the wording agreed upon at Bali. First, Members “note” that the programs listed “could” be part of the “non-exhaustive list” of Annex 2, paragraph 2, always “subject to Annex 2 of the Agreement on Agriculture.” Therefore, the list in Annex 2, paragraph 2 has not been modified, as it was in the 2008 Modalities; only a notional list of potential programs “could” be considered as part of General Services, if they comply with the Annex 2 of the AoA. In the 2008 Modalities, the language was more direct, inserting a final item (h) in the enumeration of Annex 2, paragraph 2. The language from the Bali Ministerial Decision on this topic also reiterates that the list in Annex 2, paragraph 2 of AoA is not limiting (“non-exhaustive”). The second point to be noted is that the list in the 2008 Modalities included infrastructural services and nutritional food security, which were not included in the text agreed on in Bali. The issue of nutritional food security had its own separate debate, to be discussed below. The notion of infrastructural services without any qualification potentially provid- ed developing countries with an option in addition to the previous section (g), which refers specifically to infrastructural 12 SUMMARY | APRIL 2010 services limited “to the provision or construction of capital works only, and shall exclude the subsidized provision of on farm facilities other than for the reticulation of generally available public utilities. It shall not include subsidies to inputs or operating costs, or preferential user charges.” Significance The usefulness of developing countries’ choice to act as “demandeurs” regarding this specific section of the 2008 Modali- ties is debatable. Although clarifications that could avoid future disputes are always useful, these countries may have used up a “bargaining chip” in asking for something that there was already available under any reasonable interpretation of Annex 2. First, the pre-Bali language of Annex 2, paragraph 2 clearly indicates that the list in paragraph 2 is not limiting, as highlighted previously. The Bali Ministerial Decision on this topic further refers to the list as “non-exhaustive.” Second, several of the items in the new list can be interpreted as variations of the programs already enumerated in Annex 2, paragraph 3 or elsewhere. Third, even without the new enumeration in the Bali Ministerial Decision, there is basically zero likelihood of a Member country bringing a case regarding the programs in the new list to the Dispute Settlement mecha- nism if they conform to the general principles of Annex 2. Finally, and most importantly, the language negotiated in Bali does not change the list itself; it only states that other things “could” be added to the “non-exhaustive list,” provided these other things comply with the criteria in Annex 2 of the AoA. In our opinion, developing countries were asking for something that they already had under any reasonable interpre- tation of the AoA; further, the new language does not change what was there already. It was for this reason that this ostensible expansion of the list of GS was readily accepted by the Members that were not “demandeurs.” In the peculiar logic of trade negotiations, it looks as if the latter group conceded something to the developing countries requesting the change, while in fact those developing countries, as “demandeurs,” may have just used up a bargaining chip and gotten little of substance in return. PUBLIC STOCKHOLDING FOR FOOD SECURITY PURPOSES15 Background16 Public stockholding for the purpose of food security was a controversial issue in the negotiations before, during, and after the Bali Ministerial, leading to the breakdown of the talks at the WTO General Council in July 2014. The agreement reached at the WTO General Council in November of that year meant that the steps that needed to be completed before ratifica- tion of the ATF could commence were done. On the other hand, the compromise reached on public food stocks only clarified the period of operation of the “peace clause,” as sought by India; the substance of the original controversy remains to be sorted out. The legal debate revolves around two sections of Annex 2 of the AoA (aka. Green Box): food security stocks (Annex 2, paragraph 3) and domestic food subsidies (Annex 2, paragraph 4). Initially, a group of developing countries known at the WTO as the G-3317 presented a proposal based on the 2008 Modalities that included new language for paragraph 3 (Public Stockholding for Food Security Purposes) and paragraph 4 (Domestic Food Aid). To understand the suggested modifications by the G-33, it is necessary to look first at the current language. The provi- sion on food security stocks (Annex 2, paragraph 3) declares “expenditures (or revenue foregone) in relation to the accu- mulation and holding of stocks of products which form an integral part of a food security program identified in national 15 The relevant document from Bali is WT/MIN(13)/W/10 and the one from the November General Council is WT/L/939. 16 This section draws significantly from Díaz-Bonilla, 2014, where a more detailed discussion of this topic can be found. That work is here updated with the new language approved on November 27, 2014. 17 As is the case with many international groups, the G-33 has a different number of members (46) than the name suggests. The countries are Antigua and Barbuda, Barbados, Belize, Benin, Bolivia, Botswana, Côte d’Ivoire, China, Congo, Cuba, Dominica, Dominican Republic, El Salvador, Grenada, Guatemala, Guyana, Haiti, Honduras, India, Indonesia, Jamaica, Kenya, Korea, Madagascar, Mauritius, Mongolia, Mozambique, Nicaragua, Nigeria, Pakistan, Panama, Peru, Philippines, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Senegal, Sri Lanka, Suriname, Tanzania, Trinidad and Tobago, Turkey, Uganda, Venezuela, Zambia, and Zimbabwe. 13 SUMMARY | APRIL 2010 legislation” to be Green Box measures. It also adds conditions, such as that “the volume and accumulation of such stocks shall correspond to predetermined targets related solely to food security,” and “the process of stock accumulation and disposal shall be financially transparent.” However, a footnote simplifies the criteria for developing countries: food security stocks are considered in conformity with Annex 2, paragraph 3, if the operation “is transparent and conducted in accord- ance with officially published objective criteria or guidelines.” Annex 2, paragraph 3, also indicates that “food purchases by the government shall be made at current market prices and sales from food security stocks shall be made at no less than the current domestic market price for the product and quality in question.” Note that purchases for food security stocks, and sales from them, must be made at market prices. This is again modified for developing countries on at least one account: footnotes 5 and 6 (a combined footnote so num- bered) applies both to Public Stockholding for Food Security Purposes (Annex 2, paragraph 3) and Domestic Food Aid (another Green Box measure in Annex 2, paragraph 4) and allows the selling of products at subsidized prices “with the objective of meeting food requirements of urban and rural poor in developing countries on a regular basis.” Although this stipulation allows developing countries to subsidize the selling price, footnote 5 to the food security stock provision does not permit other than market prices for purchases to be part of the Green Box. If purchases are at “administered prices,” then the “difference between the acquisition price and the external reference price is accounted for in the AMS [Aggregate Measure of Support].” 18 The AMS is a residual category of domestic support after different types of domestic support, either presumably non-distorting (e.g. Annex 2 of the AoA aka. Green Box), or distorting but exempted from being counted (e.g. Article 6.4 of the AoA, aka. Blue Box, and Article 6.2, which applies only for developing countries). The remaining domestic support measures not included in these three categories are a residual category, usually called the Amber Box. These measures must be added in an AMS for each product (that is, product-specific AMS) and for agricultural producers in general (non-product-specific AMS) (Brink 2011). Product-specific support includes an estimation of market price support (MPS), which “shall be calculated using the gap between a fixed external reference price and the applied administered price multiplied by the quantity of production eligible to receive the applied administered price” (Annex 3, paragraph 8). It is important to note the three crucial concepts in that statement—the fixed external reference price (FERP), the applied administered price (AP), and the idea of “produc- tion eligible”—because they feature prominently in the legal issues discussed below. Product-specific AMS must also include other nonexempt production-related payments and support to producers. The sum of MPS and other nonexempt payments is then compared with the value of production. If the sum is more than a de minimis level (5 percent for developed countries and 10 percent for developing countries),19 then the value is computed in its entirety (that is, not only the excess over the de minimis) in the current total aggregate measurement of support (CTAMS). The non-product-specific support (usually measured through budgetary data) also needs to be compared to the entire agricultural production; if it exceeds the de minimis value, it must be added to the CTAMS. Finally, the CTAMS is compared to, and cannot exceed, the ceiling commitment (sometimes called the final bound total AMS, or FBTAMS), which is negotiated during the Uruguay Round or defined later during the accession process for new WTO members (Brink 2011). Most developing countries have not declared domestic support in the negotiations-- that is, the amount considered in the FBTAMS-- so the de minimis limits these countries’ level of domestic support.20 The small number of developing countries that have declared domestic support have the possibility of offering domestic support up to the FBTAMS; however, the value of that domestic support is small compared to the allowances negotiated by developed countries during the Uruguay Round.21 18 19 Countries like China have accepted a different de minimis (in China’s case, 8.5 percent) as part of the accession agreement. 20 The relevant text in Article 7(b) reads: “Where no Total AMS commitment exists … the Member shall not provide support to agricultural producers in excess of the relevant de minimis level set out in paragraph 4 of Article 6.” 21 A total of 33 countries have FBTAMS, but the European Union, Japan, the United States, Canada, Switzerland, and Norway represent 87 percent of the total value (the European Union alone represents about 49 percent of the total). 14 SUMMARY | APRIL 2010 Some developing countries, particularly India, believed that if they had to account for the gap between administered prices and fixed external reference prices, then they would be bumping against, and probably exceeding, the product- specific limit of 10 percent de minimis of total production in some key products.22 Further, they argued, given the current high international prices, it did not make sense to compare buying prices to the external reference prices that were specifically defined under the AoA as those prevailing in 1986–1988. In fact, if purchases were done at administered prices that closely tracked current world prices (and therefore would not be distortionary in an economic sense), the AoA com- parison with the 1986–1988 levels would still show (largely imaginary) levels of market price support, as can be inferred when comparing the lower nominal values for 1986–1988 compared to the higher current prices. Based on those concerns, the language proposed by the G-33 exempted the difference between administered prices and the FERPs from the obligation of being included in the AMS when the governments of developing countries have bought products for food security stocks (paragraph 3) and domestic aid (paragraph 4) from a specific type of producer— that is, those that are “low income or resource poor” (LIRP; this category is already considered by the AoA for some special treatment in Article 6.223). This approach generated two basic objections. First, it appeared to go against the conditions established for the Green Box (Annex 2, paragraph 1)—in particular, the second basic criteria (point b in the paragraph), which indicates that “the support in question shall not have the effect of providing price support” (Annex 2, paragraph 1). The G-33 proposal, on the other hand, clearly provided price support, at least to a certain type of producer; the concern was that once a loophole was created in the Green Box chapeau, then anything could happen with the rest of the programs listed there. Furthermore, other developing countries were concerned about the leeway granted under the current Annex 2 to provide income support that is, in theory but not clearly in fact, decoupled from prices. Offering price support to LIRP producers would significantly undermine the possibility of enforcing other WTO disciplines considered in the Green Box measures which are currently used mostly by industrialized countries24 and which may create more than the minimal trade distortions required to be considered a Green Box measure (Annex 2, paragraph 1). The second objection arose from concern that the stocks allegedly accumulated for food security reasons may end up being sold on world markets. WTO members using the new allowance to provide price support to LIRP producers could accumulate products in excess of some desired ratio of stock-to-consumption and then decide to sell those surpluses in external markets to help finance the program’s fiscal cost. In fact, during the 2011/2012 and 2012/2013 periods, about 20 percent of Indian exports of wheat where drawn from public stocks.25 In general, any public stockholding programs involv- ing administrated prices, or at least price management within some range, may require the use of variable trade policy instruments (export subsidies/restrictions for exported commodities, import duties adjustments for imported commodi- ties) as demonstrated in the previous version of the EU CAP. These operations may create large negative effects in trade partners if there are no strict rules limiting those operations.26 22 This limit does not affect other options, such as the rest of the Green Box measures, Blue Box measures of support, and, for developing countries only, those considered in Article 6.2. 23 Note that the relevant section of Article 6.2 of the AoA says to refer to “investment subsidies,” which are “generally available” to LIRPs. An issue to be considered is whether “generally available” means only for LIRP producers. Also, the wording refers to “low-income or resource-poor producers” (emphasis added), which seems to expand the scope of the category when compared to “low-income and resource-poor producers” (Lars Brink, personal communication; emphasis added). 24 Some developing countries, such as China, use that type of support. 25 Based on data available at http://dfpd.nic.in/ 26 India has enough trade policy space in this regard, considering that tariffs are bound at high levels, and export taxes and restrictions are not disciplined (and the country has used them in the recent past). For instance, some countries, such as Pakistan, argued during the Bali negotiations that India was exporting rice from food stocks, thus affecting global rice markets and their own domestic markets and food security. India has become the first or second world rice exporter in recent years. Other people have argued that the rice exported by India and the rice used in its food security stocks are of different qualities (basmati rice being the exported product and common rice the one for domestic consumption); therefore, the postulated impact on global markets of domestic food stocks would not exist. However, Dorosh and Rashid (2012) showed that rice prices in Bangla- desh and subsidized prices for that product from India’s public stocks appear highly correlated. This was the result of the operation of private-sector importers that helped stabilize the operation of the rice market in Bangladesh to the benefit of poor consumers but with negative impact on produc- ers. 15 SUMMARY | APRIL 2010 Therefore, negotiators looked at other options, including the possibility of changing the FERPs of 1986–1988, adjusting the definition of the production eligible to receive administered prices, and imposing a temporary standstill to challenges under the Dispute Settlement Mechanism (DSM), which may arise if a country breached its allowed levels of domestic support because of government procurement of products for food security stocks at administered prices. However, changing the FERPs would have opened an entirely new set of difficult issues; for instance, the valuation of the commitments by countries, most of them developed countries, with declared domestic support in the base year. The notion of eligible production was also a key variable because, according to the AoA, the gap between the administered price and the FERP for 1986-1988 must be multiplied by the quantity of all production of a WTO member that is eligible to be bought at the administered prices. Some Members have argued that this gap should be multiplied by total quantity of a country’s production, while others made the case that it should be the quantity actually bought by the government. The interpretation of the term was clarified by the decision in the dispute about Korean beef (WTO 2000a); although the ruling left out other options to define eligible production (discussed below), this line of thinking was considered too complicated to be sorted out in time for the Bali Ministerial Declaration. In the end, the approach followed in Bali was to implement the peace clause as an interim solution27 (see the discus- sion of different options in Díaz-Bonilla, 2014, and Matthews, 2014a). Legal Aspects: the Bali Ministerial Decision (WT/MIN(13)/W/10) India considered the language of the draft Ministerial Decision that negotiators took to Bali to be inadequate, leading to a series of iterations before a final text was accepted (although, as the dispute in the July General Council showed, India was still not completely in agreement). To understand these disagreements, it is useful to distinguish three different issues: (1) the substance and coverage of the peace clause, (2) the conditions for its application, and (3) the period and conditions under which the peace clause is operational. The substance (point 1) and the conditions (point 2) for the peace clause did not change from the original language (more on this below). What did change was the definition of the period during which the peace clause was operational (point 3). In the original language, the clause lasted until the Tenth Ministerial Conference (about two years after Bali), at which time WTO members “will decide on next steps.” The new language indicates that “members agree to put in place an interim mechanism . . . , and to negotiate on an agreement for a permanent solution … for adoption by the 11th Ministerial Conference” (paragraph 1), or about four years after Bali. The new language thus changed the period covered by the peace clause and the conditions to end this mechanism, as discussed later. Opening with the substance and coverage of the peace clause (point 1), paragraph 2 of the Ministerial Decision goes on to state that WTO members cannot challenge developing country Members regarding compliance with the obligations of not exceeding their AMS (Article 6.3) or the de minimis (Article 7.2. b) when the following conditions apply:  It is “support provided for traditional staple food crops in pursuance of public stockholding programs for food security purposes.”  The programs protected from challenges are only those that exist as of the date of the decision;  Those programs must be consistent with the rest of the criteria of Annex 2 for food security stocks (other than the issue of price support; see the earlier discussion);  They must comply with other conditions established in the ministerial decision.28 27 There is a non-trivial difference between “temporary” (which seems to imply a specific date as a deadline) and “interim” (which does not necessarily have a specific termination date). While some people referred to the Ministerial Decision as a “temporary” solution, some of the countries requesting the Decision call it an “interim” one. The clarification sought by India focuses on this issue (see below). 28 The complete language is as follows: “2. In the interim, until a permanent solution is found, and provided that the conditions set out below are met, Members shall refrain from challenging through the WTO Dispute Settlement Mechanism, compliance of a developing Member with its obligations 16 SUMMARY | APRIL 2010 One footnote to the Ministerial Decision indicates that if and when a permanent solution is found, it will apply to all developing countries. Another footnote mentions that developing countries can initiate new programs that comply with Annex 2, paragraph 3-- but, as noted, only existing programs as of the date of the decision are covered by the peace clause).29 As mentioned above, the Ministerial Decision establishes additional conditions (point 2) in paragraphs 3, 4, and 5 for those developing countries that currently operate food stock programs and want to be protected from legal challenges by the peace clause. Those conditions did not change from the original draft, and are in addition to the relevant conditions in Annex 2, paragraph 3. They include the need for countries to:  notify the Committee on Agriculture “that it is exceeding or is at risk of exceeding either or both30 of its Aggre- gate Measurement of Support (AMS) limits (the Member’s Bound Total AMS or the de minimis level)”;  be current on notifications of its domestic support;  provide timely information for each public stockholding program maintained for food security purposes, plus other related information according to a template included in the annex of the decision;  ensure that stocks procured under food security programs do not distort trade or adversely affect the food se- curity of other Members (this reference was also an addition to the original language, which referred only to trade distortions); and  make sure that the potential increase in domestic support in excess of the allowed levels is only the amount notified to the Committee on Agriclture (see the first point mentioned above). Furthermore, any developing country benefiting from this decision must accept the requests for consultations by any other WTO member countries that may be interested in the operation of the notified public stockholding program or programs (paragraph 6). The decision also instructs the Committee on Agriculture to monitor the information submitted under this decision (paragraph 7) and indicates that WTO members must agree to establish a work program in which the committee can make recommendations for a permanent solution no later than the Eleventh Ministerial Conference. Advances will be reported to the General Council during the Tenth Ministerial Conference (paragraphs 8, 9, and 10). The main point to be noticed is the change in the period considered and the conditions for ending that period. In the original formulation, the peace clause had a specific end--the Tenth WTO Ministerial (about two years from Bali, consider- ing the usual time between ministerial meetings)—after which it would lapse and WTO members would decide on next steps. With the new language, Member countries commit to finding a permanent solution by the Eleventh Ministerial about four years from now, and the peace clause will remain in place “until a permanent solution is found.” There remains, however, some ambiguity in the text: on the one hand, it ostensibly defines a deadline (the Eleventh Ministerial), but on the other, it refers to the peace clause as being in effect until a “permanent solution is found” without specifying when that may happen. By the first interpretation, the day of reckoning is extended from the two years in the original language to four years in the text finally agreed upon; after that, the protection of the “peace clause” lapses and challenges in the under Articles 6.3 and 7.2 (b) of the AoA in relation to support provided for traditional staple food crops in pursuance of public stockholding programs for food security purposes existing as of the date of this Decision, that are consistent with the criteria of paragraph 3, footnote 5, and footnote 5&6 of Annex 2 to the AoA when the developing Member complies with the terms of this Decision.” 29 Also there may be some discussion as to the product scope, depending on the interpretation of words such as “traditional staple food crops,” and “primary agricultural products” (which may go beyond crops) (see Díaz-Bonilla, 2014) 30 Some observers have noted that the use of both is a drafting mistake, considering that a WTO member is constrained only by the bound total AMS or the de minimis under the Agreement (Lars Brink, personal communication; see also Matthews 2014). 17 SUMMARY | APRIL 2010 Dispute Settlement mechanism may take place.31 The latter interpretation, on the other hand, would strengthen the protection of the countries involved (e.g. India, or any other country following the same approach) to the extent that the country will avoid challenges under the AoA until a solution is found. In addition, under the WTO’s consensus approach, potential solutions may be blocked until there is one that those countries consider acceptable. This ambiguity is what was clarified in the November General Council. Legal Aspects: the Decision at the November General Council (WT/L/939). As noted, the breakdown in the WTO process at the July 2014 General Council was related to India’s discomfort with the ambiguity regarding whether the “peace clause” was operational for a specified time limit (determined to be until the Eleventh Ministerial Conference, or about four years) or for an undefined period until a permanent solution is found. The latter was India’s preferred interpretation and this is what was approved in the Ministerial Decision on “Public Stockholding for Food Security Purposes” of November 27, 2014 (document WT/L/939). The first paragraph states that “Until a permanent solution is agreed and adopted [emphasis added], and provided that the conditions set out in paragraphs 3 to 6 of the Bali Decision are met, Members shall not challenge through the WTO Dispute Settlement Mechanism compliance of a developing Member with its obligations under Articles 6.3 and 7.2(b) of the Agreement on Agriculture (AoA) in relation to support provided for traditional staple food crops in pursuance of public stockholding programmes for food security purposes existing as of the date of the Bali Decision, that are consistent with the criteria of paragraph 3, footnote 5, and footnote 5 and 6 of Annex 2 to the AoA.” To make things even clearer, the second paragraph indicates that “If a permanent solution for the issue of public stockholding for food security purposes is not agreed and adopted by the 11th Ministerial Conference, the mechanism referred to in paragraph 1 of the Bali Decision, as set out in paragraph 1 of this Decision, shall continue to be in place until a permanent solution is agreed and adopted.” With this language the “peace clause” is operational until a “permanent solution” is “agreed and adopted.” Note that the permanent solution has to be not only “agreed” upon but “adopted” as well. The Decision also urges that “the negotiations on a permanent solution on the issue of public stockholding for food security purposes shall be pursued on priority” (paragraph 3). To ensure that this happens, it sets out a proposed deadline (31 December 2015), defines the institutional framework for the negotiations (the Committee on Agriculture in Special Session, “in dedicated sessions and in an accelerated time-frame,”), and clarifies that it must take place as “distinct from the agriculture negotiations under the Doha Development Agenda ("DDA").” In paragraph 5 it establishes that the “General Council shall regularly review the progress of these dedicated sessions.” While the Bali Ministerial Decision instructed WTO members to establish a work program in the Committee on Agricul- ture to make recommendations for a permanent solution no later than the Eleventh Ministerial Conference (paragraphs 8, 9, and 10), the November Decision shortens the time frame by about two years (December 2015) and creates a distinct institutional track (the CoA in special, separated sessions) and conceptual negotiating framework (discrete from the Doha negotiations on Agriculture). Significance The notification and transparency requirements are not mere formalities when one considers that many WTO members-- including India but also other developed and developing countries as well-- are extremely behind schedule in their notifica- tions under the current obligations of the AoA (see the discussion in Orden, Blandford, and Josling 2011 in general, and Gopinath 2012 in the case of India). In order to invoke the peace clause, therefore, WTO members will have to complete notifications of domestic support and be open to consultations and questions regarding the actual operation of the food 31 WTO members could still mount challenges under other legal texts, such as the Agreement on Subsidies and Countervailing Measures, if they consider themselves affected by the operation of governmental purchases under this scheme. Whatever the legal issues involved, there is always the public relation issue of challenging a developing country on a program that is allegedly aimed at helping the poor. 18 SUMMARY | APRIL 2010 stock programs. Doing so would allow for more transparency and facilitate closer scrutiny and monitoring of the different programs of domestic support in the countries using the peace clause option. Additionally, only those food security programs existing at the time of the Ministerial Decision are covered. Matthews (2014) identified 16 developing countries under the WTO definition.32 It is also clear that any developing country (and developed countries as well) can provide subsidized food to its own population under paragraph 4 of Annex 2 of the AoA. The claim of some NGOs that it is unfair that the US can have a food stamp program while denying India the right to have a similar program is mistaken. The question is not if a country can give subsidized food to its poor, which is allowed under the AoA, but how governments procure that food; for which, as dis- cussed in greater detail in Díaz-Bonilla (2014), a proper clarification of the links between administered prices and market prices would be needed. More generally, the issue is whether a significant policy space to support food stocks in a country may affect its trading partners, which should receive guarantee that such programs will not lead to ad-hoc adjustments in export and import policies that generate instability and other negative externalities for the rest of the world. Now that the ambiguity of the Bali Decision has been eliminated, it is clear that the peace clause will be operational until a permanent solution is agreed and adopted. As noted before, it remains to be seen whether challenges under other legal texts, such as the Agreement on Subsidies and Countervailing Measures, could be mounted by countries believing that they are affected by the operation of governmental purchases under this scheme. Finally, finding a permanent solution is now under an accelerated time frame and separate institutional and conceptual frameworks. Therefore, the WTO work program during 2015 will be dedicated to find a permanent solution. The latter may take different forms, including some of the options discussed in the pre-Bali negotiations that were discarded because of a lack of time to complete adequate negotiations (see a full discussion of options in Díaz-Bonilla, 2014 and Matthews, 2014). In particular, as indicated before, we believe that it would be very useful to clarify the link between administered and market prices, considering the peculiar conditions of agricultural production in developing countries and the issue of inflation (see Díaz-Bonilla 2013 and 2014). UNDERSTANDING ON TARIFF RATE QUOTA ADMINISTRATION PROVISIONS OF AGRICULTURAL PRODUCTS, AS DEFINED IN ARTICLE 2 OF THE AGREEMENT ON AGRICULTURE33 Background Tariff-rate quotas (TRQs) were implemented during the Uruguay Round agreements to maintain or increase some minimal market access in countries with closed domestic markets for specific agricultural products, mainly those that were trans- forming quantitative restrictions into tariffs. TRQs have lower taxes on the quantities imported within the quota and higher taxes (usually high enough to inhibit trade) for quantities outside the quota. Another WTO negotiating group (the WTO G- 2034, not to be confused with the economic and financial G-20), which includes several developing countries that are producers and exporters of agricultural products, raised concerns about TRQs remaining substantially under-filled and argued that this was mainly because of manipulations in the way TRQs are administered by importing countries, and not because of valid market reasons. For instance, in 1996, the simple average of the filling rate of TRQ negotiated during the Uruguay Round was 66%; in 1999 this average had fallen to 50% (G/AG/NG/S/8, May 26, 2000). In the 2002-2004 period, at the beginning of the Doha Round, only 44% of the TRQs where filled by 80% or more.35 These figures show that the 32 Those countries are Botswana, Brazil, China, Costa Rica, India, Indonesia, Israel, Kenya, Saudi Arabia, Republic of Korea, Namibia, Nepal, Pakistan, Philippines, South Africa, and Sri Lanka. Note the presence of Israel and Republic of Korea, which, according to the WTO rules, have defined them- selves as developing countries. 33 The relevant document is WT/MIN(13)/W/11. 34 The WTO G-20 has in fact 23 members: Argentina, Bolivia, Brazil, Chile, China, Cuba, Ecuador, Egypt, Guatemala, India, Indonesia, Mexico, Nigeria, Pakistan, Paraguay, Peru, Philippines, South Africa, Tanzania, Thailand, Uruguay, Venezuela, Zimbabwe. 35 Based on the MAcMapHS6v1.1 database. 19 SUMMARY | APRIL 2010 concerns of existing and potential exporters are relevant. The G-20 presented a proposal to address the under-filling also extracted from the 2008 Modalities, which formed the basis for the agreement at Bali. Legal Aspects The agreed-upon language works mainly as a modification/clarification of the Agreement on Import Licensing Procedures (AILP) (paragraph 1). In this sense, it has more teeth than several of the other texts agreed upon at Bali. The new text can be divided into three conceptually different groups of issues. The first one includes language that re- stricts some of the more common practices that an importing country may follow to limit the use of TRQs; the second group establishes procedures and parameters to define the reasons for the under-filling of TRQs; and the third group focuses on possible solutions when under-filling has been identified and when it does not result from market conditions. The monitoring of the obligations and the processing of complaints is done by the Committee on Agriculture (paragraph 11). Within the first group we can find some of the following practices:  Opening a TRQ to imports with a very short notice -- Importing countries now have to inform within 90 days prior to the opening; paragraph 2 of the Understanding on TRQ Administration modifies paragraph 4(a) of Article 1 of AILP.  Asking applicants to apply to several bodies -- Now they must apply to one administrative body only; paragraph 3 modifies paragraph 6 of Article 1 of AILP.  Taking a long time to process applications to TRQ licenses -- Paragraph 4 now says that the government must take no longer than 30 days for "as and when received" cases and no longer than 60 days for "simultaneous" considera- tion cases. Now “the issuance of licenses shall, therefore, take place no later than the effective opening date of the tariff quota concerned.” A qualification that exists in the current paragraph 5(f) of Article 3 of AILP, and that indicat- ed that the obligation to follow such time frames were valid “except when that was impeded for reasons not de- pending on the Member country” was removed by paragraph 4 as well.  Granting licenses on time but in quantities that do not make economic sense -- Paragraph 5 now indicates that “scheduled tariff quotas shall be issued in economic quantities.”  Discouraging applicants with burdensome administrative procedures -- Now “importing Members shall ensure that unfilled tariff quota access is not attributable to administrative procedures that are more constraining than an ‘abso- lute necessity’ test would demand” (paragraph 7, referring to Article 3.2 of AILP).  Issuing licenses to weak and/or phantom operators so that TRQs may not be fully utilized -- In paragraphs 8-10, there are now several obligations for importing Member countries issuing licenses, including the analysis of private sector operators that exhibit a pattern of not fully utilizing their licenses “for reasons other than those that would be expected to be followed by a normal commercial operator.” If such is the case, Members are obligated to ask those operators whether they would make the licenses available to other users and to consider this situation for the allo- cation of new licenses (related to Article 3.5 (j) of ALIP).  Not informing the rate of utilization of the TRQs -- Paragraph 6 now indicates that TRQ fill rates must be notified;  Not informing who is the holder of the TRQs so that exporters do not know whom to contact to export the prod- uct -- Now “Members shall make available the contact details of those importers holding licenses for access to scheduled agricultural tariff quotas,” subject to certain restrictions about confidentiality and consent of the holders of those licenses (paragraph 10). Annex A of the Understanding on TRQ Administration explains the second and third groups of obligations, jointly known as the “under-fill mechanism.” They are, respectively, a time frame and a procedure to determine whether a TRQ’s 20 SUMMARY | APRIL 2010 low rate of utilization is due to the manner in which the TRQ is administered, as opposed to valid market reasons, and determine the remedy that a country will then have to apply in order to eliminate or reduce the under-fill. The second group of obligations (related to time frame and procedures) is explained in Annex A, paragraphs 1-3. The process is divided into three years. If, in the first year, a Member has not notified the TRQ fill rate or if that rate is less than 65 percent, other Members can raise specific concerns and there will be an exchange of information about procedures, market circumstances, and related considerations; at the end of that exchange, Members can decide whether the matter has been resolved and the case is closed. If a Member country believes that the matter has not been resolved, it must provide the Committee on Agriculture with a document explaining the reasons why the matter requires further discussion. If for two consecutive years, the importing Member has not notified the TRQ fill rate or if that rate less than 65 percent, a concerned Member may request the CoA that the importing Member change its TRQ administrative procedures; if, as a result of the changes, the TRQ fill goes above 65 percent (or if the complaining Member informs the CoA that is satisfied with the explanations/changes), the case is closed. Finally, in the third year, the next phase of the under-fill mechanism continues if several conditions apply: first, if the fill rate is less than 65 percent or if no notification has been submitted for three consecutive years; second, if the fill rate has not increased in each of the preceding three years by more than some pre-specified annual increments (defined in the Annex); third, if interested Members have not concluded that the under-fill is due to market circumstances; and fourth, if an interested Member informs the Committee on Agriculture that it wishes to initiate the final stage of the under-fill mechanism. In other words, if after three years of information-sharing and consultations, a TRQ continues to be under-filled, then the importing country will reach the third group of obligations and will have to follow some method for administering the TRQs, as defined in paragraph 4 of Annex A, that would remedy the under-fill. The obligations are separated for developed and developing countries (the latter of which receive SDT on the operation of this Agreement, as explained below). An importing Member which is not a developing country must, if this third stage is reached, “provide unencumbered access via one of the following tariff quota administration methods: a first-come, first-served only basis (at the border); or an automatic, unconditional license on demand system within the tariff quota.” The method selected will be maintained by the importing Member for a minimum of two years, and if “timely notifications for the two years have been submitted,” the case will be closed. Developing country Members do not have to choose one of the two methods mentioned above; rather, they have the alternative of choosing other TRQ administration methods or even maintaining the current method in place. The method elected by the developing country Member must be notified to the Committee on Agriculture and that method must be maintained by the importing Member for a minimum of two years. If after that, the fill rate has increased by certain amounts prescribed in the Annex, the case is closed. Once a case is closed, the process has to go through the three-year cycle described above to reopen it (paragraph 2 in Annex A). Significance First, we must note the trade-off in the SDT provisions. Developed countries are obligated to follow one of two procedures, but they are not then asked to show results, provided they maintain the method for two years; the presumption is that if either of the methods (“first-come, first-served” on the border or granting licenses automatically and unconditionally on demand) does not increase fill rates, it must be because of valid market reasons. Developing countries, on the other hand, may select other methods for TRQ administration, but they must then show increases in the TRQ fill (although the percent- ages considered my end up being less than the 65% that triggered the revisions in the first place). These SDT provisions allegedly led to complaints by the United States against some large developing countries like China regarding whether it was fair for the latter (some of which are very competitive exporters for a range of agricultural and non-agricultural products) to avail themselves of the specific SDT provisions considered in Annex A (ICTSD, 2013). This 21 SUMMARY | APRIL 2010 issue is also related to the presence of state trading enterprises (STEs) on the import side (see Díaz-Bonilla and Harris, 2014) This apparently led to negotiations producing the confusing final paragraphs (14 and 15) of the main text of the Understanding (the WTO website, in a significant understatement, calls the results “intricate”). In paragraph 14, it says that “The General Council recommendations in relation to paragraph 4 shall provide for special and differential treatment. Unless the 12th Ministerial Conference decides to extend paragraph 4 of Annex A in its current or a modified form, it shall, subject to paragraph 15, no longer apply.” But then in paragraph 15, it indicates that “notwithstanding paragraph 14, Members shall continue to apply the provisions of paragraph 4 of Annex A in the absence of a decision to extend that paragraph, except for those Members who wish to reserve their rights not to continue the application of paragraph 4 of Annex A and who are listed in Annex B.” It is not clear whether paragraph 14 focuses on the SDT referenced at the start of that paragraph (which will lapse if the 12th Ministerial does not extend it) or on all obligations (for developed and developing countries) of paragraph 4. But then, paragraph 15 says that even if there is not an agreement to extend paragraph 4, Member countries must continue “to apply the provisions” in that paragraph, and, to confuse things even further, grants the possibility of Member countries opting out after the 12th Ministerial (about six years from now). In principle, a reasonable interpretation is that paragraph 14 refers only to the possibility that the 12th Ministerial allows the SDT provision to lapse, while paragraph 15 would maintain the other obligations. But this is still vague. This confusion may be why four developing countries (Barbados, El Salvador, Dominican Republic, and Guatemala) and one developed country (US) simultaneously asked to be included in Annex B, which allows them to not comply with paragraph 4 in the absence of a decision by the 12th Ministerial. This Understanding’s potential for increasing market access under TRQs may be seriously affected if large developing countries aggressively use the SDT provision in paragraph 4 while at the same time, the US opts out of the obligations defined in that paragraph. A larger issues in this debate is whether systemically important developing countries should or should not have the right to avail themselves of those SDT provisions that some would argue are oriented mainly to smaller developing coun- tries. A related significant topic is how to address the issue of importing STEs within the WTO negotiations (Díaz-Bonilla and Harris, 2014). Of course, the designation of “developing country” is an acquired right under the negotiations, and countries with that status are understandably reluctant to relinquish it, particularly if developed countries cling to the type of SDT they receive to protect their own agricultural policies. We will return to these issues in our conclusion. EXPORT COMPETITION36 Background Many analysts and negotiators have noted the peculiar situation in which export subsidies for industrial products are prohibited under the WTO (and, previously, GATT) agreements, while export subsidies for agricultural products (several of which, such as dairy and meat products, are in fact manufactured products) were allowed under GATT37 and only partially disciplined under the AoA (see, for instance, Díaz-Bonilla and Tin, 2006). While during the Uruguay Round, export subsidies in general were considered in greater detail in the Agreement on Subsidies and Countervailing Measures (ASCM) and their prohibition was reaffirmed. Export subsidies for agriculture, however, were allowed by the Agreement on Agriculture (AoA) for countries that were using them, although they had to be capped and then cut in both value and volume. In particular, while countries were allowed to apply countervailing duties to industrial goods, agricultural subsidies were given a differ- 36 The relevant document is WT/MIN(13)/W/12. 37 From 1986–97, European and US export subsidies amounted to about USD 135 billion, or the equivalent of almost 13 per cent of the value of all agricultural exports by the developing countries of Africa, Latin American and the Caribbean and Asia (minus China) combined during the period (Díaz- Bonilla and Reca 2000). 22 SUMMARY | APRIL 2010 ent treatment, which somewhat limited the possibility of imposing those duties until 2003 if the exporting country operat- ed within the quantity limits agreed in the Uruguay Round.38 Although there are several developing countries among the WTO members that notified export subsidies (14 out of the 25 WTO members with such notifications) and can thereby use export subsidies for agricultural products, industrialized countries represent 84% of the values still allowed under the current AoA (only the European Union amounts to 62% of the total value of allowed agricultural export subsidies) (FAO 2000). The utilization of most of the export subsidies by industrial countries, along with other advantages in do