WORLD TRADE ORGANIZATION


WT/DS267/RW

18 December 2007


(07-5499)

image

Original: English


UNITED STATES – SUBSIDIES ON UPLAND COTTON

Recourse to Article 21.5 of the DSU by Brazil


Report of the Panel

WT/DS267/RW

Page i


TABLE OF CONTENTS

Page


  1. INTRODUCTION 1

  2. BACKGROUND 2

  3. FACTUAL ASPECTS 4

    1. PRODUCTS AT ISSUE 4

    2. DEVELOPMENTS CONCERNING THE MEASURES FOUND TO BE WTO-INCONSISTENT BY

      THE ORIGINAL PANEL 5

      1. User marketing (Step 2) payments, marketing loan payments, market loss

        assistance payments and counter-cyclical payments 5

        1. User marketing (Step 2) payments 5

        2. Marketing loan payments 5

        3. Market loss assistance payments 7

        4. Counter-cyclical payments 7

      2. Export credit guarantees 8

  4. PARTIES' REQUESTS FOR FINDINGS AND RECOMMENDATIONS 10

  5. ARGUMENTS OF THE PARTIES 12

  6. ARGUMENTS OF THE THIRD PARTIES 12

  7. INTERIM REVIEW 12

    1. CHANGES REQUESTED BY THE PARTIES REGARDING THE "FACTUAL ASPECTS" SECTION

      OF THE INTERIM REPORT 12

    2. CHANGES REQUESTED BY THE PARTIES WITH RESPECT TO THE PANEL'S DESCRIPTION

      OF ARGUMENTS OF THE PARTIES 12

    3. REQUESTS FOR CHANGES WITH RESPECT TO THE PANEL'S FACTUAL AND LEGAL

      ANALYSIS 13

    4. EDITORIAL CORRECTIONS 25

  8. PROCEDURAL MATTERS 25

    1. REQUEST OF BRAZIL THAT THE PANEL EXERCISE ITS DISCRETION TO SEEK CERTAIN INFORMATION FROM THE UNITED STATES UNDER ARTICLE 13 OF THE DSU 25

    2. REQUEST OF THE UNITED STATES FOR OPEN HEARING 28

    3. SERVICING OF SUBMISSIONS 29

    4. REQUEST OF THE EUROPEAN COMMUNITIES FOR A FINDING OR RULING ON THE

      PANEL'S COMPOSITION 31

    5. LEAST DEVELOPED COUNTRY MEMBERS 31

  9. GENERAL CONSIDERATIONS AND ISSUES RELATING TO THE SCOPE

    OF THIS PROCEEDING 31

    1. STANDARD OF REVIEW, BURDEN OF PROOF AND RULES OF TREATY INTERPRETATION 31

      1. Standard of review 31

        WT/DS267/RW

        Page ii

      2. Burden of proof 32

      3. Rules of treaty interpretation 34

    2. SCOPE OF THIS PROCEEDING UNDER ARTICLE 21.5 OF THE DSU 34

      1. Introduction 34

      2. Whether Brazil's claims relating to export credit guarantees for pig meat and

        poultry meat are outside the scope of this proceeding 34

        1. Main arguments of the parties 34

        2. Main arguments of third parties 36

        3. Evaluation by the Panel 37

      3. Whether Brazil's claims with respect to the marketing loan and counter-cyclical payment programmes are outside the scope of this proceeding 39

        1. Whether Brazil's claims with respect to the marketing loan payment and counter- cyclical programmes are outside the scope of this proceeding because these programmes were not subject to recommendations and rulings of the DSB in the

          original proceeding 39

          1. Main arguments of the parties 39

          2. Main arguments of third parties 42

          3. Evaluation by the Panel 44

        2. Whether Brazil's claims with respect to the marketing loan and counter-cyclical payment programmes are outside the scope of this proceeding because these programmes have not been changed in response to DSB recommendations and rulings or otherwise and therefore are not "measures taken to comply" within the meaning of

        Article 21.5 of the DSU 48

      4. Whether Brazil's claim regarding the failure of the United States to take measures between 22 September 2005 and 31 July 2006 is outside the scope of

        this proceeding 48

        1. Main arguments of the parties 48

        2. Main arguments of third parties 49

        3. Evaluation by the Panel 50

      5. Whether "payments" made after 21 September 2005 pursuant to the marketing loan and counter-cyclical payment programmes are within the scope of this

    proceeding 52

    1. Main arguments of the parties 52

    2. Evaluation by the Panel 53

  10. FINDINGS ON BRAZIL'S CLAIMS OF "PRESENT" SERIOUS PREJUDICE

    UNDER ARTICLES 5(C) AND 6.3 OF THE SCM AGREEMENT 55

    1. INTRODUCTION 55

    2. BRAZIL'S CLAIM OF "PRESENT" SERIOUS PREJUDICE WITHIN THE MEANING OF

      ARTICLES 5(C) AND 6.3 (C) OF THE SCM AGREEMENT 56

      1. Introduction – overview of the main arguments of the parties 56

      2. Arguments of third parties 58

        WT/DS267/RW

        Page iii


      3. Preliminary considerations relating to the Panel's evaluation of Brazil's claim

        under Articles 5(c) and 6.3(c) of the SCM Agreement 59

        1. Text of legal provisions 59

        2. Period to be considered by the Panel 60

          1. Main arguments of the parties 60

          2. Main arguments of third parties 60

          3. Evaluation by the Panel 60

        3. Whether marketing loan and counter-cyclical payments are specific subsidies 61

        4. Magnitude of the marketing loan and counter-cyclical payments 62

          1. Marketing loan payments 62

          2. Counter-cyclical payments 63

        5. Product at issue, like product, relevant "market" and "price" 68

        6. Significant price suppression as the effect of the subsidy 70

      4. The alleged "substantial proportionate influence" of US upland cotton

        production and exports. 72

        1. Main arguments of the parties 72

        2. Evaluation by the Panel 73

      5. "Structure, design and operation" of the marketing loan and counter-cyclical

        payments 75

        1. Main arguments of the parties 75

        2. Evaluation by the Panel 76

      6. Alleged "large magnitude" of the marketing loan payments and counter-cyclical payments 89

        1. Main arguments of the parties 89

        2. Evaluation by the Panel 90

      7. Alleged "link between high levels of US subsidies and high levels of US planted

        acreage, production and exports" 92

        1. Main arguments of the parties 92

        2. Evaluation by the Panel 95

      8. Alleged "discernible temporal coincidence" of suppressed world market prices

        and large marketing loans and counter-cyclical payments 97

        1. Main arguments of the parties 97

        2. Evaluation by the Panel 98

      9. Alleged gap between long-term costs of production and market revenues of US

        upland cotton producers 103

        1. Main arguments of the parties 103

        2. Evaluation by the Panel 107

        1. Description of the data on commodity cost and returns 108

          WT/DS267/RW

          Page iv


        2. Whether certain items (land, unpaid labour and capital recovery costs) should be

          included as variable costs 109

        3. Cash or opportunity costs 109

        4. Total or variable costs of production 111

        5. The role of off-farm income 113

        6. US upland cotton producers' long-term costs of production and market revenues 115

        7. Other arguments 117

      10. The economic simulation model submitted by Brazil in this proceeding. 120

        1. Introduction 120

        2. Main arguments of the parties 120

          1. Structure of the model 122

          2. Parameters of the model 123

        3. Evaluation by the Panel 125

      11. Impact of the elimination of Step 2 payments 125

        1. Introduction 125

        2. Main arguments of the parties 126

        3. Evaluation by the Panel 128

        1. Assessing the effects on US cotton exports 128

        2. Indirect effect on marketing loan and counter-cyclical payments 129

      12. Other factors impacting the world market price for upland cotton 131

        1. Main arguments of the parties 131

        2. Evaluation by the Panel 131

      13. Summary and conclusion of the Panel's analysis of Brazil's claim of serious

        prejudice within the meaning of Articles 5(c) and 6.3(c) of the SCM Agreement 132

    3. CLAIM OF BRAZIL OF "PRESENT" SERIOUS PREJUDICE WITHIN THE MEANING OF

      ARTICLES 5(C) AND 6.3(D) OF THE SCM AGREEMENT 134

      1. Main arguments of the parties 134

      2. Evaluation by the Panel 135

  11. BRAZIL'S CLAIM OF THREAT OF SERIOUS PREJUDICE UNDER

    ARTICLES 5(C) AND 6.3 OF THE SCM AGREEMENT 137

  12. BRAZIL'S CLAIM OF "PRESENT" SERIOUS PREJUDICE UNDER ARTICLES 5(C), 6.3(C) AND 6.3(D) OF THE SCM AGREEMENT AS A RESULT OF THE FAILURE OF THE UNITED STATES TO TAKE ANY

    ACTION BETWEEN 22 SEPTEMBER 2005 AND 31 JULY 2006 137

  13. ATTACHMENT 1: OVERVIEW OF NEW ECONOMIC LITERATURE ON COUNTER-CYCLICAL PAYMENTS 138

    1. IMPACT OF COUNTER-CYCLICAL PAYMENTS 138

      1. Studies cited by the United States 138

        WT/DS267/RW

        Page v

      2. Studies cited by Brazil 139

    2. ARE COUNTER-CYCLICAL PAYMENTS CAPITALIZED INTO HIGHER LAND VALUES? 141

      1. Studies cited by the United States 141

  14. FINDINGS ON BRAZIL'S CLAIMS UNDER ARTICLES 10.1 AND 8 OF THE AGREEMENT ON AGRICULTURE AND ARTICLES 3.1(A), 3.2 AND 4.7 OF THE SCM AGREEMENT WITH RESPECT TO EXPORT CREDIT

    GUARANTEES 142

    1. INTRODUCTION 142

      1. Description of the export credit guarantees at issue in this dispute 142

      2. Summary of the original panel and Appellate Body's findings concerning export

        credit guarantees 143

    2. BRAZIL'S CLAIMS WITH RESPECT TO EXPORT CREDIT GUARANTEES OUTSTANDING ON

      1 JULY 2005 144

      1. Introduction 144

      2. Main arguments of the parties 145

      3. Main arguments of third parties 146

      4. Evaluation by the Panel 146

        1. Introduction 146

        2. Panel and Appellate Body reports in Brazil – Aircraft (Article 21.5 – Canada) 147

        3. Whether the obligation to "withdraw the subsidy" includes an obligation to abstain

          from performing on outstanding financial commitments. 148

        4. Meaning of the term "withdraw" 151

    3. BRAZIL'S CLAIMS CONCERNING GSM 102 EXPORT CREDIT GUARANTEES ISSUED

      AFTER 1 JULY 2005 152

      1. Introduction 152

      2. Brazil's claims under Articles 8 and 10.1 of the Agreement on Agriculture 152

      3. Are GSM 102 export credit guarantees "export subsidies" – legal bases for

        Brazil's arguments 154

      4. Brazil's arguments under item (j) of the Illustrative List of Export Subsidies 156

        1. Original panel's analysis under item (j) of the Illustrative List of Export Subsidies 156

          1. Past performance of the programmes 156

          2. Structure, design and operation of the programme 158

          3. Overall conclusion by the original panel 158

          4. Appellate Body 159

        2. "Quantitative" analysis under item (j) 159

          1. Main arguments of the parties 159

          2. Evaluation by the Panel 160

        3. Comparison with MPRs under the OECD Arrangement 167

          WT/DS267/RW

          Page vi

          1. Main arguments of the parties 167

          2. Main arguments of third parties 168

          3. Evaluation by the Panel 168

        4. Structure, design and operation of the GSM 102 programme 171

          1. Main arguments of the parties 171

          2. Evaluation by the Panel 172

        5. Overall conclusion of the Panel's analysis under item (j) of the Illustrative List 180

      5. Whether the United States has applied GSM 102 guarantees in a manner that results in circumvention of its export subsidy commitments under the Agreement

        on Agriculture 181

        1. Introduction 181

        2. Unscheduled products 181

        3. Scheduled products 182

      6. Claims that GSM 102 export credit guarantees are inconsistent with Articles

        3.1(a) and 3.2 of the SCM Agreement 185

    4. CLAIMS WITH RESPECT TO LYSINE, LYOCELL AND WOOD PRODUCTS 186

  15. CONCLUSIONS AND RECOMMENDATION 188

WT/DS267/RW

Page vii


LIST OF ANNEXES


ANNEX A


FIRST WRITTEN SUBMISSIONS OF PARTIES AND THIRD PARTIES


Contents

Page

Annex A-1

Executive Summary of the First Written Submission of Brazil

A-2

Annex A-2

Executive Summary of the First Written Submission of the United States (Including Request for Preliminary Rulings)

A-9

Annex A-3

Third Party Submission of Argentina

A-18

Annex A-4

Third Party Submission of Australia

A-22

Annex A-5

Third Party Submission of Canada

A-31

Annex A-6

Third Party Submission of Chad

A-38

Annex A-7

Third Party Submission of the People's Republic of China

A-52

Annex A-8

Third Party Submission of the European Communities

A-57

Annex A-9

Third Party Submission of Japan

A-72

Annex A-10

Third Party Submission of New Zealand

A-79


ANNEX B


REBUTTAL SUBMISSIONS OF PARTIES


Contents

Page

Annex B-1

Executive Summary of the Rebuttal Submission of Brazil

B-2

Annex B-2

Executive Summary United States

of

the

Rebuttal

Submission

of

the

B-8

Annex B-3

Executive Summary of Brazil's Submission Regarding US Requests for Preliminary Ruling

B-16


ANNEX C


ORAL STATEMENTS OF PARTIES AND THIRD PARTIES AT THE MEETING WITH THE PANEL


Contents

Page

Annex C-1

Opening Statement of Brazil (Executive Summary)

C-2

Annex C-2

Closing Statement of Brazil

C-10

Annex C-3

Opening Statement of the United States (Executive Summary)

C-18

Annex C-4

Closing Statement of the United States

C-26

Annex C-5

Third Party Oral Statement of Argentina

C-29

Annex C-6

Third Party Oral Statement of Australia

C-34

Annex C-7

Third Party Oral Statement of Canada

C-37

Annex C-8

Third Party Oral Statement of Chad

C-40

Annex C-9

Third Party Oral Statement of the People's Republic of China

C-45

Annex C-10

Third Party Oral Statement of the European Communities

C-47

Annex C-11

Third Party Oral Statement of India

C-56

Annex C-12

Third Party Oral Statement of Japan

C-58

Annex C-13

Third Party Oral Statement of New Zealand

C-61

Annex C-14

Third Party Oral Statement of Thailand

C-64

WT/DS267/RW

Page viii


ANNEX D


RESPONSES OF PARTIES TO THE PANEL'S QUESTIONS AND OTHER COMMENTS AND DOCUMENTS RECEIVED FROM PARTIES


Contents

Page

Annex D-1

Brazil's Request for the Panel to Seek Production of Documents and Information Pursuant to Article 13 of the DSU

D-2

Annex D-2

Responses of Brazil to the Panel's Questions on the DSU Article 13 Issue

D-18

Annex D-3

Responses of the United States to the Panel's Questions on the DSU Article 13 Issue

D-21

Annex D-4

Comments of Brazil on the Oral Statements of the United States at the Meeting with the Panel

D-24

Annex D-5

Comments of the United States on the Oral Statements of Brazil at the Meeting with the Panel

D-46

Annex D-6

Responses of Brazil to the Panel's First Set of Questions (Sections A-C)

D-78

Annex D-7

Responses of Brazil to the Panel's First Set of Questions (Sections D&E)

D-114

Annex D-8

Responses of the United States to the Panel's First Set of Questions (Sections A-C)

D-134

Annex D-9

Responses of the United States to the Panel's First Set of Questions (Sections D&E)

D-152

Annex D-10

Brazil's Comments on the Responses of the United States to the Panel's First Set of Questions

D-177

Annex D-11

United States' Comments on the Responses of Brazil to the Panel's First Set of Questions

D-220

Annex D-12 Responses of Brazil to the Panel's Second Set of Questions

D-269

Annex D-13

Responses of the United States to the Panel's Second Set of Questions

D-355

Annex D-14 Brazil's Comments on the Responses of the United States to the Panel's Second Set of Questions

D-443

Annex D-15

United States' Comments on the Responses of Brazil to the Panel's Second Set of Questions

D-518


ANNEX E


COMMUNICATIONS FROM PARTIES


Contents

Page

Annex E-1

Letter of the United States dated 7 November 2006

E-2

Annex E-2

Letter of the United States dated 20 November 2006

E-5

Annex E-3

Letter of the United States dated 21 November 2006

E-8

Annex E-4

Letter of Brazil dated 22 November 2006

E-10

Annex E-5

Letter of Brazil dated 24 November 2006

E-13

Annex E-6

Letter of Brazil dated 18 December 2006

E-15

Annex E-7

Letter of the United States dated 19 December 2006

E-16

Annex E-8

Letter of Brazil dated 22 January 2007

E-19

WT/DS267/RW

Page ix


Contents

Page

Annex E-9 Letter of Brazil dated 7 February 2007

E-21

Annex E-10 Letter of the United States dated 12 February 2007

E-22


ANNEX F


COMMUNICATIONS FROM THE PANEL TO PARTIES


Contents

Page

Annex F-1

Communication of 8 November 2006

F-2

Annex F-2

Communication of 27 November 2006

F-3

Annex F-3

Communication of 28 November 2006

F-4

Annex F-4

Communication of 20 December 2006

1

F-5

Annex F-5

Communication of 17 January 2007

F-6

Annex F-6

Communication of 22 January 2007

F-7

Annex F-7

Communication of 16 February 2007

F-8

Annex F-8

Communication of 21 March 2007

F-9


ANNEX G


WORKING PROCEDURES AND TIMETABLE OF THE PANEL


Contents

Page

Annex G-1 Working Procedures

G-2

Annex G-2 Timetable (Final)

G-5


ANNEX H


REQUEST FOR THE ESTABLISHMENT OF A PANEL


Contents

Page

Annex H-1 Request for the Establishment of a Panel by Brazil

H-2


image

1 This communication was sent to the Parties as well as to the Third Parties.

WT/DS267/RW

Page x


TABLE OF CASES CITED IN THIS REPORT


Short Title

Full Case Title and Citation

Australia – Salmon (Article 21.5 – Canada)

Panel Report, Australia – Measures Affecting Importation of Salmon – Recourse to Article 21.5 of the DSU by Canada, WT/DS18/RW, adopted 20 March 2000, DSR 2000:IV, 2031

Brazil – Aircraft (Article 21.5 – Canada)

Appellate Body Report, Brazil – Export Financing Programme for Aircraft

– Recourse by Canada to Article 21.5 of the DSU, WT/DS46/AB/RW, adopted 4 August 2000, DSR 2000:VIII, 4067

Brazil – Aircraft (Article 21.5 – Canada)

Panel Report, Brazil – Export Financing Programme for Aircraft – Recourse by Canada to Article 21.5 of the DSU, WT/DS46/RW, adopted 4 August 2000, as modified by Appellate Body Report, WT/DS46/AB/RW, DSR 2000:IX, 4093

Canada – Aircraft (Article 21.5 – Brazil)

Appellate Body Report, Canada – Measures Affecting the Export of Civilian Aircraft – Recourse by Brazil to Article 21.5 of the DSU, WT/DS70/AB/RW, adopted 4 August 2000, DSR 2000:IX, 4299

Canada – Aircraft (Article 21.5 – Brazil)

Panel Report, Canada – Measures Affecting the Export of Civilian Aircraft

– Recourse by Brazil to Article 21.5 of the DSU, WT/DS70/RW, adopted

4 August 2000, as modified by Appellate Body Report, WT/DS70/AB/RW, DSR 2000:IX, 4315

Canada – Dairy

Panel Report, Canada – Measures Affecting the Importation of Milk and the Exportation of Dairy Products, WT/DS103/R, WT/DS113/R, adopted

27 October 1999, as modified by Appellate Body Report, WT/DS103/AB/R, WT/DS113/AB/R, DSR 1999:VI, 2097

Canada – Dairy

(Article 21.5 – New Zealand and US)

Panel Report, Canada – Measures Affecting the Importation of Milk and the Exportation of Dairy Products – Recourse to Article 21.5 of the DSU by New Zealand and the United States, WT/DS103/RW, WT/DS113/RW, adopted

18 December 2001, as reversed by Appellate Body Report, WT/DS103/AB/RW, WT/DS113/AB/RW, DSR 2001:XIII, 6865

Canada – Dairy

(Article 21.5 – New Zealand and US II)

Appellate Body Report, Canada – Measures Affecting the Importation of Milk and the Exportation of Dairy Products – Second Recourse to Article 21.5 of the DSU by New Zealand and the United States, WT/DS103/AB/RW2, WT/DS113/AB/RW2, adopted 17 January 2003, DSR 2003:I, 213

Chile – Price Band System (Article 21.5 – Argentina)

Appellate Body Report, Chile – Price Band System and Safeguard Measures Relating to Certain Agricultural Products – Recourse to Article 21.5 of the DSU by Argentina, WT/DS207/AB/RW, adopted 22 May 2007

EC – Bed Linen (Article 21.5 – India)

Appellate Body Report, European Communities – Anti-Dumping Duties on Imports of Cotton-Type Bed Linen from India – Recourse to Article 21.5 of the DSU by India, WT/DS141/AB/RW, adopted 24 April 2003, DSR 2003:III, 965

EC – Bed Linen (Article 21.5 – India)

Panel Report, European Communities – Anti-Dumping Duties on Imports of Cotton-Type Bed Linen from India – Recourse to Article 21.5 of the DSU by India, WT/DS141/RW, adopted 24 April 2003, as modified by Appellate Body Report, WT/DS141/AB/RW, DSR 2003:IV, 1269

EC – Export Subsidies on Sugar

Appellate Body Report, European Communities – Export Subsidies on Sugar, WT/DS265/AB/R, WT/DS266/AB/R, WT/DS283/AB/R, adopted 19 May 2005

Indonesia – Autos

Panel Report, Indonesia – Certain Measures Affecting the Automobile Industry, WT/DS54/R, WT/DS55/R, WT/DS59/R, WT/DS64/R and Corr.1, 2, 3 and 4,

adopted 23 July 1998, DSR 1998:VI, 2201

Korea – Commercial Vessels

Panel Report, Korea – Measures Affecting Trade in Commercial Vessels, WT/DS273/R, adopted 11 April 2005

WT/DS267/RW

Page xi


Short Title

Full Case Title and Citation

Mexico – Corn Syrup (Article 21.5 – US)

Appellate Body Report, Mexico – Anti-Dumping Investigation of High Fructose Corn Syrup (HFCS) from the United States – Recourse to Article 21.5 of the DSU by the United States, WT/DS132/AB/RW, adopted 21 November 2001,

DSR 2001:XIII, 6675

US – Countervailing Measures on Certain EC Products (Article 21.5 – EC)

Panel Report, United States – Countervailing Measures Concerning Certain Products from the European Communities – Recourse to Article 21.5 of the DSU by the European Communities, WT/DS212/RW, adopted 27 September 2005

US – FSC

Appellate Body Report, United States – Tax Treatment for "Foreign Sales Corporations", WT/DS108/AB/R, adopted 20 March 2000, DSR 2000:III, 1619

US – FSC

Panel Report, United States – Tax Treatment for "Foreign Sales Corporations", WT/DS108/R, adopted 20 March 2000, as modified by Appellate Body Report, WT/DS108/AB/R, DSR 2000:IV, 1675

US – FSC

(Article 21.5 – EC)

Appellate Body Report, United States – Tax Treatment for "Foreign Sales Corporations" – Recourse to Article 21.5 of the DSU by the European Communities, WT/DS108/AB/RW, adopted 29 January 2002, DSR 2002:I, 55

US – FSC

(Article 21.5 – EC)

Panel Report, United States – Tax Treatment for "Foreign Sales Corporations"

– Recourse to Article 21.5 of the DSU by the European Communities, WT/DS108/RW, adopted 29 January 2002, as modified by Appellate Body Report, WT/DS108/AB/RW, DSR 2002:I, 119

US – FSC

(Article 21.5 – EC II)

Appellate Body Report, United States – Tax Treatment for "Foreign Sales Corporations" – Second Recourse to Article 21.5 of the DSU by the European Communities, WT/DS108/AB/RW2, adopted 14 March 2006

US – FSC

(Article 21.5 – EC II)

Panel Report, United States – Tax Treatment for "Foreign Sales Corporations"

– Second Recourse to Article 21.5 of the DSU by the European Communities, WT/DS108/RW2, adopted 14 March 2006, upheld by Appellate Body Report, WT/DS108/AB/RW2

US – Oil Country Tubular Goods Sunset Reviews

Appellate Body Report, United States – Sunset Reviews of Anti-Dumping Measures on Oil Country Tubular Goods from Argentina, WT/DS268/AB/R, adopted 17 December 2004, DSR 2004:VII, 3257

US – Oil Country Tubular Goods Sunset Reviews (Article 21.5 – Argentina)

Appellate Body Report, United States – Sunset Reviews of Anti-Dumping Measures on Oil Country Tubular Goods from Argentina – Recourse to Article

21.5 of the DSU by Argentina, WT/DS268/AB/RW, adopted 11 May 2007

US – Shrimp

(Article 21.5 – Malaysia)

Panel Report, United States – Import Prohibition of Certain Shrimp and Shrimp Products – Recourse to Article 21.5 of the DSU by Malaysia, WT/DS58/RW, adopted 21 November 2001, upheld by Appellate Body Report, WT/DS58/AB/RW, DSR 2001:XIII, 6529

US – Softwood Lumber IV (Article 21.5 – Canada)

Appellate Body Report, United States – Final Countervailing Duty Determination with Respect to Certain Softwood Lumber from Canada – Recourse by Canada to Article 21.5 of the DSU, WT/DS257/AB/RW, adopted 20 December 2005

US – Softwood Lumber VI (Article 21.5 – Canada)

Appellate Body Report, United States – Investigation of the International Trade Commission in Softwood Lumber from Canada – Recourse to Article 21.5 of the DSU by Canada, WT/DS277/AB/RW, adopted 9 May 2006

US – Upland Cotton

Appellate Body Report, United States – Subsidies on Upland Cotton, WT/DS267/AB/R, adopted 21 March 2005

US – Upland Cotton

Panel Report, United States – Subsidies on Upland Cotton, WT/DS267/R, and Corr.1, adopted 21 March 2005, as modified by Appellate Body Report, WT/DS267/AB/R

WT/DS267/RW

Page xii


Short Title

Full Case Title and Citation

US – Wool Shirts and Blouses

Appellate Body Report, United States – Measure Affecting Imports of Woven Wool Shirts and Blouses from India, WT/DS33/AB/R and Corr.1, adopted 23 May 1997, DSR 1997:I, 323

WT/DS267/RW

Page 1


  1. INTRODUCTION


      1. On 18 August 2006, Brazil requested the establishment of a panel pursuant to Article 21.5 of the Dispute Settlement Understanding ("DSU") concerning the alleged failure of the United States to implement the recommendations and rulings of the Dispute Settlement Body ("DSB") in US – Upland Cotton.1


      2. At its meeting of 28 September 2006 the DSB decided, in accordance with Article 21.5 of the DSU, to refer this matter, if possible, to the original panel.


      3. The terms of reference of the Panel are the following:


        "To examine, in the light of the relevant provisions of the covered agreements cited by Brazil in document WT/DS267/30, the matter referred to the DSB by Brazil in that document, and to make such findings as will assist the DSB in making the recommendations or in giving the rulings provided for in those agreements."


      4. On 18 and 20 October 2006, Brazil and the United States respectively requested the Director- General to determine the composition of the Panel in accordance with Article 8.7 of the DSU. Article 8.7 of the DSU provides:


        "If there is no agreement on the panelists within 20 days after the date of the establishment of a panel, at the request of either party, the Director-General, in consultation with the Chairman of the DSB and the Chairman of the relevant Council or Committee, shall determine the composition of the panel by appointing the panelists whom the Director-General considers most appropriate in accordance with any relevant special or additional rules or procedures of the covered agreement or covered agreements which are at issue in the dispute after consulting with the parties to the dispute. The Chairman of the DSB shall inform the Members of the composition of the panel thus formed no later than 10 days after the date the Chairman receives such a request."


      5. On 25 October 2006, the Director-General determined the composition of the Panel as follows:

        Chairman: Mr. Eduardo Pérez Motta Members: Mr. Mario Matus

        Mr. Ho-Young Ahn.2


      6. Argentina, Australia, Canada, Chad, China, the European Communities, India, Japan, New Zealand, and Thailand reserved their rights to participate in the Panel proceedings as third parties.


      7. The Panel met with the parties to the dispute on 27-28 February 2007. The Panel met with the third parties on 28 February 2007.


      8. The Panel submitted its interim report to the parties on 27 July 2007 and submitted its final report to the parties on 15 October 2007.



    image

    1 WT/DS267/30.

    2 WT/DS267/31.

    WT/DS267/RW

    Page 2


  2. BACKGROUND


      1. This dispute concerns the disagreement between Brazil and the United States, within the meaning of Article 21.5 of the DSU, as to "the existence or consistency with a covered agreement of measures taken to comply with the recommendations and rulings" of the DSB in US – Upland Cotton.3


      2. The report of the original panel in US – Upland Cotton was circulated to Members on 8 September 2004.


      3. The original panel found that certain measures of the United States were inconsistent with the Agreement on Agriculture and/or the Agreement on Subsidies and Countervailing Measures ("SCM Agreement").


      4. First, the original panel found that export credit guarantees under the GSM 102, GSM 103 and SCGP export credit guarantee programmes4 were inconsistent with Articles 10.1 and 8 of the Agreement on Agriculture and Articles 3.1(a) and 3.2 of the SCM Agreement in respect of exports of upland cotton and other unscheduled agricultural products supported under these programmes and in respect of one scheduled product (rice).5 The original panel also found that in respect of exports of unscheduled agricultural products not supported under the programmes and other scheduled agricultural products, export credit guarantees under the GSM 102, GSM 103 and SCGP export credit guarantee programmes were not inconsistent with Articles 10.1 and 8 of the Agreement on Agriculture.6


      5. Second, the original panel found that section 1207(a) of the Farm Security and Rural Investment Act of 2002 ("FSRI Act of 2002") providing for user marketing (Step 2) payments to exporters of upland cotton was inconsistent with Articles 3.3 and 8 of the Agreement on Agriculture and Articles 3.1(a) and 3.2 of the SCM Agreement7 and that section 1207 (a) of the FSRI Act of 2002 providing for user marketing (Step 2) payments to domestic users of upland cotton was inconsistent with Articles 3.1(b) and 3.2 of the SCM Agreement.8


      6. Third, the original panel found that the effect of certain mandatory, price-contingent subsidy measures of the United States9 was significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement in the same world market constituting serious prejudice to the interests of Brazil within the meaning of Article 5(c) of the SCM Agreement.10



        image

        3 As discussed below, the United States considers that certain claims of Brazil are not properly within the scope of this proceeding because they do not pertain to a disagreement regarding "the existence or consistency with a covered agreement of measures taken to comply with the recommendations and rulings". Infra, Section IX.B.

        4 See infra, paras. 3.13-3.16.

        5 Panel Report, US – Upland Cotton, paras. 7.875, 7.881 and 8.1(d)(i).

        6 Panel Report, US – Upland Cotton, paras. 7.896 and. 8.1(d)(ii). Thus the Panel concluded that these export credit guarantees were exempt from actions based on Article XVI of the GATT 1994 and Article 3 of the

        SCM Agreement.

        7 Panel Report, US – Upland Cotton, paras. 7.749, 7.760-7.761 and 8.1(e).

        8 Panel Report, US – Upland Cotton, paras. 7.1097-7.1098 and 8.1(f).

        9 Marketing loan programme payments, user marketing (Step 2) payments, market loss assistance payments and counter-cyclical payments. See infra, paras. 3.7-3.11.

        10 Panel Report, US – Upland Cotton, paras. 7.1416 and 8.1(g)(i).

        WT/DS267/RW

        Page 3


      7. In respect of the measures found to be inconsistent with the Agreement on Agriculture, the original panel recommended pursuant to Article 19.1 of the DSU that the United States bring its measures into conformity with the Agreement on Agriculture.11


      8. In respect of the measures found to be subsidies prohibited under Articles 3.1(a) and 3.2 and 3.1(b) and 3.2 of the SCM Agreement, the original panel recommended pursuant to Article 4.7 of the SCM Agreement that the United States withdraw the prohibited subsidy without delay, i.e. at the latest within six months of the date of adoption of the panel report by the DSB or 1 July 2005, whichever was earlier.12


      9. In respect of the measures subject to the finding of significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement constituting serious prejudice to the interests of Brazil within the meaning of Article 5(c) of the SCM Agreement, the original panel stated the following:


        "[W]e recall that, in respect of the subsidies subject to our conclusion in paragraph 8.1(g)(i) above, pursuant to Article 7.8 of the SCM Agreement:


        '7.8 Where a panel report or an Appellate Body report is adopted in which it is determined that any subsidy has resulted in adverse effects to the interests of another Member within the meaning of Article 5, the Member granting or maintaining such subsidy shall take appropriate steps to remove the adverse effects or shall withdraw the subsidy.'


        Accordingly, upon adoption of this report, the United States is under an obligation to 'take appropriate steps to remove the adverse effects or ... withdraw the subsidy'."13


      10. The Appellate Body report in the original proceeding in this dispute was circulated to Members on 3 March 2005.


      11. The Appellate Body upheld the finding in paragraph 7.869 of the original panel report that the export credit guarantee programmes at issue constituted a per se export subsidy within the meaning of item (j) of the Illustrative List of Export Subsidies in Annex I of the SCM Agreement and the finding in paragraphs 7.947 and 7.948 of the original panel report that these export credit guarantee programmes were inconsistent with Articles 3.1(a) and 3.2 of the SCM Agreement.14 The Appellate Body reversed the finding in paragraph 7.881 of the original panel report that Brazil had not established that the application of export credit guarantees had resulted in actual circumvention of export subsidy commitments for pig meat and poultry meat. The Appellate Body considered that there were insufficient facts in the record to enable it to complete the legal analysis to determine whether the export credit guarantees had been applied in a manner that resulted in actual circumvention of the export subsidy commitments of the United States.15


      12. The Appellate Body upheld the findings in paragraphs 7.1088, 7.1097-7.1098 and 8.1(f) of the original panel report that Step 2 payments to domestic users of upland cotton under section 1207(a) of the FSRI Act of 2002 were inconsistent with Articles 3.1(b) and 3.2 of the SCM Agreement and the findings in paragraphs 7.748-7.749, 7.760-7.761 and 8.1(e) of the original panel



        image

        11 Panel Report, US – Upland Cotton, para. 8.3(a).

        12 Panel Report, US – Upland Cotton, paras. 8.3(b) and (c).

        13 Panel Report, US – Upland Cotton, para. 8.3(d).

        14 Appellate Body Report, US – Upland Cotton, paras. 674 and 763(e)(iv).

        15 Appellate Body Report, US – Upland Cotton, paras. 693-694 and 763(f)(i).

        WT/DS267/RW

        Page 4


        report that Step 2 payments to exporters of US upland cotton under section 1207(a) of the FSRI Act of 2002 were inconsistent with Articles 3.3 and 8 of the Agreement on Agriculture and Articles 3.1(a) and 3.2 of the SCM Agreement.16


      13. The Appellate Body upheld the finding in paragraphs 7.1416 and 8.1(g)(i) of the original panel report that the effect of certain price contingent subsidies was significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement.17


      14. On 21 March 2005, the DSB adopted the report of the Appellate Body in US – Upland Cotton

    and the report of the original panel, as modified by the Appellate Body report.


  3. FACTUAL ASPECTS


    1. PRODUCTS AT ISSUE


        1. The original panel described the products at issue as follows:


          "This dispute principally concerns alleged United States subsidies in respect of 'upland cotton' (Gossypium hirsutum). This is defined in relevant United States legislation as:


          'Planted and stub cotton that is produced from other than pure strain varieties of the Barbadense species, any hybrid thereof, or any other variety of cotton in which one or more of these varieties predominate.'


          In relation to Brazil's claims that United States export credit guarantee measures constitute export subsidies under the Agreement on Agriculture and the SCM Agreement, the dispute concerns upland cotton and other eligible agricultural products as well."18 (footnotes omitted)


        2. In this compliance dispute, there is no disagreement between the parties on the description of upland cotton, and thus, the Panel considers that the above description of upland cotton remains accurate.


        3. The original panel explained that "[u]pland cotton harvested from a field has cotton lint and cottonseed", that "[g]inning involves separating cotton lint from cottonseed", and that "[u]nless otherwise indicated, references in this report to upland cotton ordinarily refer to upland cotton lint."19 In this proceeding, there is no disagreement between the parties that the term "upland cotton" refers to upland cotton lint.


        4. While the product at issue in Brazil's claims of serious prejudice is upland cotton, with respect to Brazil's claims under the Agreement on Agriculture and the SCM Agreement relating to export credit guarantees, this compliance dispute concerns upland cotton and the following products: (i) three agricultural products for which the United States has entered export subsidy reduction


          image

          16 Appellate Body Report, US – Upland Cotton, paras. 552, 583-584 and 763(d).

          17 Appellate Body Report, US – Upland Cotton, paras. 496 and 763(c).

          18 Panel Report, US – Upland Cotton, paras. 7.197-7.198.

          19 Panel Report, US – Upland Cotton, footnote 258.

          WT/DS267/RW

          Page 5


          commitments ("scheduled products") – rice, poultry meat and pig meat20 – and (ii) "unscheduled products" for which export credit guarantees were provided after 1 July 2005.21


        5. The products described in (i) and (ii) in the preceding paragraph as well as upland cotton, all fall within the product coverage of the Agreement on Agriculture.22


        6. In addition, three other products (lysine, lyocell and wood products) are the subject of claims by Brazil that the United States provided export subsidies in the form of export credit guarantees, inconsistently with its obligations under the SCM Agreement.23


    2. DEVELOPMENTS CONCERNING THE MEASURES FOUND TO BE WTO-INCONSISTENT BY THE ORIGINAL PANEL24


      1. User marketing (Step 2) payments, marketing loan payments, market loss assistance payments and counter-cyclical payments


        1. User marketing (Step 2) payments


            1. The term "user marketing (Step 2) payments" as used in the original panel report refers to payments made under the upland cotton user marketing certificate or "Step 2" programme, a special marketing loan provision for upland cotton in section 1207(a) of the FSRI Act of 2002. This programme provided for marketing certificates or cash payments to eligible domestic users and exporters of eligible upland cotton when certain market conditions existed such that US cotton pricing benchmarks were exceeded.25 On 1 February 2006, the Congress of the United States adopted legislation repealing the Step 2 subsidy programme for upland cotton effective as of 1 August 2006.26


        2. Marketing loan payments


            1. Marketing loan payments are provided for in sections 1201-1205 of the FSRI Act of 2002. The original panel described these measures as follows:


              "Marketing loan programme payments for upland cotton began in 1986 and have continued under successive legislation, including the Federal Agriculture


              image

              20 The United States requests that the Panel rule that Brazil's claims concerning poultry meat and pig meat are outside the scope of this proceeding. US First Written Submission, paras. 24-30. The Panel addresses this request of the United States below in Section IX.B.2.

              21 Brazil has identified the following products as having benefited from GSM 102 export credit

              guarantees since 1 July 2005: cotton, oilseeds (including soybeans and soybean meal), protein meals, fresh vegetables, hides/skins, tallow, corn products. Brazil's First Written Submission, para. 450 and note 622.

              22 Article 2 and Annex 1 of the Agreement on Agriculture. In particular, the Panel notes that the

              Agreement covers products under Harmonized System Headings 52.01 to 52.03 (raw cotton, waste and cotton carded and combed).

              23 Brazil included these products in the list of unscheduled products that benefited from GSM 102

              export credit guarantees. Following an argument by the United States that these products fall outside the product coverage of the Agreement on Agriculture, Brazil indicated that, with respect to these three products, it makes no claims under the Agreement on Agriculture but maintains its claims under the SCM Agreement. See US Rebuttal Submission, para. 83, Brazil's Response to Panel Question 50 and US Comments thereon. See infra, Section XIV.D for the Panel's determination that Brazil's claims concerning these products are outside the scope of these proceedings.

              24 The inclusion of a measure in this section and the terms used to describe the measure are without prejudice to the Panel's position on the preliminary objections raised by the United States with respect to

              whether certain measures are properly within the scope of this proceeding under Article 21.5 of the DSU.

              25 Panel Report, US – Upland Cotton, para. 7.209.

              26 Section 1103 of the Deficit Reduction Act of 2005, reproduced in Exhibit Bra-435.

              WT/DS267/RW

              Page 6


              Improvement and Reform Act of 1996 ('FAIR Act of 1996') and the FSRI Act of 2002. The marketing loan programme is intended to minimize potential loan forfeitures by providing interim financing to eligible producers on their eligible production and facilitate the orderly distribution of eligible commodities throughout the year. Accordingly, rather than selling the crop at harvest when prices tend to be at their lowest, the proceeds of the interim loan enable producers to pay off their expenses when they become due, while storing their pledged harvested crop as collateral and repaying the loan when market conditions are potentially more favourable.


              Under the FAIR Act of 1996, marketing assistance loans for upland cotton were provided to producers for upland cotton harvested on a farm containing eligible cropland covered by a production flexibility contract [...]. The loans for upland cotton were made on conditions prescribed by the Secretary for a term of ten months. The legislation provided that the loan rate for upland cotton was determined by the Secretary for each year's crop at a rate that was no less than the smaller of 85 per cent of the five year Olympic average of county spot market prices, or 90 per cent of the Northern Europe-based average price but which should be not less than 50 cents per pound and not more than 51.92 cents per pound. The loan rates for the 1999, 2000, 2001 and 2002 crops of upland cotton were each determined at 51.92 cents per pound. The loans were non-recourse, so a producer had the option to deliver to the CCC [Commodity Credit Corporation] the quantity of a commodity pledged as collateral for the loan as payment in full when the loan was due to be repaid.


              Under the FSRI Act of 2002, marketing assistance loans continue to be provided, but certain features of the programme have been changed. In particular, loans are provided to producers for any upland cotton produced on a farm, the term of a marketing assistance loan for upland cotton is now nine months and the loan rate for upland cotton is fixed by the Act itself at 52 cents per pound for the 2002 through 2007 crop years.


              The repayment rate for marketing loans is the lower of the adjusted world market price and the loan rate plus interest. When the adjusted world market price is lower than the loan rate, the producer repays at less than the loan rate and the difference is referred to as a 'marketing loan gain'. Alternatively, a producer may forego a marketing loan and receive a 'loan deficiency payment' in the amount of the difference between the lower adjusted world market price and the loan rate. Additionally, in October 1999, the FAIR Act of 1996 was amended to include provisions for the issuance of 'commodity certificates', available to producers, including upland cotton producers, to use in acquiring 1998 through 2002 crop collateral pledged to CCC for a commodity loan. Marketing loan gains, loan deficiency payments and commodity certificate exchange gains are collectively referred to below as 'marketing loan programme payments'.


              When loans are repaid, the CCC charges interest at a rate 1 per cent higher than the rate it pays the Treasury. The CCC forgives interest and pays storage charges for the loan period when upland cotton under loan is forfeited, and also on redemption of upland cotton under loan where the repayment amount is too low to satisfy the loan amount and those charges."27 (footnotes omitted)


              image

              27 Panel Report, US – Upland Cotton, paras. 7.204-7.208.

              WT/DS267/RW

              Page 7


            2. It is not in dispute that the United States continues to provide marketing loan payments to US producers of upland cotton and that the legislative and regulatory provisions governing these marketing loan payments have not been changed.


        3. Market loss assistance payments


            1. Market loss assistance payments were provided for in four separate pieces of legislation.28 The original panel described these measures as follows:


              "Market loss assistance ('MLA') payments were made under four separate pieces of legislation, one each for the years 1998 through 2001. They were ad hoc emergency and supplementary assistance provided to producers in order to make up for losses sustained as a result of recent low commodity prices.


              The 1998 MLA payments were intended essentially as a 50 per cent additional [production flexibility contract] payment. The 1998, 1999 and 2001 Acts each appropriated a dollar amount to assistance which was divided among PFC payment recipients proportionately to their respective previous PFC payment. The 2000 Act provided for payments at the same contract payment rates as the 1999 Act. MLA payments were only made to recipients enrolled in the PFC programmes."29 (footnotes omitted)


        4. Counter-cyclical payments


          1. Counter-cyclical payments are provided for in sections 1101-1108 of the FSRI Act of 2002. The original panel described these measures as follows:


            "The counter-cyclical payments ('CCP') programme was also established by the FSRI Act of 2002. It provides support to producers (as defined) based on historical acreage and yields for the same commodities as [direct payments], including upland cotton.


            An eligible producer must enter into an annual agreement in order to receive payments for a crop year. The eligibility requirements and planting flexibility requirements are the same as for the DP programme. CCP payments, like DP payments, are made to producers on farms for which payment yields and base acres are established for each of the 2002 through 2007 crop years of each covered commodity.


            CCP payments depend on the current prices of commodities. They are provided to producers with base acres and yields for a covered commodity for each of the 2002 through 2007 crop years whenever the effective price falls below the target price, which is fixed by the Act at 72.4 cents per pound for upland cotton. The effective price for a commodity is the sum of the DP payment rate [...], plus the higher of the national average farm price for the marketing year or the loan rate [...]. The difference between the effective price and the target price is the CCP payment rate. Consequently, the CCP payment rate, DP payment rate and, where applicable, the



            image

            28 Market loss assistance payments were provided pursuant to the following Acts: the Agriculture, Rural Development, Food and Drug Administration and Related Agencies Appropriation Act of 1999 for the 1998 crop; the Agriculture, Rural Development, Food and Drug Administration and Related Agencies Appropriation Act of 2000 for the 1999 crop; the Agriculture Risk Protection Act of 2000 for the 2000 crop; and the Crop Year 2001 Agriculture Economic Assistance Act for the 2001 crop.

            29 Panel Report, US – Upland Cotton, paras. 7.216-7.217.

            WT/DS267/RW

            Page 8


            loan rate, are equal to the difference between the market price and 72.4 cents per pound.


            CCP payments are made on 85 per cent of the base acreage for each commodity multiplied by the corresponding payment rate multiplied by the applicable payment yield. An owner who elected to have his or her base acreage calculated on the basis of the four-year average of planted acreage during the 1998 through 2001 crops, had a one-time opportunity partially to 'update' CCP payment yields for CCP payments for all covered commodities, using one of two methods which both included the farm's average yields for the 1998 through 2001 crop years."30 (emphasis original, footnotes omitted)


          2. It is not in dispute that the United States continues to provide counter-cyclical payments to US producers of upland cotton and that the legislative and regulatory provisions governing these counter-cyclical payments have not been changed.


      2. Export credit guarantees


      1. The original panel examined three export credit guarantee programmes administered by the United States Department of Agriculture (USDA) through the Commodity Credit Corporation (CCC):31 the General Sales Manager 102 ("GSM 102") programme, the General Sales Manager 103 ("GSM 103") programme and the Supplier Credit Guarantee Programme ("SCGP"). Brazil challenged the programmes "as such" and "as applied"32 to exports of upland cotton and other eligible agricultural commodities.33 All three programmes were used to guarantee the repayment of credit made available to finance commercial export sales of US agricultural commodities.


      2. The original panel described the three programmes as follows: "General Sales Manager 102 ('GSM 102')

        Under the GSM 102 export credit guarantee programme, the CCC is authorized to guarantee the repayment of credit made available to finance commercial export sales of agricultural commodities from privately owned stocks on credit terms between 90 days and three years. The CCC generally covers 98 per cent of the principal and a portion of the interest. The CCC selects agricultural commodities and products according to market potential. The CCC does not provide financing, but rather guarantees payments due from foreign banks. To secure such a payment guarantee, once a firm export sale exists, the United States exporter must apply prior to the date of exportation. The exporter pays a fee calculated on the dollar amount guaranteed, based on a schedule of rates applicable to different credit periods. There is a statutory cap on the fee charged of 1 per cent of the guaranteed dollar value of the transaction. In terms of financing, the CCC-approved foreign bank issues an irrevocable letter of credit in favour of the United States exporter, ordinarily advised or confirmed by the financial institution in the United States which agrees to extend credit to the foreign


        image

        30 Panel Report, US – Upland Cotton, paras. 7.223-7.226.

        31 The CCC is a wholly-owned government corporation within the USDA; it has no employees and carries out the majority of its programmes through the personnel and facilities of the Farm Service Agency. See

        Panel Report, US – Upland Cotton, footnote 346.

        32 Because the original panel adopted a "programme-wide" analytical approach under item (j) of the Illustrative List of Export Subsidies in Annex I to the SCM Agreement, it did not view the distinction between

        Brazil's "as such" and "as applied" claims as a determinative one. Panel Report, US – Upland Cotton, para. 7.763.

        33 Panel Report, US – Upland Cotton, para. 7.69.

        WT/DS267/RW

        Page 9


        bank. If the foreign bank fails to make any payment as agreed, the exporter is required to submit a notice of default to the CCC. The CCC pays a valid claim for loss.


        General Sales Manager 103 ('GSM 103')


        The GSM 103 programme operates in a similar fashion to GSM 102. The main differences between the two programmes include: export credit guarantees under GSM 103 are "intermediate term credit guarantees" issued for terms from three to 10 years; there are additional statutory required determinations to be made when the CCC issues guarantees; and there is no statutory cap on the origination fees that may be charged by the CCC in connection with an export credit guarantee transaction.


        Supplier Credit Guarantee Programme ('SCGP')


        Under the SCGP, the CCC is authorized to issue guarantees for the repayment of credit made available for a period not exceeding 180 days by a United States exporter to a purchaser of United States agricultural commodities in a foreign country. These direct credits must be secured by promissory notes signed by the importers. The CCC does not provide financing, but rather guarantees payment due from the importer. Typically, the CCC guarantees a portion (65 per cent) of the value of the exports (principal only; the guarantee does not cover interest). The exporter negotiates terms of export credit sales with the importer. Once a firm export sale exists, the United States exporter must apply for a payment guarantee prior to the date of exportation. The exporter pays a fee for the guarantee calculated on the guaranteed portion of the value of the export sales. There is a statutory cap on the fee charged of 1 per cent of the guaranteed dollar value of the transaction. The importer must issue a dollar- denominated promissory note in favour of the United States exporter in the form specified in the applicable country or regional programme announcement. The United States exporter may negotiate an arrangement to be paid, in full or in part, by assigning the right to proceeds that may become payable under the CCC's guarantee to a United States financial institution. If the foreign bank fails to make any payment, the exporter or assignee is required to submit a notice of default to the CCC. The CCC pays a valid claim for loss."34 (footnotes omitted)


      3. Under the GSM 102 and GSM 103 export credit guarantee programmes, the fees charged by the CCC varied with the guaranteed dollar value of the transaction, the repayment period, and the principal repayment interval (annual or semi-annual).35 Under the SCGP, fees varied with the repayment period (up to 90 days or 90 to 180 days).36


      4. Following the adoption of the original panel and Appellate Body reports, the USDA announced, on 30 June 2005, that as of 1 July 2005, the CCC would no longer accept applications for payment guarantees under the GSM 103 programme.37 It also announced that the CCC would henceforth use a new fee structure for the GSM 102 and SCGP programmes. Under the new schedule, fees were increased and fees now vary with country risk, repayment term (tenor) and repayment frequency (annual or semi-annual).38 Countries are classified in eight risk categories (0-7);



    image

    34 Panel Report, US – Upland Cotton, paras. 7.242-7.244.

    35 Panel Report, US – Upland Cotton, para. 7.820.

    36 Panel Report, US – Upland Cotton, para. 7.821 and footnote 975.

    37 Exhibits Bra-502 and Bra-503.

    38 "Fee rates will be based on the country risk that CCC is undertaking, as well as the repayment term (tenor) and repayment frequency (annual or semi-annual) under the guarantee. For the GSM-102 programme,

    WT/DS267/RW

    Page 10


    countries in the riskiest category (category 7) are ineligible.39 Lastly, in October 2005, the United States ceased issuing export credit guarantees under the SCGP.40


  4. PARTIES' REQUESTS FOR FINDINGS AND RECOMMENDATIONS


      1. With respect to its claims relating to adverse effects within the meaning of Article 5 of the

        SCM Agreement, Brazil requests the Panel to make the following findings:


        1. The United States failed to take any measures to remove the adverse effects or to withdraw the three price-contingent upland cotton subsidy programmes under the FSRI Act of 2002 between 22 September 2005 and 31 July 2006.


        2. The US measure eventually taken to comply, i.e. the amendment of the FSRI Act of 2002 to repeal Section 1207(a), providing for Step 2 payments, is insufficient to ensure compliance with the DSB recommendations and rulings, because the resulting basket of measures is inconsistent with Articles 5 and 6 of the SCM Agreement. In particular, Brazil requests the compliance Panel to find that the marketing loan and counter-cyclical payment programmes of the FSRI Act of 2002 cause "present" serious prejudice to the interests of Brazil in the form of (a) significant price suppression in the world market for upland cotton and (b) an increase in the US world market share of upland cotton.


        3. The US measure eventually taken to comply is insufficient to ensure compliance with the DSB recommendations and rulings, because the resulting basket of measures is inconsistent with Articles 5(c) and 6.3(c) as well as footnote 13 of the SCM Agreement. In particular, Brazil requests the compliance Panel to find that the marketing loan and counter-cyclical payment programmes of the FSRI Act of 2002 cause a threat of serious prejudice to the interests of Brazil in the form of a threat of significant price suppression in the world market for upland cotton. The Panel need only decide this claim if it is unable to find that there is "present" serious prejudice from marketing loan and counter-cyclical payment programmes of the FSRI Act of 2002; and


        4. The failure of the United States to take any measure to remove the adverse effects or to withdraw the three price-contingent upland cotton subsidy programmes under the FSRI Act of 2002 between 22 September 2005 and 31 July 2006 causes "present" serious prejudice to the interests of Brazil in the form of significant price suppression in the world market for upland cotton and an increase in the US world market share, in violation of Articles 5(c), 6.3(c) and (d) of the SCM Agreement. The Panel need only reach this claim if it is unable to make findings in favour of Brazil with respect to Brazil's claims summarized under (a)-(c) above.41


      2. With regard to its claims under the Agreement on Agriculture and the SCM Agreement

        relating to export credit guarantees, Brazil requests the Panel to make the following findings:


        image

        country risk will be based on the country of the foreign obligor (opening bank), as determined by the CCC. For the SCGP, country risk will be based on the country of the foreign obligor (importer), as determined by the CCC." Exhibit Bra-504. See also Exhibits Bra-502, Bra-505 and Bra-506.

        39 A number of countries that were previously eligible were reclassified into this ineligible risk category. See US First Written Submission, footnote 15 and Exhibit US-2.

        40 See US First Written Submission, para. 20 and Exhibit Bra-513. The United States indicates that

        "[a]t present, [it] has no plan to resume the SCGP either in the same or different form." US First Written Submission, para. 20.

        41 See Brazil's First Written Submission, para. 461.

        WT/DS267/RW

        Page 11


        1. With respect to those export credit guarantees under the GSM 102, GSM 103 and SCGP programmes that were outstanding on 1 July 2005, "measures taken to comply" within the meaning of Article 21.5 of the DSU do not exist.


        2. The "measures taken to comply" by the United States with regard to the GSM 102 export credit guarantee programme are inconsistent with Articles 10.1 and 8 of the Agreement on Agriculture.


        3. The "measures taken to comply" by the United States with regard to the GSM 102 export credit guarantee programme are inconsistent with Articles 3.1(a) and 3.2 of the SCM Agreement. 42


      3. Brazil requests that, on the basis of the findings requested above, the Panel make the following recommendations:


        1. The United States has failed to comply fully with its obligation, under Article 7.8 of the SCM Agreement, to "remove the adverse effects or … withdraw the subsidy", and the original Article 7.8 recommendation of the DSB therefore remains in effect, and applies to the full extent of the findings requested above.


        2. The United States has failed to comply fully with its obligation, under Article 4.7 of the SCM Agreement, to "withdraw the subsidy", and the original Article 4.7 recommendation of the DSB therefore remains in effect, and applies to the full extent of the findings requested above.


        3. The United States has failed to bring itself into conformity with its obligations under the Agreement on Agriculture, and the original recommendation of the DSB therefore remains in effect, and applies to the full extent of the findings requested above. 43 44


      4. The United States requests the Panel to reject the claims of Brazil in their entirety and to find that the United States has complied with the recommendations and rulings of the DSB and, further, that the measures of the United States taken to comply are not inconsistent with the SCM Agreement or the Agreement on Agriculture.45


      5. The United States requests the Panel to make preliminary rulings that certain claims of Brazil are not within the scope of this proceeding.46


      6. In respect of Brazil's claims of "present" serious prejudice and threat of serious prejudice, the United States submits that: (i) Brazil fails to substantiate its arguments in support of the claim that the effects of the elimination of the Step 2 programme are modest; (ii) Brazil does not demonstrate that the marketing loan or counter-cyclical programmes mandate a breach of WTO obligations; (iii) Brazil fails to make a prima facie case of "present" serious prejudice within the meaning of Articles 5(c) and 6.3(c) (significant price suppression) of the SCM Agreement; (iv) Brazil fails to make a prima facie case of "present" serious prejudice under Articles 5(c) and 6.3(d) (increased market share) of the


        image

        42 See Brazil's First Written Submission, para. 462.

        43 See Brazil's First Written Submission, para. 463.

        44 In this connection, Brazil also observes that the deadlines for implementation of the original recommendations and rulings of the DSB continue to be 1 July 2005 (for the export credit guarantee-related

        recommendations and rulings) and 21 September 2005 (for the adverse effects-related recommendations and rulings). Brazil's First Written Submission, para. 464.

        45 US First Written Submission, para. 344.

        46 See infra, para. 9.7. The Panel addresses the issue of whether, as argued by the United States, certain claims of Brazil are not properly within the scope of this proceeding in Sections IX.B.2-5 of this Report.

        WT/DS267/RW

        Page 12


        SCM Agreement; and (v) Brazil fails to make a prima facie case of threat of serious prejudice pursuant to Articles 5(c) and 6.3(c) of the SCM Agreement.47


      7. In respect of Brazil's claims regarding export credit guarantees, the United States submits that export credit guarantees have been provided under the GSM 102 programme subsequent to 1 July 2005 consistently with WTO obligations of the United States and that the United States has taken action to withdraw the subsidy with respect to GSM 102, GSM 103 and SCGP export credit guarantees issued prior to 1 July 2005.48


  5. ARGUMENTS OF THE PARTIES


      1. The arguments of the parties, as submitted, or as summarized in their executive summaries as submitted to the Panel, are attached as Annexes.


      2. The responses of the parties to questions of the Panel, their comments on each other's responses and other relevant documents submitted by the parties are also attached as Annexes.


  6. ARGUMENTS OF THE THIRD PARTIES


    6.1 The arguments of the third parties are attached to this Report as Annexes.49


  7. INTERIM REVIEW


      1. The Panel submitted its interim report to the parties on 27 July 2007. On 3 September 2007, the parties submitted written requests for review of precise aspects of the interim report. Neither party requested that the Panel hold an interim review meeting. On 17 September 2007, the parties submitted written comments on each other's written requests for review.


      2. The Panel has modified aspects of the interim report in light of the parties' requests and comments where the Panel considered it appropriate. The Panel has also made technical corrections and changes of an editorial nature.


        1. CHANGES REQUESTED BY THE PARTIES REGARDING THE "FACTUAL ASPECTS" SECTION OF THE INTERIM REPORT


      3. The United States proposes changes to paragraphs 3.7, 3.13 and 3.16 in the "Factual Aspects" Section of the interim report (paragraphs 3.7, 3.13 and 3.16 of this Report). The Panel considers that these proposed changes correct certain factual errors or inaccuracies in the interim report and has therefore amended the concerned paragraphs in accordance with the request of the United States.


        1. CHANGES REQUESTED BY THE PARTIES WITH RESPECT TO THE PANEL'S DESCRIPTION OF ARGUMENTS OF THE PARTIES


      4. The Panel has made a number of amendments to the interim report pursuant to requests by one of the parties to amend certain paragraphs of the report in order to better reflect arguments made by that party in the course of the proceeding. The Panel has made such changes to paragraphs 4.5; 9.19 (footnote 101); 9.28 (footnote 111); 9.62 (footnote 163); 9.56; 10.6; 10.46 (footnote 254); 10.73; 10.118; 10.155; 10.156; 10.163 (footnote 443); 10.165; 10.200; 10.200 (footnote 481); 14.16 and



    image

    47 US First Written Submission, paras. 149, 196, 203 and 332 and US Rebuttal Submission, para. 405.

    48 US First Written Submission, paras. 60-142 and 143-145.

    49 India and Thailand did not make written submissions.

    WT/DS267/RW

    Page 13


    footnote 588, and 14.122 (footnote 728) of the interim report (paragraphs 4.5; 9.19 (footnote 105);

    9.29 (footnote 115); 9.64 (footnote 165); 9.58; 10.6; 10.46 (footnote 256); 10.73; 10.118; 10.155;

    10.156; 10.163 (footnote 450); 10.165; 10.205; 10.205 (footnote 497); 14.16 and footnote 604, and

    14.122 (footnote 747) of this Report).


      1. The United States requests the Panel to add a footnote to the first sentence of paragraph 8.29 of the interim report (paragraph 8.29 of this Report) with the following text: "For purposes of this dispute, the United States did not object to Chad being considered as a least-developed country Member". The United States notes in this respect that the issue of whether Chad is a least-developed country Member was not discussed by either party and that no relevant evidence was presented to support such a finding. Brazil opposes the change proposed by the United States.


      2. The Panel declines to make the change requested by the United States. The reference in the interim report to Chad as a least-developed country Member is consistent with the WTO practice to rely on the UN classification of least-developed countries.50 Moreover, the United States has not stated in this proceeding that it did not object to Chad being considered a least-developed country Member for purposes this dispute. Therefore, the sentence suggested by the United States is not an accurate reflection of an argument made in this proceeding.


      3. The United States notes that paragraph 9.7 of the interim report (paragraph 9.7 of this Report) appears to omit a summary of an additional argument of the United States and requests that this paragraph be revised to add the objection of the United States that Brazil's claims against the marketing loan and counter-cyclical programmes are outside the scope of this dispute because these measures were not measures taken to comply under Article 21.5 of the DSU. Brazil has not commented on this suggested change.


      4. The Panel agrees with the United States that the interim report does not adequately reflect the two distinct bases of the preliminary objections of the United States to Brazil's claims regarding the marketing loan and counter-cyclical payment programmes. To address this issue, the Panel has amended the title of Section IX.A.3 and has added new paragraphs 9.28 and 9.55.


        1. REQUESTS FOR CHANGES WITH RESPECT TO THE PANEL'S FACTUAL AND LEGAL ANALYSIS


      5. The United States requests that the Panel delete footnote 108 to the last sentence of paragraph 9.25 of the interim report (footnote 112 to the last sentence of paragraph 9.25 of this Report) in which the Panel notes that in interpreting the concept of "measures taken to comply" in Article 21.5 of the DSU, the Appellate Body has taken into consideration the need for prompt compliance with DSB recommendations and rulings. The United States argues that the need for prompt compliance with DSB recommendations and rulings does not somehow convert a claim outside the scope of an Article 21.5 proceeding into one that is properly before the compliance panel. The United States also observes that Brazil itself has implicitly recognized that any export credit guarantees for pig meat and poultry meat are outside the scope of this proceeding by bringing a separate case in WT/DS365 challenging US export credit guarantees in respect of all scheduled and unscheduled products.


      6. Brazil objects to the suggestion of the United States to delete this footnote. Brazil submits, in this respect, that the Panel's citation to the statements made by the Appellate Body regarding the promotion of prompt compliance is correct as a matter of law and that the Panel is entitled to consider that the argument of the United States that GSM 102 export credit guarantees for pig meat and poultry


        image

        50 It should be noted that the original panel treated Benin and Chad as least-developed country Members and that there is nothing in the original panel report to indicate that the United States had expressed a view or a reservation on that issue. Panel Report, US – Upland Cotton, para. 7.54.

        WT/DS267/RW

        Page 14


        meat are outside the scope of this proceeding will not promote prompt compliance with DSB recommendations and rulings. Brazil rejects the argument of the United States that the fact that Brazil has raised the consistency of GSM 102 export credit guarantees in a consultation request in WT/DS365 constitutes an implicit recognition by Brazil that GSM102 export credit guarantees for pig meat and poultry meat are outside the scope of this proceeding.


      7. The Panel declines to delete the footnote to the last sentence of paragraph 9.25 of the interim report (footnote 112 to the last sentence of paragraph 9.25 of this Report). In this footnote, the Panel refers to a statement of the Appellate Body in US – Oil Country Tubular Goods Sunset Reviews (Article 21.5 – Argentina), paragraph 151, which the Panel considers to provide further support for its view that Brazil's claims regarding export credit guarantees for pig meat and poultry meat are properly within the scope of this proceeding. The United States does not contest the factual accuracy of the Panel's citation to the statement of the Appellate Body. As is already clear from the fact that the Panel makes this point in a footnote, the Panel relies on this statement simply as an additional element in support of the main arguments set out in paragraphs 9.24-9.25 of the interim report (paragraphs 9.24-

    9.25 of this Report). The Panel remains of the view that the notion that the purpose of Article 21.5 of the DSU is to promote prompt compliance with DSB recommendations and rulings is a pertinent consideration that supports the Panel's conclusion that Brazil's claims are within the scope of this proceeding. Finally, the Panel sees no merit to the argument of the United States that the Panel should construe Brazil's request for consultations in WT/DS365 as constituting an implicit recognition by Brazil that Brazil's claims relating to GSM102 export credit guarantees for pig meat and poultry meat are not within the scope of this proceeding.


      1. With regard to paragraph 9.78 of the interim report (paragraph 9.80 of this Report), Brazil suggests that the Panel may wish to provide a more detailed explanation as to why the rulings in US – Softwood Lumber IV support the Panel’s conclusion that marketing loan and counter-cyclical payments made after 21 September 2005 are "measures taken to comply" within the meaning of Article 21.5 of the DSU. Brazil refers, in this regard, to the arguments that it has advanced on this issue in this proceeding. The United States disagrees with Brazil’s suggestion. The United States submits that Brazil errs in suggesting that the panel and Appellate Body reports in US – Softwood Lumber IV (Article 21.5 – Canada) support finding that the US measures at issue are measures taken to comply.


      2. The Panel agrees with Brazil that paragraph 9.78 of the interim report (paragraph 9.80 of this Report) does not sufficiently explain why the Panel considers that its reasoning is supported by the reports of the panel and the Appellate Body in previous Article 21.5 proceedings. The Panel has amended this paragraph to state more clearly why the Panel considers that its reasoning is supported by the Appellate Body's statement in US – Softwood Lumber IV (Article 21.5 – Canada) that "measures taken to comply" in Article 21.5 of the DSU include "measures with a particularly close relationship to the declared 'measure taken to comply', and to the recommendations and rulings of the DSB".


      3. Brazil requests the Panel to amend paragraph 9.79 of the interim report (paragraph 9.81 of this Report) by deleting the words "to the extent" in the first line, the word "are" in the second line, and the comma and the word "they" after "prejudice" in the fifth line. Brazil submits that these proposed changes are consistent with its comment on paragraph 9.78 of the interim report and with the reasoning underlying the Panel's conclusion that marketing-loan and counter-cyclical payments made after 21 September 2005 are within the scope of this proceeding. The United States disagrees with Brazil's proposed changes. The United States considers that the wording of paragraph 9.79 makes clear that the existing marketing loan and counter-cyclical payments could be provided by the United States on different conditions that were found to be inconsistent with the provisions of the SCM Agreement.

        WT/DS267/RW

        Page 15


      4. The Panel declines to make the changes requested by Brazil to paragraph 9.79 of the interim report (paragraph 9.81 of this Report). The Panel considers that these changes are inconsistent with the text of the immediately preceding paragraphs and that Brazil has offered no convincing explanation as to why these changes are warranted.


      5. With respect to the paragraph 10.23 of the interim report (paragraph 10.23 of this Report), the United States considers that it is inaccurate to state that there is "no disagreement" between the parties regarding the "general order of magnitude" of marketing loan payments made by the United States to upland cotton producers in MY 2004 and MY 2005. The United States notes that while Brazil and the United States have submitted identical figures for the amount of marketing loan payments in MY 2005, with respect to the amount of marketing loan payments in MY 2004 the figures submitted by the parties have continued to differ. The United States thus requests the Panel to delete the reference to MY 2004 in this paragraph. Brazil objects to the request of the United States, arguing that any difference between Brazil’s and the US figures for MY 2004 is negligible.


      6. The Panel has amended paragraph 10.23 of the interim report (paragraph 10.23 of this Report) by deleting the statement that "there is no disagreement" between the parties regarding the general order of magnitude of marketing loan payments in MY 2004 and MY 2005 and by adding language that makes clear that it is the Panel, rather than the parties, that considers that the figures supplied by the parties on marketing loan payments for MY 2004 and MY 2005 are of the same general order of magnitude.


      7. The United States requests that the Panel revise the text of footnote 236 to paragraph 10.40 of the interim report (footnote 238 to paragraph 10.40 of this Report) to make it clear that the response of the United States to Panel question 55 does not imply agreement with the "cotton-to-cotton" methodology. Brazil opposes this request, arguing that the change requested by the United States is unnecessary as the interim report already summarizes the arguments of the United States on the "cotton-to-cotton" methodology.


      8. The Panel has amended the text of footnote 236 to paragraph 10.40 of the interim report (footnote 238 to paragraph 10.40 of this Report) to clarify that in posing question 55 the Panel was not asking whether the United States agreed with "Brazil's methodology or the "cotton-to-cotton" methodology.


      9. With respect to paragraph 10.95 of the interim report (paragraph 10.95 of this Report), the United States requests the Panel either (1) to add a finding that the current economic literature supports the view that counter-cyclical payments do not appear to have had more than minimal production- and trade-distorting effects, inter alia, because the conditions under which such payments might in theory have greater effects have not been established in fact by Brazil in this proceeding, or

        (2) to explain its basis for rejecting each of the studies cited by the United States drawing the conclusion that the production effects of counter-cyclical payments are minimal.


      10. Brazil objects to this request of the United States. Brazil argues that both parties have submitted significant evidence and arguments regarding the effects of counter-cyclical payments and that in its interim report the Panel has analyzed the studies submitted by both parties in great detail. Since the Panel has explained its evaluation of these studies, there is no basis for the United States to request the Panel to explain its basis for rejecting the evidence and argument of the United States. Similarly, there is no basis for the Panel to conclude that the economic literature supports the US position since the Panel has found exactly the opposite.


      11. The Panel considers that paragraphs 10.90-10.95 of the interim report adequately explain the conclusion drawn by the Panel from its review of the economic studies referred to by the parties. However, in light of the comments of the United States, the Panel has made certain changes to

        WT/DS267/RW

        Page 16


        paragraphs 10.91 and 10.92 of the interim report (paragraphs 10.91 and 10.92 of this Report) to refer more specifically to certain studies cited by the United States.


      12. The United States requests that the text of footnote 363 to the third sentence of paragraph 10.110 of the interim report (footnote 369 to the third sentence of paragraph 10.110 of this Report) be revised to reflect that Step 2 payments were only made to users of upland cotton rather than to producers. Brazil opposes the change requested by the United States on the ground that Step 2 payments were made not only to users but also to exporters of cotton.


      13. The Panel considers that while the United States is correct that Step 2 payments were not made to producers, Brazil is correct that such payments were made not only to users but also to exporters. The Panel has therefore amended the text of footnote 363 to the third sentence of paragraph 10.110 of the interim report (footnote 369 to the third sentence of paragraph 10.110 of this Report), to reflect that Step 2 payments were made to users and exporters of cotton.


      14. With respect to paragraph 10.126 of the interim report (paragraph 10.126 of this Report), the United States requests the Panel to revise the statement in the second sentence that "neither Brazil nor the United States appear to have access to recent data on foreign planted acreage." The United States submits that foreign planted acreage data does not exist; therefore, it is not a question of access for either party. Only the United States has a consistent public source of data on both planted and harvested acreage. The United States requests that the Panel revise the second sentence to read: "However, neither Brazil nor the United States appear to have access to recent data on foreign planted acreage; indeed such data do not appear to exist." Brazil objects to the request of the United States on the ground that the United States is raising a new argument that foreign planted acreage data does not exist. Brazil submits, in this regard, that it is not appropriate for a party to make new arguments at the interim review stage that were not made during the proceedings, much less to ask a panel to reflect those new arguments in its final report.


      15. The Panel has decided not to make the change to paragraph 10.126 of the interim report proposed by the United States. The question of whether or not data on foreign planted acreage exists was not raised in this dispute.


      16. Brazil requests the Panel to amend paragraph 10.127 of the interim report (paragraph 10.127 of this Report) because, as currently formulated, the paragraph can be construed to be in contradiction with other findings made by the Panel in its interim report. Brazil refers in particular to the statement of the Panel in paragraph 10.127 that "Brazil's claim of insulation of US producers from market price signals applies when expected prices are below their intervention levels" but "this insulation does not apply when expected cotton prices are above their intervention levels". Brazil considers that, taken out of context, this statement might be construed to be in contradiction with the Panel's finding in paragraph 10.81 of the interim report (paragraph 10.81 of this Report) that marketing loan payments affect the level of acreage planted to cotton even when the expected market price for upland cotton at the time of planting is higher than the marketing loan rate. Brazil requests that the Panel revise paragraph 10.127 "to eliminate any basis for concluding that paragraph 10.127 conflicts in any way with the Panel's earlier findings regarding the role of marketing loan and counter-cyclical payments in insulating US upland cotton farmers from market price signals".


      17. The United States opposes the changes proposed by Brazil to paragraph 10.127 of the interim report. The United States considers that Brazil is attempting to substitute its own view that, regardless of the expected price level or other factors, upland cotton producers are insulated from market signals and that marketing loan and counter-cyclical payments always have effects on the production of upland cotton that lead to significant price suppression. The Panel, in Section X.7 of the interim report, has examined, inter alia, the US evidence of the relative stability of the US share of world production and exports over the entire period of the FSRI Act of 2002, and has rejected Brazil’s

        WT/DS267/RW

        Page 17


        extreme view that US upland cotton producers are always insulated from market price signals. The United States also submits that even Brazil’s proposed edits concede that producers are responding to market signals when expected prices are above the market loan rate.


      18. The Panel agrees with Brazil that certain statements in paragraph 10.127 of the interim report (paragraph 10.127 of this Report) do not entirely accurately reflect the Panel's findings regarding the issue of the insulation of US cotton producers form market prices. The Panel's finding is that whether expected market prices at the time of planting are above the levels that will trigger marketing loan and counter-cyclical payments is a factor that affects the degree of insulation of US cotton producers from market price signals but not the existence of such insulation per se. The Panel has amended the paragraph to better convey this idea.


      19. With respect to paragraph 10.146 of the interim report, (paragraph 10.146 of this Report), Brazil requests the Panel to delete the term "necessarily" from the last sentence of the paragraph. Brazil states that the inclusion of this word could be read as suggesting that, absent a clear "discernable temporal coincidence", there is an increased burden for a complaining Member to demonstrate adverse effects. Brazil considers that this is particularly troublesome in proceedings under Article 21.5 of the DSU because there is already a finding adopted by the Dispute Settlement Body that the challenged subsidies cause adverse effects. Similarly, a complaining Member that challenges a long-existing distortion of a market caused by long-term subsidization might not be able to establish that the effects of subsidies increased simply because the market distortion has continued to exist for a long time. In Brazil’s view, as in the original proceeding, the existence of a "discernable temporal coincidence" can be a useful factor to assess in an adverse effects claim. However, it is not an indispensable element for establishing causation.


      20. The United States objects to Brazil's request, arguing that there is no basis for Brazil’s argument that the word "necessarily" in the last sentence of paragraph 10.146 of the interim report relates to the question of a party’s burden of proof. The United States considers that the Panel is merely stating its own findings based on the evidence presented with respect to causation. The United States also submits that Brazil actually seems to imply that the burden of demonstrating adverse effects in proceedings under Article 21.5 of the DSU should be less "because there already is a finding adopted by the Dispute Settlement Body that the challenged subsidies cause adverse effects". In making this argument, Brazil ignores the fact that since the "challenged subsidies" in this proceeding are not the same ones as those at issue in the original proceeding, there is no finding with respect to payments made outside of MY 1999-2002. Moreover, as stated by the Panel in paragraph 9.3 of the interim report, "[t]he general rules regarding the allocation of the burden of proof in WTO dispute settlement require that a party claiming a violation . . . assert and prove its claim. These rules also apply to proceedings under Article 21.5 of the DSU." Thus, Brazil must make a showing of adverse effects, assuming that its claim with respect to the marketing loan payment and counter-cyclical payment programmes is properly before the Article 21.5 compliance panel in the first place.


      21. The Panel rejects Brazil's request to delete the term "necessarily" from the last sentence of paragraph 10.146 of the interim report (paragraph 10.146 of this Report). The Panel disagrees with Brazil's analysis of the implications of the use of this term. The word "necessarily" in the last sentence of paragraph 10.146 simply underlines the point that price suppression can exist even though "recent years have not witnessed the sharp decline in the world market price for upland cotton that occurred during the period considered by the original panel". The Panel therefore considers that there is no basis for Brazil's argument that this word can somehow be interpreted as implying a view of the Panel that "absent a clear 'discernible temporal coincidence', there is an increased burden for a complaining Member demonstrating (sic) adverse effects" and that "a complaining Member would be required to explain that the absence of a discernable temporal coincidence does not break any causal link between the subsidies and the adverse effects."

        WT/DS267/RW

        Page 18


      22. The United States requests that paragraph 10.162 of the interim report (paragraph 10.162 of this Report) be amended to indicate that the new data begins with MY 2003, based on the more recent USDA cost survey, and that the MY 2002 data originates with the 1997 USDA survey. Brazil objects to the US request, noting that USDA Survey Data is updated annually. The Panel has decided to substitute "data" for "updated information" at the beginning of the second sentence of paragraph 10.162 of the interim report.


      23. With regard to the Panel's statement in the last sentence of paragraph 10.166 of the interim report (paragraph 10.166 of this Report) that "[w]e therefore consider the inclusion of land, labour and capital recovery costs in fixed (overhead) costs as appropriate, being in accordance with the recommended methodology", the United States requests the Panel to differentiate between paid and unpaid labour costs and to treat paid labour costs as variable. The United States notes that with respect to paid labour that does not have farm ownership claims, there is a difference in opinion on whether to classify the cost as fixed or variable, and that, as a conservative measure, the United States considers paid labour cost as a variable cost. Brazil has not commented on this request of the United States.


      24. The Panel does not consider that the change requested by the United States is appropriate. Whether paid labour costs are classified as variable or fixed will have no effect on the cost of production analysis. The Panel sees no reason why it should make a ruling on this matter when such a ruling would not affect its cost of production analysis.


      25. Brazil requests the Panel to make changes to the text of paragraph 10.176 of the interim report (paragraph 10.176 of this Report), to reflect that there is no requirement that proof of long-term costs be produced over the lifetime of the challenged measure. Brazil notes that many subsidy programmes have been in place for many decades. Moreover, any suggestion that cost of production data needs to cover the lifetime of a measure would be inconsistent with claims of threat of serious prejudice under Articles 5 and 6 of the SCM Agreement in particular. Finally, the notion that the full lifetime of a measure should be the basis for assessing proof of adverse effects is inconsistent with the concept of a more recent reference period to assess the existence of present adverse effects.


      26. The United States disagrees with Brazil’s argument that the Panel is possibly implying a requirement that proof of long-term costs be produced over the lifetime of a challenged measure. Rather, the Panel simply states that the effect of marketing loan and counter-cyclical payments is "best assessed" over the lifetime of the FSRI Act, but recognizes that data was unavailable for that entire time frame, and relies instead on the data that is available for MY 2002-2006.


      27. The Panel agrees with some of the concerns that Brazil has raised and has therefore made certain of the changes proposed by Brazil. However, the Panel also agrees with the United States that other changes proposed by Brazil attempt to expand the universe of the cost data reviewed by the Panel. The Panel therefore declines to make those changes.


      28. With respect to paragraph 10.188 of the interim report (paragraph 10.193 of this Report), the United States submits that the Panel has not accounted for the argument of the United States that ginning costs should not be included in the calculation of the costs of production of upland cotton lint. The United States submits that it has demonstrated that ginning costs are appropriately excluded from the calculation of costs of production of upland cotton lint, in light of the common practice in the United States of ginning costs being paid out of the proceeds gained by the gin from the sale of the cotton seed separated out in the ginning process. The United States therefore requests the Panel to find that ginning costs should be excluded from the calculation of the costs of production of upland cotton and to make adjustments accordingly to its calculation of cotton revenue costs and returns.

        WT/DS267/RW

        Page 19


      29. Brazil objects to the United States request for exclusion of ginning costs from the calculation of the costs of production of upland cotton lint. First, Brazil observes that such an approach is inconsistent with the original panel’s approach, which, in making its cost of production-related findings, relied upon USDA cost data that included ginning costs and market returns from cottonseed. USDA cost of production data shows that the costs associated with ginning are far higher than the market value of cottonseed. An exclusion of ginning costs and revenue would result in manipulated cost of production data that would paint an inappropriately advantageous picture of the situation of US upland cotton farming. Brazil submits, in this respect, that it has explained that the US approach would cut out a large portion of ginning costs and a much smaller portion of upland cotton producers' market revenue from cottonseed. Second, Brazil notes that if one were to exclude ginning costs, for accuracy, at the very least, ginning revenues would also have to be excluded. However, the exclusion of such costs and revenues would result in an inadequate basis to assess medium-to-longer-term developments in the industry.


      30. The Panel agrees with the United States that the interim report does not address the argument made by the United States in this proceeding that ginning costs should be excluded from the calculation of the costs of production of upland cotton lint. Without taking a position on whether this argument is correct or not, the Panel has analysed the revenue and cost data in two forms: (i) without cottonseed revenues and ginning costs as the US proposes and (ii) with cottonseed revenues and ginning costs as in the original dispute. The results of this analysis show that whether or not ginning costs are excluded from the calculation of the costs of production of upland cotton lint does not alter the Panel's conclusion that market revenues exceed total costs of production. To reflect this analysis in the final report, we have amended the interim report by adding new paragraphs 10.185-10.188 and by changing the text of paragraphs 10.185-10.191 of the interim report (paragraphs 10.189-10.196 of this Report).


      31. With respect to paragraph 10.214 of the interim report (paragraph 10.219 of this Report), the United States requests that the Panel delete the last sentence, which states that "a good model does not need a lot of extraneous bells and whistles". In the view of the United States, a proper model estimating the degree of price suppression should be constructed to take into account as many relevant economic factors as possible.


      32. Brazil objects to the request of the United States, explaining that the Panel has correctly identified the advantages of Brazil’s model. Brazil agrees with the Panel that a "simpler and more accessible model can be readily focused on the specific issue at hand" and that "it should include only relevant economic features that are crucial to a determination whether certain US Government payments led to significant price suppression."


      33. Notwithstanding Brazil's objection, the Panel has accepted the request of the United States to delete the last sentence of paragraph 10.214 of the interim report (paragraph 10.219 of this Report). The Panel explains its assessment of the advantage of Brazil's model in the second and third sentences of the paragraph. The statement in the fourth sentence of the paragraph that "a good model does not need a lot of extraneous bells and whistles" is not essential in this respect and its deletion does not entail a substantive change.


      34. Brazil requests that the Panel delete the phrase "it has not undergone any peer review" from paragraph 10.215 of the interim report (paragraph 10.220 of this Report). Brazil expresses a concern that in future disputes economic models that have not undergone a "peer review" will be heavily discounted or even discarded for that reason alone, leading to a "peer review" requirement for economic models offered, for example, in disputes involving claims of adverse effects under the SCM Agreement or in proceedings under Articles 21.5 and 22.6 of the DSU. Brazil submits that from a practical standpoint peer review can be a quite lengthy and costly process. A crucial consideration in determining the issue of "peer review" is the fact that any economic model must be tailored to the

        WT/DS267/RW

        Page 20


        particular facts and circumstances of a dispute. Thus, any economic analysis presented to a panel or arbitrator in a case of adverse effects is likely to be very recent in order to reflect current conditions of competition in the relevant market. Given the relatively short time-frames of WTO dispute settlement, it may simply not be possible to secure a timely "peer review" of an economic model specifically tailored to the current conditions of competition. Brazil also submits that a requirement for "peer review" is not supported by an examination of proceedings under Article 22 of the DSU. A number of economic models have been presented to arbitrators, but there is no evidence in any of the Article 22.6 arbitration reports that these models had been "peer reviewed". Finally, Brazil claims that there is no precise definition of "peer review".


      35. The United States notes that the phrase "it has not undergone any peer review" is a simple factual statement. The United States also states, however, that the term "peer review" may not apply as well in this context so that a better means of expressing the concern over potential deficiencies inherent in the economic model would be that the model was not well-known and not widely-used for published research and analysis.


      36. The Panel agrees with Brazil that there exists no requirement that economic models submitted in WTO dispute settlement proceedings must have been the subject of a process of peer review. The Panel also notes that the United States concurs with Brazil that the term "peer review" may not be applicable in this context. In light of these considerations, the Panel has deleted the reference to peer review in the second sentence of paragraph 10.215 of the interim report (paragraph 10.220 of this Report).


      37. Regarding the Panel's statement in paragraph 10.217 of the interim report (paragraph 10.222 of this Report) that "price suppression has been the outcome of all the simulation results whether one uses the parameter values proposed by Brazil or the FAPRI and ATPSM parameter values proposed by the United States", the United States submits that the appropriate question before the Panel is whether there has been significant price suppression, not simply any price suppression. The United States argues that it has submitted the alternative parameter values solely for purposes of showing that, even if one were to employ Brazil's modelling approach, correcting its flawed parameters shows that the price effects are not significant. The United States requests the Panel to clarify that Brazil’s model has not "earn[ed] the confidence of the Panel," and that it therefore has not contributed to the Panel’s finding of significant price suppression.


      38. Brazil objects to the request of the United States. Brazil argues that in its examination of price suppression the Panel has thoroughly examined the evidence and data provided by both parties. Like the original panel, the Panel has considered several elements to be crucial in its analysis. The price suppression results of Brazil's model, applying the parameter sets both of Brazil and of the United States provided additional comfort to the Panel in its finding of price suppression because all parameter sets showed that the effect of the subsidies is to suppress prices. While the Panel has not endorsed a particular set of parameters, this is not a sufficient basis for the United States to request that the Panel find that the model itself did not earn the Panel's confidence.


      39. The Panel rejects the request of the United States, since the Panel has adequately explained the role of the model in its analysis in paragraphs 10.217 and 10.241-10.252 of the interim report (paragraphs 10.222 and 10.246-10.257 of this Report). It is clear from the report that the Panel has used the results of the economic simulation model in conjunction with other factors – the substantial proportionate influence of the United States on the world market for upland cotton, the mandatory, and price-contingent nature and revenue-stabilizing effect of marketing loan and counter-cyclical payments, the order of magnitude of the marketing loan and counter-cyclical subsidies, and the significant gap between the total costs of production of US upland cotton producers and their market revenue – to arrive at its finding that the effect of marketing loan payments and counter-cyclical payments is significant price suppression.

        WT/DS267/RW

        Page 21


      40. Brazil requests the Panel to amend paragraph 10.226 of the interim report (paragraph 10.231 of this Report). Brazil's concern with respect to this paragraph pertains to the statement of the Panel that "it is not possible for the Panel to determine with any great deal of precision or confidence just how much of the projected decline in US exports of upland cotton is due to the elimination of the Step 2 payments". Brazil considers that, if read in isolation, this statement could be taken to suggest that the Panel considers that it is unable to assess the effects of the withdrawal of the Step 2 subsidies and hence should not have concluded that the withdrawal of these subsidies is insufficient to bring the United States into conformity with Articles 5 and 6 of the SCM Agreement. Brazil proposes that the Panel add a statement to recall its finding in paragraph 10.248 of the interim report (paragraph 10.253 of this Report) that "even if the entire decline of US exports is attributable to the repeal of Step 2, we would still find that marketing loan and counter-cyclical payments significantly suppress world prices". Brazil also suggests that the Panel cite the evidence submitted by Brazil that not the entire decline in US exports is attributable to the removal of the Step 2 payments.


      41. The United States disagrees with the change proposed by Brazil. The United States considers that Brazil's proposed change mis-characterizes the Panel's finding in paragraph 10.248 of the interim report (paragraph 10.253 of this Report). The Panel there notes that the "effect of the elimination of this subsidy on US exports and on the amounts of marketing loan payments and counter-cyclical payments suggests that the elimination of this subsidy does not affect the price suppressing effects of the marketing loan payments and counter-cyclical payments in the world market for upland cotton". However, in reaching this conclusion, the Panel does not make the alternative assumption that Brazil claims it does, i.e. that "even if the entire decline of US exports is attributable to the repeal of Step 2, we would still find that marketing loan and counter-cyclical payments significantly suppress world prices".


      42. The Panel declines to make the change proposed by Brazil to paragraph 10.226 of the interim report (paragraph 10.231 of this Report). The Panel considers that the text proposed by Brazil does not correctly reflect the finding of the Panel in paragraph 10.248 of the interim report (paragraph

    10.253 of this Report).


      1. The United States requests the Panel to reiterate in paragraph 12.269 of the interim report (paragraph 12.2 of this Report) the conclusion drawn in paragraphs 9.68-9.69 of the interim report (paragraphs 9.70-9.71 of this Report) that it is not necessary for the Panel to make a finding on Brazil's claim of non-compliance by the United States in the period 22 September 31 July 2006. Brazil has not commented on this request. The Panel has amended paragraph 12.269 of the interim report (paragraph 12.2 of this Report) by inserting a reference to the panel's ruling in paragraph 9.68 of the interim report (paragraph. 9.70 of this Report).


      2. The United States requests that the Panel replace the word "underwrite" by "guarantee" in the first sentence of paragraph 14.5 of the interim report (paragraph 14.5 of this Report) to more accurately describe the GSM 102 and GSM 103 programmes. The United States also requests the Panel to amend the paragraph to reflect the fact that the GSM 103 programme no longer operates. Brazil objects to the request of the United States. The Panel has made the change requested by the United States as it considers that the term "guarantee" more precisely describes the operation of the programmes at issue than the term "underwrite". The Panel has also amended paragraph 14.5 of the interim report to reflect the fact that the GSM 103 programme no longer operates.


      3. For clarity, the United States suggests that a sentence be inserted after the first sentence of footnote 605 (paragraph 14.30) of the interim report (footnote 621 to paragraph 14.30 of this Report) to describe with more precision what was at issue in the Brazil – Aircraft (Article 21.5 – Canada) dispute. The United States further requests that the last sentence of the footnote be revised to reflect more accurately its Opening Statement. Brazil does not comment on this request. The Panel has amended footnote 605 (footnote 621 in this Report) to reflect the US suggestion.

        WT/DS267/RW

        Page 22


      4. Brazil requests that the Panel modify the first sentence of paragraph 14.67 of the interim report (paragraph 14.67 of this Report). Brazil notes that, as the next sentence indicates, the source for Brazil's argument regarding re-estimates are not US "budget" documents but are, rather, CCC Financial Statements. The United States does not oppose Brazil's request, but suggests alternative language to make it clear that Brazil exclusively referred to the CCC Financial Statements. The Panel has amended paragraph 14.67 of the interim report (paragraph 14.67 of this Report) in accordance with the language suggested by the United States.


      5. Both parties have asked that the Panel make certain changes to paragraph 14.76 of the interim report (paragraph 14.76 of this Report).


      6. Brazil's concern lies with the use by the Panel of the term "inflated". Brazil is concerned that that word is not neutral and may be misunderstood as indicating that the Panel considered the initial estimate to be flawed through exaggeration or deliberate overstatement. The United States objects to Brazil’s proposed edit. The United States considers that adopting Brazil's suggestion would unnecessarily eliminate one of the two points made by the Panel in the last sentence of paragraph 14.76 – the Panel's recognition that it is possible that the initial subsidy estimates are inflated because they do not reflect the actual use of the programme.


      7. With respect to the same paragraph, the United States is concerned that the Panel is imposing an impossible standard to achieve when it indicates that "[i]n no case has the reduction turned a positive subsidy estimate into a negative one" and requests that the Panel eliminate that sentence. The United States submits that re-estimates are not calculated from, or based on, the original subsidy estimates. Consequently, the re-estimate process can never re-capture any difference between the original subsidy estimate and estimates based on the amount of guarantees actually issued. Brazil considers that the United States misunderstands what it believes to be the intent of paragraph 14.76.


      8. The Panel has made changes to paragraph 14.76 of the interim report (paragraph 14.76 of this Report) and to footnote 665 of the interim report (footnote 681 of this Report) to make it clear that what is at issue in this paragraph is not the re-estimate process that is mandated under the FCRA. Rather, the Panel, in this paragraph, is merely describing the process by which the initial subsidy estimate is readjusted in the two subsequent budgets that follow the first budget in which a subsidy estimate for a particular cohort is provided, and increasingly takes actual use of the programme into account. The United States argued before the Panel that the initial subsidy estimate is unreliable because it is computed before any actual use is made of the GSM 102 programme. In paragraph 14.76 the Panel responds to that argument of the United States. The Panel therefore declines to delete the sentence identified by the United States. With respect to the use of the word "inflated", the Panel has replaced that term by "overstated" which, it believes, is more neutral and reflective of the Panel's intent.


      9. The United States requests that the Panel clarify its characterization of the US position concerning the initial estimates in paragraph 14.78 of the interim report (paragraph 14.78 of this Report). Brazil objects to the US request. The Panel has amended the paragraph to remove the word "indirectly" and clarify the terms "by themselves" in order to better reflect the argument of the United States. The Panel does not understand the United States to have requested any other change to the paragraph at issue.


      10. Brazil makes a number of comments with respect to the second sentence of paragraph 14.81 of the interim report (paragraph 14.81 of this Report). First, the Panel dismisses the arguments of the United States that the subsidy re-estimates demonstrate that GSM 102, GSM 103 and SCGP programmes were provided at no net cost to government. The Panel states that the evidence provided by the United States "do[es] not conclusively establish that the programmes were provided at no net

        WT/DS267/RW

        Page 23


        cost to the US Government".51 Brazil is concerned that this formulation might be read to suggest that the Panel has imposed the burden of proof on the United States, which Brazil understands is not the case. Brazil, therefore, suggests that this passage be reformulated to state that the evidence provided by the United States does not sufficiently rebut the evidence that the programmes were provided at a net cost to the US Government. Second, the Panel characterizes negative subsidy numbers as indicative of "profitability". Brazil requests that the Panel refer to the concept of "net cost", which the Panel has used elsewhere, instead of the word "profitable". Third, the Panel’s statement suggests that the re-estimated data provided by the United States uniformly shows profitability across all cohorts, whereas the data provided by the United States, however, continues to project losses for several cohorts during the period FY 1992-2006. The United States objects to Brazil's suggested edits apart from the one which corrects language that would appear to place the burden of proof on the United States.


      11. The Panel has amended the paragraph at issue (paragraph 14.81 of this Report) to adopt Brazil's suggestion with respect to the Panel's use of the term "profitability", a term which the Panel had used as shorthand for the concept of "no net cost to the government." The Panel has also taken note of Brazil's comment with respect to the fact that the re-estimates data submitted by the United States does not project a negative subsidy estimate in respect if all cohorts and has inserted a mention to that effect in a footnote (footnote 689). The Panel disagrees, however, with Brazil that the second sentence of the paragraph may be interpreted as the Panel having imposed the burden of proof on the United States. The Panel has clarified the sentence at issue to better convey the intended meaning of the sentence.


      12. The United States submits that footnote 673 of the interim report (footnote 690 of this Report) does not accurately reflect its arguments concerning the role played by original subsidy estimates in the original panel's conclusion that the programmes at issue fell within the scope of item

        (j) of the Illustrative List. Brazil objects to the US request. The Panel considers that footnote 673 of the interim report (footnote 690 of this Report) accurately reflects the argument of the United States in its Response to Panel Question 105 and therefore rejects the request of the United States.


      13. The United States considers that the use of the word "figure" in the third sentence of paragraph 14.83 of the interim report (paragraph 14.83 of this Report) is confusing, as the previous sentence references the CCC’s estimated net cash outflow for export credit guarantees as of 30 September 2006, a "figure" that was not before the original panel. Moreover, the United States indicates that while it agrees with the original panel's conclusion that the credit guarantee liability figure does not represent in any sense an actual loss, the United States in no way accepts the original panel’s ultimate reliance on the credit guarantee liability figure as an indicator of the estimated long- term loss of the programmes to the US Government. Brazil opposes the second aspect of the US request.


      14. The Panel takes note of the concern of the United States with respect to the use of the term "figure" in the sentence at issue and of its concern that it not be depicted as endorsing a view that the "credit guarantee liability" reported in the CCC's Financial Statements indicates a future loss. We have, for this reason, amended and moved the sentence at issue to a new footnote (footnote 699) which quotes from the US Response to Panel Questions 107 and 108 and therefore, in the Panel's view, accurately reflects the arguments made by the United States.


      15. The United States submits that the Panel misconstrues its criticism of Brazil’s failure to include cash received when it states, in paragraph 14.86 of the interim report (paragraph 14.86 of this Report) that "including these amounts would not allow a comparison with the re-estimates data submitted by the United States." The United States explains that its point was not to facilitate a


        image

        51 Emphasis added.

        WT/DS267/RW

        Page 24


        comparison with re-estimates data but was simply to offer a more accurate depiction of cash flows during the period of time identified by Brazil. The Panel has amended paragraph 14.86 of the interim report (paragraph 14.86 of this Report) to reflect the fact that the comment at issue was not made in response to an argument of the United States but rather reflected an additional reason why the Panel considered it inappropriate to include amounts associated with recoveries on pre-1992 guarantees, as had been suggested by the United States.


      16. Brazil requests that the Panel modify the final two sentences of paragraph 14.94 of the interim report (paragraph 14.94 of this Report). In paragraph 14.94, the Panel suggests that GSM 102 fees are not subject to the OECD Arrangement "because", "the Arrangement only applies to official support for terms of two years or more". This statement implies that GSM 102 guarantees are not subject to the OECD Arrangement because they are provided for tenors not covered by the Arrangement. Brazil notes that, as the Panel correctly observes at paragraph 14.102, however, GSM 102 provides for tenors (over 24 months) which are subject to the OECD Arrangement. The United States does not object to Brazil’s stylistic edits to this paragraph. The Panel has made the changes suggested by Brazil.


      17. With respect to paragraph 14.96 of the interim report (paragraph 14.96 of this Report), the United States asks the Panel to note that the US participation in the elaboration of the MPRs does not mean in any way that the United States views the MPRs as a proper basis for comparison to GSM 102 fees in this case. Brazil does not comment on this request. The Panel considers that there is no ambiguity in the interim report that the United States considers that the MPRs are not applicable to GSM 102 export credit guarantees (see paragraph 14.92 of the interim report, paragraph 14.92 of this Report) and therefore considers that the addition suggested by the United States is unnecessary.


      18. Brazil and the United States request that the Panel correct what they consider to be a factual inaccuracy in paragraph 14.117 of the interim report (paragraph 14.117 of this Report). They indicate that there are 6, rather than 7, eligible country risk categories under the amended GSM 102 programme. The Panel has deleted the reference to there being "7" eligible country risk categories in the body of paragraph 14.117, but has added a new footnote to that paragraph to explain that the parties have, in their submissions to the Panel, been inconsistent in designating the number of eligible country risk categories.


      19. With respect to paragraph 14.115 of the interim report (paragraph 14.115 of this Report), Brazil requests that the Panel either provide a citation to documentary evidence in support of the assertion of the United States that it "sets bank limits with respect to each foreign bank obligor" or, alternatively, that the Panel note that this assertion in this respect is unsupported by documentary evidence. The United States objects to Brazil's suggestion.


      20. The Panel considers that the main point of the last sentence of paragraph 14.115 does not concern whether the United States has established as a fact that it sets bank limits with respect to each foreign obligor. Rather, the sentence at issue indicates that whether or not such bank limits exist does not alter the Panel's conclusion, in the first sentence of the same paragraph, that "GSM 102 fees are not risk-based insofar as they do not take into account one of the main elements of risk associated with the guarantees, the risk of default that is specific to an individual foreign obligor." In any case, the Panel does not understand Brazil to have challenged, in its submissions to the Panel, the US assertion that it sets bank limits with respect to each foreign bank obligor. The Panel therefore denies Brazil's request.


      21. The United States requests that the Panel delete the parenthetical, in the last sentence of paragraph 14.119 of the interim report (paragraph 14.119 of this Report), to the effect that "meaning that, for the United States, the other programmes were primarily responsible for any net cost to the US Government in the original proceeding". The United States disagrees to the extent that the sentence

        WT/DS267/RW

        Page 25


        implies that the United States agrees that the three export credit guarantee programmes resulted in any net cost or loss. Brazil objects to the US request. The Panel considers that the use of the term "any" in the sentence at issue ensured that the reader would not ascribe to the parenthetical the meaning which the United States suggests may be implied. Nonetheless, the Panel is mindful of accurately reflecting the parties' arguments and has therefore deleted the parenthetical. In any case, the Panel does not regard that sentence as essential to the understanding of the Panel's discussion of this issue.


      22. Brazil requests that the Panel modify the second sentence of paragraph 14.120 of the interim report (paragraph 14.120 of this Report), in which the Panel indicates that it is "therefore unable to assess whether a 46 per cent average fee increase (23 per cent trade-weighted increase) is sufficient to ensure that the GSM 102 programme is not provided at a net cost to the US Government." Brazil has two concerns with this sentence. First, Brazil is concerned that this formulation might be read to suggest that the Panel has imposed the burden of proof on the United States, which Brazil understands is not the case. Second, Brazil believes that it would be prudent to put the Panel’s statement in the context of its overall conclusion, in paragraph 14.131, that the evidence as a whole demonstrates that the GSM 102 export credit guarantee programme is provided against premiums which involve a net cost to the US Government. The United States does not comment on Brazil's request with respect to this paragraph. The Panel does not consider that the language used in paragraph 14.120 of the interim report (paragraph 14.120 of this Report), gives rise to the concern expressed by Brazil. The Panel therefore rejects Brazil's request to amend this paragraph.


        1. EDITORIAL CORRECTIONS


      23. In addition to the above-mentioned changes, the Panel has made a number of editorial changes and technical corrections.52


  8. PROCEDURAL MATTERS


    1. REQUEST OF BRAZIL THAT THE PANEL EXERCISE ITS DISCRETION TO SEEK CERTAIN INFORMATION FROM THE UNITED STATES UNDER ARTICLE 13 OF THE DSU


        1. On 1 November 2006, prior to the organizational meeting of this Panel, Brazil submitted a document entitled "Brazil's Request to the Compliance Panel to Pose Questions to the United States, under Article 13.1 of the DSU", in which it requested the Panel to "exercise its discretion under Article 13.1 [of the DSU] to request the United States to produce documents and information relating to (a) payments to US producers of upland cotton under the US counter-cyclical payment programme and (b) the export credit guarantee programmes (principally the GSM 102 export credit guarantee programme)".53 Brazil asserted that all the requested documents and information were "highly relevant to Brazil's claims and the issues before this compliance Panel. Therefore, Brazil considers that it is 'necessary and appropriate' for the compliance Panel to seek the documents and information requested at this early stage of the compliance proceeding". Brazil asked the Panel to "request the United States to produce this information within three weeks of the date the Panel issues its request".


        2. On 7 November 2006, at the request of the Panel, the United States submitted comments on Brazil's request. The United States requested the Panel to decline Brazil's request stating, inter alia,


          image

          52 In the course of the interim review, the Panel has noted that the paragraph numbers cited by the Parties in their submissions are sometimes different from the paragraph numbers that appear in the paper version of the submission the parties are referring to. The Panel understands that this discrepancy is often a result of the improper functioning of the electronic version submitted by one of the parties. The Panel has endeavoured to correct any such discrepancies, but the Panel points out, for the record, that the final report may still contain such discrepancies in the references to paragraph numbering of the parties' submissions.

          53 Excerpt from Brazil's 1 November 2006 submission. See item 1 of Annex D for the full text of this submission.

          WT/DS267/RW

          Page 26


          that "Brazil fails to explain how the Panel can assess whether any of the documents and information [it is seeking] is 'essential' or 'suitable' for the Panel's objective assessment of the matter before it, and not to be used to make Brazil's prima facie case for it, before either party has even had the opportunity to present any arguments or any evidence in support of its case".54


        3. On 8 November 2006, the Panel communicated the following to the Parties:


          "The Panel declines to make a decision on Brazil's request at this point of time. The Panel will communicate to the parties its decision on this request after having received and considered Brazil's first submission. In this connection, the Panel reserves its right to amend the timetable (and the working procedures, as necessary) in relation to this issue."55


        4. On 17 November 2006, in accordance with the Panel's timetable, Brazil made its First Written Submission. In relation to its request under Article 13 of the DSU concerning counter-cyclical payments, Brazil stated the following:


          "During informal discussions held [before the establishment of the Panel], Brazil requested the United States to provide the information needed to allocate counter- cyclical payments to upland cotton production in MY 2003-2005. The United States claimed that this information was 'confidential.' In a letter to the compliance Panel dated 1 November 2006, Brazil requested the compliance Panel to request the United States, pursuant to Article 13.1 of the DSU, to produce the exact same type of information that it finally produced during the original proceeding in response to repeated requests by the original panel, updated for MY 2003-2005. Brazil reiterates its request based on the reasons set out in this submission and in its letter of 1 November 2006."56 (footnotes omitted)


        5. Brazil also stated:


          "Furthermore, there is every reason to believe that a strong positive relationship continues to exist between upland cotton base acre holders receiving annual counter- cyclical payments and upland cotton production. As during the period MY 1999- 2002, the economics of high-cost upland cotton production in the United States compel that a significant majority of US upland cotton be grown on upland cotton base acres (or other types of high-payment base acreage such as rice and peanuts) during MY 2003-2005. Brazil looks forward to receipt of this base and planted acreage information from the United States following a request by the compliance Panel to the United States pursuant to Article 13.1 of the DSU."57 (footnotes omitted)


        6. In its First Written Submission, Brazil did not refer to its DSU Article 13 request in relation to export credit guarantee programmes.


          image

          54 Excerpt from the 7 November 2006 comments of the United States. See item 1 of Annex E for the full text of these comments.

          55 Panel's communication to the parties dated 8 November 2006, para. 1. See item 1 of Annex F for the full text of this communication.

          56 Brazil's First Written Submission para. 117.

          57 Brazil's First Written Submission para. 131.

          WT/DS267/RW

          Page 27


        7. The United States further explained its position on this matter in a letter of 20 November 2006.58 On 22 November 2006, Brazil submitted comments on this letter of the United States.59


        8. On 27 November 2006, the Panel sent a communication to the parties stating that it "is of the view that it cannot conclude at this stage that it is necessary and appropriate within the meaning of Article 13 of the DSU to seek the information requested by Brazil before the Panel has had the opportunity to review the arguments and supporting evidence, if any, that will be presented by the United States in response to Brazil's argument regarding the causal relationship between certain subsidies and price suppression for upland cotton".60


        9. On 11 January 2007, Brazil filed its Rebuttal Submission. In making its argument that marketing loan and counter-cyclical payments were trade-distorting, Brazil stated the following:


          "Had the United States produced all data requested by Brazil, this data would allow Brazil to allocate counter-cyclical payments to upland cotton planted on non-upland cotton base acreage under Brazil's Methodology, as determined to be 'appropriate' and conservative by the original panel (as detailed below, Brazil requests the compliance Panel to exercise its discretion under Article 13.1 of the DSU to request that the United States produce the same data for MY 2003-2005 that was available to the original panel for MY 2002)."61 (footnote omitted)


          ...


          "As noted above, the limited data that the United States produced (in the context of its flawed argument that counter-cyclical payments have no significant effects on production) does not allow Brazil to update the amounts of counter-cyclical payments it allocates to upland cotton based on its own methodology. Therefore, Brazil reiterates its request, set out in its 1 November 2006 letter to the compliance Panel, that the Panel exercise its discretion and request that the United States produce the same data for MY 2003-2005 that was available to the original panel for MY 2002."62


        10. In its Rebuttal Submission Brazil did not refer to its DSU Article 13 request in relation to export credit guarantee programmes.


        11. On 17 January 2007, the Panel posed certain questions to the United States and to Brazil on Brazil's request under Article 13 of the DSU.63 Brazil and the United States submitted their responses to these questions on 19 January 2007.64 In relation to export credit guarantee programmes65, Brazil stated that it "continues to believe that the information, documents and data requested in [its request submitted on 1 November 2006] will assist the compliance Panel in making 'an objective assessment of the matter before it,' within the meaning of Article 11 of the [DSU]". In relation to counter-cyclical


          image

          58 The full text of the US letter dated 20 November 2006 is contained in item 2 of Annex E.

          59 The full text of Brazil's letter dated 22 November 2006 is contained in item 4 of Annex E.

          60 Excerpt from the Panel's communication dated 27 November 2006. The full text of this communication is contained in item 2 of Annex F.

          61 Brazil's Rebuttal Submission para. 165.

          62 Brazil's Rebuttal Submission para. 171.

          63 See item 5 of Annex F for the full text of the questions.

          64 See items 2 and 3 of Annex D for the full text of the responses of the United States and Brazil, respectively.

          65 The question read as follows: "Does Brazil deem that the questions posed [concerning export credit guarantee programs] submitted on 1 November [..] are now unnecessary?"

          WT/DS267/RW

          Page 28


          payments, the United States pointed out that it had actually provided, in an exhibit to its First Written Submission, the data (for MY 2005) sought by Brazil.


        12. On 22 January 2007, Brazil submitted another letter66 regarding the data on counter-cyclical payments. In this letter, while acknowledging that the data provided by the United States for MY 2005 corresponded to what Brazil had sought, Brazil pointed out that the United States had still not submitted the data for MY 2003-2004 and MY 2006.


        13. On 5 February 2007, the United States filed its Rebuttal Submission. In it, the United States reiterated that it had already submitted the data sought by Brazil in respect of counter-cyclical payments, and stated that MY 2006 data were not yet available and that earlier years were not relevant to Brazil's claim of "present" serious prejudice. The United States therefore considered that Brazil's request was moot.67


        14. On 21 March 2007, the Panel issued the following communication to the parties on this matter:


          "The Panel notes that Brazil's request was made prior to the filing of the written submissions in this case and that information on certain items identified in that request has since been provided to the Panel. The Panel also expects to receive information on issues raised in Brazil's request in response to the questions posed by the Panel on 15 March 2007. The Panel considers that at this juncture it is not necessary to seek more information in respect of issues identified in Brazil's request. The Panel may, however, decide to request this information at a later stage of this proceeding and may ask questions to the parties at any time, in accordance with paragraph 7 of the Panel's working procedures."68


        15. The Panel ultimately did not resort to Article 13 of the DSU concerning the above or any other issue.


    2. REQUEST OF THE UNITED STATES FOR OPEN HEARING


        1. During the organizational meeting held on 6 November 2006, the United States stated, in relation to paragraph 2 of the draft working procedures69, that it preferred the panel meeting to be opened to the public. The United States acknowledged that it could not secure agreement with Brazil on this issue, and requested that, given Brazil's refusal, the working procedures at least should indicate that the portions of the meeting where the United States presented its position be open to the public.


        2. On 8 November 2006, the Panel sent a communication in which it stated as follows:


          "In relation to paragraph 2 of the working procedures, the Panel has noted that the United States wishes to open up the portion of the meeting where it presents its own views and statements, while Brazil opposes opening the meeting to the public. The Panel draws the attention of the parties that opening up of any portion of the Panel meeting entails logistical considerations. The Panel invites the United States to submit comments, by close of business of Tuesday 21 November, on how exactly it


          image

          66 See item 8 of Annex E for the full text of Brazil's letter.

          67 See US Rebuttal Submission, para. 278.

          68 Excerpt from the Panel's communication dated 21 March 2007. See item 8 of Annex F for the full text of this communication.

          69 The Panel's draft working procedures paragraph 2 read as follows: "The Panel shall meet in closed

          session. The parties to the dispute, and interested third parties, shall be present at the meetings only when invited by the Panel to appear before it." (This paragraph remained unchanged in the final working procedures.)

          WT/DS267/RW

          Page 29


          envisages such opening to be logistically meaningful. The Panel reserves its right to amend the working procedures in relation to this matter." 70


        3. On 21 November 2006, the United States submitted, in response to the Panel's invitation above, its comments on the possible logistics of opening a specific part of the meeting. The United States explained inter alia that it "has invoked its right under Article 18.2 of the [DSU] to make statements of its own positions public and, as an appropriate means to do this, has requested that the Panel exercise its authority to open to the public the US statements and US answers to questions presented in the panel meeting".71


        4. In a letter of 24 November 2006, Brazil indicated that it did not agree with the assertions of the United States, noting inter alia, that "nothing in Article 18.2 mandates or implies that to effect a Member's right to disclose its own positions or statements, panels must agree to open meetings to public observation" and that "a Member's right to disclose its own positions cannot prejudice another Member's right to confidentiality".72


        5. On 28 November 2006, the Panel informed the parties as follows:


          "The Panel notes, first, that there is no precedent in WTO dispute settlement practice for a decision to open a (portion of) a panel meeting to the public based on a request by one party to a dispute that is opposed by another party. Second, the Panel has carefully considered the argument of the United States in its letter dated 21 November that in requesting the Panel to open to the public US statements and US answers to questions presented in the panel meeting the United States is exercising its right under Article 18.2 of the DSU, second sentence. The Panel considers that there are various possible ways in which a party can disclose its statements to the public and that the United States has not explained why procedures the United States normally uses to disclose its statements in WTO dispute settlement proceedings to the public would not be adequate in this case. Furthermore, the right of party to disclose its statements of its positions must also be reconciled with the obligation of confidentiality in the third sentence of Article 18.2. Finally, the Panel considers that the practical arrangements proposed by the United States in its letter of 21 November 2006 pose serious practical and logistical difficulties.


          For the above reasons, we decline to grant the request made by the United States."73 (footnote omitted)


    3. SERVICING OF SUBMISSIONS


        1. On 18 December 2006, Brazil sent a letter to the Panel complaining about the late filing of the First Written Submission by the United States74 and requested that "the compliance Panel reject the United States' submission as not timely filed".75


          image

          70 Panel's communication to the parties dated 8 November 2006, paragraph 3. See item 1 of Annex F for the full text of this communication.

          71 US letter dated 21 November 2006, second paragraph. See item 3 of Annex E for the full text of this

          letter.


          72 See item 5 of Annex E for the full text of Brazil's letter of 24 November 2006.

          73 Excerpt from the Panel's communication dated 28 November 2006. See item 3 of Annex F for the

          full text of this letter.

          74 Brazil stated that "the United States [...] filed the electronic version of its submission 12 hours late, on Saturday, 16 December at 5:24 a.m."

          WT/DS267/RW

          Page 30


        2. On 19 December 2006, the United States sent a letter to the Panel stating that the Panel should reject Brazil's request, asserting inter alia that "the DSU does not authorize the 'rejection' of submissions for the inability of a Member to meet precisely the time by which parties 'should provide their submissions' according to a panel's working procedures. To the contrary, the DSU provides that '[p]anel procedures should provide sufficient flexibility so as to ensure high-quality panel reports, while not unduly delaying the panel process'".76 (footnotes omitted)


        3. On 20 December 2006, the Panel communicated the following to the parties:


          "The Panel notes that Brazil requests the Panel to 'reject the United States' submission as not timely filed'. The Panel declines to do so. However, the Panel regrets this delay by the United States, and requests the United States to abide by the deadlines established by the Panel concerning its future submission(s)."77


        4. On 7 February 2007, Brazil sent a letter to the Panel complaining about the late filing of the Rebuttal Submission by the United States.78 Brazil requested that the compliance Panel "reject the United States' submission as not timely filed".79


        5. On 12 February 2007, the United States sent a letter to the Panel commenting on this letter from Brazil stating inter alia that "Brazil has no basis for its request that the Panel reject the US Rebuttal Submission".80


        6. On 16 February 2007, the Panel communicated the following to the parties:


          "The Panel regrets that the United States has missed the deadline for the second time in this Article 21.5 proceeding. The Panel finds it extraordinary that the United States has missed so many deadlines in the original proceeding as well as in this Article 21.5 proceeding.


          On the other hand, the Panel considers that there is no provision in the DSU that allows a panel to simply reject an entire submission as not timely. The Panel does not consider that the two cases cited by Brazil support Brazil's request. Thus, we decline to grant the request made by Brazil."81 (footnote omitted)


          image


          letter.

          75 Excerpt from Brazil's letter dated 18 December 2006. See item 6 of Annex E for the full text of this


          76 Excerpt from US letter dated 19 December 2006. See item 7 of Annex E for the full text of this letter.

          77 Excerpt from the Panel's communication to parties dated 20 December 2006. See item 4 of Annex F

          for the full text of this communication.

          78 Brazil stated that "the United States [...] filed the electronic version of its submission more than 6 hours late - at 11:39 p.m. on Monday, 5 February".

          79 Excerpt from Brazil's letter dated 7 February 2007. (footnote omitted.) See item 9 of Annex E for the full text of this letter.

          80 Excerpt from US letter dated 12 February 2007. See item 10 of Annex E for the full text of this

          letter.


          81 Excerpt from the Panel's communication dated 16 February 2007. See item 7 of Annex F for the full

          text of this communication.

          WT/DS267/RW

          Page 31


    4. REQUEST OF THE EUROPEAN COMMUNITIES FOR A FINDING OR RULING ON THE PANEL'S COMPOSITION


        1. The European Communities, a third party to this proceeding, requests the Panel to make "a finding or ruling" with respect to "the propriety of [the Panel's] composition".82


        2. The Panel considers that it is not within its authority under the DSU to make a "finding or ruling" on an issue that has not been raised by any of the parties to the dispute and which concerns the application by the WTO Director-General of the DSU provisions regarding panel composition.83 The Panel fails to see how the "finding or ruling" requested by the European Communities would contribute to a positive solution to this dispute within the meaning of Article 3.7 of the DSU.


    5. LEAST DEVELOPED COUNTRY MEMBERS


      1. Chad, a least-developed country Member, participated in this proceeding as a third party. In accordance with Article 24.1 of the DSU, particular consideration was given to the special situation of this Member. Chad made a detailed written submission and made an oral statement at the meeting of the Panel with the third parties on 28 February 2007.84 Chad was represented by a delegation which was headed by its Ambassador to the United Nations and Other International Organisations in Geneva and which also included staff members of the Advisory Centre on WTO Law. In its Submission and Oral Statement, Chad stressed that it had a special interest in this proceeding in light of the critical importance of the cotton sector to its economy and of the commitments that WTO Members have undertaken with respect to cotton in the July 2004 Framework and at the Hong Kong Ministerial Conference. Chad emphasized the benefits for Chad and other Sub-Saharan cotton-producing countries of full implementation by the United States of the DSB recommendations and rulings in this dispute and drew attention, in this regard, to recent studies that provide empirical evidence of the magnitude of these benefits. Chad also submitted detailed legal arguments in support of its view that the United States had not complied with the DSB recommendations and rulings. The Panel refers to the arguments of Chad in the summaries of arguments of the third parties.


  9. GENERAL CONSIDERATIONS AND ISSUES RELATING TO THE SCOPE OF THIS PROCEEDING


    1. STANDARD OF REVIEW, BURDEN OF PROOF AND RULES OF TREATY INTERPRETATION


      1. Standard of review


          1. In respect of the standard of review that this Panel must apply, Article 11 of the DSU provides, in relevant part:


            "The function of panels is to assist the DSB in discharging its responsibilities under this Understanding and the covered agreements. Accordingly, a panel should make an objective assessment of the matter before it, including an objective assessment of the facts of the case and the applicability of and conformity with the relevant covered agreements, and make such


            image

            82 EC Third Party Submission, paras. 6-30; EC Third Party Oral Statement, paras. 2-4.

            83 While the European Communities asserts that there is precedent in WTO dispute settlement practice

            that supports its position that the Panel has the authority to rule on the propriety of its own composition, and that the Panel is under a duty to address this issue because "it is of fundamental importance for the correct interpretation of the DSU and for the smooth and equitable operation of the WTO dispute settlement system" (EC Third Party Submission, paras. 28-29), the Panel finds the arguments of the European Communities in this regard unpersuasive.

            84 Chad's Written Submission and Oral Statement were prepared with the cooperation of the Advisory Centre on WTO Law.

            WT/DS267/RW

            Page 32


            other findings as will assist the DSB in making the recommendations or in giving the rulings provided for in the covered agreements."


          2. The Panel is cognizant of the need to give due consideration to the findings of the original panel in this dispute. Proceedings under Article 21.5 of the DSU do not occur in isolation but are part of a "continuum of events". The Panel conducts its analysis of the claims of Brazil "against the background of the original proceedings, and with full cognizance of the reasons provided by the original panel".85 86 The parties have expressed views on the issue of the weight that this Panel should accord to the legal and factual findings and reasoning of the panel in the original proceeding.87 The Panel does not consider that it is necessary or appropriate for it to make a general statement on this question in the abstract.


      2. Burden of proof


          1. The general rules regarding the allocation of the burden of proof in WTO dispute settlement require that a party claiming a violation of a provision of a WTO agreement by another Member assert and prove its claim.88 These rules also apply to proceedings under Article 21.5 of the DSU.89 Brazil


            image

            85 Appellate Body Report, Mexico – Corn Syrup (Article 21.5 – US), para. 121. In the same report, the Appellate Body stated that "a panel acting pursuant to Article 21.5 of the DSU would be expected to refer to the initial panel report, particularly in cases where the implementing measure is closely related to the original measure, and where the claims made in the proceedings under Article 21.5 closely resemble the claims made in the initial proceedings". Appellate Body Report, Mexico – Corn Syrup (Article 21.5 – US), para. 109.

            86 We note, in this regard, that in reviewing the findings of a panel in a proceeding under Article 21.5 involving a determination of injury by a domestic investigating authority, the Appellate Body has observed that:


            "doubts could arise about the objective nature of an Article 21.5 panel's assessment if, on a specific issue, that panel were to deviate from the reasoning in the original panel report in the absence of any change in the underlying evidence in the record and explanations given by the investigating authority in a redetermination. These concerns are not, however, based on the binding effect of the adopted findings of the original panel".


            Appellate Body Report, US – Softwood Lumber VI (Article 21.5 – Canada), para. 103. The Appellate Body made these observations in addressing an argument of Canada that the panel had "failed to make an objective examination of the matter before it because it failed to apply findings from the original panel report even though these findings had been adopted by the DSB and are binding upon the parties to the dispute." Appellate Body Report, US – Softwood Lumber VI (Article 21.5 – Canada), para. 101.

            87 E.g., Brazil's Response to Panel Question 2; US Response to Panel Question 2.

            88 Appellate Body Report, US – Wool Shirts and Blouses, p. 14. With respect to these general rules on allocation of the burden of proof, the Appellate Body recently observed:


            "... in WTO dispute settlement, as in most legal systems and international tribunals, the burden of proof rests on the party that asserts the affirmative of a claim or defence. A complaining party will satisfy its burden when it establishes a prima facie case by putting forward adequate legal arguments and evidence. The nature and scope of arguments and evidence required 'will necessarily vary from measure to measure, provision to provision, and case to case'. When a claim is brought against a WTO Member's legislation or regulation, a panel may, in some circumstances, consider that the text of the relevant legal instrument is sufficiently clear to establish the scope and meaning of the law. However, in other cases, a panel may consider that additional evidence is necessary to do so. Once the complaining party has established a prima facie case, it is then for the responding party to rebut it.


            A panel errs when it sustains a claim for which the complaining party has failed to make out a prima facie case. Nevertheless, a panel does not commit legal error merely by omitting to specify which party bears the burden of proof in respect of each claim or defence. Moreover, a panel is not obliged, in every instance, to make an explicit finding that a complaining party

            WT/DS267/RW

            Page 33


            as the complaining party must therefore make a prima facie case of violation of the relevant provisions of the WTO agreements invoked by Brazil, which the United States must refute. We also note, however, that it is generally for each party asserting a fact, whether complainant or respondent, to provide proof thereof.90 In this respect, therefore, it is also for the United States to provide evidence for the facts which it asserts. A prima facie case is one which, in the absence of effective refutation by the other party, requires a panel, as a matter of law, to rule in favour of the party presenting the prima facie case.


          2. The Appellate Body recently discussed the application of the general rules on the allocation of burden of proof in the context of Article 21.5 of the DSU as follows:


            "The text of Article 21.5 expressly links the 'measures taken to comply' with the recommendations and rulings of the DSB concerning the original measure. A panel's examination of a measure taken to comply cannot, therefore, be undertaken in abstraction from the findings by the original panel and the Appellate Body adopted by the DSB. Such findings identify the WTO-inconsistency with respect to the original measure, and a panel's examination of a measure taken to comply must be conducted with due cognizance of this background. Thus, the adopted findings from the original proceedings may well figure prominently in proceedings under Article 21.5, especially where the measure taken to comply is alleged to be inconsistent with WTO law in ways similar to the original measure. In our view, these considerations may influence the way in which the complaining party presents its case, and they may also be relevant to the manner in which an Article 21.5 panel determines whether that party has discharged its burden of proof and established a prima facie case."91


            image

            has met its burden to establish a prima facie case in respect of each element of a particular claim, or that the responding party has effectively rebutted a prima facie case. Thus, a panel is not required to make an explicit ruling that a complaining party has established a prima facie case of inconsistency before examining the responding party's defence and evidence. At the same time, the mere articulation by a panel of the correct rules as to the burden of proof is not sufficient if the panel does not, in fact, properly allocate that burden in the case before it."


            Appellate Body Report, Chile – Price Band System (Article 21.5 – Argentina), paras. 134-135.

            89 Appellate Body Report, Chile – Price Band System (Article 21.5 – Argentina), para. 136.

            New Zealand argues in its Third Party Submission that Article 7.8 of the SCM Agreement affects the distribution of burden of proof in this proceeding under Article 21.5 of the DSU. Specifically, New Zealand asserts that:


            "[s]ince an implementing Member has the option of either withdrawing the subsidising measure or removing the adverse effects, a Member that fails to withdraw the measure in question then should have the burden of showing that it has removed the adverse effects. Thus, in the present case proof by Brazil that the measures found by the original panel to cause adverse effects have not been withdrawn should establish a prima facie case that implementation has not occurred and the burden should then shift to the United States to prove that the adverse effects caused by those subsidizing measures have been removed."


            New Zealand's Third Party Submission, para. 5.06. The Panel finds no support in the text of Article 7.8 of the SCM Agreement and Article 21.5 of the DSU for the proposition that in a proceeding under Article 21.5 involving an alleged failure to comply with Article 7.8 of the SCM Agreement, a Member which has not withdrawn the subsidy has the burden of showing that it has removed the adverse effects. The Panel also notes Brazil's statement that, in the present case, "Brazil has the burden of demonstrating under Article 7.8 that the United States did not take appropriate steps to remove the adverse effects caused by the original basket of measures". Brazil's Response to Panel Question 26, para. 20.

            90 Appellate Body Report, US – Wool Shirts and Blouses, p. 14.

            91 Appellate Body Report, Chile – Price Band System (Article 21.5 – Argentina), para. 136.

            WT/DS267/RW

            Page 34


      3. Rules of treaty interpretation


          1. Article 3.2 of the DSU requires the Panel to apply customary rules of public international law on the interpretation of treaties. It is generally accepted that these rules can be found in Articles 31-33 of the Vienna Convention on the Law of Treaties of 1969 ("Vienna Convention"). Article 31(1) of the Vienna Convention provides:


            "A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose."


    2. SCOPE OF THIS PROCEEDING UNDER ARTICLE 21.5 OF THE DSU


      1. Introduction


          1. Article 21.5 of the DSU provides:


            "Where there is disagreement as to the existence or consistency with a covered agreement of measures taken to comply with the recommendations and rulings such dispute shall be decided through recourse to these dispute settlement procedures, including wherever possible resort to the original panel. The panel shall circulate its report within 90 days after the date of referral of the matter to it. When the panel considers that it cannot provide its report within this time frame, it shall inform the DSB in writing of the reasons for the delay together with an estimate of the period within which it will submit its report."


          2. The United States requests that the Panel make preliminary rulings that certain claims of Brazil are outside the scope of Article 21.5 of the DSU because they are not claims relating to the "existence or consistency with a covered agreement of measures taken to comply with the recommendations and rulings" of the DSB. The United States makes this request in respect of: (i) claims of Brazil relating to export credit guarantees regarding exports of pig meat and poultry meat;

            (ii) claims of Brazil relating to the marketing loan and counter-cyclical payment programmes in addition to payments made under those programmes; and (iii) claims of Brazil regarding non- compliance in a past period.92


          3. The United States also considers that payments made under the marketing loan and counter- cyclical payment programmes after 21 September 2005 are not properly within the scope of this proceeding.93


      2. Whether Brazil's claims relating to export credit guarantees for pig meat and poultry meat are outside the scope of this proceeding


        1. Main arguments of the parties


            1. The United States argues that since export credit guarantees for pig meat and poultry meat were not found to be WTO-inconsistent in the original proceeding they were not subject to DSB recommendations and rulings. Therefore, GSM 102 export credit guarantees in respect of exports of pig meat and poultry meat are not "measures taken to comply" with the recommendations and rulings


              image

              92 US First Written Submission, paras. 22-56; US Rebuttal Submission, paras. 16-71.

              93 The United States did not make the argument that payments made after 21 September 2005 are not within the scope of this proceeding as part of its request for preliminary rulings in its First Written Submission

              but presented this argument at a later stage of the proceeding. US Comments on Brazil's Response to Panel Question 15; US Response to Panel Question 47; US Comments on Brazil's Response to Panel's Second Set of Questions, paras. 8-14.

              WT/DS267/RW

              Page 35


              of the DSB and these measures and Brazil's claims relating to these measures are not within the scope of this proceeding.94


            2. The United States considers that Brazil confuses the proper standard as to when claims are within the scope of Article 21.5 proceedings and incorrectly assumes that the standard is one of whether there has been a "final resolution" of the issue in the original proceeding. Brazil misinterprets the Appellate Body report in EC – Bed Linen (Article 21.5 – India) and confuses two distinct issues: the issue of the scope of a compliance proceeding under Article 21.5 of the DSU, on the one hand, and the issue of when a claim against a specific measure or aspect of a measure can be considered to be "finally resolved" for purposes of WTO dispute settlement, on the other.95


            3. The United States submits that compliance proceedings under Article 21.5 of the DSU may address two types of matters. First, that measures taken to comply with DSB recommendations and rulings do not exist. Second, that measures taken to comply with DSB recommendations and rulings are not consistent with a covered agreement. If a measure is not subject to any DSB recommendations and rulings because the measure has not been found to be WTO-inconsistent, there is no logical basis for any claim that a Member has not implemented the DSB recommendations and rulings in respect of that measure. Moreover, unless the original measure itself is considered to be a measure taken to comply, there is no basis for claims to be made against the measure alleging that it is inconsistent with a covered agreement. Thus, measures are outside the scope of Article 21.5 of the DSU when they have not been found to be WTO-inconsistent and are not themselves measures taken to comply.96


            4. The United States submits that specific export credit guarantees are "measures" and those guarantees provided in support of exports of pig meat, poultry meat and other scheduled products other than rice have never been subject to any DSB recommendations and rulings and are not measures taken to comply with any DSB recommendations and rulings. Prior WTO dispute settlement reports do not support the proposition that the scope of this proceeding under Article 21.5 can be extended to include claims against measures that were never subject to any DSB recommendations and rulings and that are not measures taken to comply with any such recommendations and rulings.97


            5. Brazil requests the Panel to reject the preliminary objection of the United States because this objection is contrary to Appellate Body jurisprudence on the scope of claims properly brought in compliance proceedings.98 While the Appellate Body held in EC – Bed Linen (Article 21.5 – India) that an unappealed finding in a panel report adopted by the DSB must be treated as a final resolution to the dispute in respect of a particular claim and specific component of a measure, this limitation does not apply to Brazil's claims relating to export credit guarantees for pig meat and poultry meat because there has been no such final resolution with respect to these claims.99 Thus, Brazil argues that it may bring these claims in this proceeding because the original proceeding did not lead to a "final resolution" with respect to these claims.


            6. Brazil argues that its claims relating to export credit guarantees for pig meat and poultry meat relate to a measure taken to comply within the meaning of Article 21.5 of the DSU. While Brazil's


              image

              94 US First Written Submission, para. 24-30. The United States emphasizes that the relevant question is not whether, as argued by Brazil, Brazil may "re-assert" its claims against export credit guarantees for exports of pig meat and poultry meat, but whether Brazil may do so in the context of a proceeding under Article 21.5 of the DSU.

              95 US Rebuttal Submission, paras. 10-15; US Response to Panel Question 10.

              96 US Response to Panel Question 6.

              97 US Comments on Brazil's Response to Panel Question 6.

              98 Brazil's Submission Regarding US Requests for Preliminary Ruling, para. 7.

              99 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 9-12.

              WT/DS267/RW

              Page 36


              claims are product-specific, the measure at issue, the amended GSM 102 programme, applies in the same manner to all products covered by the programme. There is no factual basis for the US argument that the measures at issue are the export credit guarantees for pig meat and poultry meat.100


            7. Further, Brazil submits that the measure subject to its claims is a new measure and that a Member may present any claim in a proceeding under Article 21.5 with respect to a new measure. While panels have determined that a Member may not bring a new claim in a proceeding under Article 21.5 in respect of an aspect of a measure that remains unchanged from the original proceeding and that was not challenged in the original proceeding, that limitation does not apply to the present case because Brazil presents the same claim with respect to a new aspect of a new measure. Referring to the Appellate Body report in EC – Bed linen (Article 21.5 – India), Brazil asserts that it should not be held responsible for the absence of a final resolution in the original proceeding.101


        2. Main arguments of third parties


            1. Australia submits that Brazil's claim relating to GSM 102 export credit guarantees issued in respect of pig meat and poultry meat may properly be considered by this Panel. The situation in the present case is not that of a measure that was not challenged in the original proceedings or, if challenged, was addressed in those proceedings and not found to be WTO-inconsistent. The claim specifically made by Brazil in its request for establishment relates to the WTO-consistency of a new measure – guarantees issued under the revised GSM 102 programme – the revisions to which were implemented in response to the rulings and recommendations of the DSB. Brazil is entitled to request, and the Panel is required to consider, the new measure in its totality, including the measure's application.102


            2. The European Communities considers that in addressing the question whether GSM 102 export credit guarantees relating to pig meat and poultry meat are properly within the scope of this proceeding, the key issue before the Panel is the nexus or the degree of interrelatedness or interdependence between different elements of the measure. Thus, the Panel should examine the original measure at issue and the measure taken to comply and, with particular reference to the elements of the measure that the United States argues are outside the terms of reference of the Panel, enquire into the extent to which these elements are interrelated or interdependent with the measures or elements of measures that the United States accepts as being within the Panel's terms of reference. At first sight, the question of whether US export credit guarantees relating to pig meat and poultry meat circumvent US commitments under the Agreement on Agriculture and are inconsistent with the SCM Agreement may not be necessarily interdependent with and not even necessarily interrelated with the same question regarding upland cotton. However, there are other factors that merit careful consideration, including the fact that the SCM Agreement makes clear that a "programme" can be a measure, that the United States has admitted that the measure at issue is a "programme" within the meaning of the SCM Agreement, that the original panel consistently referred to the GSM 102 "programme", and that the United States has elected to amend the GSM 102 programme generally, including with respect to pig meat and poultry meat.103


            3. Japan submits that the claim of Brazil regarding GSM 102 export credit guarantees for pig meat and poultry meat is within the scope of this proceeding. The inability of the Appellate Body to complete the analysis with respect to a panel finding reversed by the Appellate Body is similar to the exercise of judicial economy in that in both cases there is no "final resolution" with respect to the


              image

              100 Brazil's Response to Panel Question 6, para. 21; Brazil's Comments on US Response to Panel Question 6.

              101 Brazil's Response to Panel Question 6, paras. 42-47.

              102 Australia's Third Party Submission, paras. 8-11.

              103 EC Third Party Oral Statement, para. 5-13.

              WT/DS267/RW

              Page 37


              claim in question. The recent panel report in US – Oil Country Tubular Goods Sunset Reviews (Article 21.5 – Argentina) makes clear that this Panel has the authority to identify and assess the new measure taken by the United States to comply with the DSB recommendations and rulings, including the GSM 102 programme as applied to pig meat and poultry meat.104


        3. Evaluation by the Panel


          1. The United States request the Panel to make a preliminary ruling that Brazil's claims relating to GSM 102 guarantees in respect of exports of pig meat and poultry meat are outside the scope of this proceeding under Article 21.5 of the DSU.105


          2. The Panel recalls that the original panel found that in respect of twelve scheduled products, including pig meat and poultry meat, it had not been established that the United States had applied export credit guarantees constituting export subsidies within the meaning of Article 10.1 of the Agreement on Agriculture in a manner resulting in circumvention of export subsidy commitments of the United States.106 The Appellate Body reversed this finding of the original panel, but considered that there were insufficient undisputed facts in the record to enable it to complete the legal analysis to determine whether the United States had applied export credit guarantees with respect to pig meat and poultry meat in a manner resulting in circumvention of export subsidy commitments of the United States. 107


          3. Thus, the issue before the Panel is whether in this proceeding under Article 21.5 of the DSU Brazil may present claims relating to export credit guarantees for pig meat and poultry meat considering that: (i) Brazil presented such claims in the original proceeding; (ii) the original panel rejected these claims; (iii) the Appellate Body reversed the original panel's rejection of these claims, but found itself unable to complete the analysis; such that (iv) the original proceeding did not result in a finding of WTO-inconsistency in respect of export credit guarantees for pig meat and poultry meat.


          4. The phrase "measures taken to comply with the rulings or recommendations" in Article 21.5 of the DSU limits the scope of Article 21.5 proceedings to a particular category of measures; "[i]f a claim challenges a measure which is not a 'measure taken to comply', that claim cannot properly be raised in Article 21.5 proceedings".108 Thus, "[i]n an Article 21.5 proceeding, the measure or measures taken to comply go to the heart of [a] panel's mandate".109 110


          5. The Panel considers that the fact that export credit guarantees for pig meat and poultry meat were not found to be WTO-inconsistent in the original proceeding and were therefore not subject to DSB recommendations and rulings does not necessarily mean that the claims of Brazil with respect to export credit guarantees for pig meat and poultry meat should be found to be outside the scope of


            image

            104 Japan's Third Party Oral Statement, paras. 3-4.

            105 The United States does not argue that these claims are not covered by Brazil's request for the establishment of a panel (WT/DS267/30). The Panel notes that paragraph 27 of this request refers to "upland

            cotton and other unscheduled products; and rice, pig meat and poultry meat".

            106 Panel Report, US – Upland Cotton, para. 7.881.

            107 Appellate Body Report, US – Upland Cotton, paras. 692-694 and 763(f)(i).

            108 Appellate Body Report, EC – Bed Linen (Article 21.5 – India), para. 78. See also Appellate Body

            Report, Canada – Aircraft (Article 21.5 – Brazil), para. 36; Appellate Body Report, US – Softwood Lumber IV (Article 21.5 – Canada), para. 65.

            109 Panel Report, US – Countervailing Measures on Certain EC Products (Article 21.5 – EC);

            para. 7.35.

            110 Whether a measure identified in a panel request under Article 21.5 is a "measure taken to comply" within the meaning of Article 21.5 is a matter to be decided by the Article 21.5 panel, not by the parties to the

            dispute. Appellate Body Report, EC – Bed Linen (Article 21.5 – India), para. 78; Panel Report, US – Countervailing Measures on Certain EC Products (Article 21.5 – EC), para. 7.7.

            WT/DS267/RW

            Page 38


            Article 21.5 of the DSU as not relating to "measures taken to comply". The Panel finds support for this view in the reasoning of several recent Appellate Body and panel reports in proceedings under Article 21.5 of the DSU.


          6. The Panel notes, in particular, the observations of the Appellate Body in US – Softwood Lumber (Article 21.5 – Canada):


            "Some measures with a particularly close relationship to the declared 'measure taken to comply', and to the recommendations and rulings of the DSB, may also be susceptible to review by a panel acting under Article 21.5. Determining whether this is the case requires a panel to scrutinize these relationships, which may, depending on the particular facts, call for an examination of the timing, nature, and effects of the various measures. This also requires an Article 21.5 panel to examine the factual and legal background against which a declared 'measure taken to comply' is adopted. Only then is a panel in a position to take a view as to whether there are sufficiently close links for it to characterize such an other measure as one 'taken to comply' and, consequently, to assess its consistency with the covered agreements in an Article 21.5 proceeding."111


          7. The United States amended its GSM 102 programme in order to comply with the DSB recommendations and rulings in the original proceeding. The amendments made to the programme apply to exports of all products covered under the programme, including pig meat and poultry meat. The Panel does not agree with Brazil's argument that the measure that is the subject of its claims is the amended GSM 102 programme itself and that export credit guarantees with respect to pig meat and poultry meat are not themselves "measures". In the Panel's view, that the amended programme applies to all products in the same manner does not alter the fact that the application of the programme to an individual product constitutes a "measure". Nevertheless, while accepting the argument of the United States that GSM 102 export credit guarantees in respect of exports of pig meat and poultry meat are measures, we consider that these measures have a "particularly close relationship to the declared measure taken to comply and to the recommendations and rulings of the DSB". As noted above, the amended programme applies in the same manner to all products covered by the programme. Furthermore, the Panel notes that most of the legal and factual issues raised by Brazil's claims relating to export credit guarantees pertain to the amended GSM programme in general without distinction between individual products to which the programme applies. In addition, the United States has strongly argued that the Panel should examine Brazil's claim solely under the standard contained in item (j) of the Illustrative List of Export Subsidies, a standard that requires the Panel to analyze the GSM programme as a whole rather than in relation to particular products.112


          8. The Panel also notes that Brazil's claims with respect to export credit guarantees for pig meat and poultry meat are not claims that were rejected in the original proceeding and that relate to an unchanged measure. Rather with respect to a new measure – the application of the amended GSM 102 programme to exports of pig meat and poultry meat – Brazil makes a claim in respect of which the original proceeding did not result in a finding as to whether or not the United States was acting


            image

            111 Appellate Body Report, US – Softwood Lumber IV (Article 21.5 – Canada), para. 77.

            112 We note in this regard that in interpreting the scope of the phrase "measures taken to comply" in

            Article 21.5, the Appellate Body has taken into consideration the fact that the purpose of Article 21.5 of the DSU is to promote prompt compliance with DSB recommendations and rulings and the consistency of measures taken to comply with the covered agreements by making it unnecessary for complainants to initiate new proceedings. (Appellate Body Report, US – Oil Country Tubular Goods Sunset Reviews (Article 21.5 – Argentina), para. 151). The logic of the argument of the United States is that in order to obtain a finding on whether export credit guarantees for pig meat and poultry meat are WTO-inconsistent, Brazil must commence a new panel proceeding in which Brazil will need to make the same legal and factual arguments as in this case.

            WT/DS267/RW

            Page 39


            inconsistently with its WTO obligations.113 No previous panel or Appellate Body proceeding has addressed the precise question before us in this case of whether a Member may assert a claim in an Article 21.5 proceeding in respect of which the Appellate Body found itself unable to complete the analysis in the original proceeding. We find it significant, however, that the Appellate Body recently held that a claim relating to an aspect of a measure on which the panel in the original proceeding had exercised judicial economy was properly within the scope of Article 21.5 of the DSU.114 In our view, this demonstrates that a claim relating to a measure that has a sufficiently close nexus with the measure taken to comply or with the DSB recommendations and rulings in the original proceeding can be within the scope of Article 21.5 even where that measure itself has not been the subject of DSB recommendations and rulings in the original proceeding.


          9. In light of the foregoing considerations, the Panel finds that the claims of Brazil relating to export credit guarantees for exports of pig meat and poultry meat are within the scope of this proceeding under Article 21.5 of the DSU.


      3. Whether Brazil's claims with respect to the marketing loan and counter-cyclical payment programmes are outside the scope of this proceeding


          1. In support for its request for a preliminary ruling that Brazil's claims regarding the marketing loan and counter-cyclical payment programmes are outside the scope of this proceeding, the United States argues, first, that these programmes were not subject to recommendations and rulings of the DSB in the original proceeding and, second, that these programmes are not "measures taken to comply" within the meaning of Article 21.5 of the DSU because they have not been changed in response to DSB recommendations and rulings or otherwise.


            1. Whether Brazil's claims with respect to the marketing loan payment and counter-cyclical programmes are outside the scope of this proceeding because these programmes were not subject to recommendations and rulings of the DSB in the original proceeding


              (i) Main arguments of the parties


          2. The United States requests the Panel to make a preliminary ruling that Brazil's claims of (threat of) serious prejudice in respect of the marketing loan and counter-cyclical payment programmes are outside the scope of this proceeding because these programmes are not measures to which the recommendations and rulings of the DSB in the original proceeding were addressed.115


          3. The United States submits that Brazil made a clear distinction in the original proceeding between its actionable subsidy claims against programmes per se and actionable subsidy claims against payments under specific programmes, and that the original panel only made a finding of


            image

            113 The facts before us are in this respect significantly different from the facts before the panel and the Appellate Body in EC – Bed Linen (Article 21.5 – India). In that case, the original panel had rejected a claim of India that a certain aspect of a measure taken by the European Communities was WTO-inconsistent. However, in the Article 21.5 proceeding India raised the same claim with respect to the same unchanged aspect of the original measure.

            114 Appellate Body Report, US – Oil Country Tubular Goods Sunset Reviews (Article 21.5 – Argentina).

            115 The United States does not argue that these claims are not covered by Brazil's request for the

            establishment of a panel (WT/DS267/30), but argues that they are outside the Panel's terms of reference by virtue of the limited terms of reference contained in Article 21.5 of the DSU. The Panel notes that in paragraphs 18 and 20 of this request Brazil refers to the "US marketing loan, and counter-cyclical payment programs under the FSRI Act of 2002, as amended, as well as payments made under those programs" and to "the US marketing loan and counter-cyclical payment programs under the FSRI Act of 2002, as amended, as well as payments mandated to be made thereunder".

            WT/DS267/RW

            Page 40


            WTO-inconsistency with regard to the claim of "present" serious prejudice against certain payments made in MY 1999-2002. According to the United States, the original panel specifically rejected or declined to address Brazil's other serious prejudice claims.116 117


          4. The United States submits that Brazil's claims are outside the scope of this proceeding, regardless of whether they are against the marketing loan and counter-cyclical payment programmes alone or against the programmes in addition to payments under the programmes. The original panel did not make a finding of WTO-inconsistency against the Step 2, marketing loan and counter-cyclical payment programmes or against the programmes in addition to all payments thereunder. In this regard, the United States submits that there is no legal or factual basis for Brazil's assertion that "in the circumstances of this dispute a subsidy payment cannot be divorced from a subsidy programme". As a factual matter, the marketing loan and counter-cyclical payment programmes, i.e. the statutory and regulatory provisions authorizing marketing loan and counter-cyclical payments are distinct from the payments that they authorize. Either the statutory/regulatory provisions themselves or the measures resulting from their application, i.e. particular payments, may be challenged in dispute settlement, but each is a distinct measure. The Appellate Body has clarified that "as such" and "as applied" claims are distinct, and there is no support in any Appellate Body report for Brazil's argument that the real world difference between a statutory/regulatory provision and the measures that it authorizes must be disregarded. With respect to Brazil's argument that because an analysis under Articles 5 and 6 of the SCM Agreement focuses on the effects of the subsidy, no distinction can be made between a programme and the payments made under the programme, the United States submits that Brazil does not explain why the focus on effects in Part III of the SCM Agreement would require that the statutory and regulatory provisions authorizing payments be conflated with the payments themselves. This argument of Brazil is also in contradiction with the fact that in the original proceeding Brazil itself made separate claims against the payments and the statutory and regulatory provisions.118


          5. The United States asserts that Brazil did not challenge, the original panel did not find, and the DSB did not make any recommendation or ruling against the marketing loan and counter-cyclical payment programmes or against the programmes in addition to the payments under the programmes. The measures that were the subject of Brazil's claims, and the panel's finding, of present serious prejudice were subsidies provided during MY 1999-2002. The absence of any express reference to any legal/regulatory provisions in paragraphs 8.3(d) and 8.1(g)(i) of the original panel report confirms that the actionable subsidy-related conclusion and recommendation laid out therein do not apply to any programme as such, either alone or as part of any invented amalgam of programmes and payments. As illustrated by Section VII:D of the panel report, where the original panel wanted to refer to programmes, as such, and payments together, it expressly stated it was doing so.119 The United States considers that in arguing that the original panel did make findings of WTO- inconsistency against the marketing loan and counter-cyclical payment programmes as such, Brazil relies on selectively cited language from the panel report taken out of its proper context.120


          6. Brazil argues that the preliminary objection of the United States rests on a mischaracterization of the claims of Brazil in this proceeding and of the findings of the original panel.



            image

            116US First Written Submission, paras. 31-33.

            117 The United States points out that the statement in paragraph 7.1501 of the original panel report that "the United States is obliged to take action concerning its present statutory and regulatory framework as a result of our 'present' serious prejudice finding" is not a recommendation. US First Written Submission, para. 39.

            118 US Rebuttal Submission, paras. 20-28.

            119 US Rebuttal Submission, paras. 30-43.

            120 US Rebuttal Submission, paras. 44-52. See also US Comments on Brazil's Responses to Panel Questions 11 and 15.

            WT/DS267/RW

            Page 41


            Brazil further argues that under established jurisprudence, Brazil's claims fall squarely within the scope of this proceeding.


          7. Brazil submits that in this proceeding it is not asserting a claim against US marketing loan and counter-cyclical payment programmes "as such". The measures that constitute the subsidies at issue in this proceeding are the statutory and regulatory provisions of the FSRI Act of 2002 relating to upland cotton and the mandatory and price-contingent payments that have been made and will continue to be made during the lifetime of the FSRI Act of 2002, and Brazil makes no distinction between serious prejudice caused by the provisions of the FSRI Act and serious prejudice caused by the payments mandated by those provisions. It follows from the particular nature of the obligations in Articles 5 and 6 of the SCM Agreement, which focus on the effects of subsidies, that a payment of a subsidy cannot be divorced from the subsidy programme pursuant to which the payment is made.121 Brazil argues that the idea that programmes and payments together constitute the subsidy is consistent with the approach of the original panel. Brazil refers to statements in the panel report that indicate that the panel considered that its findings in Section VII:G and its conclusion in paragraph 8.1(g)(i) pertained to both the subsidy programmes under the FSRI Act and the payments mandated under those programmes. Brazil also submits that a review of the submissions of the parties in the original proceeding shows that they both referred to the effects of the programmes as well as the effects of payments under the programmes. Since the original panel necessarily considered all forms of the subsidy, i.e. programmes and specific payments made under the programmes, and since the United States did not appeal the original panel's finding that it was obliged to take action with respect to the statutory and regulatory framework under which subsidies were provided, the United States is precluded from challenging this Panel's authority to review Brazil's claim regarding the marketing loan and counter-cyclical payment programmes.122


          8. Brazil argues that while recent jurisprudence has clarified that the original proceeding imposes limits on the scope of the claims that may be presented in an Article 21.5 proceeding, those limits do not apply to Brazil's claims regarding the marketing loan and counter-cyclical payment programmes.123 Brazil's claims against the adverse effects flowing from the programmes and the payments mandated under the programmes were identified in Brazil's request for establishment and involve measures taken to comply within the meaning of Article 21.5 of the DSU.124 Brazil further submits that the United States is factually incorrect in suggesting that the original panel rejected the serious prejudice claims of Brazil in respect of the marketing loan and counter-cyclical programmes.125 Brazil also submits that the original panel exercised judicial economy with respect to Brazil's claims of threat of serious prejudice and with respect to Brazil's per se claims because the panel considered that its finding regarding "present" serious prejudice would oblige the United States to take action concerning its statutory and regulatory framework. Therefore, the exercise of judicial


            image

            121 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 31-34.

            122 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 36-40.

            123 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 41-47. Brazil points out

            in this connection that the fact that a panel in an original proceeding declines to address certain claims does not prevent a party from raising the same claims in an Article 21.5 proceeding.

            124 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 49-50.

            125 In this regard, Brazil refers in particular to the statement in paragraph 7.1501 of the panel report. Brazil also refers to para. 8.1(g)(i) of the report (particularly the phrase "mandatory price-contingent

            United States subsidy measures"), the definition of the measures at issue in para. 7.1107 of the report, statements in para. 7.1499 of the report, findings of the Panel regarding the "statutory and regulatory framework" providing for the subsidies at issue (for example, the Panel's examination of the "structure, design and operation" of the programmes) and statements reflecting that the Panel assessed the effects caused by statutory and regulatory instruments that provide for the payment of the subsidies along with the magnitude of the payments. Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 52-62.

            WT/DS267/RW

            Page 42


            economy with respect to these claims confirms that the Panel's finding with respect to "present" serious prejudice covered the subsidy programmes as well the payments made thereunder.126


          9. Brazil summarizes as follows the factors that support its view that the original panel's finding of "present" serious prejudice included the subsidy programmes as well as the payments mandated by the subsidy programmes: (i) the identification by the panel of the measures at issue in the "present" serious prejudice section; (ii) the panel's reasoning regarding the "effect" of those measures; (iii) the panel's reasoning in explaining its decision not to examine Brazil's claims of "as such" violations, and

            (iv) the panel's description of the implementation action that the United States was obliged to take.127


            1. Main arguments of third parties


          10. Argentina submits that the marketing loan and counter-cyclical payment programmes were subject to the findings and recommendations of the original panel. The panel specifically stated that the measures at issue included "legislative and regulatory provisions currently providing for the payment of measures". Moreover, paragraph 8.1(g)(i) of the panel report refers to "mandatory price- contingent United States subsidy measures". The panel stated that its findings pertained to measures in force and subsidies paid in MY 2002 and that, as a result of its finding of "present" serious prejudice, the United States was "obliged to take action concerning its present statutory and regulatory framework".128


          11. Australia considers that the finding of the original panel in paragraph 8.1(g)(i) of its report relates to certain payments rather than to the subsidy programmes as such. However, Brazil is not seeking to challenge the marketing loan and counter-cyclical payment programmes per se in this proceeding. The focus of Brazil's case under Articles 5 and 6 of the SCM Agreement is the continuing effects of the unamended programmes, i.e. payments made under those programmes, rather than the programmes as such. In this context, it is impossible to analyze the price-contingent subsidy measures still in existence in isolation from the programmes that govern those measures. Insofar as a consideration of those programmes is necessary to determine whether the price-contingent subsidy measures cause significant price suppression, the programmes may properly be considered by this Panel.129


          12. China argues that the text of the report of the original panel makes clear that the serious prejudice finding of the panel applies to the marketing loan and counter-cyclical payment programmes. China mentions in this regard references made by the panel to the "legislative and regulatory provisions currently providing for the payment of" certain subsidies, including marketing loan and counter-cyclical subsidies, to "the legislation pursuant to which the granting authority operates" and to the "structure, design and operation of these measures". In addition, the panel specifically stated that, as a result of the panel's finding of serious prejudice, the United States was "obliged to take action concerning its present statutory and regulatory framework".130


          13. Chad argues that the original panel described the measures at issue as including "the legislative and regulatory provisions currently providing for the payment of measures". Moreover, the panel explained that, as a consequence of its finding of serious prejudice, the United States was "obliged to take action concerning its present statutory and regulatory framework", and found it unnecessary to address Brazil's claim of threat of serious prejudice. Thus, the original panel clearly considered that its finding of serious prejudice included not simply the payments themselves but also



            image

            126 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 63-71.

            127 Brazil's Comments on US Response to Panel Question 45.

            128 Argentina's Third Party Oral Statement, paras. 5-12.

            129 Australia's Third Party Submission, paras. 14-15.

            130 China's Third Party Submission, paras. 12-13.

            WT/DS267/RW

            Page 43


            the statutory and regulatory measures under which the payments were made. Chad considers that by describing the measures at issue in paragraph 8.1(g)(i) of its report as "mandatory" payments, the panel necessarily included the statutory and regulatory framework that made those payments mandatory. Chad also argues that the United States appears to confuse the issue of whether a statutory provision is challenged as such with the altogether separate issue of whether a statutory provision mandating certain payments can be found to cause serious prejudice. Brazil's claims in the original proceeding were "as applied" claims regarding measures that included legislative and regulatory provisions. The distinction sometimes made in WTO proceedings between "as such" and "as applied" claims does not mean that statutory or regulatory provisions cannot be challenged on an "as applied" basis or that "as applied" claims are necessarily limited to the payments themselves to the exclusion of measures mandating those payments.131


          14. The European Communities agrees with the United States that there is a meaningful distinction between an "as such" claim and an "as applied" claim and between a subsidy "programme" within the meaning of the SCM Agreement and the provision of subsidies in individual cases under the programme. A claim under Part III of the SCM Agreement can pertain to either the subsidy programme, the provision of subsidy in individual instances or both. These measures should normally not be conflated. However, the European Communities considers that where an entitlement is conditional upon the fulfilment of essentially factual requirements and there is no discretion in the hands of a granting authority, it may be justifiable to treat the subsidy programme and payments made under the programme as indivisible for purposes of Article 21.5 proceedings. The European Communities agrees with Brazil that the fact that the United States implemented the DSB recommendations and rulings by withdrawing the Step 2 programme supports the view that there was at least a sufficient nexus between the Step 2 payments and the Step 2 programme. Moreover, that the original panel assessed a bundle of payments under the three different programmes implies that it assumed that there was a sufficient nexus between the payments under the three different programmes. If there is a sufficient nexus between payments and programme for one programme and if there is a sufficient nexus between all payments, this in turn suggests that there is a sufficient nexus between the three programmes, such that the whole bundle can and should be treated as indivisible for the purposes of this compliance panel.132


          15. India submits that the original panel clearly described the measures at issue as comprising the "legislative and regulatory provisions currently providing for the payment of measures", including the marketing loan and counter-cyclical payments, and that the panel clearly considered that its findings regarding "present" serious prejudice applied not only to payments but also to the statutory and regulatory framework of the FSRI Act of 2002 and its implementing obligations. Therefore, the obligations of the United States in the implementation period were not limited to payments made in a period in the past.133


          16. New Zealand submits that, read properly in the context of the panel's decision as a whole, the finding of the original panel in paragraph 8.1(g)(i) of its report is not limited to the specific payments made in MY 1999-2002 but applies to the measures under which those payments are authorized and hence to the marketing loan and counter-cyclical payment programmes themselves. The panel made a finding with respect to "the mandatory price-contingent United States subsidy measures". Payments are not mandatory subsidy measures; they are simply the means by which the subsidy measures are implemented. Moreover, it is clear from the report that the panel's analysis of the subsidy measures that caused price suppression was of the programmes under the FSRI Act of 2002 under which payments were made. Finally, the statement of the panel in paragraph 7.1501 of its report that "the United States is obliged to take action concerning its present statutory and regulatory framework as a


            image

            131 Chad's Third Party Submission, paras. 22-31. 132 EC Third Party Oral Statement, paras. 15-21. 133 India's Third Party Oral Statement, para. 6.

            WT/DS267/RW

            Page 44


            result of our 'present' serious prejudice finding" gives meaning to the finding in paragraph 8.1(g)(i) relating to the "United States subsidy measures" that constituted significant price suppression.134


            1. Evaluation by the Panel


          17. The Panel interprets the scope of the original panel's finding of "present" serious prejudice based on the terms used by the panel to express that finding. The original panel's finding of "present" serious prejudice is contained in paragraphs 7.1416 and 8.1(g)(i) of the panel report.135


          18. Paragraph 7.1416 of the report reads as follows:


            "In conclusion, in light of all of these considerations, we find that the effect of the mandatory, price contingent United States subsidies at issue – that is, marketing loan programme payments, user marketing (Step 2) payments and MLA payments and CCP payments – is significant price suppression in the same world market for upland cotton in the period MY 1999-2002 within the meaning of Articles 6.3(c) and 5(c) of the SCM Agreement."


          19. Paragraph 8.1(g)(i) of the report reads as follows:


            "the effect of the mandatory price-contingent United States subsidy measures - marketing loan programme payments, user marketing (Step 2) payments, MLA payments and CCP payments - is significant price suppression in the same world market within the meaning of Article 6.3(c) of the SCM Agreement constituting serious prejudice to the interests of Brazil within the meaning of Article 5(c) of the SCM Agreement;"


          20. In the Panel's view, there is no basis in the text of these paragraphs to interpret the measures subject to the original panel's finding as including anything other than "subsidies" and "subsidy measures" defined in terms of the four categories of "payments" identified by the panel. The term "payments" in its ordinary meaning does not include the legal provisions or programmes pursuant to which the payments are made.136 The explicit statement by the panel that the subsidies and subsidy


            image

            134 New Zealand's Third Party Submission, paras. 4.03-4.07.

            135 Paragraph 7.1416 is the concluding paragraph in section VII:G.3 of the report ("Claims under Articles 6.3(c) and 5(c) of the SCM Agreement"). Paragraph 8.1(g)(i) appears in the "Conclusions and

            recommendations" section of the report. We note that the Appellate Body Report in the original proceeding refers to the "finding [of serious prejudice] in paragraphs 7.1416 and 8.1(g)(i) of the Panel Report." Appellate Body Report, US – Upland Cotton, paras. 249(c)(i), 395, 496 and 763(c)(i). Thus, for example, in paragraph 763(c)(i), the Appellate Body upheld "the Panel's finding, in paragraphs 7.1416 and 8.1(g)(i) of the Panel Report, that the effect of the marketing loan program payments, Step 2 payments, market loss assistance payments, and counter-cyclical payments (the 'price-contingent') subsidies is significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement".

            136 A careful reading of the original panel report as a whole makes it clear that the panel generally

            equated "subsidies" with "payments" and that the term "payments" in this context does not include legal and regulatory provisions. For example, one of the issues addressed by the panel in response to preliminary objections raised by the United States was whether production flexibility contract payments and market loss assistance payments made under legislation that no longer existed were within the panel's terms of reference. In its discussion of this issue the panel observed inter alia:


            "The Panel notes that Brazil pursues claims only in respect of the subsidies and domestic support provided under the expired programmes and authorizing legislation, in other words, the payments themselves. Brazil does not seek any relief in respect of the PFC and MLA programmes or authorizing legislation 'as such'. Therefore, the Panel only considers whether the payments are within its terms of reference." (emphasis added)

            WT/DS267/RW

            Page 45


            measures at issue are certain kinds of payments, and the absence of any reference to legal provisions or subsidy programmes providing for the payments, therefore necessarily mean that the subsidies and subsidy measures do not include such legal provisions or subsidy programmes.137 We also note, in this regard, that Section III of the panel report ("Parties' Requests for Findings and Recommendations") identifies the measures subject to Brazil's claims of "present" serious prejudice as "subsidies provided during MY 1999–2002" without mentioning the legal provisions or programmes providing for these subsidies. The panel's finding of "present" serious prejudice in paragraphs 7.1416 and 8.1(g)(i) logically cannot have been broader in scope than the claims of Brazil in respect of which the panel made this finding. Finally, the panel specified the temporal scope of its finding when it stated in paragraph 7.1416 that the effect of the mandatory, price-contingent subsidies at issue was significant price suppression in the period MY 1999-2002.


          21. In support of its argument that the original panel's finding of "present" serious prejudice applied to payments and programmes, Brazil emphasizes the original panel's explanation as to why the panel considered it not necessary or appropriate to address Brazil's claims of threat of serious prejudice under Articles 5(c) and 6.3(c) of the SCM Agreement.138

          22. We find Brazil's emphasis on these paragraphs in the panel report139 to interpret the scope of the panel's finding of "present" serious prejudice misplaced. First, the finding of "present" serious prejudice is contained in paragraphs 7.1416 and 8.1(g)(i), not in paragraphs 7.1499-7.1503. As discussed above, consistent with Brazil's claims, paragraphs 7.1416 and 8.1(g)(i) of the original panel report contain a finding that the effect of certain subsidies and subsidy measures, defined in terms of certain kinds of payments, is significant price suppression constituting serious prejudice to the interests of Brazil in the period MY 1999-2002. These paragraphs do not refer in any way to programmes or legislation. We do not see how the observations made by the panel in paragraphs 7.1499-7.1503 could be used to support an interpretation of the panel's finding of "present" serious prejudice that is inconsistent with the natural meaning of the terms used by the panel in paragraphs 7.1416 and 8.1(g)(i). Second, there is no explicit statement in paragraphs 7.1499-7.1503 that the panel had found that serious prejudice to the interest of Brazil was the effect of subsidy programmes or legal provisions establishing subsidy programmes as well as payments made under those programmes. The panel stated in paragraphs 7.1499 and 7.1501 that "[o]ur finding of 'present' serious prejudice thus pertain[s] also to measures in force and subsidies paid in MY 2002..." and that "the Panel's 'present' serious prejudice findings include findings of inconsistency that deal with the FSRI Act of 2002 and subsidies granted thereunder". The panel did not state that it had found that, in addition to subsidies paid in MY 1999-2002, the FSRI Act itself was inconsistent with Articles 5 and 6 of the SCM Agreement.


            image

            Panel Report, US – Upland Cotton, para. 7.108. Thus, the panel understood "the subsidies and domestic support..." to refer to "the payments themselves", as distinguished from the "... programmes or authorizing legislation 'as such'". Where the panel used the concept of payments in a broader sense to encompass regulatory provisions, it made an explicit statement to that effect. See Panel Report, US – Upland Cotton, para. 7.337, footnote 466: "In this Section of our report, the Panel will consider the current programmes 'as applied' and 'as such' together. Therefore, references to marketing loan programme, user marketing (step 2), direct, counter-cyclical and crop insurance 'payments' include the legislative and regulatory provisions authorizing those payments, unless otherwise indicated." No such statement can be found in the section of the original panel report that addresses Brazil's claims of serious prejudice.

            137 By using the words "mandatory" and "price contingent" to qualify "United States subsidy measures"

            the panel characterized the subsidies at issue in terms of certain conditions set out in the legal provisions governing those subsidies, but this does not alter the fact that the panel only made a finding of "present" serious prejudice with respect to subsidies (defined as payments) and not with respect to both subsidies and the legal provisions governing those subsidies.

            138 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 36-39, 52, 59-60, 64-71;

            Brazil's Responses to Panel Questions 11 and 13, paras. 99-102, 105-106, 119, 126-128.

            139 Panel Report, US – Upland Cotton, paras. 7.1499-7.1503.

            WT/DS267/RW

            Page 46


          23. Similarly, the statement in paragraph 7.1501 of the report of the original panel that "the United States is obliged to take action concerning its present statutory and regulatory framework as a result of our 'present' serious prejudice finding" is an expression of a view about the steps to be taken by the United States under Article 7.8 of the SCM Agreement and does not mean that the Panel had found that the "present statutory and regulatory framework" in addition to subsidies paid under that framework had itself caused serious prejudice.140


          24. While the legal provisions or subsidy programmes pursuant to which the subsidies were provided were not themselves part of the measures subject to the original panel's finding in paragraphs 7.1416 and 8.1(g)(i), it is clear that in making this finding the original panel took into consideration these legal provisions or subsidy programmes. The panel undertook "an analysis focusing on the existence and nature of the subsidies in question by examining their structure, design and operation with a view to discerning their effects".141 The panel considered that "the nature of the United States subsidies at issue – in terms of their structure, design and operation – is relevant in assessing whether or not they have price suppressing effects".142 In finding that there was "significant" price suppression, the panel relied inter alia upon the "the market-price contingent, counter-cyclical, nature of the marketing loan programme [and] the user marketing (Step 2) programme".143 The panel observed that "the structure, design and operation of the marketing loan programme has enhanced production and trade-distorting effects".144 145 The panel found that "the structure, design and operation of these three measures constitutes evidence supporting a causal link with the significant price suppression we have found to exist".146


          25. In our view, it is self-evident that in order to properly analyze the effect of a subsidy for purposes of an analysis under Articles 5 and 6 of the SCM Agreement, the subsidy must be examined in light of the conditions and criteria contained in the legal and regulatory framework that governs the granting of that subsidy. Therefore, in this sense we concur with the argument of Brazil that it is difficult to divorce a payment of a subsidy from the programme or legal provisions pursuant to which the subsidy is provided. As we understand the position of the United States, the United States does not contest that in order to evaluate whether a subsidy is inconsistent with a Member's obligations it may be necessary to analyze that subsidy in the context of the legal and regulatory framework that defines the conditions under which the subsidy is granted.147 It would appear that the preliminary objection of the United States reflects a concern that Brazil challenges the programmes as such, as



            image

            140 Contrary to what Brazil argues, this statement cannot be treated as part of the panel's "recommendation". As the panel recalled in paragraph 8.3(d), the obligations of the United States arising from the finding of serious prejudice are defined in Article 7.8 of the SCM Agreement. There is no legal basis in the text of the SCM Agreement or in the DSU to suggest that where a panel has made a finding of serious prejudice under Articles 5 and 6 of the SCM Agreement it has the authority to make a "recommendation" other than in terms of Article 7.8 of the SCM Agreement.

            141 Panel Report, US – Upland Cotton, para. 7.1194. 142 Panel Report, US – Upland Cotton, para. 7.1289. 143 Panel Report, US – Upland Cotton, para. 7.1332. 144 Panel Report, US – Upland Cotton, para. 7.1295.

            145 Similarly, in discussing the influence of counter-cyclical payments on production decisions, the

            Panel observed inter alia: "The FSRI Act of 2002 provides, on an on-going basis, for payments under certain market conditions. This contrasts with the ad hoc nature of the MLA payments. Under the FSRI Act of 2002, United States producers are currently aware that such payments are available under certain conditions and that there is no need to be uncertain as to whether [the] United States Congress may or may not decide to authorize such payments in a given year. This certain claim on future payments under certain conditions increases United States producers' economic stability." Panel Report, US – Upland Cotton, para. 7.1302, footnote 1410.

            146 Panel Report, US – Upland Cotton, para. 7.1349.

            147 US Comments on Brazil's Response to Panel Question 11, para. 50.

            WT/DS267/RW

            Page 47


            that concept is commonly understood in WTO dispute settlement practice.148 The references in Brazil's First Written Submission to "programme" and "programmes" and certain statements made by Brazil during this proceeding could indeed be interpreted to suggest that Brazil is challenging programmes as such.149 However, Brazil has stated explicitly that it is not challenging the marketing loan and counter-cyclical payment programmes as such.150 While Brazil has emphasized that it is not possible to distinguish between the adverse effects of the statutory and regulatory provisions of the FSRI Act relating to upland cotton and the adverse effects of the payments mandated by these provisions151, Brazil does not request the Panel to find that, independently of their application in individual cases, the marketing loan and counter-cyclical payment programmes "will necessarily be inconsistent with"152 Articles 5 and 6 of the SCM Agreement.


          26. Given that Brazil is not seeking an "as such" finding, we consider that the preliminary objection of the United States is without object insofar as it would appear to be based on an interpretation of Brazil's claims as claims against programmes per se.153


          27. The Panel finds that, although the original panel's finding of "present" serious prejudice did not apply to legal provisions or subsidy programmes in addition to subsidies and subsidy measures, it is not necessary for the Panel to make a ruling on the preliminary objection of the United States because Brazil does not request the Panel to find that the marketing loan and counter-cyclical payment programmes are WTO-inconsistent as such. The Panel also concludes that the original panel's finding of serious prejudice was based on an analysis that took into consideration the legal provisions or subsidy programmes pursuant to which the subsidies were provided. The Panel considers that it is


            image

            148 As stated by the Appellate Body: "[b]y definition, an 'as such' claim challenges laws, regulations, or other instruments of a Member that have general and prospective application, asserting that a Member's conduct—not only in a particular instance that has occurred but in future situations as well—will necessarily be inconsistent with that Member's WTO obligations. In essence, complaining parties bringing 'as such' challenges seek to prevent Members ex ante from engaging in certain conduct." Appellate Body Report, US – Oil Country Tubular Goods Sunset Reviews, para. 172.

            149 For example, Brazil has stated that "[t]he measures that constitute these 'subsidies' are the statutory and regulatory provisions of the FSRI Act of 2002 that relate to upland cotton, i.e. the marketing loan and

            counter-cyclical payment provisions, and the mandatory and price-contingent payments to US upland cotton farmers that have been and will continue to be made during the lifetime of the FSRI Act of 2002, i.e. until MY 2007, unless the measure is extended". Brazil's Submission Regarding US Requests for Preliminary Ruling, para. 31.

            150 Brazil's Rebuttal Submission, Section 2.1.3 ("Brazil's Serious Prejudice Claims Do Not Constitute what the United States Calls 'As Such' Claims"); Brazil's Submission Regarding US Requests for Preliminary

            Ruling, para. 31 ("The United States claims that Brazil is asserting a claim against US marketing loan and counter-cyclical payment programs 'as such'. This is erroneous. Rather, Brazil is challenging in this proceeding the US subsidies inasmuch as they cause adverse effects."); Brazil's Response to Panel Question 12, para. 112 ("While Brazil has challenged the marketing loan and counter-cyclical 'programs' and their mandatory, price- contingent 'payments', to the extent that they cause adverse effects, it has not asserted an 'as such' claim regarding the marketing loan and counter-cyclical payment programs in the FSRI Act of 2002."). At the meeting of the Panel with the parties Brazil confirmed, in response to a question of the Panel, that it was not seeking a finding of the Panel under Articles 5 and 6 of the SCM Agreement against the marketing loan and counter-cyclical payments programmes themselves, as distinguished from the payments made pursuant to those programmes.

            151 See, e.g., Brazil's Rebuttal Submission, paras. 34-35; Brazil's Submission Regarding the US Requests for Preliminary Ruling, paras. 33-34.

            152 Supra, footnote 148.

            153 The question of whether a measure can, as such, i.e. independently of its application, be inconsistent with the effects-based disciplines in Articles 5 and 6 raises serious conceptual issues. As Brazil is not seeking a

            finding that the subsidy programmes are as such inconsistent with Articles 5 and 6 of the SCM Agreement, the Panel does not need to address these issues.

            WT/DS267/RW

            Page 48


            appropriate in this proceeding to conduct a similar analysis of subsidies in the context of the legal provisions or subsidy programmes pursuant to which the subsidies are granted.


            1. Whether Brazil's claims with respect to the marketing loan and counter-cyclical payment programmes are outside the scope of this proceeding because these programmes have not been changed in response to DSB recommendations and rulings or otherwise and therefore are not "measures taken to comply" within the meaning of Article 21.5 of the DSU


          28. As discussed above, Brazil does not request the Panel to find that the marketing loan and counter-cyclical payment programmes are WTO-inconsistent as such. For this reason, the Panel does not find it necessary to rule on the objection of the United States that since the marketing loan and counter-cyclical programmes are not "measures taken to comply" within the meaning of Article 21.5 of the DSU, Brazil's claims with respect to these programmes are not within the scope of this proceeding.


      4. Whether Brazil's claim regarding the failure of the United States to take measures between 22 September 2005 and 31 July 2006 is outside the scope of this proceeding


        1. Main arguments of the parties


            1. Brazil requests the Panel to find that the United States failed to take any measures to remove the adverse effects or to withdraw the three price-contingent upland cotton subsidy programmes under the FSRI Act of 2002 between 22 September 2005 and 31 July 2006 and that, as a consequence, measures taken to comply with the adverse effects-related recommendations and rulings of the DSB did not exist in this period.154 Brazil also requests that the Panel conclude that the recommendation under Article 7.8 has not been complied with entirely and remains in effect.155


            2. Brazil argues that under Articles 7.8 and 7.9 of the SCM Agreement the United States had until 21 September 2005 (six months after the date of adoption of the reports in the original proceeding) to withdraw the subsidies or to take appropriate steps to remove their adverse effects. Since the United States repealed the Step 2 programme only on 1 August 2006 and did not take any action to amend the marketing loan and counter-cyclical payments provisions of the FSRI Act of 2002, the United States has failed to adopt measures to comply within the implementation period. This means that the Panel can conclude that no measures to comply existed between 22 September 2005 and 31 July 2006.156 Brazil submits that the panel report in Australia – Salmon (Article 21.5 – Canada) provides relevant precedent for a finding in an Article 21.5 proceeding that measures to comply did not exist in a past period.157


            3. The United States asserts that Brazil's claim that there were no measures taken to comply in the period between 22 September 2005-31 July 2006 is not within the scope of this proceeding. First, assuming (incorrectly) that the DSB actionable subsidy-related recommendations are related to the Step 2, marketing loan, and counter-cyclical programmes, Brazil asserts that it is 'undisputed' that the United States failed to take measures to comply in a timely fashion, before the repeal of the Step 2 programme on 1 August 2006. Thus, in the view of the United States, Brazil does not even assert a 'disagreement' as to the absence of measures taken to comply prior to 1 August 2006.158 Second, the issue of the timeliness of implementation is not within the scope of a panel's mandate under Article 21.5 of the DSU. Specifically, Article 21.5 does not direct a panel to examine whether a



              image

              154 Brazil's First Written Submission, paras. 40 and 461.

              155 Brazil's First Written Submission, para. 46.

              156 Brazil's First Written Submission, paras. 39-46.

              157 Brazil's First Written Submission, para. 46, footnote 37.

              158 See US First Written Submission, para. 50.

              WT/DS267/RW

              Page 49


              Member has implemented actionable subsidy-related recommendations within the six-month period provided for in Article 7.9 of the SCM Agreement. The United States submits that not only is there no textual basis for the kind of finding requested by Brazil, but that a finding with respect to a factual situation that no longer exists is also inconsistent with the DSU objective of securing a positive solution to a dispute. The United States submits that prior panel reports in Article 21.5 proceedings support its position that compliance must be examined on the basis of the facts as they stand when a matter is referred to a compliance panel.159


            4. Brazil submits that by arguing that there is no disagreement between the parties with respect to the existence of measures taken to comply the United States acknowledges that it has not taken any measures to comply with the adverse effects-related rulings and recommendations of the DSB by the end of the implementation period and requests the Panel to make a finding to that effect. Brazil also argues that Articles 7.9 and 7.10 of the SCM Agreement constitute a legal basis for the kind of finding requested by Brazil. Brazil argues that whether measures taken to comply existed within the time- frame specified in Article 7.9 is relevant to an analysis under Article 21.5 of the "consistency" of such measures with the provisions of the covered agreements and that "Brazil's right to pursue the United States' failure to remove the adverse effects or withdraw the subsidy, and the assessment of whether any proposed countermeasures are commensurate, are tied to the date on which the implementation period expires".160


        2. Main arguments of third parties


            1. Australia argues that the Panel should refuse the request of the United States for a preliminary ruling regarding Brazil's claim concerning the failure of the United States to take measures to comply within the period 22 September 2005-1 July 2006. Australia considers that an examination under Article 21.5 of the DSU of the "existence" of measures taken to comply can encompass a finding that no measures taken to comply exist. Australia refers, in this respect, to the panel in Australia – Salmon (Article 21.5 – Canada), which found that because the date of entry into force of measures taken to comply by Australia occurred subsequent to the date upon which the reasonable period of time for implementation expired, for the period of time that the new measures did not apply subsequent to that date, "no measures taken to comply existed ... in the sense of Article 21.5". Australia also argues that a preliminary ruling cannot be granted in a way that would void rights under Article 22 of the DSU. Article 22 rights would logically ensue from findings of inconsistency with Articles 7.8 and 7.9 of the SCM Agreement. Thus, if the Panel were to accept the preliminary ruling requested by the United States, this would diminish the rights of Brazil under Article 22 of the DSU.161


            2. The European Communities disagrees with the view of the United States that Brazil's claim regarding non-compliance in a past period is not within the scope of this proceeding under Article 21.5 of the DSU. The European Communities considers that in making this argument the United States overlooks the fact that any act or omission may be a measure for purposes of WTO dispute settlement. The European Communities also submits that there is nothing in the text of Article 21.5 of the DSU to limit the temporal application of that provision in the manner advocated by the United States.162



              image

              159 US First Written Submission, paras. 49-56.

              160 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 77-87.

              161 Australia's Third Party Submission, paras. 18-20.

              162 EC Third Party Oral Statement, paras. 25-30. With respect to the merits of Brazil's claim, the European Communities argues that Members enjoy a certain measure of discretion in the implementation of the

              obligation under Article 7.8 of the SCM Agreement to take appropriate steps to remove the adverse effects of the subsidy or withdraw the subsidy. Article 7.8 does not necessarily require that a Member take positive action. Whether a Member has complied with its obligation under Article 7.8 requires a careful analysis of the

              WT/DS267/RW

              Page 50


            3. Japan submits that the Panel has the authority to make a finding on Brazil's claim regarding the non-compliance of the United States with the DSB recommendations and rulings in the period 22 September 2005-31 July 2006. Japan specifically rejects, in this regard, the argument of the United States that since there is no disagreement between the parties, Brazil's claim is outside the scope of this proceeding under Article 21.5 of the DSU. Japan agrees with Australia's argument that a failure of the Panel to address Brazil's claim with respect to the alleged non-compliance of the United States in the period 22 September 2005-31 July 2006 would diminish rights of Brazil under Article 22 of the DSU.163


            4. New Zealand submits that the United States cannot claim that since it took no implementing action until 1 August 2006, that period is outside the scope of this proceeding. Article 21.5 of the DSU applies to the "existence" of measures, not just to the consistency of measures actually taken to comply with the recommendations and rulings of the DSB.164


        3. Evaluation by the Panel


          1. The United States requests the Panel to make a preliminary ruling that Brazil's claim that the United States failed to take any measures to remove the adverse effects or to withdraw the three price- contingent upland cotton subsidy programmes under the FSRI Act of 2002 between 22 September 2005 and 31 July 2006 and that, as a consequence, measures taken to comply with the adverse effects-related recommendations and rulings of the DSB did not exist in this period, is not properly within the scope of this proceeding under Article 21.5 of the DSU.165


          2. It is not in dispute that the factual situation referred to by Brazil in this request no longer existed at the time of the establishment of the Panel as the United States repealed the Step 2 programme effective 1 August 2006. In this regard, the parties disagree as to whether it is within this Panel's mandate under Article 21.5 of the DSU to make a finding as to whether the United States had complied with its obligations under Article 7.8 of the SCM Agreement at the end of the six-month period provided for in Article 7.9 of the SCM Agreement, or whether the Panel's mandate under Article 21.5 only allows the Panel to determine whether the United States had complied with its obligations under Article 7.8 of the SCM Agreement at the time of the establishment of the Panel.


          3. Brazil's request is based on two important premises. First, the request assumes that under Article 21.5 of the DSU this Panel is required to make findings in relation to factual situations at two different points in time. Specifically, according to the logic of Brazil's argument, the Panel is required to make a finding as to the "existence or consistency with a covered agreement of measures taken to comply" as of the end of the period referred to by Brazil as the implementation period166 (22 September 2005) in addition to a finding as to "existence or consistency with a covered agreement of measures taken to comply" as of the time of the establishment of the Panel. Second, Brazil's request also assumes that the fact that the United States did not take any action in the period 22 September 2005-31 July 2006 with respect to the measures subject to the DSB recommendations and rulings in the original proceeding is a sufficient basis for this Panel to find that the United States


            image

            particular facts and circumstances of each individual case. Therefore, the fact that between 22 September 2005 and 31 July 2006 the United States took no positive action to comply with the recommendations and rulings of the DSB does not necessarily warrant the conclusion that the United States is in breach of Article 7.8. EC Third Party Submission, paras. 34-48.

            163 Japan's Third Party Oral Statement, paras. 8-9.

            164 New Zealand's Third Party Submission, para. 4.02.

            165 The Panel notes that in its request for the establishment of a panel Brazil asserts that "[a]t least during the period 21 September 2005 through 31 July 2006, measures taken to comply did not exist, within the

            meaning of Article 21.5 of the DSU". WT/DS267/30, para. 12.

            166 Brazil refers to the period of six months mentioned in Article 7.9 of the SCM Agreement as the "implementation period". See e.g. Brazil's Response to Panel Question 22, para. 159.

            WT/DS267/RW

            Page 51


            failed to comply with its obligation under Article 7.8 of the SCM Agreement to "take appropriate steps to remove the adverse effects or …withdraw the subsidy"167 and that, as a consequence, no measure taken to comply within the meaning of Article 21.5 of the DSU existed at the end of that period.


          4. With respect to the first premise of Brazil's argument, the Panel considers that a finding as to the "existence or consistency with a covered agreement of measures taken to comply" as of the end of the "implementation period" of Article 7.9 of the SCM Agreement or of the reasonable period of time referred to in Article 21.3 of the DSU, in addition to a finding as to the "existence or consistency with a covered agreement of measures taken to comply" as of the time of the establishment of a panel, would be of little relevance to the effective resolution of disputes. Where a panel makes a finding under Article 21.5 of the DSU that a Member has not complied with the DSB recommendations and rulings in the original dispute, the consequence of that finding is that the Member remains subject to obligations that flow from the recommendation issued in the original proceeding168 and is thus required to take steps to bring itself into compliance with that recommendation. A finding by the panel that the Member also failed to comply with the DSB recommendations and rulings in the original proceeding at an earlier point in time would have no additional operative value in terms of the nature of the obligations of the Member in question. On the other hand, if a panel under Article 21.5 finds that the Member has brought itself into compliance with the DSB recommendations and rulings as of the time of the establishment of the panel, such a finding logically would supersede and render irrelevant any finding that the Member was not in compliance with those recommendations and rulings at an earlier point in time. Thus, in both cases a finding of a violation at the end of the reasonable period of time would be of a declaratory nature and without practical implications as to the obligations of the Member in question, unless one interprets the DSU to mean that a proceeding under Article 21.5 can create obligations for a Member to take steps that must be effective as of the end of the reasonable period of time. We see no textual support for such a retroactive interpretation of the DSU.


          5. Brazil argues that the provisions on countermeasures in Articles 7.9 and 7.10 of the SCM Agreement provide support for its request that the Panel make a finding that the United States failed to take any measure at the end of what Brazil characterizes as the implementation period.169 We disagree with this argument. We see nothing in the text of Article 21.5 or of other provisions of the DSU that suggests that, because of considerations relating to the application of Articles 7.9-7.10 of the SCM Agreement or Article 22.6 of the DSU, a panel must make a finding as to the existence or consistency of measures taken to comply as of the end of the six-month period set out in Article 7.9 of the SCM Agreement or the reasonable period of time rather than as of the time of the establishment of the panel. Nothing in the text or context of Article 21.5 of the DSU indicates that one of the functions of a panel in a proceeding under Article 21.5 of the DSU is to assist arbitrators170 in proceedings under Articles 7.10 of the SCM Agreement or Article 22.6 of the DSU to determine the amount of countermeasures or the amount of suspension of concessions or obligations and that, to this end, the panel must make findings as to the existence or consistency with a covered agreement of measures taken to comply other than as of the date of the establishment of a panel. Brazil also argues that a right to apply countermeasures arises under Article 7.9 of the SCM Agreement if the implementing Member fails to comply with Article 7.8 within six months of the date of adoption of the panel report or the Appellate Body report in the original proceeding, and that "[b]y pursuing its claim regarding



            image

            167 Brazil submits that Article 7.8 of the SCM Agreement requires affirmative action to remove the adverse effects or withdraw the subsidy and that inaction is not sufficient in this respect. Brazil's Responses to Panel Questions 22, 24-25 and 27.

            168 Panel Report, US – FSC (Article 21.5 – EC).

            169 Brazil's Submission Regarding US Requests for Preliminary Ruling, paras. 79-85; Brazil's Response to Panel Question 21; Brazil's Comments on US Response to Panel Question 23; Brazil's Comments on US Response to Panel Question 43.

            170 Brazil's Response to Panel Question 21.

            WT/DS267/RW

            Page 52


            the non-existence of measures on 21 September 2005, Brazil seeks a multilateral basis for the DSB to authorize countermeasures against the United States for its failure to take any implementation measures by the date required in Article 7.9 of the SCM Agreement". However, while a finding by a panel under Article 21.5 of the DSU may be required to provide "a multilateral basis" for DSB authorization of countermeasures under Article 7.9 of the SCM Agreement, it does not logically follow that such a panel must make a finding of non-compliance as of the end of the six-month period refered to in Article 7.9 of the SCM Agreement.


          6. Finally, we note that in response to a Panel question as to whether Brazil's request for a finding of non-compliance by the United States as of the end of the six-month period provided for in Article 7.9 of the SCM Agreement is supported by prior panel practice in Article 21.5 proceedings, Brazil identifies only one panel report, Australia – Salmon (Article 21.5 – Canada). While we acknowledge that this panel report contains findings of non-compliance in past periods171, the United States points to other panel reports that have concluded that it is not appropriate to make findings of non-compliance under Article 21.5 as of the end of the reasonable period of time.172


          7. Thus, we disagree with Brazil's argument that Article 21.5 of the DSU contemplates a finding on the existence or consistency with a covered agreement of measures taken to comply at the end of the reasonable period of time in addition to a finding on the existence of a measure taken to comply as of the date of the establishment of a panel. We therefore do not need to address the question whether the fact that the United States did not take action before the expiry of the period of six months referred to in Article 7.9 of the SCM Agreement is a sufficient basis to find that the United States had not complied with Article 7.8 of the SCM Agreement by that date.


          8. In light of the foregoing considerations, the Panel finds that it is not appropriate to make a finding on Brazil's claim that the United States failed to take any measures to remove the adverse effects or to withdraw the three price-contingent upland cotton subsidy programmes under the FSRI Act of 2002 between 22 September 2005 and 31 July 2006 and that, as a consequence, measures taken to comply with the adverse effects-related recommendations and rulings of the DSB did not exist in this period.


      5. Whether "payments" made after 21 September 2005 pursuant to the marketing loan and counter-cyclical payment programmes are within the scope of this proceeding


      1. The Panel now turns to the question of whether payments made under the marketing loan and counter-cyclical payment programmes after 21 September 2005 are properly within the scope of this proceeding.


        1. Main arguments of the parties


      2. The United States argues that the subsidies subject to the obligation in Article 7.8 of the SCM Agreement to "take appropriate steps to remove the adverse effects or... withdraw the subsidy" were the subsidies provided during MY 1999-2002 which the original panel had found to cause "present" serious prejudice in MY 1999-2002.173 Since the only measure subject to the obligation under Article 7.8 were payments made until 31 July 2003 and since the payments made after 21 September 2005 are not measures taken to comply the United States considers that these payments are not properly within the scope of this proceeding.174 The United States considers that there is no



        image

        171 Panel Report, Australia – Salmon (Article 21.5 – Canada), paras. 7.30, 7.35 and 8.1(i).

        172 Panel Report, US – Shrimp (Article 21.5 – Malaysia), paras. 5.12-5.13; Panel Report, EC – Bed Linen (Article 21.5 – India), para. 6.28.

        173 US Response to Panel Question 18; US Comments on Brazil's Response to Panel Question 15.

        174 US Response to Panel Question 47.

        WT/DS267/RW

        Page 53


        merit to the argument of Brazil that, even assuming that the original panel's findings of "present" serious prejudice were "as applied" findings, subsequent payments under the same programme would still be subject to the implementation obligations of the United States. The Appellate Body report that Brazil cites in support of that argument (US – Softwood Lumber IV (Article 21.5 – Canada)) said nothing about whether a complaining party could in post hoc fashion attempt to add to the measures found to be WTO-inconsistent by asserting that other measures not found to be WTO-inconsistent are similar. Rather that dispute addressed the distinct issue of what measures could be considered to be part of the measure taken to comply. Future payments under the marketing loan and counter-cyclical payment programmes in MY 2003-2007 are not measures taken to comply with any recommendations and rulings. To the contrary, they were original measures that were challenged by Brazil but against which the original panel made no finding of WTO-inconsistency.175


      3. Brazil argues that the subsidies subject to the obligation in Article 7.8 of the SCM Agreement include the legislative and regulatory subsidy programmes in the FSRI Act and the payments made under those programmes. However, even assuming arguendo that the original panel's finding of serious prejudice only applied to payments made in MY 1999-2002, the reasoning of the Appellate Body in US – Softwood Lumber IV (Article 21.5 – Canada) supports the view that subsequent payments made under these programmes are subject to the obligation in Article 7.8 because such subsequent payments have "a particularly close relationship" to the recommendations and rulings of the DSB.176 177 Brazil also submits that Article 7.8 requires a full and permanent removal of adverse effects and that this means that a Member may not replace a subsidy found to cause adverse effects with another subsidy that continues to cause the same adverse effects. Therefore, even if in this dispute the original panel had made findings limited to payments made in MY 1999-2002, the obligation of the United States under Article 7.8 would not be limited to the removal of the adverse effects of payments made in that period.178 Brazil argues that if a panel makes a finding of "present" adverse effects, such a finding applies beyond the historical period considered by the panel and that the Member whose subsidies cause such adverse effects is therefore obligated under Article 7.8 to take appropriate steps to fully withdraw the present, ongoing and future effects of the subsidies.179 Finally, Brazil also argues that the fact that the original panel exercised judicial economy with respect to future payments in 2003-2007 does not prevent Brazil from challenging those payments in this proceeding.


        1. Evaluation by the Panel


      4. In order to determine whether payments made under the marketing loan and counter-cyclical payment programmes after 21 September 2005 are within the scope of this proceeding, we must consider whether such payments fall within the scope of the phrase "disagreement as to the existence or consistency with a covered agreement of measures taken to comply with the recommendations and rulings" of the DSB in the original proceeding. In this respect, the Panel's examination of the scope of "measures taken to comply" must also involve an examination of the recommendations and rulings of the DSB.180


      5. With respect to its finding of "present" serious prejudice the original panel recalled in paragraph 8.3(d) of its report that Article 7.8 of the SCM Agreement provides as follows:



        image


        para. 4.

        175 US Comments on Brazil's Response to Panel Question 15.

        176 Brazil's Response to Panel Question 15, paras. 145-150; Brazil, Response to Panel Question 24,


        177 See also Brazil's Comments on US Response to Panel Question 44, paras. 41-47.

        178 Brazil's Response to Panel Question 24, paras. 13-14.

        179 Brazil's Comments on US Response to Panel Question 24, para. 39.

        180 Appellate Body Report, US – Softwood Lumber IV (Article 21.5 – Canada), para. 68; Appellate

        Body Report, US – Oil Country Tubular Goods Sunset Reviews (Article 21.5 – Argentina), para.142.

        WT/DS267/RW

        Page 54


        "Where a panel report or an Appellate Body report is adopted in which it is determined that any subsidy has resulted in adverse effects to the interests of another Member within the meaning of Article 5, the Member granting or maintaining such subsidy shall take appropriate steps to remove the adverse effects or shall withdraw the subsidy."


      6. Under Article 7.8 of the SCM Agreement the United States was obligated, with respect to the subsidies subject to the "present" serious prejudice finding of the original panel, to "take appropriate steps to remove the adverse effects or ... withdraw the subsidy".


      7. It is clear from the context that the adverse effects that must be removed are the adverse effects of the subsidy that has been determined to have resulted in adverse effects. Since the original panel made a finding of present serious prejudice in respect of subsidies provided during MY 1999- 2002, the question arises whether the obligation to take appropriate steps to remove the adverse effects only applies to payments of subsidies made in those years.


      8. It is not in dispute that the United States presently provides marketing loan and counter- cyclical payments on the same legal basis and subject to the same conditions and criteria as the marketing loan payments and counter-cyclical payments that were subject to the panel's finding of "present" serious prejudice. In a situation where the subsidy in question has been found to be a prohibited one, the continued use of the subsidy under the same conditions and criteria is inconsistent with a Member's obligation to "withdraw" the subsidy under Article 4.7 of the SCM Agreement.181 Thus, the concept of "withdrawal" must in any event be interpreted to mean that a Member must cease to act in a WTO-inconsistent manner with respect to that subsidy. If a failure to cease conduct inconsistent with a Member's obligations under Article 3 of the SCM Agreement is inconsistent with the obligation to withdraw the subsidy in Article 4.7 of the SCM Agreement, we see no logical reason why the same concept should not also apply to the obligation that arises under Article 7.8 of the SCM Agreement to withdraw the subsidy or to take appropriate steps to remove the adverse effects of a subsidy that has been determined to result in adverse effects. In our view, the remedy under Article 7.8 must be viewed in its relationship to the obligation in Article 5 not to cause through the use of any subsidy referred to in Articles 1.1 and 1.2 of the SCM Agreement adverse effects to the interests of other Members. It must serve to restore conformity with the Member's obligation to avoid causing adverse effects through the use of any subsidy. As a consequence, a Member does not take appropriate steps to remove adverse effects of a subsidy if it continues to provide payments under the same conditions and criteria as the original subsidy in a manner that causes adverse effects. The interpretation advocated by the United States, whereby the obligation under Article 7.8 of the SCM Agreement is limited to the removal of the adverse effects caused by subsidies granted in a particular period of time, implies that it would not be possible to review in an Article 21.5 proceeding whether a Member causes adverse effects by continuing to grant subsidies under the same conditions and criteria as the subsidies found to have caused adverse effects. Such an interpretation fails to take into account the relationship between Article 7.8 and Article 5 of the SCM Agreement and thus fails to interpret Article 7.8 in its proper context.


      9. We find further support for this approach in panel and Appellate Body reports on the scope of Article 21.5 of the DSU in which it has been held that "measures taken to comply" include " measures with a particularly close relationship to the declared 'measure taken to comply', and to the recommendations and rulings of the DSB".182 As explained above, our interpretation of Article 7.8 implies that if a Member continues to make payments of a subsidy on the same legal basis and under


        image

        181 See Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 45; Panel Report, Brazil Aircraft (Article 21.5 – Canada), para. 6.8; Panel Report, Canada – Aircraft (Article 21.5 – Brazil), para. 5.10.

        182 E.g. Appellate Body Report, US – Softwood Lumber IV (Article 21.5 – Canada), para. 77. Supra,

        para. 9.24.

        WT/DS267/RW

        Page 55


        the same conditions and criteria as the subsidy found to have caused serious prejudice in the original proceeding, it must be possible for a panel in a proceeding under Article 21.5 of the DSU to examine a claim that the continued payment of that subsidy causes adverse effects within the meaning of Article 5 of the SCM Agreement. To hold that in such a situation a Member that considers itself to be adversely affected by the subsidy must initiate an entirely new proceeding under the DSU would in our view undermine the effectiveness of Article 7.8. Therefore, in light of our interpretation of Article 7.8, we consider that, where payments of a subsidy have been found to cause serious prejudice and the Member in question continues to provide the same subsidy under the same conditions and criteria, there exists "a particularly close relationship" between the subsidy that the Member continues to provide and the recommendations and rulings of the DSB in the original proceeding.


      10. In light of these considerations, the Panel concludes that to the extent marketing loan payments and counter-cyclical payments made by the United States after 21 September 2005 are provided under the same conditions and criteria as the marketing loan payments and counter-cyclical payments subject to the original panel's finding of "present" serious prejudice, they are subject to the obligation of the United States under Article 7.8 of the SCM Agreement to take appropriate steps to remove the adverse effects of the subsidy. As a consequence, we also consider that Brazil's claim that the United States has failed to comply with its obligations under Article 7.8 with respect to these payments is properly before this Panel. In our view this claim pertains to a disagreement between the parties as to the "existence or consistency with a covered agreement of measures taken to comply" with the recommendations and rulings of the DSB.


  10. FINDINGS ON BRAZIL'S CLAIMS OF "PRESENT" SERIOUS PREJUDICE UNDER ARTICLES 5(C) AND 6.3 OF THE SCM AGREEMENT


    1. INTRODUCTION


        1. Brazil requests the Panel to find that the measure taken by the United States to comply with the recommendations and rulings of the DSB, i.e. the amendment of the FSRI Act of 2002 to repeal section 1207(a) providing for Step 2 payments, is insufficient to ensure compliance with those recommendations and rulings because the remaining price-contingent subsidies are inconsistent with Articles 5 and 6 of the SCM Agreement. In particular, Brazil requests the Panel to find that the marketing loan and counter-cyclical payment programmes183 of the FSRI Act of 2002 cause "present" serious prejudice to the interests of Brazil within the meaning of Article 5(c) of the SCM Agreement in the form of significant price suppression and an increase in the US world market share of upland cotton within the meaning of, respectively, Articles 6.3(c) and 6.3(d) of the SCM Agreement.


        2. The United States requests that the Panel find that the United States has complied with the recommendations and rulings of the DSB. The United States submits that Brazil fails to substantiate its arguments in support of the claim that the effects of the elimination of the Step 2 programme are


          image

          183 As formulated in para. 461 of Brazil's First Written Submission, Brazil's request is with respect to the marketing loan and counter-cyclical programmes. The Panel notes, however, that Brazil has clarified that it does not request the Panel to make a finding that the programmes or legal provisions pursuant to which marketing loan and counter-cyclical payments are made are, as such, inconsistent with Articles 5 and 6 of the SCM Agreement. Supra, footnote 150. In light of this clarification the Panel will hereinafter refer to the subsidies subject to Brazil's serious prejudice claims as marketing loan and counter-cyclical payments. The Panel emphasizes, however, that as discussed above (supra, para. 9.52) a proper analysis of the effect of the marketing loan and counter-cyclical payments for purposes of Articles 5 and 6 of the SCM Agreement must examine them in the context of the legal and regulatory framework that defines the nature, conditions and criteria under which these payments will be made.

          WT/DS267/RW

          Page 56


          modest and that Brazil fails to make a prima facie case of "present" serious prejudice within the meaning of Articles 5(c) and 6.3(c) and 6.3(d) of the SCM Agreement.184


        3. The Panel addresses Brazil's claim of "present" serious prejudice within the meaning of Articles 5(c) and 6.3(c) of the SCM Agreement in Section X.B of this Report. The Panel addresses Brazil's claim of "present" serious prejudice within the meaning of Articles 5(c) and 6.3(d) of the SCM Agreement in Section X.C of this Report.


    2. BRAZIL'S CLAIM OF "PRESENT" SERIOUS PREJUDICE WITHIN THE MEANING OF ARTICLES 5(C)

      AND 6.3 (C) OF THE SCM AGREEMENT


      1. Introduction – overview of the main arguments of the parties


          1. Brazil submits that marketing loan and counter-cyclical payments made pursuant to the FSRI Act of 2002 cause significant price suppression in the world market for upland cotton within the meaning of Article 6.3(c) of the SCM Agreement and thereby cause "present" serious prejudice to the interests of Brazil within the meaning of Article 5(c) of the SCM Agreement. The United States submits that Brazil has not made a prima facie case under Articles 5(c) and 6.3(c) of the SCM Agreement.


          2. In support of its claim that US marketing loan and counter-cyclical payments cause significant price suppression in the world market for upland cotton, Brazil submits that:185


            1. US upland cotton production and exports continue to expand and exercise a substantial proportionate influence on prices in the world market for upland cotton.


            2. The level of US planted acreage, production and exports heavily impacts world market prices of upland cotton, as evidenced by data showing broad similarities in the trends of the five major price indices relied upon by the original panel. Moreover, upland cotton prices during MY 2002-2005 have remained well below historic levels.


            3. The magnitude of the marketing loan and counter-cyclical payments remains very large.


            4. The structure, design and operation of the marketing loan and counter-cyclical payments186 stimulate US upland cotton production by insulating US producers from market forces.



              image

              184 The United States also submits that Brazil has failed to demonstrate that the marketing loan payment and counter-cyclical payment programmes mandate a breach of WTO obligations. US First Written Submission, paras. 196-202. The Panel understands that the United States made this argument in its First Written Submission on the assumption that Brazil was challenging the marketing loan and counter-cyclical payment programmes as such. Because Brazil has made it clear that it does not claim in this proceeding that the marketing loan and counter-cyclical payment programmes at issue are WTO-inconsistent as such (supra, footnote 150), we do not need to address this argument of the United States.

              185 Brazil also submits that marketing loan and counter-cyclical payments are specific subsidies, that US and Brazilian upland cotton are like products, that there exists a world market for upland cotton and that the A-Index represents the world market price. See infra, paras. 10.20 and 10.42-10.44.

              186 We note that in the title of the relevant section of its First Written Submission, Brazil refers to the

              structure, design and operation of "US marketing loan and counter-cyclical payments", but Brazil elsewhere in the same section also uses the phrase "nature of marketing loan and counter-cyclical payment programs, in terms of their structure, design and operation". Brazil's First Written Submission, Section 7.8.

              WT/DS267/RW

              Page 57


            5. There remains a strong link between high levels of US subsidies and high levels of US planted acreage, production and exports during MY 2002-MY 2005, which is due to the fact that marketing loan and counter-cyclical payments insulate US cotton producers from market price signals.


            6. There remains a discernible temporal coincidence of suppressed world market prices and large marketing loan and countercyclical payments.


            7. Marketing loan and counter-cyclical payments continue to cover a significant portion of US upland cotton producers' long-term costs of production.


            8. Economic studies support a finding that marketing loan and counter-cyclical payments continue to cause significant price suppression.


            9. The factors relied upon by the original panel as the basis for its conclusion regarding the "significance" of the price suppression caused by the subsidies provided by the United States continue to exist.187


          3. Brazil argues that the effects of the US marketing loan and counter-cyclical payments alone cause significant price suppression. Although Step 2 payments caused significant price suppression in and of themselves during MY 2005, this does not detract from the significant, independent price suppressing effects of the marketing loan and counter-cyclical payments. The removal of Step 2 payments has had a relatively modest impact on US production, US exports and world market prices for upland cotton, compared to the effects of marketing loan and counter-cyclical payments.188 Other factors impacting the world market price of upland cotton do not account for the significant price suppressing effects caused by marketing loan and counter-cyclical payments.189


          4. In support of its argument that Brazil fails to make a prima facie case of WTO inconsistency under Articles 5(c) and 6.3(c) of the SCM Agreement, the United States submits that:


            1. The structure, design and operation of the marketing loan and counter-cyclical payments do not support Brazil's claim of significant price suppression.


            2. The facts demonstrate that US producers and exporters have reacted to market signals and are not "insulated" by the marketing loan and counter-cyclical payments. Brazil fails to demonstrate the alleged strong link between marketing loan and counter- cyclical payments and US planted acreage, production and exports.


            3. "Absolute" increases in US production are the effect of improvements in yields, not the marketing loan and counter-cyclical payments.


            4. Brazil fails to demonstrate a "temporal coincidence" between marketing loan and counter-cyclical payments and the alleged price suppression.


            5. Brazil fails to demonstrate that US producers would "switch to alternative crops" in the absence of marketing loan and counter-cyclical payments.


            6. Brazil attempts to attribute the price effects of other factors to the marketing loan and counter-cyclical payments.



              image

              187 Brazil's First Written Submission, paras. 88-209. 188 Brazil's First Written Submission, paras. 191-209. 189 Brazil's First Written Submission, paras. 210-211.

              WT/DS267/RW

              Page 58


            7. The econometric modeling cited by Brazil is flawed and greatly exaggerates any effects of removing the marketing loan and counter-cyclical payment programmes.


            8. Brazil has not demonstrated that "the effect" of the marketing loan and counter- cyclical payments is "significant" price suppression.190


      2. Arguments of third parties


          1. Argentina considers that by maintaining the marketing loan and counter-cyclical payment programmes the United States has failed to comply with its obligation under Article 7.8 of the SCM Agreement to take appropriate steps to remove the adverse effects or withdraw the subsidy. The evidence submitted by Brazil demonstrates that the marketing loan and counter-cyclical payment programmes continue to cause adverse effects on the interests of Brazil and other cotton-producing countries such as Argentina in the form of significant price suppression.191


          2. Australia submits that the Panel should consider whether the United States, by repealing the Step 2 payment programme while maintaining the two other price-contingent subsidies (payments under the marketing loan and counter-cyclical payment programmes), has taken appropriate steps to remove the adverse effects or withdraw the subsidies within the meaning of Article 7.8 of the SCM Agreement in accordance with the DSB recommendations and rulings. Australia recalls, in this respect, that in making its findings the original panel examined each price-contingent subsidy measure in turn before considering the collective operation of those measures, and that it is clear from the panel's analysis that each of the price-contingent measures individually was found to have a trade- distorting effect.192


          3. Chad argues that the United States has not properly implemented the original panel's findings regarding the marketing loan and counter-cyclical payment programmes. As the only action undertaken by the United States is the repeal of the Step 2 programme with effect from 1 August 2006, it is clear that the United States has not withdrawn or amended the marketing loan or counter-cyclical payment programmes found to be WTO-inconsistent. Moreover, the repeal of the Step 2 programme cannot be considered to have removed the adverse effects of the marketing loan and counter-cyclical payments. The marketing loan and counter-cyclical payment programmes accounted for the vast majority of the adverse effects found by the original panel. In this respect, Chad points to empirical studies that have found that the measures most detrimental to Chad are the US domestic support programmes, rather than export subsidies under the Step 2 programme. Chad submits that since the United States has not changed the marketing loan and counter-cyclical programmes that were subject to the original panel's finding, this Panel should rely on the legal and factual conclusions drawn by the original panel concerning the existence, nature and potential for price suppression of those measures. The price effects of these measures have continued and indeed gotten worse since the original panel made its finding and Chad and other West African cotton- producing countries have continued to lose market share.193


          4. India submits that the original panel examined carefully and separately the individual nature, operation and magnitude and price-suppressing effects of the marketing loan payment programme, the counter-cyclical payment programme and the Step 2 payment programme and found that the vast majority of the price-suppressing and adverse effects were caused by the marketing loan and counter- cyclical payment programmes, rather than by the Step 2 payments. Therefore, the withdrawal of the



            image

            190 US First Written Submission, paras. 204-331.

            191 Argentina's Third Party Submission, paras. 5-8. 192 Australia's Third Party Submission, paras. 25-31. 193 Chad's Third Party Submission, paras. 22-46.

            WT/DS267/RW

            Page 59


            Step 2 programme for upland cotton cannot be said to have also removed the adverse effects of the marketing loan and counter-cyclical payment programmes.194


          5. New Zealand submits that the various arguments of the United States on the removal of the Step 2 payment programme leave intact the fundamental point made by Brazil, that even in the absence of the Step 2 payment programme the marketing loan and counter-cyclical payment programme still result in significant price suppression within the meaning of Article 6.3.(c) of the SCM Agreement and thus cause serious prejudice to the interests of Brazil. Regarding the argument of the United States that Brazil has failed to make a prima facie case of WTO-inconsistency of the marketing loan and counter-cyclical payment programmes, New Zealand argues that an approach that would put the burden on Brazil to prove once again what Brazil already proved before the original panel would risk depriving the provisions in Part III of the SCM Agreement on actionable subsidies of their useful effect. Similarly, New Zealand considers that the attempt of the United States to reargue what it had argued before the original panel is not consistent with the objective of Article 7.8 of the SCM Agreement and of Article 21.5 of the DSU. New Zealand argues that counter-cyclical payments are not real decoupled payments because they are linked to prices. The proposition that marketing loan payments are distorting is self-evident because these payments are amber box measures, the category in which are included the non-prohibited measures with the most distorting effect on production and trade. Moreover, the arguments of the United States on marketing loan payments ignore that it is the expectations of producers that are important, and that those expectations are generated because of the existence of the applied administrative prices under the marketing loan programme, not because of the financial transfers in any year. New Zealand considers that the United States does not and cannot demonstrate that planting decisions of US cotton producers were shaped by market factors. With respect to the arguments of the United States on the econometric modelling provided by Brazil, New Zealand submits that these arguments are either an attempt to reargue what was litigated before, and disposed of, by the original panel and the Appellate Body, or are inconsequential.


      3. Preliminary considerations relating to the Panel's evaluation of Brazil's claim under Articles 5(c) and 6.3(c) of the SCM Agreement


    1. Text of legal provisions


        1. Article 5(c) of the SCM Agreement provides:


          "No Member should cause, through the use of any subsidy referred to in paragraphs 1 and 2 of Article 1, adverse effects to the interests of other Members, i.e.:


          [...]


          (c) serious prejudice to the interests of another Member.13



          image


          13 The term "serious prejudice to the interests of another Member" is used in this Agreement in the same sense as it is used in paragraph 1 of Article XVI of GATT 1994, and includes threat of serious prejudice."


        2. Article 6.3(c) of the SCM Agreement reads as follows:



          image

          194 India's Third Party Oral Statement, para. 5.

          WT/DS267/RW

          Page 60


          "6.3 Serious prejudice in the sense of paragraph (c) of Article 5 may arise in any case where one or several of the following apply:


          [...]


          (c) the effect of the subsidy is a significant price undercutting by the subsidized product as compared with the price of a like product of another Member in the same market or significant price suppression, price depression, or lost sales in the same market."


    2. Period to be considered by the Panel


      1. Main arguments of the parties


          1. Brazil submits that in analyzing Brazil's claims of "present" serious prejudice the Panel should use as reference period MY 2005 (1 August 2005-31 July 2006), which is the most recently completed marketing year for which complete data is available. The Panel can also use partial data for MY 2006 if such data is credible and reliable. Brazil submits that by conducting a counter-factual analysis the Panel can isolate the effects of the marketing loan and counter-cyclical payments made in MY 2005 from the effects of Step 2 payments made in MY 2005. Brazil argues that the Panel can also use data for the period MY 2002-MY 2005 in order to assess the continued relevance of the findings of the original panel.195


          2. The United States argues that Brazil's claims of "present" serious prejudice pertain to the present marketing year, MY 2006. The use of the term "is" in Article 6.3(c) of the SCM Agreement ("the effect of any subsidy is ... significant price suppression") means that the relevant period is the immediate present rather than a historical period. While it may be appropriate to use historical data as a proxy for the present period when complete data is not available for that period, where relevant data is available for MY 2006 or part thereof the Panel should consider that data.196


      2. Main arguments of third parties


          1. The European Communities argues that the Panel must examine recent relevant data reasonably available, including, where possible, data for MY 2006.197


      3. Evaluation by the Panel


        1. The claim of Brazil under Article 6.3(c) of the SCM Agreement requires an analysis by the Panel as to whether "the effect of the subsidy … is … significant price suppression". While the SCM Agreement does not contain a specific provision on the period to be considered for this purpose198, the


          image

          195 Brazil's First Written Submission, para. 49; Brazil's Rebuttal Submission, paras. 22-25; Brazil's Oral Statement, paras. 19-20; Brazil's Responses to Panel Questions 28 and 74.

          196 US First Written Submission, para. 229; US Rebuttal Submission, paras. 215-216; US Response to Panel Question 28.

          197 EC Third Party Submission, paras. 53-58; EC Third Party Oral Statement, para. 36. The European

          Communities considers that in establishing claims of "present" serious prejudice, Members are required to rely on the most recent relevant data reasonably available to them.

          198 The original panel addressed the issue of what it termed the "reference period" to be used in its analysis as follows:


          "The Panel concurs with the United States assertion that MY 2002 is a relevant year for our serious prejudice inquiry. It represents a recent period for which essentially complete data exists. The identification of 'significant price suppression' flowing

          WT/DS267/RW

          Page 61


          use of the present tense logically implies the need to make a determination with respect to the present period. The United States has put before us some data relating to MY 2006. Given that our task is to decide whether or not significant price suppression "is" the effect of the marketing loan and counter- cyclical payments at issue in this proceeding we see no reason to exclude data relating to MY 2006 to the extent that it is available.199 We note that while Brazil has argued that the Panel should rely primarily on data for MY 2005, Brazil does not contest that it is appropriate for the Panel to use data pertaining to MY 2006 to the extent that such data is reliable and credible.200 An important consideration reinforcing the need to take into account data pertaining to the period since July 2006 is that the "measure taken to comply" by the United States – the elimination of the Step 2 programme – was effective as of the beginning of MY 2006.


        2. Moreover, while our determination relates to the present period, we agree with the observation of the original panel that consideration of developments over a longer period of time "provides a more robust basis for a serious prejudice evaluation than merely paying attention to developments in a single year".201


    3. Whether marketing loan and counter-cyclical payments are specific subsidies


        1. The Panel considers that marketing loan payments and counter-cyclical payments to US upland cotton producers pursuant to the relevant provisions of the FSRI Act of 2002 are "subsidies" within the meaning of Article 1.1 of the SCM Agreement, which are "specific" within the meaning of Article 1.2 of the SCM Agreement and that Part III of the SCM Agreement therefore applies to these measures.202 We recall, in this regard, that the original panel found that the measures at issue in that case, including marketing loan payments and counter-cyclical payments, constituted "subsidies"



          image

          from 'the effect of the subsidy' calls for an evaluation of this effects-based phenomenon that cannot be conducted in the abstract. Rather, discerning adverse effects of subsidies seems to us to require reference to a recent historical period. We believe, however, that it is important for the establishment of "current" serious prejudice that such prejudice would be established to exist up to, and including, a recent point in time.


          We also believe that subsidies granted prior to MY 2002 are relevant to our evaluation. Consideration of developments over a period longer than one year, while not necessarily required (at least in Articles 5(c) and 6.3(c)) provides a more robust basis for a serious prejudice evaluation than merely paying attention to developments in a single recent year. It may put such developments in a broader temporal context. In addition, having found that the subsidies exist and have been provided over a longer period of time than one year, it would be inappropriate to look at trends only in the last year to arrive at any conclusion on serious prejudice. It may, for example, be that the market may well already be distorted in a given year due to subsidies."


          Panel Report, US – Upland Cotton, paras. 7.1198-7.1199.

          199 Indeed, the failure to take into account relevant and available data placed before us pertaining to the

          period since July 2006 would not be consistent with the requirement under Article 11 of the DSU that a panel "make an objective assessment of the matter before it, including an objective assessment of the facts of the case…"

          200 We note that the original panel, which was established on 18 March 2003, considered data covering the period MY 2002, which ended on 31 July 2003. Thus, the panel took into account data covering a period of

          4 ½ months following its establishment.

          201 Supra, footnote 198.

          202 Article 5 of the SCM Agreement applies to "any subsidy referred to in paragraphs 1 and 2 of Article 1" of the SCM Agreement. Thus in order for any measure to be an "actionable subsidy" subject to the

          provisions in Articles 5-7 of the SCM Agreement the measure must be a "specific subsidy" within the meaning of Articles 1.1 and 1.2 of the SCM Agreement.

          WT/DS267/RW

          Page 62


          within the meaning of Articles 1.1(a) and 1.1(b) of the SCM Agreement and that these subsidies were "specific" within the meaning of Article 2 of the SCM Agreement.203 It is not disputed that the provisions of the FSRI Act of 2002 concerning marketing loan and counter-cyclical payments have not been changed since the original panel conducted its analysis. The United States has not contested the argument made by Brazil in this proceeding that marketing loan and counter-cyclical payments constitute specific subsidies.204


    4. Magnitude of the marketing loan and counter-cyclical payments


        1. The original panel considered that while a serious prejudice analysis under Part III of the SCM Agreement does not require a precise quantification of the amount of a subsidy205, "the general order of magnitude of a subsidy may be a relevant consideration in a serious prejudice analysis in a particular case (where such information is probative and readily available)".206 The Appellate Body held that the magnitude of a subsidy and its relationship to prices is a relevant factor to consider in an analysis of whether the effect of a subsidy is significant price suppression207, but that Article 6.3(c) of the SCM Agreement does not oblige a panel "to quantify precisely the amount of a subsidy benefiting the product at issue in every case. A precise, definitive quantification of the subsidy is not required".208 However, the Appellate Body also considered that the original panel could have been more explicit as to what it meant by referring to "very large amounts" of subsidies.209


        2. The Panel provides below an overview of the factual information submitted by the parties with regard to the magnitude of the marketing loan and counter-cyclical payments to US upland cotton producers.


          1. Marketing loan payments


        3. The Panel considers that the figures provided by the parties regarding the marketing loan payments made by the United States to upland cotton producers in MY 2004 and MY 2005210 are in the same "general order of magnitude."211


        4. The following table compares the figures provided by Brazil and the United States on the magnitude of marketing loan payments:


          image

          203 Panel Report, US – Upland Cotton, paras. 7.1120 and 7.1154.

          204 Brazil's First Written Submission, paras. 64-78.

          205 Panel Report, US – Upland Cotton, paras. 7.1167, 7.1177 and 7.1179.

          206 Panel Report, US – Upland Cotton, footnote 1289.

          207 Appellate Body Report, US – Upland Cotton, para. 461. 208 Appellate Body Report, US – Upland Cotton, para. 467. 209 Appellate Body Report, US – Upland Cotton, para. 468.

          210 Brazil's First Written Submission, paras. 108 and 111 (Table 6) and Exhibit Bra-471; US Response

          to Panel Question 4, paras. 13-14; Brazil's Comments on US Response to Panel Question 4, para. 14 and Exhibit Bra-639.

          211 Supra, para. 10.21.

          WT/DS267/RW

          Page 63


          Table 1: Marketing Loan Payments, MY 2002-2005

          (million US$)

          Source

          MY 2002

          MY 2003

          MY 2004

          MY 2005

          Brazil's First Written Submission, Table 6

          898

          184

          1,842

          1,257

          US Response to Panel Question 4

          1,780

          1,269

          Brazil's Comments on US Response to Panel Question 4

          1,812

          1,269


        5. Brazil has also provided data on projected marketing loan payments in the period MY 2006- 2010. This data shows that the projected amount of marketing loan payments in MY 2006 is US$650 million based on USDA Budget Estimates212, US$475 million based on data from the Food and Agricultural Policy Research Institute ("FAPRI")213, and US$757 million based on February 2007 USDA stochastic projections.214 215


          1. Counter-cyclical payments


        6. The parties disagree on the magnitude of counter-cyclical payments made by the United States to upland cotton producers. This disagreement is the result of different approaches to



          image

          212 According to Table 26 of the First Written Submission of Brazil, which is based on USDA Budget Estimates, marketing loan payments are expected to amount to US$650 million in MY 2006, US$541 million in MY 2007, US$433 million in MY 2008, US$407 million in MY 2009 and US$407 million in 2010. Brazil's First Written Submission, paras. 281-282 and Exhibit Bra-456.

          213 In Table 27 of its First Written Submission Brazil provides estimates of marketing loan payments in the period MY 2006-2010 based on the July 2006 FAPRI Baseline Update for US Agricultural Markets. This

          Table indicates that the amount of marketing loan payments will be US$475 million in MY 2006, US$190 million in MY 2007, US$50 million in MY 2008, US$0 in MY 2009 and US$0 in MY 2010. Brazil's First Written Submission, para. 286 (Table 27) and Exhibit Bra-479.

          214 In response to a question of the Panel, Brazil explains that the figures in Table 27 of its First Written Submission reflect deterministic projections of outlays, which by their very nature tend to underestimate

          outlays. The most up to date (February 2007) USDA deterministic and stochastic projections of upland cotton budget outlays show that the degree of this underestimation is very significant. According to the February 2007 USDA stochastic projections, the amounts of marketing loan payments will be US$757 million in MY 2006, US$798 million in MY 2007, US$515 million in MY 2008, US$468 million in MY 2009 and US$477 million in 2010. Brazil's Response to Panel Question 89, para. 189 (Table 4). The United States submits that Brazil's response is incorrect because the figures on government outlays in Table 27 of Brazil's First Written Submission were not projected by FAPRI but were estimated by Brazil. Moreover, the values reported by FAPRI are, in fact, based on a stochastic approach. US Comments on Brazil's Response to Panel Question 89.

          215 The United States considers that the FAPRI data provided by Brazil on the projected amounts of

          marketing loan payments are entirely consistent with a market report of the International Cotton Advisory Committee and recent projections by the US Congressional Budget Office of higher world market prices for cotton in MY 2006 and MY 2007. At these higher prices, any marketing loan payments are likely to be small. US Response to Panel Question 82. Brazil contests this argument of the United States. Brazil considers that in making this argument the United States is making the mistake of relying on expected prices without taking into account the probability that actual prices will be lower than expected prices. Brazil refers in this respect to the USDA February 2007 stochastic projections of marketing loan payments, which are higher than the July 2006 FAPRI projections. Brazil argues that the February 2007 USDA projections are more reliable than the July 2006 FAPRI projections. Brazil's Comments on US Response to Panel Question 82.

          WT/DS267/RW

          Page 64


          the question of how the counter-cyclical payments at issue in this case should be allocated216 to actual production of upland cotton.


        7. Two main issues have arisen in this regard. First, as in the original proceeding, Brazil has used two methodologies for allocating counter-cyclical payments to actual production of upland cotton, referred to as "Brazil's methodology" and the "cotton-to-cotton" methodology.217 Whereas Brazil asserts that "Brazil's methodology" was endorsed in the original proceeding, the United States argues that this allocation methodology was specifically rejected by the Appellate Body. Second, the parties disagree on whether counter-cyclical payments should be allocated in accordance with the methodology found in Annex IV of the SCM Agreement.


        8. According to the First Written Submission of Brazil, "counter-cyclical payments allocated to actual upland cotton production" amounted to US$869 million in MY 2002, US$260 million in MY 2003, US$913 million in MY 2004 and US$914 million in MY 2005.218 In its Rebuttal Submission, Brazil asserts that, calculated in accordance with the "cotton-to-cotton" methodology, the amounts of counter-cyclical payments allocated to upland cotton were US$851.5 million in MY 2002, US$237.2 million in MY 2003, US$823.5 million in MY 2004 and US$832.0 million in MY 2005.219 In its Opening Statement, Brazil, referring to Exhibit Bra-625, submits that applying (an updated) "Brazil's methodology", counter-cyclical payments allocated to cotton production amounted to US$868 million in MY 2005 and that applying (an updated) "cotton-to-cotton" methodology, the amount of counter-cyclical payments allocated to cotton production was US$838 million in MY

      2005.220 221 Brazil asserts, in this regard, that the original panel described "Brazil's methodology" as


      image

      216 The counter-cyclical payments at issue in this case are calculated on the basis of historical cotton base acres and are not tied to actual production of upland cotton. We recall in this respect the observations of the original panel:


      "The PFC, MLA, DP and CCP programmes permit planting flexibility and do not require production. Eligibility for these payments depends on production of specific covered commodities in a base period, not the year of the payment. Payments are calculated with respect to these 'base acres', at a rate specific to each covered commodity, as described in Section VII:C of this report. Recipients who hold upland cotton base acres, in respect of which payments are calculated at the upland cotton rate, could have since ceased production of upland cotton. Recipients who hold other base acres, in respect of which payments are calculated at the rate for other specific covered commodities, could have since commenced production of upland cotton." (footnote omitted, emphasis added)


      Panel Report, US – Upland Cotton, para. 7.635.

      217 For a description of "Brazil's methodology" see Panel Report, US – Upland Cotton, para. 7.642 and footnote 836. For a description of the "cotton-to-cotton" methodology see Panel Report, US – Upland Cotton

      para. 7.641 and footnote 835.

      218 Brazil's First Written Submission, para. 111 (Table 6) and Exhibit Bra-471. Brazil explains that

      these allocated amounts of counter-cyclical payments are derived by multiplying (a) the ratio of allocated upland cotton counter-cyclical payments in MY 2002 (869.4 million) to total upland cotton counter-cyclical payments in MY 2002 (US$1.309 billion) by (b) total upland cotton counter-cyclical payments in MY 2003-2005. Brazil's First Written Submission, paras. 116-118. The figure of US$869.4 million for allocated counter-cyclical payments in MY 2002 is based on the application of "Brazil's methodology".

      219 Brazil's Rebuttal Submission, paras. 173-174, Table 5. Brazil observes that the limited data

      provided by the United States in this proceeding does not enable Brazil to update the amounts of counter- cyclical payments allocated to upland cotton based on "Brazil's methodology".

      220 Brazil's Opening Statement, para. 40.

      221 With regard to the difference between the figures mentioned in Brazil's First Written Submission, Rebuttal Submission and Opening Statement, Brazil has explained, in response to Panel Question 66, that: (i)

      the figures in Table 6 of Brazil's First Written Submission reflect an allocation according to "Brazil's Methodology" using data available at the time of the preparation of that submission; (ii) the figures in Table 5 of Brazil's Rebuttal Submission reflect an allocation according to the "cotton-to-cotton methodology" using the

      WT/DS267/RW

      Page 65


      "appropriate" and that the Appellate Body endorsed "Brazil's methodology" when it affirmed the original panel's findings.222 223


        1. Brazil submits that the argument of the United States on the allocation of counter-cyclical payments on the basis of the methodology set out in Annex IV of the SCM Agreement is inconsistent with the findings of the panel and the Appellate Body in the original proceeding. Brazil considers that the calculation provided by the United States is in any event incorrect and that, correctly applied, the amount of counter-cyclical payments allocated to cotton production in accordance with Annex IV of the SCM Agreement is US$577 million in MY 2005.224


        2. The United States submits that there is no basis for Brazil's argument that counter-cyclical payments should be allocated according to "Brazil's methodology" since that methodology was only addressed in the original panel's Peace Clause analysis and was not relied upon by the panel. Moreover, the methodology was specifically rejected by the Appellate Body.225 In response to a question of the Panel, the United States indicates that it has "tested" the figures in Table 5 of Brazil's Rebuttal Submission and that applying the "cotton-to-cotton" methodology the figures for counter- cyclical payments allocated to cotton production are US$820 million in MY 2004 and US$825 million in MY 2005.226


        3. The United States submits that under Article 6.3(c) of the SCM Agreement Brazil must identify the subsidized product for each of the types of subsidies from which it claims serious prejudice.227 In this regard, the United States considers that in the case of counter-cyclical payments, which are not tied to the production or sale of any particular product, Annex IV of the SCM Agreement contains an appropriate, economically neutral, allocation methodology.228 The amount of counter-cyclical payments allocated to upland cotton production in accordance with this methodology is US$386.6 million in MY 2005.229


        4. The following tables compare the results of the various calculations provided by the parties of the amounts of counter-cyclical payments to US upland cotton producers.


          image

          data in the table in paragraph 224 of the US First Written Submission; (iii) the figures in paragraph 40 of Brazil's Opening Statement reflect allocations according to, respectively, "Brazil's methodology" and the "cotton-to-cotton methodology" using the updated data provided by the United States in Exhibit US-64; and (iv) the figures in Table 5 of Brazil's Rebuttal Submission do not supersede the figures in Table 6 of Brazil's First Written Submission.

          222 Brazil's Comments on US Response to Panel Question 3, para. 7; Brazil's Opening Statement, para.

          40. Brazil refers to paragraph 469 of the Appellate Body Report as support for this contention.

          223 Brazil argues that while the methodology applied by the Appellate Body might have related to

          counter-cyclical payments with respect to upland cotton base only, the finding that "support to" upland cotton exceed support decided in 1992 includes any support for upland cotton and is not limited to counter-cyclical payments for upland cotton base only. Brazil's Comments on US Response to Panel Question 3, para. 9.

          224 Brazil's Comments on US Response to Panel Question 3.

          225 US Rebuttal Submission, paras. 278-280.

          226 US Response to Panel Question 4, para. 16.

          227 US Response to Panel Question 3, para. 7. The United States refers in this regard to the statement of the Appellate Body in the original proceeding that the "subsidized product" must be properly identified for

          purposes of significant price suppression under Article 6.3(c) of the SCM Agreement and that, if challenged payments do not in fact subsidize the "subsidized product", this may undermine the conclusion that the effect of the subsidy is significant price suppression. Appellate Body Report, US – Upland Cotton, para. 472.

          228 US Response to Panel Question 3, paras. 7-11. The United States notes that Annex IV of the SCM

          Agreement is not directly applicable to Articles 5(c) and 6.3(c), but that it provides essential context as it is the only allocation methodology that Members have agreed in the SCM Agreement and that specifically deals with the question of how to allocate subsidies that are not tied to the production or sale of a given product. US Response to Panel Question 3, para. 10.

          229 US Response to Panel Question 3, para. 12.

          WT/DS267/RW

          Page 66


          Table 2A: Counter-Cyclical Payments (non-Annex IV Methodology), MY 2002-2005


          (million US$)

          Source

          MY 2002

          MY 2003

          MY 2004

          MY 2005

          Brazil's First Written Submission Table 6

          869.0

          260.0

          913.0

          914.0

          Brazil's Rebuttal Submission Table 5

          851.5

          237.2

          823.5

          832.0

          US Response to Panel Question 4

          237.0

          820.0

          825.0

          Brazil's Opening Statement

          868/838

          Brazil's Comments on US Response to Panel Question 4

          868.0


          Table 2B: Counter-Cyclical Payments (Annex IV Methodology), MY 2004-2005


          (million US$)

          Source

          MY 2004

          MY 2005

          US Response to Panel Question 3 (Annex IV methodology)

          437.3

          386.6

          Brazil's Comments on US Response to Panel Question 3 (Annex IV methodology (a))

          540

          468

          Brazil's Comments on US Response to Panel Question 3 (Annex IV methodology (b))

          577


        5. Brazil submits data on projected counter-cyclical payments in MY 2006-2015. Based on USDA Budget Estimates, allocated counter-cyclical payments to US upland cotton producers are projected to amount to US$715 million in MY 2006, US$516 million in MY 2007 and US$449 million in MY 2008, MY 2009 and MY 2010.230 Based on the July 2006 FAPRI Baseline Update for US Agricultural Markets, projected allocated counter-cyclical payments are US$912 million in MY 2006, US$825 million in MY 2007, US$732 million in MY 2008, US$679 million in MY 2009 and US$686 million in MY 2010.231


        6. The Panel first addresses the argument of the United States that counter-cyclical payments calculated on the basis of cotton base acres should be allocated to cotton production in accordance with Annex IV of the SCM Agreement.


        7. We recall that the original panel and the Appellate Body rejected the view that Article 6.3 of the SCM Agreement requires a precise quantification of the subsidy and that the allocation methodology provided for in Annex IV of the SCM Agreement should be used in this connection.232


          image

          230 Brazil's First Written Submission, para. 282 (Table 26).

          231 Brazil's First Written Submission, para. 286 (Table 27).

          232 Panel Report, US – Upland Cotton, paras. 7.1159-7.1190; Appellate Body Report, US – Upland Cotton, paras. 461-473. We note that the original panel discussed the issue of whether and how subsidies

          should be quantified and allocated in an analysis under Article 6 of the SCM Agreement in response to an argument of the United States that Brazil had failed to properly identify "the subsidized product for each of the types of subsidies from which it claims serious prejudice". Panel Report, US – Upland Cotton, para. 7.1159. In

          WT/DS267/RW

          Page 67


          We see no basis to accept in this proceeding the same argument of the United States that was rejected by the panel and the Appellate Body in the original proceeding. We also note that the methods actually relied upon by the panel and the Appellate Body to determine the amounts of counter-cyclical payments allocated to upland cotton production were not based on the methodology set out in Annex IV of the SCM Agreement. Thus, we reject the argument of the United States that for the purposes of this proceeding the amount of counter-cyclical payments benefiting upland cotton production should be determined by using the allocation methodology provided for in Annex IV of the SCM Agreement.


        8. With respect to the disagreement between the parties as to the appropriateness of using "Brazil's methodology" to allocate the cotton base acres counter-cyclical payments to actual production of upland cotton, we consider that it is clear from the Appellate Body report in the original proceeding that the Appellate Body did not endorse that methodology. First, in the context of its analysis of issues raised with respect to the application of Article 13 (b)(ii) of the Agreement on Agriculture, the Appellate Body made the following observations:


          "We underline that these Panel findings do not pertain to all payments to current producers of upland cotton, but rather are limited to payments to producers with respect to historic upland cotton base acres. Indeed, we see little in the Panel's finding or on the record that would allow us to discern a link between the support- conferring measures with respect to non-cotton historical base acres and current production of upland cotton. We do not, therefore, accept the methodology submitted by Brazil that included, in the Article 13(b)(ii) calculation, payments with respect to both cotton and non-cotton base acres flowing to current production of upland cotton. We believe that only the 'cotton to cotton' methodology, included by the Panel in 'Attachment to Section VII:D' to its Report as an 'appropriate' alternative calculation, sufficiently demonstrates a discernible link between payments under base acre dependent measures (related to upland cotton) and upland cotton."233 (footnotes omitted, original emphasis)


        9. Thus, in upholding the panel finding that the US domestic support measures in question granted support to upland cotton in excess of that decided during MY 1992234, the Appellate Body


          image

          our view, the considerations which led the original panel to reject the argument that Brazil had failed to properly identify the product benefiting from the subsidies at issue remain relevant.

          233 Appellate Body Report, US – Upland Cotton, para. 380. In paragraph 384 of the Report, the Appellate Body observed:


          "For the reasons stated above, we conclude that payments with respect to upland cotton base acres to producers currently growing upland cotton under the production flexibility contract, market loss assistance, direct payment and counter-cyclical payment measures, calculated in accordance with the 'cotton to cotton' methodology, are support granted to the specific commodity upland cotton in the sense of Article 13(b)(ii) of the Agreement on Agriculture."


          Finally, the Appellate Body stated that


          "In our review above, we have concluded that, for purposes of the comparison envisaged by Article 13(b)(ii), the values of the four measures, namely, production flexibility contract payments, market loss assistance payments, direct payments and counter-cyclical payments in the years 1999, 2000, 2001, and 2002 are properly determined by using the 'cotton to cotton' methodology, and we have therefore modified the Panel's findings in this regard...."


          Appellate Body Report, US – Upland Cotton, para. 392.

          234 Appellate Body Report, US – Upland Cotton, para. 393.

          WT/DS267/RW

          Page 68


          calculated the amount of counter-cyclical payments to upland cotton in accordance with the "cotton- to-cotton" methodology.235


        10. Second, in reviewing the serious prejudice finding of the original panel, the Appellate Body upheld the panel's conclusion regarding the amount of the subsidies at issue. In this respect, the Appellate Body referred inter alia to paragraph 7.641 of the original panel report, which is the paragraph that contains figures reflecting the "cotton-to-cotton" methodology.236 The Appellate Body referred to the same figures as support for its observation that even using certain alternative allocation methodologies, the panel's conclusion regarding the general magnitude of price-contingent subsidies stood.237


        11. The Panel considers that Brazil has not presented any persuasive argument as to why it would be appropriate to apply in this proceeding an allocation methodology that was specifically rejected by the Appellate Body in the original proceeding.


        12. In light of the above considerations, the Panel concludes that the amount of counter-cyclical payments allocated to upland cotton production must be calculated based on the "cotton-to-cotton" methodology. The information before us indicates that in MY 2005 this amount was US$838 million.238


        13. Finally, the Panel notes that the parties have also made arguments on the likely impact of the removal of Step 2 payments on the amounts of marketing loan and counter-cyclical payments in future years. The Panel addresses that issue below in Section X.B.11 of this Report.


    5. Product at issue, like product, relevant "market" and "price"


        1. The Panel notes that the parties agree that the product at issue for purposes of Brazil's claims of "present" serious prejudice is upland cotton lint.239 Brazil recalls that the original panel had found that US upland cotton is like Brazilian upland cotton and submits evidence to demonstrate that this finding remains valid.240 The United States has not contested Brazil's position on this issue.241


          image

          235 Appellate Body Report, US – Upland Cotton, Annex 2, Tables 1-3.

          236 Appellate Body Report, US – Upland Cotton, para. 468 and footnote 631.

          237 Appellate Body Report, US – Upland Cotton, para. 469 and Annex 2, Table 3. The notes to the

          Tables in Annex 2 state explicitly that "[t]he values of production flexibility contract payments, market loss assistance payments, direct payments, and counter-cyclical payments are based on the 'cotton to cotton' methodology, discussed supra, paras. 377-380. Figures are drawn from Panel Report, para. 7.641". Thus, we disagree with Brazil that the Appellate Body's observations in paragraph 469 amount to an endorsement of "Brazil's methodology". Brazil's Opening Statement, para. 40.

          238 Exhibit Bra-625 and US Response to Panel Question 55. The Panel notes that this question was

          posed to the United States with the explanation that the Panel "is not asking whether the US agrees with these [i.e. 'Brazil's methodology' and the 'cotton-to-cotton'] methodologies".

          239 US Response to Panel Question 3, para. 9; Brazil's Comments on US Response to Panel Question 3,

          para. 4. The Panel recalls in this respect that the original panel explained that, as used in the panel's report, the term "upland cotton" ordinarily refers to upland cotton lint. Panel Report, US – Upland Cotton, footnote 258. It is the Panel's understanding that in this proceeding Brazil and the United States also use the term "upland cotton" to refer to upland cotton lint.

          240 Brazil's First Written Submission, para. 80 and Annex II ("Declaration by Andrew Macdonald").

          241 The original panel observed that it was not entirely clear from the text of Article 6.3(c) of the SCM

          Agreement whether the requirement to establish a "subsidized product" and a "like product" applies to a price suppression analysis. The panel nevertheless decided to address these concepts because its "examination of 'prices' in the world market necessarily relate[d] to 'prices' of certain 'products' ". Panel Report, US – Upland Cotton, footnote 1333. The Appellate Body noted that since the panel and the participants agreed that US upland cotton and Brazilian upland cotton must be "in the same market" it was not necessary to decide whether

          WT/DS267/RW

          Page 69


        2. Under Article 6.3(c) of the SCM Agreement, the Panel must determine whether the effect of the subsidy is significant price suppression "in the same market". Brazil submits that in this case the relevant market for purposes of Article 6.3(c) is the world market for upland cotton. We note that the original panel found that the term "same market" in Article 6.3(c) can be interpreted "as including a geographical area which may embrace the entire world"242 and that such a "world market" did, in fact, exist in the case of upland cotton.243 The Appellate Body upheld these conclusions of the panel.244 We note that the United States does not contest that the relevant market for purposes of our analysis under Article 6.3(c) is the world market for upland cotton. Therefore, in determining whether the effect of the subsidies at issue is significant price suppression "in the same market", we will treat the world market for upland cotton as the relevant market.


        3. As pointed out by Brazil, the original panel found that the relevant "price" for purposes of its analysis under Article 6.3(c) was the world market price for upland cotton, as represented by the A- Index.245 246 Brazil submits that this Panel should adopt the same approach and use the A-Index as reflecting the world market price for upland cotton. The Panel finds the arguments of Brazil in this regard persuasive.247 The Panel also notes that the United States has not contradicted Brazil's arguments regarding the existence of a world market price for upland cotton and the A-Index as representing this world market price.248 In light of these considerations, the Panel finds that a world "market" exists for upland cotton and that the A-Index represents the "price" relevant to the analysis of price suppression under Article 6.3(c) of the SCM Agreement.249


          image

          the "like product" requirement applied to a price suppression analysis. Appellate Body Report, US – Upland Cotton, para. 407 and footnote 453. The panel in Korea – Commercial Vessels concluded that the concept of "like product" does not apply to an analysis of price suppression or price depression under Article 6.3(c) of the SCM Agreement. Panel Report, Korea – Commercial Vessels, paras. 7.538-7.560.

          242 Panel Report, US – Upland Cotton, para. 7.1243. See also Panel Report, Korea – Commercial Vessels, paras. 7.564-7.565.

          243 Panel Report, US – Upland Cotton, paras. 7.1246-7.1247.

          244 Appellate Body Report, US – Upland Cotton, paras. 410-411.

          245 Panel Report, US – Upland Cotton, para. 7.1274.

          246 The original panel described the "A-Index" as "a composite of an average of the five lowest price

          quotes from a selection of the principal upland cottons traded in the world market obtained by Cotlook, a private UK-based organization". The panel clarified in a footnote that "the prices are quoted c.i.f. Northern Europe". Panel Report, US – Upland Cotton, paras. 7.1264 and footnote 1374. Brazil has explained in this proceeding that in March 2003 the basis for the calculation of the A-Index was changed to reflect the fact that Asian importers now account for three quarters of total imports of upland cotton. The A-Index proper now reflects values for delivery c.i.f. to Far Eastern ports (A(FE)Index). Cotlook continues to publish an A(NE)Index which reflects values for delivery c.i.f. Northern European ports. Brazil submits that the A(FE)Index and the A(NE)Index are almost perfectly aligned and that the technical change in the A-Index does not affect the conclusion that the A-Index represents the world market price for upland cotton. Brazil's First Written Submission, paras. 85-86 and Annex II (Declaration by Andrew Macdonald), footnote 4.

          247 Brazil's First Written Submission, para. 84.

          248 The United States has argued in this proceeding that the fact that the A-Index had to be changed to

          reflect the growing importance of China in the world market supports the US argument that price trends for upland cotton cannot be explained without taking into account the influence of China. US First Written Submission, paras. 312-313. However, the United States has not contested that the A-Index represents the world market price of upland cotton.

          249 We also note that the Appellate Body held in the original proceeding that it was sufficient for the original panel to analyze the price of upland cotton in general in the world market and that the panel was not

          required to find significant price suppression specifically with respect to the price of Brazilian upland cotton in the world market. Appellate Body Report, US – Upland Cotton, paras. 415-418.

          WT/DS267/RW

          Page 70


    6. Significant price suppression as the effect of the subsidy


      1. The original panel defined "price suppression" as referring to "the situation where 'prices' – in terms of the 'amount of money set for sale of upland cotton' or the 'value or worth' of upland cotton – either are prevented or inhibited from rising (i.e. they do not increase when they otherwise would have) or they do actually increase, but the increase is less than it otherwise would have been".250 The Appellate Body ruled that this description of the meaning of "price suppression" reflected the ordinary meaning of that term.251 252 In this proceeding neither party has contested this definition of "price suppression".


      2. The Panel considers that, as discussed in the Appellate Body report in the original proceeding253, a determination of whether price suppression exists is difficult to separate from an analysis of the effect of the subsidy. Brazil has explained that in light of the comments of the Appellate Body in the original proceeding254, it has adopted a "unitary" approach, which does not separate the question of the existence of significant price suppression from the issue of the causal relationship between significant price suppression and the subsidies at issue.255 The United States has not questioned this "unitary" approach in this proceeding.256 Therefore, while we note that the original panel determined in three separate analytical steps that (i) there was price suppression in the world market for upland cotton, (ii) this price suppression was significant and (iii) a causal relationship existed between this significant price suppression and certain price-contingent subsidies257, in this proceeding we will not make a determination as to whether significant price


        image

        250 Panel Report, US – Upland Cotton, para. 7.1277.

        251 Appellate Body Report, US – Upland Cotton, para. 424.

        252 Similarly, the panel in Korea – Commercial Vessels noted that the parties in that dispute understood

        the term "price suppression" in Article 6.3(c) of the SCM Agreement "to refer to the situation where prices have not increased when, or have increased less than, they otherwise would have". Panel Report, Korea – Commercial Vessels, para. 7.533.

        253 Appellate Body Report, US – Upland Cotton, para. 433. ("...the ordinary meaning of the transitive verb 'suppress' implies the existence of a subject (the challenged subsidies) and an object (in this case, prices in

        the world market for upland cotton). This suggests that it would be difficult to make a judgment on significant price suppression without taking into account the effect of the subsidies.") See also Panel Report, Korea – Commercial Vessels, para. 7.537:


        "...the existence of a flat or declining price trend, on its own, would not be a sufficient basis on which to conclude that prices were 'suppressed' or 'depressed'. For such a conclusion to be reached, the causes of these observed trends would need to be examined. In other words, price depression is not simply a decline in prices but a situation where prices have been 'pushed down' by something. Price suppression is where prices have been restrained by something. In other words, for a finding of 'price suppression' or price depression' in the sense of SCM Article 6.3(c), there must not only be a flattened or downward price trend as a prerequisite, but in addition this trend must be the result of an exogenous factor, namely the subsidy or subsidies in question. Thus, the analysis that seems to be called for by the Agreement (by virtue of the concepts of price suppression and price depression themselves), concerns what the price movements for the relevant ships would have been in the absence of (i.e. 'but for') the subsidies at issue." (emphasis original)

        254 Appellate Body Report, US – Upland Cotton, paras. 432-434.

        255 Brazil's First Written Submission, paras. 55-58.

        256 In the original proceeding the United States argued before the Appellate Body that the panel had erred in adopting a bifurcated approach under which the original panel made a finding of price suppression

        before making a prior finding of the effect of the subsidy. Appellate Body Report, US – Upland Cotton, para.

        430. The Appellate Body rejected that argument and held that "the text of Article 6.3 (c) does not, however, preclude the approach taken by the Panel to examine first whether significant price suppression exists and then, if it is found to exist, to proceed further to examine whether the significant price suppression is the effect of the subsidy." Appellate Body Report, US – Upland Cotton, para. 431.

        257 Panel Report, US – Upland Cotton, paras. 7.1275-7.1363.

        WT/DS267/RW

        Page 71


        suppression exists separately from whether significant price suppression is the effect of the subsidies at issue.258


      3. Closely related to this, the Panel notes the argument of Brazil that an analysis of price suppression under Article 6.3(c) is necessarily of a counterfactual nature, focusing on the situation that would exist but for the subsidies at issue. The United States has not contested this argument of Brazil. We note that the original panel did not explicitly use the concept of a counterfactual analysis and that the panel did to some extent rely on an analysis of trends in prices.259 At the same time, it is clear that key elements of the panel's analysis involved an examination of the situation that would have existed but for the subsidies at issue. Thus, for example, the panel found that the fact that prices had declined from 1996 through the beginning of 2002 and increased from 2002 was not conclusive as to whether or not there was price suppression: "[w]e need to examine whether these prices were suppressed, that is, lower than they would have been without the United States subsidies in respect of upland cotton".260 In its analysis of the design, structure and operation of the subsidies at issue the panel found that certain price-contingent subsidies resulted in "lower world market prices than would prevail in their absence".261 Based on an examination of the costs of production and market revenues of US upland cotton producers, the panel concluded that:


        "United States upland cotton producers would not have been economically capable of remaining in the production of upland cotton had it not been for the United States subsidies at issue and that the effect of the subsidies was to allow United States producers to sell upland cotton at a price lower than would otherwise have been necessary to cover their total costs."262


      4. As noted above, the panel in Korea – Commercial Vessels has interpreted Article 6.3(c) of the SCM Agreement as requiring a "but for" approach to the analysis of whether the effect of a subsidy is significant price suppression.263


        image

        258 We note that the Appellate Body has stated that "Article 6.3(c) does not set forth any specific methodology for determining whether the effect of a subsidy is significant price suppression" and that "Article 6.3(c) is silent as to the sequence of steps to be followed in assessing whether the effect of a subsidy is significant price suppression". Appellate Body Report, US – Upland Cotton, paras. 427 and 431. While the Appellate Body saw no legal error in the approach taken by the original panel, it is clear that the Appellate Body did not consider that Article 6.3(c) required that approach.

        259 Panel Report, US – Upland Cotton, paras. 7. 1286-7.1288.

        260 Panel Report, US – Upland Cotton, para. 7.1288.

        261 Panel Report, US – Upland Cotton, paras. 7.1291 and 7.1295.

        262 Panel Report, US – Upland Cotton, paras. 7.1353. The panel concluded its discussion of costs of production in relation to market revenue by observing that "the existence of this gap between upland cotton

        producers' total costs and market revenue, on the one hand, and the effect of the subsidies, on the other hand, was to sustain a higher level of output than would have occurred in the absence of the United States subsidies at issue". Panel Report, US – Upland Cotton, para. 7.1354.

        263 Supra, footnote, 253. The panel also observed:


        "As noted supra, we believe that the text of Article 6.3(c) implies a 'but for' approach to causation in respect of price suppression/price depression. Price suppression is the situation where prices have been restrained by something, and price depression is the situation where prices have been pushed down by something. So the question to be answered is whether the 'something' is subsidization. Looking at a counterfactual situation, i.e. trying to determine what prices would have been in the absence of the subsidy, seems to us the most logical and straightforward way to answer this question."


        Panel Report, Korea – Commercial Vessels, para. 7.612. The panel considered that this "but for" approach was consistent with the approaches taken by the panels in Indonesia – Autos and in US – Upland Cotton.

        WT/DS267/RW

        Page 72


      5. Thus, in order to determine whether the effect of marketing loan payments and counter- cyclical payments to upland cotton producers is significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement, the Panel will determine whether, but for these subsidies, the world market price264 for upland cotton "would have increased [significantly], or would have increased by [significantly] more than was in fact the case".265


      6. The Panel concurs with the original panel that the word "significant" in "significant price suppression" in Article 6.3(c) of the SCM Agreement relates to the degree of price suppression in the context of the price that has been affected and that the significance of any degree of price suppression may vary from case to case, depending upon the factual circumstances.266 As stated by the original panel,


        "… for a basic and widely traded commodity, such as upland cotton, a relatively small decrease or suppression of prices could be significant because, for example, profit margins may ordinarily be narrow, product homogeneity means that sales are price sensitive or because of the sheer size of the market in terms of the amount of revenue involved in large volumes traded on the markets experiencing the price suppression."267


      7. Finally, like the original panel, we conduct an examination of the collective effects of the marketing loan and counter-cyclical payments.268 We recall, in this respect, that the original panel made a finding that certain price-contingent subsidies (marketing loan payments, counter-cyclical payments, Step 2 payments and market loss assistance payments) collectively were a cause of significant price suppression in the world market for upland cotton. The panel did not find that each of these four subsidies, taken individually, was a cause of significant price suppression nor did it pronounce on the relative importance of each of these subsidies in causing significant price suppression.


        1. The alleged "substantial proportionate influence" of US upland cotton production and exports


          1. Main arguments of the parties


      8. Brazil submits that US upland cotton production and exports continue to expand and exercise a "substantial proportionate influence" on prices in the world market for upland cotton. In the view of Brazil, the United States is "the single most important market influencing world market prices" of upland cotton.269 In this connection, Brazil presents data showing that in MY 2005 the US share of world upland cotton production was 20.9 per cent and the US share of the world market for exports of


        image

        264 We recall that the Appellate Body found that the original panel had not erred by not conducting a separate analysis of Brazilian prices in the world market. Appellate Body Report, US – Upland Cotton, paras. 415-418.

        265 Panel Report, Korea – Commercial Vessels, para. 7.615.

        266 Panel Report, US – Upland Cotton, paras. 7.1328-7.1329. The Panel also notes the observation made by the panel in Korea – Commercial Vessels, para. 7.571 that "only price suppression or price depression

        of sufficient magnitude or degree, seen in the context of the particular product at issue, to be able to meaningfully affect suppliers should be found to be 'significant' in the sense of SCM Article 6.3(c)".

        267 Panel Report, US – Upland Cotton, para. 7.1330.

        268 Panel Report, US – Upland Cotton, para. 7.1192. See also Panel Report, Korea – Commercial Vessels, para. 7.616.

        269 Brazil's First Written Submission, para. 89.

        WT/DS267/RW

        Page 73


        upland cotton was 39.6 per cent. During the period MY 2002-2005 the volume of US production increased by 40.7 per cent and the volume of US exports increased by 54.8 per cent.270


      9. The United States submits that Brazil's argument that the United States is "the single most important market influencing world market prices" of upland cotton is overly simplistic and inconsistent with the realities of the world cotton market. In particular, Brazil fails to account properly for the important role of China's trade in cotton.271


        1. Evaluation by the Panel


      10. The original panel relied upon what it termed a substantial proportionate influence of the United States in the world upland cotton market as one of four factors that cumulatively supported a finding of a causal link between certain of the challenged subsidies and significant price suppression.272 This substantial proportionate influence of the United States resulted inter alia from the magnitude of US production and export of upland cotton.273 The United States was the world's second largest producer, and the world's largest exporter, of upland cotton.274


      11. The Appellate Body considered that, when viewed in conjunction with other findings made by the original panel, the "substantial proportionate influence" of the United States in the world market for upland cotton seemed to support the panel's conclusion of the existence of a causal relationship between the price-contingent subsidies and significant price suppression.275

      12. The data provided by the United States276 indicates that in MY 2005 the US share of world cotton production was 20.9 per cent and that the US share of world upland cotton production averaged

    19.7 per cent during the period MY 2002-2005 (see Table 3). The US share of world exports was

    40.1 per cent in MY 2005. During the period MY 2002-2005 the US share of world exports averaged

    40.5 per cent.


    image

    270 Brazil's First Written Submission, paras. 92-93, Figures 2 and 3.

    271 US First Written Submission, paras. 308-318. US Rebuttal Submission, paras. 387-396.

    272 Panel Report, US – Upland Cotton, paras. 7.1347-7.1348.

    273 Panel Report, US – Upland Cotton, paras. 7.1285 and 7.1348.

    274 In the period MY 1999–2002, the United States accounted for approximately 20 per cent of world production of upland cotton. The share of the United States in world production of upland cotton was 19.2 per

    cent in MY 1999, 19.3 per cent in MY 2000, 20.6 per cent in MY 2001 and 19.6 per cent in MY 2002. The share of the United States in world exports of upland cotton was 23.5 per cent in MY 1999, 24.5 per cent in MY 2000, 37.3 per cent in MY 2001 and 39.9 per cent in MY 2002. Panel Report, US – Upland Cotton, paras. 7.1282-7.1283.

    275 Appellate Body Report, US – Upland Cotton, para. 449:


    "We agree that, in and of itself, the degree of influence of the United States in the world market for upland cotton may not be conclusive as to the effect of the price-contingent subsidies on prices in that market. However, if the price-contingent subsidies increased United States production and exports or decreased prices for United States upland cotton, then the fact that United States production and exports of upland cotton significantly influenced world market prices would make it more likely that the effect of the price-contingent subsidies is significant price suppression. Accordingly, this fact seems to support the Panel's conclusion when read in conjunction with its other findings."

    276 Exhibit US-27.

    WT/DS267/RW

    Page 74


    Table 3: US and World Cotton Exports and Production, MY 2002-2005


    MY

    United States

    (in 1000 480 lb bales)

    World

    (in 1000 480 lb bales)

    US Share of World Exports and Production

    Exports

    Production

    Exports

    Production

    Exports

    Production

    2002

    11,900

    17,209

    30,325

    88,251

    39.2%

    19.5%

    2003

    13,758

    18,255

    33,221

    95,267

    41.4%

    19.2%

    2004

    14,436

    23,251

    35,022

    120,394

    41.2%

    19.3%

    2005

    18,036

    23,890

    44,935

    114,142

    40.1%

    20.9%

    Average

    14,533

    20,651

    35,876

    104,514

    40.5%

    19.7%

    Source: Exhibit US-27.


      1. Given that complete data for MY 2006 is not available, the Panel has relied on projections of production and exports provided by the United States.277 These projections of US and world production and exports were made in February 2007, halfway through MY 2006. According to these projections, the absolute volumes of US cotton production and exports in MY 2006 will decline by 9 per cent and 19.6 per cent, respectively, from MY 2005 levels. Since the average "error" of USDA projections is only 0.1 per cent for US cotton production and 4 per cent for US exports278, the Panel considers that these USDA projections of a decline in production and exports in MY 2006 are very likely to materialize. Taken together with projections of world production and exports, they indicate that the US share of world upland cotton production is projected to fall from 20.9 per cent in MY 2005 to 18.6 per cent in MY 2006. Further, the US share of world exports of upland cotton is projected to fall from 40.1 per cent in MY 2005 to 36.3 per cent in MY 2006.


        image

        277 The source of the US projections of production and exports is the USDA's World Agricultural Supply and Demand Estimates (WASDE) report, which provides forecasts of supply and demand for major US and global crops and US livestock. The February 2007 projection of US production and exports for MY 2006 is contained in Exhibit US-114 while corresponding figures for world production and exports are taken from WASDE 443-27.

        278 Given that a lower degree of certainty attaches to these projections, the Panel asked the United States to provide the information on how final marketing year data on the variables of interest would

        differ from projections made in February of the same marketing year. The idea is to compare projections made in the middle of a given marketing year with the final estimate for the complete marketing year. The closer February projections are to final estimates of production and export for the same marketing year, the more reliable we judge these projections to be. The United States provided the Panel with this data spanning the period from MY 1992-2005. (Exhibit US-145) Using this information, we have estimated how much these February projections have differed from the final marketing year data:


        Table 4: Difference Between Final Estimates And Projections, MY 1992-2005


        Item

        Production

        Exports

        Average difference: (Final estimate minus February projection)

        0.1%

        4.0%

        Standard deviation of difference

        0.6%

        6.6%

        Source: Exhibit US-145.


        Over the period MY 1992-2005, projections of US production made in February underestimated the final production number by only an average of 0.1 per cent (with a standard deviation of 0.6 per cent). The corresponding underestimate for US exports is 4 per cent (with a standard deviation of 6.6 per cent). Projections of US cotton production have been more accurate than projections of US exports.

        WT/DS267/RW

        Page 75


      2. In our view, even taking into account the projections for MY 2006, the figures on the relative magnitude of US production and exports of upland cotton at present are not significantly different from the figures that formed the basis for the original panel's finding that the United States exerted a substantial proportionate influence on the world market for upland cotton. The US share of world upland cotton production was 19.6 per cent in MY 2002, 20.9 per cent in MY 2005 and is projected to be 18.6 per cent in MY 2006. The US share of world exports of upland cotton was 39.9 per cent in MY 2002, 39.8 per cent in MY 2005 and is projected to be 36.3 per cent in MY 2006. In light of these figures and taking into account other evidence presented by Brazil279, the Panel considers that the United States exerts a substantial proportionate influence on the world market for upland cotton. This substantial proportionate influence of the United States, analyzed in light of the totality of the evidence before us, as discussed in the next sub-sections, is an element that supports a finding that the marketing loan and counter-cyclical payments to US upland cotton producers cause significant price suppression in the world market for upland cotton.


        1. "Structure, design and operation" of the marketing loan and counter-cyclical payments


          1. Main arguments of the parties


      3. Brazil argues that the "structure, design and operation" of the marketing loan and counter- cyclical payments280 stimulate US upland cotton production by insulating US producers from market forces. In support of this argument, Brazil refers to the magnitude of the marketing loan payments in MY 2004 and MY 2005, the fact that the legislative provisions in question have not been changed, economic studies presented in the original proceeding, and the fact that during most of the period MY 2002-2005 the adjusted world price was below the marketing loan rate. In Brazil's view, these factors show that the conclusion of the original panel that marketing loan payments stimulate production and exports and suppress prices continues to apply.281 Similarly, the findings of the original panel on the effects of counter-cyclical payments on production, exports and prices of upland cotton also remain applicable. In this regard, Brazil refers to the magnitude of the counter-cyclical payments in MY 2004 and MY 2005, the fact that the legislative provisions in question have not been changed, economic studies presented in the original proceeding, and the continued existence of a strong positive relationship between upland cotton base acre holders receiving annual counter-cyclical payments and upland cotton production.282 Referring to the original panel's findings that price- contingent US subsidies isolated US cotton farmers from market forces, Brazil argues that data on the importance of marketing loan and counter-cyclical payments in stabilizing US upland cotton farmers' revenues demonstrates that this insulating effect continues today.283


      4. The United States argues that Brazil has not provided any persuasive evidence that the structure, design and operation of the marketing loan and counter-cyclical payments programme support Brazil's claims of significant price suppression. With respect to marketing loan payments, the United States submits that an analysis of planting decisions made by US upland cotton producers in light of the conditions that exist at the time of planting (January-March) shows that for the present marketing year, MY 2006, payments under the marketing loan programme did not distort production decisions. Since the average New York futures price in January-March 2006 was well above the marketing loan rate, US upland cotton producers did not expect to receive marketing loan payments


        image

        279 The Panel notes, in particular, the evidence provided by Brazil on the similarity in trends in certain US domestic prices and the world market price for upland cotton (Brazil's First Written Submission, Figures 4 and 5), a factor also relied upon by the original panel. Panel Report, US – Upland Cotton, para. 7.1311. The Panel also notes the analysis prepared by Brazil's consultant of the functioning of the world market for upland cotton. Brazil's First Written Submission, Annex II (Declaration by Andrew Macdonald).

        280 Supra, footnote 186.

        281 Brazil's First Written Submission, paras. 123-127; Brazil's Rebuttal Submission, paras. 101-110. 282 Brazil's First Written Submission, paras. 128-131. Brazil's Rebuttal Submission, paras. 111-118. 283 Brazil's First Written Submission, paras. 133-136 and Figure 8.

        WT/DS267/RW

        Page 76


        upon harvest, and these payments could not have figured in their planting decisions for MY 2006. 284 The United States submits, in this respect, that even in earlier marketing years planting decisions of US upland cotton producers were not shaped by expectations about marketing loan payments but by market factors. With respect to counter-cyclical payments, the United States argues that (i) research based on recent empirical data does not support Brazil's claim that counter-cyclical payments have significant trade-distorting effects; (ii) significant amounts of decoupled payments are capitalized into higher land values; and (iii) data comparing base acres to planted acreage does not demonstrate that counter-cyclical payments have significant effects on production.285 The United States also submits that Brazil does not demonstrate that the size of the outlays under the marketing loan and counter- cyclical payment programmes support the existence of a causal link.286


        1. Evaluation by the Panel


      5. The question before the Panel is whether the structure, design and operation of marketing loan payments and counter-cyclical payments provided by the United States pursuant to the FSRI Act of 2002 support a finding that the effect of these subsidies is "present" significant price suppression in the world market for upland cotton within the meaning of Article 6.3(c) of the SCM Agreement.


      6. Because Brazil asserts that, based on the evidence which it has submitted in this proceeding, this Panel should reach the same conclusions as the original panel regarding the price suppressing effects of the structure, design and operation of the marketing loan payments and counter-cyclical payments, we begin by providing a summary of the original panel's analysis that formed the basis for those conclusions.


      7. The original panel first examined the "nature" of the subsidies in determining whether price suppression existed in the world market for upland cotton.287 The panel considered that the "nature of the US subsidies at issue – in terms of their structure, design and operation – is relevant in assessing whether or not they have price suppressing effects".288 The panel attached particular importance to the "directly price-contingent" nature of certain subsidies.289


      8. Regarding marketing loan payments, the original panel pointed out that under the marketing loan programme benefits to producers resulted when the adjusted world market price was below the marketing loan rate and that the payments were mandatory290, that the magnitude of the marketing loan payments was counter-cyclical and increased with the difference between the adjusted world price and the marketing loan rate291, and that the marketing loan payments insulated the revenue of US


        image

        284 US First Written Submission, paras. 226-235; US Rebuttal Submission, paras. 268-275. 285 US First Written Submission, paras. 207-225; US Rebuttal Submission, paras. 224-267. 286 US First Written Submission, paras. 236-241; US Rebuttal Submission, paras. 267-280.

        287 Panel Report, US – Upland Cotton, paras. 7.1289-7.1307. In determining whether price suppression

        existed in the world market for upland cotton, the panel also considered the relative magnitude of the US production and exports and the general price trends. Panel Report, US – Upland Cotton, paras. 7.1281-7.1288.

        288 Panel Report, US – Upland Cotton, para. 7.1289.

        289The panel explained that in light of certain guiding considerations outlined earlier in its report, the

        distinction between "directly price-contingent" subsidies and other subsidies before it was critical "in terms of the nexus which the subsidies have to any price suppression and to the subsidized product". Thus, the panel considered it highly relevant that certain US subsidies under consideration, including marketing loan programme payments and counter-cyclical programme payments were "directly linked to world market prices". Panel Report, US – Upland Cotton, paras. 7.1289-7.1290.

        290 Panel Report, US – Upland Cotton, para. 7.1291.

        291 Panel Report, US – Upland Cotton, paras. 7.1292-7.1294. The panel also observed that the data before it showed that "the marketing loan subsidy by the United States government, at certain points, was

        greater than the market value of the product" and that "it accounted for more than half of the difference between the adjusted world price and the marketing loan rate (and thus of United States upland cotton producer revenue)". Panel Report, US – Upland Cotton, para. 7.1294 (original emphasis).

        WT/DS267/RW

        Page 77


        upland cotton producers from a decline in the adjusted world price.292 The panel considered that the marketing loan payments "stimulate production and exports and result in lower world market prices than would prevail in their absence"293 and that the "structure, design and operation of the marketing loan programme has enhanced production and trade-distorting effects".294 The panel also observed that data for the period MY 1999-2002 showed that, because of the impact of marketing loan payments on production of upland cotton, a decline in US prices did not result in a decline in US production or export, that the structure of the measure affected world market prices generally, and that for most of the period under consideration the adjusted world market price was below the marketing loan rate.295 Finally, the panel noted that the amount of the subsidies was "very large".296


      9. Also as part of its examination of whether price suppression existed, the original panel stated the following with respect to counter-cyclical payments:


        "We agree with the view of USDA economists that, due to their market-price contingency, [counter-cyclical payments] may influence production decisions indirectly by reducing total and per unit revenue risk associated with price variability in some situations. In the price range from the loan rate up to the target price minus the [direct payments] rate, changes in producer revenues due to changes in market prices are partly offset by the countercyclical payments if the base acreage crop is planted, thereby reducing total revenue risk associated with price variability. We have confirmed a strong positive relationship between upland cotton (base acre) producers receiving annual payments and upland cotton production. Moreover, the payments are mandatory, under certain market conditions."297 (footnotes omitted)


      10. It is clear from these statements that a key element in the original panel's analysis of the structure, design and operation of marketing loan payments and counter-cyclical payments was the idea that the price-contingent nature of these subsidies made US upland cotton production less responsive to market prices by insulating the revenue of US upland cotton producers from declines in prices. It is also clear, however, that closely related to this, the panel attached importance to the magnitude of these subsidies in the period under consideration.


      11. The original panel highlighted these two themes in its conclusion on the existence of price suppression:


        "As we have just indicated, several of the United States subsidies are directly linked to world prices for upland cotton, thereby numbing the response of United States producers to production adjustment decisions when prices are low. We have readily available information on the record showing us that the price-contingent subsidies in question involve very large amounts of United States government money benefiting United States upland cotton. In our view, the collective operation of these subsidies was akin to a very large, counter cyclical, deficiency payment laced with additional enhancements. We believe that the structure, design and operation, particularly of the


        image

        292 Panel Report, US – Upland Cotton, para. 7.1294: "The further the adjusted world price drops, the greater the extent to which United States upland cotton producers' revenue is insulated from the decline, numbing United States production decisions from world market signals."

        293 Panel Report, US – Upland Cotton, paras. 7.1291 and 7.1295.

        294 Panel Report, US – Upland Cotton, paras. 7.1295.

        295 Panel Report, US – Upland Cotton, paras. 7.1294 and 7.1296.

        296 Panel Report, US – Upland Cotton, para. 7. 1297.

        297 Panel Report, US – Upland Cotton, para. 7.1302.

        WT/DS267/RW

        Page 78


        price-contingent subsidies, constitutes strong evidence supporting a finding of price suppression."298 299 (emphasis added)

      12. Having found that price suppression existed300 and that this price suppression was significant301, the original panel also referred to the structure, design and operation of the price- contingent subsidies at issue in its examination of whether this significant price suppression was the effect of certain of the subsidies at issue. The panel again emphasized the "insulating" effect and the magnitude of the price-contingent subsidies at issue:


        "Second, we recall our examination of the nature of the United States subsidies at issue. In particular, we recall that several of the United States subsidies – marketing loan programme payments, the user marketing (Step 2) payments, MLA payments and CCP payments – are directly linked to world prices for upland cotton, thereby insulating United States producers from low prices. We believe that the structure, design and operation of these three measures constitutes evidence supporting a causal link with the significant price suppression we have found to exist. Furthermore, while we do not believe that it is strictly necessarily to calculate precisely the amount of the subsidies in question, we observe that we have readily available information on the record showing us that the price-contingent subsidies in question involve very large amounts of United States government money benefiting United States upland cotton production."302 (emphasis added, footnote omitted)


      13. By contrast, the original panel considered that Brazil had not established that significant price suppression was the effect of certain subsidies – production flexibility payments, direct payments and crop insurance payments – that were not price-contingent. The panel considered that these non-price contingent subsidies were "more concerned with income support than directly with world price effects" and were "of a different nature, and thus effect" than the price-contingent subsidies.303


      14. The Appellate Body upheld the original panel's finding that the effect of certain price- contingent subsidies was significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement constituting serious prejudice to the interests of Brazil within the meaning of Article 5(c) of the SCM Agreement, and in so doing rejected various arguments of the United States, including in respect of the panel's analysis of the structure, design and operation of the marketing loan and counter-cyclical subsidies.304


      15. It is not in dispute that the conditions of application of the marketing loan and counter- cyclical payments, as defined in the relevant provisions of the FSRI Act of 2002, are the same at present as when the original panel analyzed these measures. Therefore, there has been no change with respect to the mandatory, price-contingent and counter-cyclical nature of these subsidies.


      16. Apart from the fact that the provisions of the FSRI Act of 2002 regarding marketing loan payments and counter-cyclical payments have remained unchanged, Brazil refers to several other factors, including the magnitude of these subsides in recent years, in light of which Brazil considers that this Panel should reach the same conclusions as the original panel regarding the effects of the


        image

        298 Panel Report, US – Upland Cotton, para. 7.1308.

        299The panel considered that the parallelism between declining US price and world market prices in the period under consideration was evidence that increased US production and supply had caused price suppression.

        Panel Report, US – Upland Cotton, paras. 7.1309-7.1311.

        300 Panel Report, US – Upland Cotton, para. 7.1312. 301 Panel Report, US – Upland Cotton, para. 7.1333. 302 Panel Report, US – Upland Cotton, para. 7.1349.

        303 Panel Report, US – Upland Cotton, paras. 7.1307 and 7.1350.

        304 Appellate Body Report, US – Upland Cotton, paras. 395-496.

        WT/DS267/RW

        Page 79


        structure, design and operation of marketing loan payments and counter-cyclical payments on production, exports and prices of US upland cotton.305


      17. The United States does not contest that the effect of marketing loan and counter-cyclical payments on revenues of US upland cotton producers could affect acreage and production of upland cotton. However, the United States argues that the relevant question is whether, taking into account the particular structure and design of the marketing loan and counter-cyclical payment programmes and the way that they operate under marketing conditions at present, payments under the programmes are in fact having production effects, and if so, whether the degree of the effect is so significant that it is resulting in "present" significant price suppression. The United States submits that in this proceeding Brazil has failed to demonstrate, through a proper analysis of the factors actually considered by US upland cotton producers in their annual planting decisions and of the way that marketing loan and counter-cyclical payments interact with these planting decisions, that the effect of marketing loan and counter-cyclical payments is to induce US upland cotton producers to increase the level of upland cotton acreage above what it would have been in the absence of those payments. The United States emphasizes in this respect that a proper analysis of annual planting decisions of upland cotton producers must take into account the fact that producers must base their planting decisions on expectations with regard to factors such as weather, demand, supply, prices of upland cotton and other crops, and variable costs. Moreover, the United States considers that Brazil has not demonstrated that the degree of any effect of marketing loan and counter-cyclical payments on production is such as to cause "present" significant price suppression.306


      18. In our view, the United States made the same arguments in the original proceeding. For example, before the original panel the United States argued, with regard to marketing loan payments, that "Brazil has failed to make a proper analysis of the conditions actually faced by producers in making production decisions using futures prices, which reveals no expected impact from marketing loans except for MY 2002".307 In this connection, the United States also submitted that "[i]n no marketing year from 1999-2001 would the marketing loan rate have impacted planting decisions and in marketing year 2002 the marketing loan rate did not prevent price-responsive acreage adjustment".308


      19. As noted above, the original panel considered, in respect of marketing loan payments, that "the payments stimulate production and exports and result in lower world market prices than would prevail in their absence" and that "the structure, design and operation of the marketing loan programme has enhanced production and trade-distorting effects".309 It is clear from these statements that the panel did not consider that the fact that in certain years expected market prices at the time of planting exceeded the marketing loan rate meant that marketing loan payments had only minimal effects on production in those years.310 However, while the United States argued before the Appellate Body that "the Panel failed to address the relevant economic decision faced by United States upland


        image

        305 Supra, para. 10.59.

        306 E.g. US Rebuttal Submission, paras. 222 and 287-291; US Opening Statement paras. 62-64; US Reponses to Panel Questions 51, 53 and 54. US Comments on Brazil's Response to Panel Question 62.

        307 US Further Rebuttal Submission (18 November 2003), Section IV.G.

        308 US Further Rebuttal Submission (18 November 2003), Section IV.G.6.

        309 Panel Report, US – Upland Cotton, paras. 7.1291 and 7.1295.

        310 The United States interprets these statements to mean that the panel noted "that the particular design and operation of the marketing loan program is such that it may have more than non- or minimally-trade-

        distorting effects given the particular market conditions that prevail." US Response to Panel Question 54, para. 105. While we do not disagree that the panel's analysis must be understood in the context of the "particular market conditions" that prevailed in the period considered by the panel, it is clear that the panel did not adopt the particular methodology for analyzing the relevance of market conditions advocated by the United States in the original proceeding.

        WT/DS267/RW

        Page 80


        cotton farmers at the time of planting, namely, the decision of whether to plant upland cotton or alternative crops (and how much of each)" 311, the Appellate Body rejected this argument.312


      20. We also note that specifically with regard to the argument of the United States that marketing loan payments have minimal effects on planting decisions if the expected market price at the time planting decisions are made is above the marketing loan rate, the Appellate Body observed:


        "During the oral hearing, the United States presented data to show that, when planting decisions were made for the 1999, 2000, 2001, and 2003 upland cotton crops, the expected upland cotton price upon harvest was higher than the marketing loan rate. Accordingly, the United States contends, the marketing loan program payments would have had only a minimal effect on planting decisions, because farmers would have expected to receive a higher price from the sale of their upland cotton and no marketing loan program payments.


        We note, based on the evidence provided by the United States, that, for four of the five upland cotton crops between 1999 and 2003, the expected harvest price at the time of making planting decisions was always substantially higher than the actual price realized at the time of harvest of the crop. This suggests that although farmers had expected higher prices in making their planting decisions, they were also aware that if actual prices were ultimately lower, they would be 'insulated' by government support, including not only marketing loan program payments but also counter- cyclical payments, which were based on a target upland cotton price of 72.4 cents per pound. We are therefore satisfied that the Panel adopted a plausible view of the facts in connection with expected prices and planting decisions, even though it attributed to these factors a different weight or meaning than did the United States. As the Appellate Body has said, it is not necessary for panels to 'accord to factual evidence of the parties the same meaning and weight as do the parties'."313 (emphasis original, footnotes omitted)


      21. Thus, even if expected market prices at the time of planting are higher than the marketing loan rate and producers do not expect to receive marketing payments at that time314, the availability of marketing loan payments can be considered to influence planting decisions. While US upland cotton producers may be uncertain at the time of planting as to whether actual harvest prices will be below the marketing loan rate, there is no uncertainty as to the fact that, if actual harvest prices are below the marketing loan rate, they will receive marketing loan payments. Therefore, the logic of the original panel's analysis regarding the revenue-stabilizing impact of the price-contingent, counter-cyclical nature of marketing loan payments and the consequent impact on the level of production applies even


        image

        311 Appellate Body Report, US – Upland Cotton, para. 440. In arguing that the original panel had ignored the relevant production decision faced by US cotton farmers at the time of planting, the United States referred to the panel's conclusion that the price-contingent subsidies are linked to world prices for upland cotton "thereby numbing the response of United States producers to production adjustment decisions when prices are low". Appellate Body Report, US – Upland Cotton, para. 38.

        312 Appellate Body Report, US – Upland Cotton, paras. 440-445. The Appellate Body considered that

        the original panel had, in fact, addressed upland cotton planting decisions, expected prices and expected market revenue in paragraph 7.1362 of its report. Appellate Body Report, US – Upland Cotton, para. 441.

        313 Appellate Body Report, US – Upland Cotton, paras. 444-445.

        314 We note that Brazil disagrees with the United States that in most recent years US upland cotton

        producers have not expected to receive marketing loan payments. See Brazil's Rebuttal Submission, paras. 104-105.

        WT/DS267/RW

        Page 81


        if one accepts that, as argued by the United States, it is necessary to take into account producers' price expectations.315


      22. We note that the Appellate Body also upheld the original panel's conclusion in paragraph 7.1349 of the panel report that the fact that certain subsidies were directly related to the world market prices for upland cotton supported a finding of the existence of a causal relationship between these subsidies and significant price suppression. In that context, the Appellate Body reviewed and affirmed the conclusions drawn by the original panel from its earlier assessment of the structure, design and operation of the price-contingent subsidies, including the panel's statements concerning the effects of these subsidies on production and exports.316


      23. The United States argues that Article 11 of the DSU requires this Panel to carry out an "objective assessment of the matter before it" and that "there is no basis to automatically adopt for purposes of this proceeding the same factual conclusions drawn by the original panel about the effects of the marketing loan, counter-cyclical payment, and Step 2 payments in a different period".317


      24. We note, in this regard, that in its discussion of the argument of the United States concerning the role of expected market prices, the Appellate Body referred to the fact that in four of the five years under consideration the actual market price at the time of harvest was below the expected market price at the time of planting.318 The evidence shows that this pattern has persisted in more recent years.319 The Panel also has evidence before it, provided by Brazil but not contested by the


        image

        315 Brazil does not disagree that it is "important to look at the way that the marketing loan and counter- cyclical payments programs actually operate and interact with production decisions". Brazil's Comments on US Response to Panel Question 51, para. 71. Brazil also agrees that the expectations of US cotton producers are a relevant factor to consider in an assessment of the effects of the marketing loan and counter-cyclical payments. Brazil's Comments on US Response to Panel Question 51, paras. 70-80.

        316 Appellate Body Report, US – Upland Cotton, para. 450:


        "The second reason the Panel provided for finding a 'causal link' was its view that the price- contingent subsidies 'are directly linked to world prices for upland cotton'. This conclusion flowed from the Panel's earlier assessment—in connection with its analysis of significant price suppression—of the nature of the price-contingent subsidies. The nature of a subsidy plays an important role in any analysis of whether the effect of the subsidy is significant price suppression under Article 6.3(c). With respect to marketing loan program payments, the Panel found that '[t]he further the adjusted world price drops, the greater the extent to which United States upland cotton producers' revenue is insulated from the decline'. As a result, during the 1999-2002 marketing years, United States production and exports remained stable or increased, even though prices of United States upland cotton decreased. The Panel found that Step 2 payments stimulate domestic and foreign demand for United States upland cotton by 'eliminating any positive difference between United States internal prices and international prices of upland cotton'. The Panel stated that Step 2 payments 'result in lower world market prices than would prevail in their absence'. Finally, the Panel found that market loss assistance payments and counter-cyclical payments are made in response to low prices for upland cotton and stimulate United States production of upland cotton by reducing the 'total and per unit revenue risk associated with price variability'. The United States contends that the Panel's analysis of the price-contingent subsidies was 'deficient'. However, the Panel found that the price-contingent subsidies stimulated United States production and exports of upland cotton and thereby lowered United States upland cotton prices. This seems to us to support the Panel's conclusion that the effect of the price-contingent subsidies is significant price suppression." (original emphasis, footnotes omitted)

        317 US Response to Panel Question 54, para. 102.

        318 Supra, para. 10.76.

        319Brazil submits that a comparison of the adjusted world price and the average December price of the futures contract at the time of planting shows that in the period 1996-2005 the adjusted world price was on

        WT/DS267/RW

        Page 82


        United States, that in most recent years the adjusted world price has been below the marketing loan rate.320


      25. Given that in most recent years actual market prices have been lower than expected market prices at the time of planting and that the adjusted world price has been below the marketing loan rate, the Panel considers that it is reasonable to conclude that the fact that US cotton producers know that they will receive marketing loan payments whenever the adjusted world price is below the marketing loan rate continues to be an important factor affecting the level of acreage planted to cotton (and thus the level of production), even when, as in MY 2006, the expected market price for upland cotton at the time of planting is higher than the marketing loan rate.321 The Panel also notes in this regard that according to FAPRI and USDA data, the projected amount of marketing loan payments in MY 2006 is very significant.322


      26. Thus, in arguing that because the expected market price for MY 2006 was above the marketing loan rate, marketing loan payments did not significantly affect the planting decisions of upland cotton producers, the United States makes an argument that was already rejected by the panel and the Appellate Body in the original proceeding.323 Nothing in the facts presented to the Panel in this proceeding would appear to warrant a conclusion different from the conclusion drawn by the panel and the Appellate Body in the original proceeding.


      27. Moreover, while the discussion in the preceding paragraphs has focused on the impact of the structure, design and operation of marketing loan payments on acreage and production when analyzed from the short-term perspective of how these payments interact with annual planting decisions, the Panel considers that this impact should also be analyzed from the long-term perspective of how these subsidies influence the economic viability of US upland cotton farming. Article 6.3(c) of the SCM Agreement does not prescribe a particular methodology for determining whether significant price suppression is the effect of a subsidy. It is clear to us that whether production of a particular product is higher than would have been the case in the absence of the subsidy is often a critical issue in establishing whether the effect of the subsidy is significant price suppression. However, nothing in the text of Article 6 suggests that this can only be examined from a short-term perspective. In our view, the type of effect of a subsidy on production relevant to the analysis under Article 6.3(c) can also be demonstrated on the basis of a longer-term perspective that focuses on how the subsidy affects decisions of producers to enter or to exit a given industry. The Panel considers, in this regard, that the evidence on the record, notably the evidence regarding the role of marketing loan and counter-cyclical payments in covering the difference between the market revenue of US upland cotton producers and their costs of production, supports the view that these subsidies have a long-term impact on acreage and production of upland cotton by affecting decisions of US cotton farmers to enter or exit cotton farming. The Panel addresses this issue below in section X.B.9.


        image

        average 17.3 per cent below the December futures contract price. Brazil's Rebuttal Submission, para. 105 and Exhibit Bra-564.

        320 Data submitted by Brazil shows that for most of the period from August 1999 to February 2007, the

        adjusted world price was below the marketing loan rate. Brazil's Response to Panel Question 62, Figure 1, and Exhibit Bra-673.

        321 In order to assess the significance of the effect of marketing loan payments on acreage and

        production under current conditions, we obviously must take into consideration other factors, such as the magnitude of these subsidies and their importance to the revenues of US upland cotton producers. We discuss these factors below.

        322 Supra, para. 10.25 and accompanying footnotes.

        323 We recall, in this respect, the observation made by the Appellate Body that "doubts could arise about the objective nature of an Article 21.5 panel's assessment if, on a specific issue, that panel were to deviate

        from the reasoning in the original panel report in the absence of any change in the underlying evidence in the record…" Appellate Body Report, US - Softwood Lumber VI (Article 21.5 – Canada), para. 103.

        WT/DS267/RW

        Page 83


      28. The Panel now turns to the issues that have been raised in this proceeding specifically with respect to the design, structure and operation of counter-cyclical payments. In this regard, the parties disagree on (i) whether recent economic literature referred to by the United States contradicts Brazil's claim that counter-cyclical payments have significant trade-distorting effects and whether significant amounts of decoupled payments are capitalized into higher land values, and (ii) whether the existence of what Brazil characterizes as a strongly positive relationship between US upland cotton farmers and upland cotton production demonstrates that counter-cyclical payments have significant effects on production.


      29. The Panel first considers the recent economic literature referred to by the United States.


      30. The United States argues that recent economic literature that reflects the results of research based on empirical data finds no evidence that counter-cyclical payments are the cause of any significant changes in plantings.324 In the view of the United States, these new studies and reviews show that counter-cyclical payments have no effects on crop-specific production decisions and minimal effects on overall production.


      31. Brazil, for its part, criticizes the studies referred to by the United States and submits other new studies.325 Brazil contends that these studies are consistent with the original panel's finding that counter-cyclical payments sustain high levels of acreage, production and exports.326


      32. The United States also argues that the question of whether counter-cyclical payments are capitalized into higher land values has a bearing on the extent to which counter-cyclical payments provide incentives to farm operators to increase production of upland cotton. If a significant part of upland cotton farm land is rented by operators instead of being owned by them and if most of the counter-cyclical payments end up as higher land values, and thus in landowners' rather than operators' hands, then there may be little production effect from the payments. In this regard, the United States refers to studies that the United States claims find that decoupled payments, such as counter-cyclical payments, are capitalized into land values and land rents.327


      33. Brazil argues that the price-contingent nature of counter-cyclical subsidies means that they can be very large in one year and small or non-existent in another year and that, hence, such payments are not as easily capitalized into land values as decoupled payments. Brazil submits that a comparison of actual land rents and the amount of direct and counter-cyclical payments per acre demonstrates that only a relatively small portion of the counter-cyclical payments could possibly be shifted to land rents


        image

        324 The studies referred to by the United States include: "Risk Considerations in Supply Response: Implications for Counter-Cyclical Payments' Production Impact" by William Lin and Robert Dismukes (Exhibit US-34); "Counter-Cyclical Payments Under the 2002 Farm Act: Production Effects Likely to be Limited" by Paul Westcott (Exhibit US-35); "US Agricultural Policy: Overview and Recent Analyses" by Edwin C. Young, Anne Effland, Paul Westcott and Demcey Johnson (Exhibit US-40); and "Another Look at Decoupling: Additional Evidence on the Production Effects of Direct Payments" by B. Goodwin and A. Mishra (Exhibit US- 41).

        325 The new studies submitted by Brazil include: "Are 'Decoupled' Farm Payments Really Decoupled?

        An Empirical Evaluation" by B. K. Goodwin and A. Mishra (Exhibit BRA-618); "Farm-Level Production Effect From Participation in Government Commodity Programs: Did the 1996 Federal Agricultural Improvement and Reform Act Make a Difference?" by Nigel Key, Ruben Lubowski and Michael Roberts (Exhibit BRA-566); "Supply Response to Counter-Cyclical Payments and Base Acre Updating under Uncertainty: An Experimental Study" by Christopher R. McIntosh, Jason F. Shogren and Erik Dohlman (Exhibit BRA-565); and "How Distorting Are Direct Payments?" by Erik J. O'Donoghue and James Whitaker (Exhibit BRA-568).

        326 Brazil's Rebuttal Submission, para. 144.

        327 These include the USDA ERS study "Structure and Finances of US Farms: 2005 Family Farm Report" reproduced in Exhibit US-67 and the OECD study "A Review of Empirical Studies of the Acreage and

        Production Response to US Production Flexibility Contract Payments Under the Fair Act and Related Payments Under Supplementary Legislation" reproduced in Exhibit US-32.

        WT/DS267/RW

        Page 84


        or capitalized into the value of land.328 While Brazil recognizes the possibility that certain landowners might capture a portion of counter-cyclical payments, this assessment does not change the fact that upland cotton counter-cyclical payments are tied to prices and increase US upland cotton production and exports.


      34. The Panel recalls that in assessing the production impact of counter-cyclical payments, the original panel relied upon several USDA studies329 to conclude that, due to their market-price contingency, counter-cyclical payments "may influence production decisions indirectly by reducing total and per unit revenue risk associated with price variability in some situations".330 As the United States points out in this proceeding, these studies did not reflect experience with the counter- cyclical payment programme itself.331 The Panel therefore considers that the studies cited by the United States and Brazil in this proceeding are relevant insofar as they provide new information not available to the original panel reflecting the experience with the operation of counter-cyclical payments to upland cotton. Both academics and USDA staff have contributed to this recent literature. This is we believe the most recent set of studies available. The Panel has reviewed this new literature and weighed the findings of the experts carefully.332 The conclusion that we draw from our review of this literature is the same as that of the original panel i.e. that "due to their market-price contingency, [counter-cyclical payments] may influence production decisions indirectly by reducing total and per unit revenue risk associated with price variability in some situations". 333


      35. Turning to the literature cited by the United States, the Panel notes that none of the recent studies submitted by the United States specifically analyzes the experience with counter-cyclical payments to growers of upland cotton. We find that:


        1. Instead of counter-cyclical payments, fully decoupled payments or programmes that antedated the counter-cyclical programme (such as the production flexibility contract payments) were studied. For instance, the Goodwin and Mishra study published in 2005334 examines the extent to which production flexibility contract or Agricultural Market Transition Act (AMTA) payments under the FAIR Act affected the acreage decisions of a set of farms in the US Corn Belt who grew corn, wheat, and soybeans.


        2. Instead of empirical studies of the effects of counter-cyclical payments:


          1. A survey of other studies was undertaken. The Young, Effland, Westcott and Johnson paper335 presents an overview of recent US agricultural policies under the 1996 and 2002 Farm Acts. The paper surveys recent analyses regarding the effects of those policies;


            image

            328 Brazil's Rebuttal Submission, paras. 146-148.

            329 The studies include the USDA paper "The 2002 Farm Act, Provisions and Implications for

            Commodity Markets" (November 2002), reproduced in Exhibit BRA-42 and a 2002 USDA ERS report to the United States Congress contained in "Farm Program Payments and the Economic Viability of Production Agriculture", reproduced in Exhibit BRA-130.

            330 Panel Report, US – Upland Cotton, para 7.1302.

            331 The FSRI Act of 2002, which authorized countercyclical payments for the first time, entered into

            force only on 13 May 2002. The period examined by the panel in determining "present" serious prejudice was MY 1999-2002.

            332 Attachment 1 (Section XIII), contains a summary of the studies submitted by the parties.

            333 Panel Report, US – Upland Cotton, para. 7.1302.

            334 See supra, footnote 324.

            335 See supra, footnote 324.

            WT/DS267/RW

            Page 85


          2. A simulation of the effects of these payments was undertaken. The Lin and Dismukes study336 attempts to simulate the acreage effects of counter-cyclical payments for three crops – corn, soybeans and wheat. The simulation is based on an acreage model, which was estimated using state-level data during 1991-2001 (before counter-cyclical payments came into effect), drawn from farms in the Northcentral region of the United States; and


          3. A conceptual analysis of the effects of counter-cyclical payments was attempted. The Westcott paper337 considers the case of a hypothetical corn farmer who plants all of his hundred acres of corn base with corn. It examines how the farmer's revenues are affected by counter-cyclical payments (among other subsidies) depending on realizations of the market price.


            Thus, the relevance of these studies to the present dispute is somewhat limited. We also note some variation in the type and quality of the studies: some are published or slated to be published in academic journals while others are articles in newsletters or unpublished manuscripts.


      36. Many of these observations also apply to the studies provided by Brazil. While the parties have drawn widely differing conclusions from these studies, the Panel has also found that the studies tend to converge on some important points. First, there is no disagreement in the recent literature cited by both parties that counter-cyclical payments can or may have production effects. Second, the studies cited by the parties are also in some agreement about the mechanisms by which these payments can encourage additional production. Increased wealth can reduce risk aversion, leading farmers to increase production. The presence of direct payments will allow credit-constrained farmers to finance the purchase of more farm inputs. Farmers may be willing to produce more output since counter-cyclical payments, which vary with the output price, can reduce revenue risk. Payments may allow producers to cover fixed costs and thus may allow marginal farmers who would otherwise be forced to shut down to remain in production. If farmers believe that future payments are based on current production, because of the possibility of updating base acres, they may increase the current production of those crops for which they expect to receive future payments.338


      37. Where the studies differ is in their assessment of the magnitude of these production effects. The studies cited by the United States argue that the production effects of counter-cyclical payments are likely to be small, while the papers cited by Brazil contend that the effects are substantial. Almost all of the papers have focused on the acreage response by farmers to the payments. There appears to have been a surge of interest in the research community on the question of the production effects of decoupled payments. It would seem that this recent literature is fairly divided on the magnitude of the production impact of counter-cyclical payments. Academics and USDA economists can be found on both sides of the question.


      38. With respect to the issue of whether counter-cyclical payments are capitalized into higher land values, the Panel considers that this in no way implies that the payments have no production effects. While the OECD study cited by the United States concluded that most of the impact of production flexibility contract and market loss assistance payments pass-through to higher land values, enough seems to have been left in the hands of farm operators to create a modest production effect. Given that counter-cyclical payments, in contrast to production flexibility contract payments, are price-contingent, there is a presumption that the output effect of counter-cyclical payments will be higher than that of the production flexibility payments.


        image

        336 See supra, footnote 324.

        337 See supra, footnote 324.

        338 The Panel understands that the issue of base acreage updating is not an issue before it.

        WT/DS267/RW

        Page 86


      39. Thus, in our assessment, the current state of this academic and policy research is such that even though the experts disagree among themselves about the exact magnitude of the production effects of counter-cyclical payments, there is no disagreement about the direction of those effects nor on the mechanisms by which the production effects are transmitted. Thus, counter-cyclical payments can increase production and these production effects occur because increased wealth reduces risk aversion, the availability of the payments relieves credit constraints and provides some insurance from fluctuations in revenues and the payments allow farmers to cover fixed costs of production.


      40. The Panel now turns to the arguments of Brazil and the United States on the issue of whether the relationship between upland cotton farmers receiving counter-cyclical payments and upland cotton production supports a finding that the effect of counter-cyclical payments, by virtue of their structure, design and operation, is significant price suppression in the world market for upland cotton.


      41. In support of its argument that the structure, design and operation of the counter-cyclical payments insulate US cotton producers from market forces and thereby stimulate production, Brazil submits that there continues to be a "strong positive relationship" between upland cotton base acre holders receiving annual counter-cyclical payments and upland cotton production.339 The overwhelming majority of US upland cotton is grown on farms holding upland cotton base. Moreover, the evidence also shows that the amount of upland cotton acreage of recipients who over- planted their cotton base remains at a low level compared to 1999-2001. Brazil considers that it is unimportant that, as argued by the United States, some 40 per cent of upland cotton base acreage is not planted to upland cotton and that an average of 17 per cent of upland cotton acreage is on farms with cotton planted acreage that exceeded cotton base acreage or on farms with no cotton base acreage.340 Brazil submits that its claim pertains to the effect of counter-cyclical subsidies in maintaining production of US upland cotton by keeping current producers of upland cotton in business and that it is not arguing that counter-cyclical subsidies induce production. Therefore, the focus of the analysis should be on the relationship between current US upland cotton producers and upland cotton counter-cyclical subsidies. From this perspective, the key fact is that the overwhelming majority of current US upland cotton production is undertaken by historic upland cotton producers that receive upland cotton-counter-cyclical payments. The percentage of US upland cotton acreage on farms with upland cotton base acres is essentially the same as in the period considered by the original panel. Similarly, the fact that cotton producers planted 40 per cent fewer acres to cotton than in the base period is irrelevant to Brazil's claim that counter-cyclical subsidies received by current producers of upland cotton are critical to ensure a profit for those farms that currently do produce upland cotton. In any event, this percentage is essentially the same as in the period considered by the original panel. Brazil submits that the fact that, as argued by the United States, US upland cotton was grown historically and continues to be grown today in certain regions of the United States supports Brazil's argument that by providing for counter-cyclical payments based on historical base acres the US Congress intended to support current production of upland cotton.341


        image

        339 Brazil's First Written Submission, para. 131. Brazil refers to the statement of the original panel that the evidence before the Panel showed "a strongly positive relationship between those recipients who hold upland cotton base acres and those who continue to plant upland cotton..." Panel Report, US – Upland Cotton, para. 7.637.

        340 Brazil's Rebuttal Submission, paras. 153-165.

        341 Brazil's Comments on US Responses to Panel Questions 29 and 54. Brazil submits that in MY 2005, 83 per cent of cotton planted acreage was on cotton base acreage. This is the more significant portion

        of US upland cotton acreage. Moreover, of the remaining 17 per cent, 12 per cent was on farms that "overplanted" their cotton base acreage and 5 per cent was on farms that did not have any upland cotton base acreage. In other words, 95 per cent of US upland cotton acreage is on farms that hold upland cotton base. These facts, according to Brazil, demonstrate that there continues to be a strong positive relationship between upland cotton (base acre) producers receiving annual payments and upland cotton production. In regard to the argument of the United States that 40 per cent of upland cotton base acres are not currently planted to upland

        WT/DS267/RW

        Page 87


      42. The United States argues that data comparing base to planted acreage does not demonstrate that counter-cyclical payments have significant effects on production. The fact that there is some overlap between current and historic production of upland cotton in the United States is not remarkable. What is notable is that the data shows that counter-cyclical payments do not induce recipients to continue planting cotton. Over the period MY 2002-2005 an average of 17 per cent of US cotton planted acreage was on farms with cotton planted acreage that exceeded cotton base acres or on farms with no cotton base acres at all. Moreover, traditional US cotton producers receiving counter-cyclical payments planted approximately 40 per cent fewer cotton acres over MY 2002-2005 than in the period used to calculate cotton base acres.342 Brazil's attempts to dismiss these facts as unimportant are inexplicable and baseless. Given that Brazil is arguing that counter-cyclical payments induce US producers to produce cotton even though they are not required to do so to receive the payments, it would seem to be very important whether in fact, US producers holding upland cotton base acres are continuing to produce upland cotton and if so, whether they are producing cotton in such a way as to evidence the significant production effects that Brazil has alleged. Brazil fails to explain what the fact that 95 per cent of US upland cotton planted acreage was on farms that held upland cotton base acreage has to do with its claim that counter-cyclical payments have significant production inducing effects. The United States notes in this respect that the same "positive relationship" exists between upland cotton production and recipients of direct payments, which were not found to have significant price suppressing effects in the original proceeding.343 The facts show that US farmers use the planting flexibility afforded by the direct and counter-cyclical payments and that counter-cyclical payments do not induce producers to grow cotton where they would not otherwise have done so. Upland cotton is grown in a limited number of regions in the United States where the weather and other conditions are ideal for upland cotton production. The fact that many farmers who planted upland cotton in the historical period continue to produce some cotton today simply means that it makes sense for these producers to grow cotton given their experience and expertise, their equipment and the particular growing conditions. Therefore, what matters in order to assess the effects of counter-cyclical payments on production is not the percentage of farms producing upland cotton that received upland cotton counter-cyclical payments but the level of acreage planted to cotton vis-à-vis upland cotton base acres.344


      43. The Panel notes that the original panel found a "strongly positive relationship between recipients who currently plant upland cotton and those whose payments are calculated with respect to upland cotton base acres and whose payments are therefore calculated at the upland cotton–specific payments rate".345 More specifically, the panel determined on the basis of data provided by the United States that:


        "a very large proportion of farms with upland cotton base acres continue to plant upland cotton in the year of payment. The data provided also shows that the overwhelming majority of farms enrolled in the programmes which plant upland cotton also hold upland cotton base."346 (footnote omitted)


        image

        cotton, Brazil notes that at no time in recent history has US upland cotton acreage been 18.4 million acres – the amount of upland cotton base. The high level of upland cotton base relative to production reflects the fact that with consecutive base updates, US farms elected to update only if that would increase their high payment upland cotton base. In case an update would have decreased high payment upland cotton base in favour of lower payment crop base, US farms declined to update. Thus, US upland cotton base acres are not a reliable measure to compare to annual US upland cotton acreage.

        342 US First Written Submission, paras. 220-225.

        343 US Rebuttal Submission, paras. 263-266.

        344 US Response to Panel Question 29; US Response to Panel Question 54, paras. 109-111.

        345 Panel Report, US – Upland Cotton, para. 7.583.

        346 Panel Report, US – Upland Cotton, para. 7.636. The panel provided precise figures on these relationships in Tables A-1 and A-2.

        WT/DS267/RW

        Page 88


      44. The original panel provided figures on this relationship in Tables A-1 and A-2 of its report and observed that:


        "These figures show a strongly positive relationship between those recipients who hold upland cotton base acres and those who continue to plant upland cotton, despite their entitlement to plant other crops, which is indicative of the relationship between payments calculated with respect to upland cotton base acreage and recipients who plant upland cotton. These figures also show that upland cotton production within these four programmes is almost exclusively taking place on farms which hold upland cotton base acres."347


      45. The original panel referred to this positive relationship in support of certain aspects of its findings on serious prejudice.348 With respect to the structure, design and operation of counter- cyclical payments the panel observed:


        "We agree with the view of USDA economists that, due to their market-price contingency, CCPs may influence production decisions indirectly by reducing total and per unit revenue risk associated with price variability in some situations. In the price range from the loan rate up to the target price minus the DP payment rate, changes in producer revenues due to changes in market prices are partly offset by the countercyclical payments if the base acreage crop is planted, thereby reducing total revenue risk associated with price variability. We have confirmed a strong positive relationship between upland cotton (base acre) producers receiving annual payments and upland cotton production. Moreover, the payments are mandatory, under certain market conditions."349 (footnotes omitted, emphasis added)


      46. The evidence provided by Brazil in this proceeding shows that the facts regarding the relationship between base acres and planted acreage are not significantly different from the facts that formed the basis for the Panel's conclusion regarding the existence of a "strongly positive relationship" between upland cotton base acre holders receiving annual counter-cyclical payments and upland cotton production.350 The United States has not questioned the factual accuracy of the data submitted by Brazil on the existence of this relationship. In other words, the United States does not contest that, as argued by Brazil, the overwhelming majority of current US upland cotton production is undertaken by historic upland cotton producers that receive upland cotton counter-cyclical payments. Rather, the United States contests that this relationship supports Brazil's claim of significant price suppression. However, the Panel concurs with Brazil that this relationship between upland cotton base acre holders and upland cotton production is significant in that it suggests that cotton counter-cyclical payments play a role in maintaining the level of acreage and production at a higher level than would otherwise be the case. In addition, as noted above, this relationship is essentially the same as in the period considered by the original panel. We therefore see no basis in this proceeding to arrive at a conclusion different from that of the original panel with respect to this issue.


    image

    347 Panel Report, US – Upland Cotton, para. 7.637.

    348 Panel Report, US – Upland Cotton, paras. 7.1119, 7.1226, 7.1302, footnote 1417 and para. 7.1362.

    349 Panel Report, US – Upland Cotton, para. 7.1302.

    350 Table 3 of Brazil's Rebuttal Submission shows that in MY 2005, 95 per cent of US upland cotton planted acreage was planted on farms that received upland cotton counter-cyclical payments. This table is an

    updated version of a table in the original panel report that showed that the percentage of upland cotton planted acres on farms with upland cotton base acres was 92.9 in MY 1999, 92.0 in MY 2000, 91.3 in MY 2001 and

    96.1 in MY 2002. Panel Report, US – Upland Cotton, para. 7.636, Table A-2. In addition, the Panel notes that while the United States emphasizes that 40 per cent of upland cotton base acreage is not planted to upland cotton, Brazil has pointed out that during the period considered by the original panel this percentage was similar.

    WT/DS267/RW

    Page 89


      1. The Panel therefore finds that a strongly positive relationship exists between recipients of upland cotton counter-cyclical payments who hold upland cotton base acres and those who continue to plant upland cotton and that this factor is an element that provides support for a finding that the counter-cyclical payments, because of their structure, design and operation, have price-suppressing effects.


      2. In conclusion, based on an analysis of the arguments and evidence presented to the Panel in this proceeding regarding the structure, design and operation of the marketing loan and counter- cyclical payments to US upland cotton producers, the Panel considers that these subsidies affect the level of US upland cotton acreage and production as a result of their mandatory and price-contingent nature and their revenue-stabilizing effect. In this respect, the Panel concurs with the view of the original panel that these subsidies protect or "insulate" revenues of US upland cotton producers when prices are low. It is important to emphasize, however, that in determining whether the structure, design and operation of these subsidies support a finding of significant price suppression under current factual conditions, we need to consider this factor in conjunction with other facts. The Panel considers, in this regard, that viewed together with other evidence on the record, including the fact that the adjusted world price in most recent years has been below the marketing loan rate, the magnitude of the marketing loan and counter-cyclical payments, the importance of these subsidies as a share of the revenues of US upland cotton producers and their role in covering a significant part of the costs of production of these producers351, it is reasonable to conclude that without these subsidies the level of US upland cotton acreage and production would likely be significantly lower.


      3. As discussed above352, the original panel relied upon the design, structure and operation of the marketing loan and counter-cyclical payments in conjunction with the magnitude of the subsidies. We therefore now turn to the argument of Brazil that the magnitude of the marketing loan and counter-cyclical payments supports a finding that the effect of marketing loan and counter-cyclical payments is significant price suppression.


        1. Alleged "large magnitude" of the marketing loan payments and counter-cyclical payments


          1. Main arguments of the parties


      4. Brazil argues that the magnitude of US marketing loan and counter-cyclical payments under the FSRI Act of 2002 remains very large and that this is evidence supporting the existence of a causal link between these two subsidies and significant price suppression in the world market for upland cotton. The elimination of Step 2 payments has not significantly reduced the overall magnitude of the price-contingent subsidies.353 Brazil considers that the significance of the magnitude of the marketing loan and counter-cyclical payments must be assessed in light of the fact that, as found by the original panel, these subsidies are provided in respect of a fungible, widely-traded product. Because of the high subsidization rate from the marketing loan payments and counter-cyclical payments, US producers and exporters have a significant advantage over their competitors in securing sales in the world market. Brazil submits that the large magnitude of the marketing loan payments is highly significant in light of their trade-distorting nature. The panel and Appellate Body reports in the original proceeding and the panel reports in Indonesia – Autos and Korea – Commercial Vessels



        image

        351 Infra, Sections X.B.6 and X.B.9.

        352 Supra, paras. 10.66-10.67.

        353 Brazil asserts that while the ad valorem rate of subsidization from Step 2 payments was 6.7 per cent

        in MY 2005, the ad valorem subsidization rate from marketing loan payments and counter-cyclical payments in MY 2005 was 42 per cent.

        WT/DS267/RW

        Page 90


        support the view that the amount of a subsidy is an important factor to consider in a price suppression analysis.354


      5. The United States submits that Brazil does not demonstrate that the size of the outlays under the marketing loan and counter-cyclical payment programmes supports the existence of a causal link between these payments and the existence of significant price suppression.355 Labelling the annual amounts of marketing loan and counter-cyclical payments as "large" or "very large" says nothing about the effects of these payments on world market prices. The size of outlays under the marketing loan and counter-cyclical payment programmes cannot be meaningfully addressed apart from an actual examination of their design, structure and operation. Brazil provides no evidence to substantiate its contention that marketing loan and counter-cyclical payments provide an "enormous advantage" to US producers and exporters "to secure sales" in the world market. While Brazil appears to suggest that US exporters use marketing loan and counter-cyclical payments to undercut foreign cotton prices, Brazil provides no evidence to support this assertion. Moreover, Brazil has failed to explain the relevance of this assertion to its claim, which is a claim of price suppression rather than price undercutting.356


        1. Evaluation by the Panel


      6. The original panel considered that a precise quantification of the amount of a subsidy was not required in a serious prejudice analysis under Part III of the SCM Agreement357, but that "the general order of magnitude of a subsidy may be a relevant consideration in a serious prejudice analysis in a particular case (where such information is probative and readily available)".358 The panel noted that as it had readily available information on the general order of magnitude of many of the subsidies at issue, it would take this into account in determining whether or not significant price suppression and/or serious prejudice existed.359 The panel relied upon the magnitude of the subsidies at issue, which it characterized as "very large", as a factor supporting its findings that price suppression existed360, that this price suppression was significant361 and that a causal link existed between certain price-contingent subsidies and the significant price suppression which it had found to exist.362


      7. The Appellate Body held that the original panel had not erred in its assessment of the amount of the subsidies for the purposes of its analysis under Article 6.3(c) of the SCM Agreement.363 The Appellate Body also held that the magnitude of a subsidy and its relationship to prices is a relevant factor to consider in an analysis of whether the effect of a subsidy is significant price suppression 364,



        image

        354 Brazil's First Written Submission, paras. 107-119; Brazil's Rebuttal Submission, paras. 167-179.

        355 The United States addresses this issue as part of its arguments on the design, structure and operation

        of the subsidies at issue. US First Written Submission, paras. 236-241; US Rebuttal Submission, paras. 276- 280.

        356 US First Written Submission, paras. 236-241; US Rebuttal Submission, paras. 276-280.

        357 Panel Report, US – Upland Cotton, paras. 7.1167, 7.1177 and 7.1179.

        358 Panel Report, US – Upland Cotton, footnote 1289.

        359 Panel Report, US – Upland Cotton, para. 7.1194. The "readily available information" referred to by the panel was information contained in section VII:D and in footnote 895 of the panel report.

        360 Panel Report, US – Upland Cotton, paras. 7.1297 and 7.1308.

        361 Panel Report, US – Upland Cotton, para. 7.1332.

        362 Panel Report, US – Upland Cotton, para. 7.1349.

        363 Appellate Body Report, US – Upland Cotton, para. 473.

        364 The Appellate Body stated:


        "Beginning with the text of Article 6.3(c), we note that this provision does not state explicitly that a panel needs to quantify the amount of the challenged subsidy. However, in assessing whether 'the effect of the subsidy is ... significant price suppression', and ultimately serious prejudice, a panel will need to consider the

        WT/DS267/RW

        Page 91


        but that Article 6.3(c) of the SCM Agreement does not oblige a panel "to quantify precisely the amount of a subsidy benefiting the product at issue in every case. A precise, definitive quantification of the subsidy is not required".365 The Appellate Body also considered that although the panel could have been more explicit as to what it meant by the "very large amounts" of subsidies, the information before the panel clearly supported the panel's general statements on the magnitude of the subsides.366


      8. We have summarized the data on record with respect to the magnitude of marketing loan payments and counter-cyclical payments in Tables 1 and 2.367 The evidence before the Panel indicates that in MY 2005 the total amount of marketing loan payments to upland cotton producers and of counter-cyclical payments allocated to upland cotton was higher than in MY 2002.368 We also note that the amount of marketing loan payments and counter-cyclical payments in MY 2005 is of the same order of magnitude as the total amount in MY 2002 of the price-contingent subsidies subject to the original's panel finding of "present" serious prejudice.369 At the same time, we also note that the amount of projected marketing loan payments in MY 2006 is significantly less than the amount of marketing loan payments in MY 2005.370


      9. The Panel considers, in light of the approach taken by the panel and Appellate Body in the original proceeding, that in determining whether the magnitude of the subsidies at issue in this proceeding supports a finding that the effect of these subsidies is significant price suppression, it is necessary to examine this factor in relation to other factors. Thus, the relevance of the magnitude of the subsidies in this connection must also be assessed in light of our analysis above of the structure, design and operation of the subsidies. We recall, in this respect, our conclusion regarding the important revenue-stabilizing effect of these mandatory, price-contingent subsidies. In the latter regard, the evidence before us indicates that marketing loan and counter-cyclical payments play a



        image

        effects of the subsidy on prices. The magnitude of the subsidy is an important factor in this analysis. A large subsidy that is closely linked to prices of the relevant product is likely to have a greater impact on prices than a small subsidy that is less closely linked to prices. All other things being equal, the smaller the subsidy for a given product, the smaller the degree to which it will affect the costs or revenue of the recipient, and the smaller its likely impact on the prices charged by the recipient for the product. However, the size of a subsidy is only one of the factors that may be relevant to the determination of the effects of a challenged subsidy. A panel needs to assess the effect of the subsidy taking into account all relevant factors."


        Appellate Body Report, US – Upland Cotton, para. 461 (footnote omitted).

        365 Appellate Body Report, US – Upland Cotton, para. 467. 366 Appellate Body Report, US – Upland Cotton, para. 468. 367 Supra, paras. 10.24 and 10.32.

        368 As can be seen from Tables 1 and 2, the total amount of marketing loan payments in MY 2005 was US$1.269 billion and the amount of counter-cyclical payments allocated to upland cotton was US$838 million.

        The total amount of marketing loan and counter-cyclical payments to US upland cotton producers was US$2.107 billion in MY 2005. In MY 2002, the amount of marketing loan payments was US$898 million while the amount of counter-cyclical payments allocated to upland cotton was US$851.5 million.

        369 According to Table 6 in Brazil's First Written Submission, the amount of Step 2 payments in

        MY 2002 was US$429 million. No market loss assistance payments were made in MY 2002. Thus, the total amount of marketing loan and counter-cyclical payments to upland cotton producers and of Step 2 payments to upland cotton users and exporters was US$2.179 billion in MY 2002.

        370 As noted above, Brazil has pointed to three different projections of the amount of marketing loan payments MY 2006: US$650 million, US$475 million and US$757 million. Brazil argues that the more recent

        and reliable of these projections is the February 2007 USDA projection (US$757 million). However, even if we take the lowest of the three figures mentioned by Brazil we note that the total amount of marketing loan and counter-cyclical payments in MY 2006 is projected to be well above US$1 billion.

        WT/DS267/RW

        Page 92


        significant role in stabilizing the revenues of US upland cotton producers.371 The share of marketing loan payments and counter-cyclical payments in total revenues of US upland cotton producers was 35 per cent in MY 2004 and 27 per cent in MY 2005.372 This data on the share of marketing loan and counter-cyclical payments in US upland cotton producers' revenues is especially important in light of our finding regarding the gap between costs of production and market revenues of US upland cotton producers.373 In our view, when considered in conjunction with other factors, the order of magnitude of the marketing loan payments and counter-cyclical payments supports a finding of significant price suppression, even when account is taken of the decline in the amount of marketing loan payments projected for MY 2006.374


        1. Alleged "link between high levels of US subsidies and high levels of US planted acreage, production and exports"


          1. Main arguments of the parties


      10. Brazil argues that there remains a strong link between high levels of US subsidies and high levels of US planted acreage, production and exports during MY 2002-2005, which is due to the fact that marketing loan and counter-cyclical payments insulate US cotton producers from market price signals.375 Brazil asserts in this connection that, regardless of the type of price examined, US planted acreage, production and exports do not respond significantly to market prices and have increased significantly in recent years.376 377 Brazil provides graphs demonstrating that while US acreage does not show any relationship to futures prices, foreign acreage reacts to changes in expected prices.378 Brazil also provides graphs demonstrating that US upland cotton production and exports do not respond significantly to changes in either movements or absolute levels of the US farm price.379


      11. Brazil provides a table to show the amounts of US planted and harvested acreage, yields, US production and US exports of upland cotton over the MY 1999-2005 period that were achieved under the revenue protection afforded US upland cotton producers by the effects of the marketing loan and counter-cyclical payment programmes.


        image

        371 Brazil's First Written Submission, para. 133, Figure 8; Brazil's Comments on US Response to Panel Question 51, para. 72, Figure 1.

        372 Brazil's First Written Submission, para. 135.

        373 Infra, Section X.B.9

        374 Supra, footnote 370.

        375 Brazil's First Written Submission, para. 137.

        376 Brazil's First Written Submission, paras. 139-149; Brazil's Rebuttal Submission, paras. 209-231.

        377 Brazil asserts that US planted acreage, after falling slightly in MY 2002 and 2003, has increased every year since. MY 2006 acreage, planted in the spring of 2006, is estimated to be the highest in five years,

        and, at 14.96 million acres, only a half-million acres below the ten-year high reached in MY 2001. Overall, US planted acreage during MY 2002-2005 has remained at levels that are similar to 10- and 15-year averages – periods during which US upland cotton producers also benefited from large US subsidies. Brazil's First Written Submission, para. 140 and Exhibit Bra-447 (Upland Cotton Supply and Use).

        378 Brazil's First Written Submission, Figure 9; Brazil's Rebuttal Submission, Figures 8 and 9.

        379 Brazil's First Written Submission, Figures 10 and 11.

        WT/DS267/RW

        Page 93


        Table 5: US Upland Cotton Production, Exports and Acreage, MY 1999-2006


        MY

        1999

        2000

        2001

        2002

        2003

        2004

        2005

        2006

        Production (million bales)

        16.294

        16.799

        19.603

        16.531

        17.823

        22.505

        23.260

        Exports (million bales)

        6.303

        6.303

        10.603

        11.266

        13.220

        13.618

        17.437

        Planted acres (million acres)


        14.584


        15.347


        15.499


        13.714


        13.301


        13.409


        13.975


        14.995

        Harvested acres (million acres)


        13.138


        12.884


        13.560


        12.174


        11.826


        12.809


        13.534

        Yield (bales per acre)

        1.240

        1.304

        1.446

        1.358

        1.507

        1.757

        1.719


        Source: Brazil's First Written Submission, Table 12.380


      12. Brazil asserts that the record amount of US upland cotton production in MY 2001 was surpassed in MY 2004 by 15 per cent and that MY 2005 saw yet another record level of production. Production in MY 2005 was 43 per cent higher than in MY 1999.381 Since MY 1999, US exports have increased even more rapidly than production. In MY 2005, US upland cotton exports were 177 per cent higher than in MY 1999. MY 2001, 2002, 2003, 2004 and 2005 each established new record volumes of US exports. Exports in MY 2005 were, in fact, larger than total US upland cotton production in MY 1999, 2000 and 2002.382


      13. Brazil submits that in arguing that US cotton producers are not insulated from market signals the United States makes arguments that were already rejected in the original proceeding. According to Brazil, the United States effectively asks this Panel to overturn the finding of the panel and Appellate Body that the structure, design and operation of the marketing loan and counter-cyclical subsidies insulate US upland cotton producers from market forces by stabilizing their revenue at very high levels. Adverse effects within the meaning of Articles 5 and 6 of the SCM Agreement result not from increases as such but from production and exports in excess of what would have been the levels of production and exports in the absence of the subsidies. But for the marketing loan and counter- cyclical payments, US producers would have been more responsive to market price signals. The very fact that the US shares of world production and exports have remained stable is evidence of the effect of the marketing loan and counter-cyclical subsidies.383


      14. Brazil submits that the argument of the United States that US cotton planting decisions should be explained by the ratio of upland cotton futures prices to soybeans futures prices was already rejected by the panel and Appellate Body in the original proceeding and is not supported by the evidence submitted by the United States.384


      15. The United States argues that US upland cotton producers and exporters have reacted to market signals and are not insulated by the marketing loan and counter-cyclical payment programmes. First, stable US shares of world production and exports confirm that the marketing loan and counter-


        image

        380 Brazil compiled this table from Exhibit Bra-447 (Upland Cotton Supply and Use) and, for planted acreage in MY 2006, Exhibit Bra-449 (Cotton and Wool Outlook, USDA, October 2006, p. 9, accessed October 2006 at http://usda.mannlib.cornell.edu/usda/current/CWS/CWS-10-13-2006.pdf). Brazil notes that evidence on harvested acres, yield per harvested acre, production, and exports is not yet available for MY 2006.

        381 Brazil's First Written Submission, para 147.

        382 Brazil's First Written Submission, para 148.

        383 Brazil's Rebuttal Submission, paras. 195-208. Brazil also submits that the US share of world production has in fact increased in MY 2005.

        384 Brazil's Rebuttal Submission, paras. 214-215.

        WT/DS267/RW

        Page 94


        cyclical payment programmes are not insulating US producers from market signals.385 The fact that these shares have been stable shows that US upland cotton producers are affected by market signals in the same ways as foreign producers.386 According to the United States, Brazil has no credible answer to a fundamental question in this regard if US producers continue to plant and produce, and US exporters continue to export, in circumstances where "any rational non-subsidized producer" would have "reduced plantings, production, and exports," why has US share of world production and world exports not increased over the life of the FSRI Act of 2002?387 The United States submits that it has not argued that adverse effects within the meaning of Articles 5 and 6 of the SCM Agreement can only be found if there are increases in production, acreage and exports, but that increased US share of production is a necessary implication of Brazil's theory of market insulation.


      16. Second, the United States argues that Brazil fails to demonstrate the alleged strong link between marketing loan and counter-cyclical payments and US planted acreage, production and exports. Brazil's analysis of the sensitivity of US planted acreage, production and exports to prices is flawed in that it does not properly reflect the factors taken into consideration by US farmers when making their production decisions. Since farmers do not only consider the cotton futures price but also take into account other factors, such as weather conditions, prices of competing crops, changes in technology, labour and input prices and marketing efforts, it is not surprising that planted acreage does not correlate with the cotton futures price.388 Regarding Brazil's comparison of US cotton production to the farm price, the United States submits that Brazil firstly assumes, incorrectly, that planting decisions could be explained through an examination of cotton prices alone. Moreover, in comparing US upland cotton production in a marketing year to the average US farm price for the same marketing year, Brazil assumes that US farmers know at the time that they plant what the actual farm price will be in the upcoming marketing year, which does not start until August. Finally, producers do not decide on production, but on plantings. Ultimate production is affected by weather and other factors affecting yields.389

      17. The United States argues that it is apparent from Brazil's chart390 that foreign acreage (even harvested acreage) is many orders of magnitude greater than US planted acreage. Plotted against each other in a chart, US planted acreage will inevitably appear to be flat relative to foreign acreage.391 The United States also submits that a comparison of US acreage and foreign acreage to futures prices of upland cotton shows that neither US nor foreign harvested acreage moves closely in line with futures prices of cotton alone.392


        image

        385 US First Written Submission, Section VI.C.2(a). The United States provides charts to show that the US shares of world production and exports of upland cotton have remained stable over the life the FSRI Act of 2002 in paras. 243 and 247 of its First Written Submission. In paras. 283-284 of its Rebuttal Submission, the United States provides charts to show that the US share of world production has been stable from well before either the marketing loan or the counter-cyclical payment programme came into effect.

        386 With respect to the increase in the US share of world exports in the period MY 1999-2002, the United States argues that this increase did not reflect an increase in US share of world production. In other

        words, US producers were not increasing their production relative to foreign producers in order to gain market share, as Brazil has suggested. Rather, a greater proportion of US production was being exported. The reason for this was the decline in US mill use resulting from the declining competitiveness of the US textile industry and increasing US imports of textiles and clothing in the same period. US First Written Submission, para 249.

        387 US Rebuttal Submission, para 283.

        388 US First Written Submission, paras. 255-261.

        389 US First Written Submission, para 262-263. The United States also argues that, if one compares

        both US and foreign production to movements in the A-Index, the same alleged "absence" of a "link" between absolute production levels and prices is apparent in both cases. US First Written Submission, para. 265.

        390 Brazil's Rebuttal Submission, para. 221, Figure 8.

        391 US Rebuttal Submission, para 303.

        392 US Rebuttal Submission, para 308.

        WT/DS267/RW

        Page 95


      18. The United States argues that "absolute" increases in US production are "the effect" of improvements in yields, not the marketing loan and counter-cyclical payment programmes.393


      19. Brazil submits that there is no legitimate basis for the United States to claim that increased yields alone accounted for increased US production and increased US exports in MY 2004 and 2005. Brazil agrees that for the portion of US upland cotton planted acreage that would have been planted even without the effects of the US subsidies, the increased yields alone accounted for an increase in resulting production and exports. However, for the portion of US upland cotton planted acreage that would not have been planted but for the marketing loan and counter-cyclical payment revenue, the increased yields secured on such acreage – and hence also the resulting production and exports – are linked to the effects of the US subsidies.394


        1. Evaluation by the Panel


      20. The Panel notes that in support of its argument that there exists a strong link between high levels of US subsidy payments and high levels of US planted acreage, production and exports, Brazil argues that US cotton producers are insulated from market price signals.395 In Brazil's view, the price- contingent marketing loan and counter-cyclical payments act to ensure that US upland cotton producers' total revenue per unit remains largely unaffected by price movements. Brazil refers in this respect to the findings of the panel and the Appellate Body in the original proceeding.


      21. Brazil has also sought to provide factual support for its argument that US producers do not react to market price signals by comparing US planted acreage and futures prices of cotton. Brazil argues that, as during the time period assessed by the original panel, US planted acreage continues to show no obvious relationship to market price signals. 396 By contrast, the United States asserts that the futures price of cotton is not the sole basis for a farmer's planting decision as many growers, in the United States and elsewhere have several alternative crops to consider. In this respect, the United States submits that a comparison between, on the one hand, US cotton planted acreage and, on the other hand, the ratio of cotton futures prices to soybeans futures prices rebuts Brazil's argument that US cotton producers are insulated from market price signals.397 Brazil contests that this comparison supports the argument of the United States.398


        image

        393 US First Written Submission, Section VI.C.3, page 87. See also US Rebuttal Submission, Section IV.C.2, page 108.

        394 Brazil's Rebuttal Submission, paras. 229-231.

        395 Brazil's First Written Submission, para. 137.

        396 Figure 9 in Brazil's First Written Submission compares the average price of the December cotton

        futures contract in February (a proxy for farmers' pre-planting expectations of market prices during the beginning of the marketing year) with US upland cotton planted acreage. Brazil argues that this figure shows that US planted acreage does not move with futures prices in at least three of the last four planting seasons. Brazil argues that the graph demonstrates that, in 2003, US planted acreage declined even though futures prices indicated much higher prices for the December 2003 harvest. Similarly, in February 2004, planted acreage only increased slightly (108,000 acres) despite the highest February futures prices for the December contract in the previous five years. And with futures prices falling significantly in February 2005, US upland cotton farmers planted 566,000 more acres than the previous year. Brazil's First Written Submission, para. 143.

        397 Thus, during January, February and March, US growers would be looking at the levels of the

        upcoming December futures contract on cotton, September futures contract for corn, and November futures contract for soybeans. To demonstrate this, the United States presents another figure providing a comparison of US planted acreage to the ratio of cotton futures and soybean futures prices. The United States claims that the figure shows that, when the expected cotton futures price is high relative to the expected soybean futures price, planted acreage tends to increase. When the opposite is true – and the ratio is small – cotton planted acreage tends to decrease. US First Written Submission, paras. 257-259.

        398 Brazil's Rebuttal Submission, para. 215. Brazil asserts that a close examination of the figure presented by the United States demonstrates that US acreage and the cotton to soybean futures price ratio move

        WT/DS267/RW

        Page 96


      22. The Panel finds that the data presented by Brazil supports the argument of Brazil that the cotton futures prices does not seem to explain US cotton planted acreage. The Panel also agrees with Brazil that the evidence provided by the United States regarding the relationship between, on the one hand, the ratio of cotton futures prices to soybeans future prices and, on the other hand, US planted acreage does not provide an adequate explanation for the observed changes, particularly for the period MY 2003 to 2006. The Panel notes, in this regard, that the ratio of cotton futures prices to soybean futures prices and US cotton planted acreage only moved in the same direction in MY 2006. In MY 2003, the ratio of cotton futures prices to soybeans futures prices increased sharply while cotton planted acreage declined. The ratio decreased in both MY 2004 and MY 2005, while planted acreage increased in both years. Thus, even if we agree with the argument of the United States that futures prices of alternative crops to cotton need to be taken into account, the data does not support an important role for this ratio of cotton futures prices to soybeans futures prices.


      23. The major argument raised by the United States to rebut Brazil's argument that US upland cotton producers do not react to market price signals concerns the relative stability of the US share of world production and exports over the entire period of the FSRI Act of 2002. In the view of the United States, the stable US share of world production reflects the fact that US production has increased and decreased in much the same way as production elsewhere in the world, undermining Brazil's claims that the marketing loan and counter-cyclical payment programmes have "insulated" US producers from world market price signals.399 As the United States puts it, whatever signal the "world market" is sending, US producers are receiving it clearly and responding to it in much the same way as their foreign counterparts. 400 Similarly, increases in the volume of US exports during MY 2002-2005 were not disproportionately greater than the increase in the volume of exports from foreign suppliers. In the view of the United States, this indicates that, contrary to Brazil's allegations, US exports were affected by the changes in the marketplace in much the same way as exports of foreign producers.401


      24. Brazil has responded to this by arguing that the relevant economic decision of cotton producers to consider is not production nor exports but planted acreage. However, neither Brazil nor the United States appear to have access to recent data on foreign planted acreage. Brazil presents a figure which compares US planted acreage and foreign harvested acreage to movements in futures prices of cotton.402 Referring to this figure, Brazil contends that changes in foreign harvested acreage and futures prices moved in the same direction in seven out of eight years. By contrast, changes in US planted acreage and futures prices move in the same direction in only three out of eight years. In the remaining five years, they move in different directions.403 The United States objects to the comparison because Brazil compares foreign harvested acreage (instead of foreign planted acreage) with US planted acreage.


      25. The fact that the US share of world cotton production and exports has remained relatively constant in MY 2002-2005 suggests to us that US producers have increased production and exports in proportionately the same way as foreign cotton producers. In our view, Brazil's claim of insulation of US producers from market price signals applies in particular when expected prices are below their intervention levels (the market loan rate or the target price under the counter-cyclical programme).



        image

        in parallel in only 3 of the 8 years identified. In the other five years, the development is either in the opposite direction (3 years) or in the same direction, but to very different degrees (2 years). In Brazil's view, this is hardly evidence of US upland cotton acreage being "generally" responsive to the cotton to soybean futures price ratio.

        399 US First Written Submission, para. 244.

        400 US Rebuttal Submission, para. 285.

        401 US First Written Submission, para. 248.

        402 Brazil's Rebuttal Submission, Figure 9.

        403 Brazil's Rebuttal Submission, para. 224.

        WT/DS267/RW

        Page 97


        However, the degree of insulation is less when expected cotton prices are above their intervention levels. In such a situation, US cotton producers should respond to higher expected cotton prices much in the same way as foreign cotton producers. Thus, the Panel is of the view that the stable US share of world production and exports does not mean an absence of insulation of US producers from market price signals. What it means is that the degree of price insulation that the original panel found is now weaker possibly because prices are not as depressed as during the period examined by the original panel.


        1. Alleged "discernible temporal coincidence" of suppressed world market prices and large marketing loans and counter-cyclical payments


          1. Main arguments of the parties


      26. Brazil argues that the existence of a "discernible temporal coincidence" of suppressed world market prices and large marketing loan payments and counter-cyclical payments is one of many factors that cumulatively support a finding of a causal relationship between marketing loan and counter-cyclical payments and significant price suppression in the world market for upland cotton. Brazil refers to the finding of the original panel of a "discernible temporal coincidence of suppressed world market prices and the price-contingent United States subsidies" and submits that developments in MY 2001-2005 regarding the same factors relied upon by the original panel show that there remains such a discernible temporal coincidence between suppressed world market prices and large marketing loan and counter-cyclical payments. Brazil refers, in this respect, to the increase in US production of upland cotton; the high US share in world production for upland cotton; the historically low level of upland cotton prices; the increase in US exports and the record US share in world exports of upland cotton. Brazil also asserts that during MY 2002-2005, lower world and US prices for upland cotton are matched with higher levels of marketing loan and counter-cyclical payments and that the original panel's finding regarding the influence of the New York futures price and the A-Index price continues to be valid.404


      27. The United States submits that Brazil fails to show any discernable temporal coincidence between the marketing loan and counter-cyclical payment programmes and any alleged significant price suppression. None of the factors considered by the original panel in reaching its conclusion regarding the discernible temporal coincidence between US subsidies and significant price suppression supports such a finding with respect to the marketing loan and counter-cyclical programmes now. First, US planted acreage has been stable for the entire period that the FSRI Act has been in effect and is now substantially lower than during the period examined by the original panel. The fact that production in MY 2005 was larger than in MY 1999 and MY 2001, even though producers actually planted fewer acres in MY 2005, reflects the dramatic improvements in yields. Second, over the period of the FSRI Act the US share of world production has remained stable. Third, prices received by US upland cotton producers have not decreased since the FSRI Act came into effect. Fourth, the fact that the A-Index in MY 2002-MY 2005 was an average of 16 per cent lower than the 1980-2001 average says nothing about whether the marketing loan and counter-cyclical payments programmes are causing present significant price suppression as a decline in prices does not establish price suppression. Moreover, the fact the A-Index has trended downward for more than 25 years and has gone up from the levels prevailing before the entry into effect of the FSRI Act suggests that if there is price suppression, it is not the effect of the marketing loan and counter-cyclical payment programmes. Fifth, the share of the United States in world exports has not increased over the period of the FSRI Act and has even declined in MY 2002. Finally, the low level of imports of upland cotton has little to do with the marketing loan and counter-cyclical payments as US imports of


        image

        404 Brazil's First Written Submission, paras. 153-154; Brazil's Rebuttal Submission, para. 233.

        WT/DS267/RW

        Page 98


        upland cotton have been at low levels for many decades and as US demand for cotton generally has declined as the textile industry shrunk.405


        1. Evaluation by the Panel


      28. In support of its finding of a causal link between certain of the challenged US subsidies and the significant price suppression in the world market for upland cotton, the original panel relied upon the existence of "a discernible temporal coincidence" of suppressed world market prices and the price- contingent US subsidies.


      29. The original panel explained its finding as to the existence of this discernible temporal coincidence as follows:


        "Third, there is a discernable temporal coincidence of suppressed world market prices and the price-contingent United States subsidies. Looking at the period, we see the following:


        • United States production of upland cotton increased from MY 1998 to MY 2001 and, while production dropped in MY 2002, there was still an overall increase in MY 2002 compared to MY 1998;


        • the United States' share of world upland cotton production increased to and remained at a level of approximately 20 per cent;


        • United States prices received by United States upland cotton producers decreased by 34 per cent between MY 1998 and MY 2001;


        • the A-Index in MY 1999 – MY 2002 was, on average, 29.5 per cent below its 1980- 1998 average;


        • United States exports increased by approximately 160 per cent from MY 1998 to MY 2001 and by an even greater percentage from MY 1998-MY 2002;


        • United States share of world exports of upland cotton increased; and


        • United States imports of upland cotton remained at comparatively low levels.


          These data reveal that, over the same period that the subsidies in question were being granted, the United States market generated large supplies of upland cotton. Over this same time period, market revenue of United States upland cotton producers decreased. So did world market prices. There was also a marked increase in United States exports, both absolute and in terms of relative share of world exports. Even taking into account that in 1998, production may have been driven downward by drought and high levels of crop abandonment and that, in 2001, production may have been driven upward by high yields, we see a strong temporal coincidence between the United States subsidies and the drop in United States prices, the drop in – and suppression of – world market prices, the increase in United States exports."406 (footnotes omitted)


      30. The Appellate Body addressed this finding of the original panel as follows:


        image

        405 US First Written Submission, paras. 288-290; US Rebuttal Submission, para. 316.

        406 Panel Report, US – Upland Cotton, paras. 7.1351-7.1352.

        WT/DS267/RW

        Page 99


        "The third reason the Panel provided for finding a 'causal link' was that 'there is a discern[i]ble temporal coincidence of suppressed world market prices' and the price- contingent subsidies. The United States describes this as 'an exercise in spurious correlation'. However, in our view, one would normally expect a discernible correlation between significantly suppressed prices and the challenged subsidies if the effect of these subsidies is significant price suppression. Accordingly, this is an important factor in any analysis of whether the effect of a subsidy is significant price suppression within the meaning of Article 6.3(c). However, we recognize that mere correlation between payment of subsidies and significantly suppressed prices would be insufficient, without more, to prove that the effect of the subsidies is significant price suppression."407


      31. In light of these comments of the Appellate Body, the Panel considers that the existence of a correlation between a subsidy and a particular level of prices is not in and of itself sufficient to establish that the subsidy causes significant price suppression.408 Moreover, the data before the Panel indicates that with respect to several key factors relied upon by the original panel in finding "a discernible temporal coincidence" of suppressed world market prices and the price-contingent US subsidies, the current situation is significantly different from the situation considered by the original panel.


      32. Regarding the evolution of prices of upland cotton, the original panel found that there had been a 34 per cent decrease in prices received by US upland cotton producers between MY 1998 and MY 2001. With respect to exports, the original panel found that there had been "a marked increase in United States exports, both absolute and in terms of relative share of world exports". From MY 1998 to MY 2002, US exports had increased by more than 160 per cent and the share of the United States in world exports had increased from 23.5 per cent in MY 1999 to 39.9 per cent in MY 2002.


      33. For the period MY 2002 to MY 2005, we are unable to detect a similar "strong discernable coincidence" between the US subsidies, the increase in US cotton exports and the drop in world market prices. Instead, what could be discerned is the continuing and sharp rise in volumes of US production and exports. However, these increases in volumes did not translate into a higher share of US cotton in world production or world exports. Nor have the absolute increases in US cotton production and exports been associated with a "broad decline" in cotton prices.


      34. We shall first examine the pattern of US subsidy payments using data from Tables 1 and 2A above409, relating them to the behaviour of the world market price of cotton. Except for MY 2003, when cotton prices saw a sharp increase, the sum of marketing loan and counter-cyclical payments averaged more than US$2 billion dollars annually. Over this four-year period, marketing loan payments made up more than 60 per cent of total payments to cotton producers. The amount of subsidy payments varied inversely with cotton prices. When cotton prices were low, as in MY 2002, 2004 and 2005, significant marketing loan and counter-cyclical payments were triggered. When cotton prices were high, as in MY 2003, both marketing loan and counter-cyclical payments were substantially reduced. This pattern of subsidy payments results in a stabilization of US cotton producers' revenues. Figure 1 below illustrates this inverse relationship of subsidy payments and the A-index for MY 2002-2005.



        image

        407 Appellate Body Report, US – Upland Cotton, para. 451.

        408 We also recall our observations (supra, paras. 10.45-10.49) on the concept of price suppression and on the need for a counterfactual analysis in determining whether price suppression is the effect of a subsidy.

        409 Supra, paras. 10.24 and 10.32.

        WT/DS267/RW

        Page 100


        Figure 1: Subsidy (Marketing Loan and Counter-Cyclical) Payments and World Cotton Prices, MY 2002-2005


        2,000


        1,800


        1,600


        1,400


        Millions of dollars

        Millions of dollars

        1,200


        1,000


        800


        600


        400


        200


        0


        image

        80


        70


        60


        Cents per pound

        Cents per pound

        50


        40


        30


        20


        10


        0

        2002 2003 2004 2005

        Marketing Year


        image

        Marketing Loan CCP A-Index


        Sources: The A-Index is from Exhibit US-25. Total marketing loan payments are taken from rows 2 and 3 of Table 1 above. Counter-cyclical payments are taken from row 3 of Table 2A above.


      35. During the period MY 2002-2005, when these large marketing loan and counter-cyclical payments continued to be made by the United States, the absolute level of US cotton production increased by 38.8 per cent with a much bigger increase in the absolute level of US cotton exports of

    51.6 per cent. Cotton production and exports in MY 2005 were at the highest levels ever achieved by the United States. However, these increases in US production and exports did not translate into an increase in the US share of the world cotton market. As can be observed in Table 3 above, the US shares of world exports and world production have stayed at about 40 per cent and 20 per cent, respectively, during the period analyzed. This means that cotton producers and exporters in the rest of the world were increasing their levels of production and exports at the same time. The increases were to such an extent that, collectively, the foreign producers and exporters maintained their share of the world market.


    10.138 Also, unlike the situation found by the original panel, these increases in the absolute levels of US production and exports did not take place during a period when the market price of US upland cotton producers was declining. The original panel noted that prices received by US upland cotton producers decreased by 34 per cent between MY 1998 and MY 2001. The implication was that US subsidy payments could explain why US cotton production was increasing despite declining market revenues. Our examination of the prices received by US cotton producers in the more recent period does not show an equally dramatic decline. Instead, the average US farm price during the period MY 2002-2005 was 14.3 per cent higher than the average US farm price during the period MY 1999-

    2002.410 In the case of upland cotton spot market prices, the average for MY 2002-2005 was 9.6 per cent higher than the average for MY 1999-2002. No doubt this higher average price of cotton in MY 2002-2005 has been driven by the spike in MY 2003.


    image

    410 This is based on the unweighted marketing year averages of US farm and spot market prices compiled by Brazil in Exhibit Bra-452.

    WT/DS267/RW

    Page 101


      1. Nevertheless, these two factors – first, that other major cotton producers were increasing their production and exports at the same time as US producers and second, that prices received by US cotton producers were not declining dramatically during this period – make it more difficult to discern a pronounced temporal coincidence between the US subsidies, the increase in US cotton exports and the drop in world market prices.


      2. With respect to the trend in world cotton prices, the original panel found "a broad decline in the overall level of these price trends from 1996 through the beginning of calendar year 2002, with intermittent peaks and valleys".411 The original panel qualified this observation of price trends by noting that this is not, in and of itself, conclusive for a determination of price suppression. However, in the panel's causality analysis, the decline in world cotton prices was one of the elements it relied on in finding a temporal coincidence of suppressed world market prices and the price-contingent US subsidies.


      3. An examination of recent world price trends does not show a similar broad decline. Instead, over the period MY 2002-2005, there have been intermittent peaks and troughs. Prices rose strongly from MY 2002 onwards with the A-index reaching a peak of 76.7 cents per lb in November 2003.412 Prices started a sharp decline thereafter with the A-index falling to a low of 48.6 cents per lb. in December 2004. Prices have since climbed with the A-index reaching about 60 cents per lb. at the start of MY 2006. The absence of a broad decline in prices persists even if one takes a slightly longer period of analysis – from MY 1999-2005 – as Brazil does in its Submission (see Figure 2).


        Figure 2: Monthly US and World Upland Cotton Prices


        image

        90


        80


        70


        cents per pound

        cents per pound

        60


        50


        40


        30


        20


        10


        Aug-99

        Aug-99

        Dec-

        Dec-

        Apr-00

        Apr-00

        Aug-00

        Aug-00

        Dec-00

        Dec-00

        Apr-01

        Apr-01

        Aug-01

        Aug-01

        Dec-01

        Dec-01

        Apr-02

        Apr-02

        Aug-02

        Aug-02

        Dec-02

        Dec-02

        Apr-03

        Apr-03

        Aug-03

        Aug-03

        Dec-03

        Dec-03

        Apr-04

        Apr-04

        Aug-04

        Aug-04

        Dec-04

        Dec-04

        Apr-05

        Apr-05

        Aug-05

        Aug-05

        Dec-05

        Dec-05

        Apr-06

        Apr-06

        Aug-06

        Aug-06

        99

        99

        0


        Adjusted World Price

        A (NE) Index

        NY Nearby Futures

        Brazil ESALQ

        U.S. Upland Spot Market

        Farm Price


        image

        image

        image

        image

        image

        image

        Source: Brazil's First Written Submission, Figure 4.


        image

        411 Panel Report, US – Upland Cotton, para. 7.1288.

        412 The price data is from Exhibit Bra-452.

        WT/DS267/RW

        Page 102


      4. In its analysis of price trends, the original panel noted that the world price (the A-Index) during MY 1999-2002 was, on average, 29.5 per cent below its 1980-1998 average. In the current dispute, Brazil extends the price comparison to the MY 2002-2005 period and notes that the A-index remains well below historic price levels. The average A-Index price during MY 1980-1998 was 73.36 cents per pound, while the average during MY 2002-2005 was 58.92 cents per pound.413 However, this historical comparison is subject to an important objection. The United States has argued that the A-Index has trended downwards for more than a quarter of a century, well before the FSRI Act came into effect.414 The long-term decline in non-oil commodity prices is in fact one of the recurring themes of development and international economics. Thus, it is not clear to us that the comparison between the average price during the 1980-1998 period and the average price afterwards is as important as Brazil makes it out to be.


      5. Therefore, while the original panel appears to have attached particular significance to the coincidence of a substantial increase in the US share of world exports and the decline in world market price of upland cotton415, such a coincidence is not clearly discernible from the data for the years since the period considered by the panel.


      6. Brazil attempts to discount some of these facts by arguing that the very fact that the level of planted acreage is stable and that the United States has a consistently high share of world production and of world exports is evidence of a coincidence of subsidies and price suppression because, but for those subsidies, these shares of world production and exports would be significantly lower. Similarly, Brazil argues that the increase in absolute volume of production and exports is evidence of a coincidence between subsidies and price suppression because, absent the subsidies, this increase would not have occurred.416


      7. The question before the Panel is whether, as argued by Brazil, "a discernible temporal coincidence" of subsidies and significantly suppressed prices exists that supports a finding of a causal relationship between subsidies and significant price suppression. To posit that where prices and US shares of production and exports have not increased but remained relatively stable this is the result of the effect of subsidies is to assume the very conclusion that must be demonstrated. The fact that stable shares of world production and exports coincide with the payment of subsidies cannot in and of itself establish the existence of a causal relationship between the subsidies and significant price suppression.


      8. The Panel concludes from its analysis of developments during the period MY 2002-2006 concerning the factors relied upon by the original panel in reaching its conclusion regarding the existence of a temporal coincidence between subsidies and significantly suppressed world market prices for upland cotton, that it is more difficult to find support in these factors for a finding that the effect of the subsidies is significant price suppression under current conditions than at the time of the


        image

        413 Brazil's First Written Submission, para. 101.

        414 US First Written Submission, para. 287.

        415 Panel Report, US – Upland Cotton, para. 7.1352:


        "Even taking into account that in 1998, production may have been driven downward by drought and high levels of crop abandonment and that, in 2001, production may have been driven upward by high yields, we see a strong temporal coincidence between the United States subsidies and the drop in United States prices, the drop in – and suppression of – world market prices, the increase in United States exports." (emphasis added)


        We also note the panel's statement in paragraph 7.1303 of its report that "[t]he effects of these three price- contingent subsidies are, in our view manifest in the movements in upland cotton prices in the same world market during the reference period."

        416 Brazil's Rebuttal Submission, paras. 202-203.

        WT/DS267/RW

        Page 103


        original panel's analysis. At the same time, the Panel recalls that the purpose of its inquiry under Article 6.3(c) of the SCM Agreement is to determine whether, but for the subsidies at issue, the world market price for upland cotton "would have increased [significantly], or would have increased by [significantly] more than was in fact the case".417 The fact that recent years have not witnessed the sharp decline in the world market price for upland cotton that occurred during the period considered by the original panel does not necessarily mean that there is currently no price suppression within the meaning of Article 6.3(c) of the SCM Agreement.


        1. Alleged gap between long-term costs of production and market revenues of US upland cotton producers


          1. Main arguments of the parties


      9. Brazil submits that the widening gap between the total long-term costs of production and the market revenue of US upland cotton producers demonstrates that much of US upland cotton production would not be economically viable without the marketing loan and counter-cyclical payments. During the period MY 2000-2005 US upland cotton producers would have lost US$663 per planted acre, or US$9.84 billion in aggregate, without the marketing loan and counter-cyclical payments.418 Brazil also argues that a survey conducted by the International Cotton Advisory Committee (ICAC) shows that the United States is a high-cost producer relative to other countries producing upland cotton. Brazil considers that the data on costs of production of US upland cotton producers shows that, in the absence of the marketing loan and counter-cyclical payments, many US upland cotton producers would have had to discontinue growing upland cotton and switch to alternative crops. US supply would have been, on average, 15-16 per cent lower in the absence of marketing loan and counter-cyclical payments over the period MY 2000-2005.419


      10. Brazil submits that the United States cannot sidestep the gap between upland cotton revenue and total costs of production by reference to whole-farm or off-farm revenues. Brazil recalls that the United States made a similar argument in the original proceeding and submits that, like the original panel, this Panel should reject the argument of the United States that whole-farm costs and income, including off-farm income, are relevant to an assessment of the effects of the marketing loan and counter-cyclical payments.420 The United States has not answered the question why a farmer would cross-subsidize the production of upland cotton. The fact that US upland cotton producers supplement their income with off-farm work is irrelevant to an examination of costs and returns in respect of upland cotton.421


      11. Brazil also submits that while variable costs of production may play a role in the short-term, US upland cotton producers must meet both short-term variable costs of production and long-term


        image

        417 Supra, para. 10.49. The Panel also recalls the observation of the original panel that "[t]he fact that prices dropped from 1996 through the beginning of 2002 is relevant. However, it is not, in and of itself, conclusive for a determination of price suppression. Nor is the increase from 2002 conclusive that a determination that any price suppression necessarily ceased at that point. We must examine whether these prices were suppressed, that is, lower than they would have been without the United States subsidies in respect of upland cotton". Panel Report, US – Upland Cotton, para. 7.1288.

        418 The data and calculations provided by Brazil in support of this assertion are contained in Exhibit Bra-477. Brazil illustrates this widening gap in a graph in Figure 13 of its First Written Submission. Brazil

        argues that without the Step 2 programme, the effect of which was to raise the domestic price of upland cotton, this gap between costs and revenues would have been even larger. Brazil's First Written Submission, para. 161.

        419 Brazil's First Written Submission, paras. 156-164.

        420 Brazil's Rebuttal Submission, paras. 249-253; Brazil's Oral Statement, para. 83.

        421 Brazil's Comments on US Response to Panel Question 59, para. 141.

        WT/DS267/RW

        Page 104


        total costs of production to be economically viable and continue upland cotton farming.422 Therefore the evidence provided by Brazil, which shows that US upland cotton producers do not cover their long-term costs of production, is sufficient to establish a causal relationship between the subsidies at issue in this proceeding and significant price suppression. Brazil also submits that the United States mischaracterizes certain cost items – land, labour and capital recovery costs – as fixed costs when these costs, which vary with the amount of production in any given year, are in fact variable costs. If these items are correctly treated as variable costs, a significant number of US upland cotton producers are not able to cover their variable costs of production, nor do they expect to be able to cover their variable costs of production.423 424


      12. Based on a comparison of expected returns and total costs of growing cotton, corn or soybeans on a base acre of cotton in MY 2005, Brazil submits that without marketing loan and counter-cyclical payments US upland cotton producers would have had every incentive to switch to production of the other two crops. Similarly, a comparison of expected returns and total costs of growing cotton, corn or soybeans on a base acre of cotton in MY 2007 shows that the effect of the marketing loan and counter-cyclical payments is to mute significantly the incentives that would otherwise exist for upland cotton producers to switch to production of soybeans and corn. A comparison of average returns and total costs associated with planting cotton, corn or soybeans on a cotton base acre in MY 2002-2006 indicates that only US subsidies made growing cotton during MY 2002-2006 economically rational and very profitable and that in the absence of cotton marketing loan and counter-cyclical payments, cotton farmers would have sustained significant losses whereas growing corn or soybeans would have been profitable.425 Brazil argues that the conclusion that in the absence of the marketing loan and counter-cyclical payments it would be far more profitable for US upland cotton farmers to switch to production of corn and soybeans holds even if one makes a comparison of expected and actual returns with cash costs or with variable costs.426


      13. Brazil submits that the cost data relied upon by the United States in support of the proposition that most US cotton producers are able to cover both variable and total costs of production is based on MY 2003 when prices were at their highest level in the past ten years. The fact that even with these high prices eight per cent of US production did not cover variable and total costs of production suggests that in MY 2004 and MY 2005, when prices were much lower, a far greater percentage of US producers were unable to cover variable and total costs of production. Brazil refers in this regard to data on losses made by US upland cotton producers in MY 2004-2005 and to projections of lower yields and increased costs for MY 2006. In any event, instead of breaking cotton farms into arbitrary



        image

        422 Brazil considers that long-term total costs are highly relevant to an assessment of the effects of the marketing loan and counter-cyclical payments even if the Panel uses only MY 2005 as a reference period. Brazil's Response to Panel Question 71.

        423 Brazil's Rebuttal Submission, paras. 257-277 and Figure 10. Brazil submits in this respect that US variable costs of production are expected to be significantly higher in MY 2006-2008 than in MY 2003-2005

        and that expected variable costs are projected to exceed expected market revenue in MY 2006-2008. Brazil's Rebuttal Submission, paras. 276-277. See also Brazil's Oral Statement, paras. 87-89.

        424 In addition, Brazil considers that the United States has provided a misleading presentation of costs

        of production data insofar as certain data offered by the United States excludes the costs of ginning and the revenues associated with cottonseed. USDA costs of production shows that the costs associated with ginning are far higher than the market value of cottonseed. The USDA data also does not make a distinction between costs associated with producing upland cotton lint and costs associated with cottonseed. Nor did the original panel make such a distinction. Brazil's Rebuttal Submission, paras. 279-281.

        425 Brazil's Oral Statement, paras. 63-78, Figures 1-6 and Exhibits Bra-634 and Bra-648.

        426 Brazil's Comments on US Oral Statement, paras 15-17, Figures 7-12, Exhibit Bra-667 and Exhibit

        Bra-668.

        WT/DS267/RW

        Page 105


        cost groups, the Panel should adopt the same approach as the original panel and rely upon national average cost of production data.427


      14. Brazil considers that in criticizing the use made by Brazil of the total cost of production data published by USDA the United States makes an argument that was already rejected by the original panel. The argument that total cost data is not an appropriate indicator of the financial viability of upland cotton farmers because of the inclusion of non-cash costs is in contradiction with basic economic and accounting principles. As explained by USDA, non-cash costs are real economic costs and are part of the total costs that are taken into account by farmers when making their production decisions. The criticism by the United States also ignores the fact that the USDA total cost data reflects valuation methodologies that were conceived by the American Agricultural Economics Association and implemented by USDA.428


      15. Brazil also submits that the evidence does not support the US argument that low-cost producers would expand production and thus offset the decline of production caused by the exit of high-cost producers.429


      16. The United States argues that Brazil does not demonstrate that US producers would switch to alternative crops in the absence of payments under the marketing loan and counter-cyclical payment programmes. The gap between total costs and revenues of growing cotton does not support the argument that cotton plantings and production would have been lower but for the marketing loan and counter-cyclical payments. First, total costs and returns of growing cotton do not explain short-term planting decisions and do not alone explain long-term farming decisions. Short-term decisions on whether to plant cotton or an alternative crop are not made by reference to long-term costs of production but by reference to variable costs of production. While in the long run total costs may be relevant to decisions on whether to continue cotton farming, such decisions are made taking into consideration whole-farm costs and revenues, including off-farm revenues, and are not based only on the costs of producing cotton.430 Second, US cotton producers have not only covered their variable costs of production but also a substantial portion of their total costs from cotton revenues in MY 2002-2005.431 The evidence also shows that over the period of the FSRI Act of 2002, costs per pound of cotton have declined steadily.432


      17. The United States submits that Brazil has not shown that US upland cotton producers are unable to meet their variable costs of production without marketing loan and counter-cyclical



        image


        142-149.

        427 Brazil's Oral Statement, para. 84; Brazil's Comments on US Response to Panel Question 59, paras.


        428 Brazil's Comments on US Response to Panel Question 59, paras. 150-160.

        429 Brazil's Oral Statement, para. 85.

        430 The United States asserts that Brazil also fails to take into account the diversity of US cotton

        producers and the possibility that any exit of high-cost producers would be offset by the expansion of more efficient low-cost producers. US First Written Submission, para. 298.

        431 US First Written Submission and Exhibit US-47. The United States explains that the graph in paragraph 300 of its First Written Submission presents costs of production in two ways:


        "For both variable and total costs, the upper line is the cost for both cotton and cottonseed (i.e. for 'raw ' or seedcotton). The farm price shown above, however, is the price for cotton lint, which is the product at issue here. To calculate costs for cotton lint only that can properly be compared to the upland cotton farm price, the United States has subtracted ginning costs from variable costs of production. This is consistent with the practice in the United States of ginning costs being paid for out of the proceeds gained by the gin from sale of the cottonseed that is separated out in the ginning process. The bottom line for both variable and total costs shows costs net of ginning." US First Written Submission, para. 301 (original emphasis)

        432 US First Written Submission, paras. 302.

        WT/DS267/RW

        Page 106


        payments programmes. First, the economic literature does not support Brazil's argument that the opportunity costs of unpaid labour, land costs and capital recovery costs should be treated as variable cash costs.433 Second, prices were sufficiently high to cover variable costs of production and, in many cases, total costs of production both on a national and a regional basis. A comparison of upland cotton lint costs and returns shows that even classifying paid labour as a variable cost, US upland cotton producers have been able to more than cover their operating costs and much of their overhead costs over the period 1999-2005.434 In addition, US producers are forecast to more than cover operating costs as well as a large share of total costs.435 Contrary to Brazil's argument, the fact that the cost of production data represents averages understates the cost effectiveness of US producers because, as shown by data for 2003, the overwhelming majority of US production took place on farms that not only met their variable costs but also their total costs and only high-cost producers, accounting for approximately 8 per cent of US production, did not cover their operating costs.436 Third, ginning costs are appropriately excluded from the calculation of costs of production for US upland cotton lint in light of the common practice in the United States of ginning costs being paid out of the proceeds gained by the gin from the sale of the cotton seed separated out in the ginning process.437


      18. The United States argues that Brazil has no basis to contest the economic literature showing that long-term decisions about upland cotton production take total costs of producing upland cotton into account as one of many other factors including off-farm income and income from other sources. Brazil's argument that income from other sources is not legally or factually relevant to an analysis of US upland cotton production costs and returns misses the point because the question before the Panel is not one regarding the balance of upland cotton costs of production and returns. The question before the Panel is whether the marketing loan and counter-cyclical payments are causing significant price suppression of the world market price for upland cotton. An analysis, such as Brazil's, that purports to show that US producers would make different long-term production decisions affecting price in the absence of marketing and counter-cyclical payments is not valid if it disregards how US producers actually make these decisions.438


      19. The United States also submits that Brazil has yet to provide evidence that even if some US upland cotton producers were to decide to exit cotton farming in the absence of marketing loan and counter-cyclical payments, this would necessarily result in any significant decline in US cotton plantings.439


      20. With respect to the comparisons offered by Brazil of expected market revenue and total costs of production of growing cotton, corn or soybeans in MY 2005 and MY 2007 and of the average market revenue and total costs of production in MY 2002-2006, the United States submits, first, that variable costs, not total costs of production, are the relevant measure in assessing year-to-year


        image

        433 US Rebuttal Submission, paras. 326-332.

        434 US Rebuttal Submission, paras. 334-335 and Exhibits US-89 and US-90.

        435 US Rebuttal Submission para. 336 and Exhibits Bra-575 and Bra-479.

        436 US Rebuttal Submission, para. 337. As part of its Opening Statement, the United States submitted a table that contains figures slightly revised from the original table in paragraph 337 of the US Rebuttal

        Submission due to a calculation error. US Opening Statement para. 72 and Exhibit US-119. The United States submits in this respect that with respect to 92 per cent of US production of upland cotton Brazil's theory of a gap between costs and revenues fails as a matter of fact.

        437 US Rebuttal Submission, paras. 338-343.

        438 US Rebuttal Submission, paras. 344-347. See also US Response to Panel Question 59, para. 136: "The fact that US upland cotton farmers actually make their decisions about whether or not to close their

        operations based on the total costs and revenues for the farm as a whole – including revenue from all crops as well as from off-farm sources – is necessarily a relevant and important consideration." (emphasis original)

        439 US Rebuttal Submission, para. 348.

        WT/DS267/RW

        Page 107


        planting decisions.440 Second, for the entire period MY 2002-2007 US cotton farmers expected to meet or exceed their variable costs of planting cotton, corn or soybeans, and in terms of the expected market revenues above variable costs, cotton was the most attractive option in at least three of these years.441 Third, while total costs may be a relevant consideration in long-run decisions, Brazil's analysis of total costs is flawed in that Brazil regards total costs of growing cotton as the only consideration for a farmer in deciding whether to continue or to exit farming altogether and in that Brazil includes imputed opportunity costs in its assessment of total costs. In most years under the FSRI Act of 2002 US farmers actually met and exceeded variable costs and even total cash costs. Examining the cost data properly, the cumulative returns over the period 2000-2005 were a positive return of US$161 per acre instead of the cumulative US$663 deficit alleged by Brazil. For the period MY 2000-2006, the cumulative returns were a positive US$131 per acre instead of the US$837 deficit per acre alleged by Brazil.442


      21. The United States considers that the use of total cost data published by USDA is not an appropriate means of assessing the financial viability of farms, let alone of an entire agricultural sector. First the total cost data is an average that obscures the fact that the vast majority of US production occurs on farms that are highly profitable.443 Second, even with respect to the small fraction of US production that is not profitable Brazil's argument is not conceptually sound. Crop- specific data published by USDA cannot be used to reach conclusions on the financial viability of cotton farmers who rely on many sources of income and incur various kinds of costs.444 Moreover, USDA data includes imputed opportunity costs, which are not necessarily costs which farmers must pay off in order to avoid having to close down their businesses.445


        1. Evaluation by the Panel


      22. The question before the Panel is whether a comparison of costs of production of upland cotton and market revenues of US cotton producers supports a finding that marketing loan and counter- cyclical payments are causing significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement. Brazil argues, in this regard, that because of the widening gap between the costs of production and the market revenues of US upland cotton producers much of US upland cotton production would not be economically viable without marketing loan and counter-cyclical payments. The United States considers that Brazil's cost of production analysis is fundamentally flawed and that a proper examination of the data supports the argument of the United States that marketing loan and counter-cyclical payments are not having any significant price suppressing effect.446


      23. The Panel considers that the principal issues on which the parties disagree include the following. First, whether variable costs or total costs should be used to measure the gap between the


        image

        440 US Comments on Brazil's Oral Statement, paras. 60-65.

        441 US Comments on Brazil's Oral Statement, paras. 66-72, Annex I and Exhibit US-139.

        442 US Comments on Brazil's Oral Statement, paras. 73-77 and Exhibit US-124.

        443 US Response to Panel Question 59, paras. 140-141. The United States refers in this respect to the USDA survey data for MY 2003 which shows that 92 per cent of US production was on farms on which total

        revenue exceeded total costs, including the items of economic opportunity cost published by USDA. The United States asserts that, in view of the substantial spread between the total cost figures and the market revenue figures for low- and mid-cost producers, the fact that prices were lower in MY 2004 and 2005 than in MY 2003 does not undermine the conclusion that 92 per cent of US upland cotton production occurs on farms that are profitable.

        444 US Response to Panel Question 59, paras. 143-146.

        445 US Response to Panel Question 59, paras. 148-149.

        446As noted above, Brazil submits that in the period MY2000-2005 US upland cotton producers would have lost US$663 per acre without marketing loan and counter-cyclical payments. The United States argues that

        the cumulative returns were a positive US$161 per acre instead of the cumulative US$663 deficit alleged by Brazil.

        WT/DS267/RW

        Page 108


        market revenues of US cotton producers and their costs of production. Second, whether the examination of total costs relevant to long term decisions, such as whether to exit cotton farming, should be limited to cotton-specific costs or whether account should also be taken of whole-farm costs and revenues. There are, in addition, disagreements between the parties about the composition of variable costs, i.e. whether certain items (land, unpaid labour and capital recovery costs) should be included as variable costs and whether total costs should include imputed values for opportunity costs or whether they should be limited to cash costs. The United States also claims that US cotton producers have not only covered their variable costs but also a substantial portion of their total costs from cotton revenues in MY 2002-2005.


        1. Description of the data on commodity cost and returns


      24. We begin our analysis with a description of the data that has been provided to the Panel. Data on cotton production cost and returns on a per acre basis covering MY 2002-2005 has been provided by both Brazil and the United States. The data is from the Commodity Costs and Returns estimates of the Economic Research Service (ERS) of USDA. The ERS describes these costs and returns estimates as: 447


        "These cost and return accounts are 'historical' accounts based on the actual costs incurred by producers. The costs and returns estimation program uses surveys conducted about every 4-8 years for each commodity as part of the annual Agricultural Resource Management Survey (ARMS), and methods that conform to standards recommended by the American Agricultural Economics Association (AAEA)."


        These standards or recommendations are contained in the Commodity Costs and Returns Estimation Handbook authored by the AAEA Task Force on Commodity Costs and Returns.448


      25. One difficulty with this data is that it decomposes costs of production into "allocated overhead" and "operating" costs rather than into the economic concepts of fixed and variable costs. This stems from the difficulty, which the AAEA Task Force recognizes, in labelling costs as fixed or variable. Instead, the AAEA Task Force recommends that costs should be categorized according to whether they are associated with expendable factors or the services of capital assets. The AAEA Task Force also recommends that all the costs of expendables be allocated to "operating costs" and that all other costs be included in "allocated overhead".449 These recommendations appear to have been followed in the ERS methodology for estimating commodity costs and returns.450


      26. In their analysis of these Commodity Costs and Returns estimates of the ERS, both Brazil and the United States use "operating" and "variable" costs interchangeably. They also use "overhead" and "fixed" costs interchangeably. While much of the debate between the parties is about the relevance of variable and total costs (the sum of variable and fixed costs) to the planting decision by farmers, the cost data that is relied upon by both parties is actually classified as "operating" and "overhead".


        image

        447 Description from the USDA's Economic Research Service website: http://www.ers.usda.gov/Data/CostsAndReturns/.

        448 Exhibit US-88.

        449 Commodity Costs and Returns Estimation Handbook, pages 1-4 and 1-5.

        450 US Response to Panel Question 58, para. 126. The United States explains that the US categorization of "operating" (or "variable") costs differs in some ways from that used by ERS. First, the

        United States, to be more conservative, includes hired labour as an operating cost. Second, the United States adjusts both ginning costs and cottonseed revenue to isolate the costs and revenues for the "subsidized" product. See US Response to Panel Question 58, para. 129.

        WT/DS267/RW

        Page 109


        1. Whether certain items (land, unpaid labour and capital recovery costs) should be included as variable costs


      27. Brazil disputes the inclusion of land, labour and capital recovery costs in fixed costs and argues that these costs are variable costs.451 The United States argues that it is the long-standing practice of USDA and agricultural economists to treat land, labour, and capital recovery costs as fixed costs.452 The United States considers that there is no basis in the economics literature for Brazil's assertion that an imputed value for the opportunity cost of unpaid labour should be treated as a variable cash outlay.453 Following the AAEA Task Force guidelines, the United States maintains that a determinative factor as to whether an input cost can be classified as a fixed or a variable cost is whether the input is an expendable or a capital input. Since land is clearly a capital input, it should be classified under fixed costs.454 The United States adopts a similar approach to the treatment of capital recovery costs.455


      28. The Panel notes the difficulty highlighted by the AAEA Task Force in classifying the many elements of commodity production costs into variable and fixed costs. While variable and fixed costs are the economically relevant cost concepts in this dispute, there appears to be no unambiguous way of arriving at such a classification. The AAEA Task Force has chosen to take a less difficult tack, which is to classify costs according to whether they are associated with expendable factors or the services of capital assets. While this classification of costs is unlikely to fully coincide with the economic concepts of variable and fixed costs, it is the classification that the agricultural economics profession has recommended. Accordingly, it is the method that has been used by the ERS in its collection of commodity costs and returns data. We therefore consider the inclusion of land, labour and capital recovery costs in fixed (overhead) costs as appropriate, being in accordance with the methodology recommended by the AAEA Task Force.


        1. Cash or opportunity costs


      29. The United States raises the issue of cash costs and their role in determining whether farmers have to "close down their businesses". Specifically, the United States argues that in the long run, the relevant comparison to make is between market return or revenue and total cash costs. The opportunity cost of factors of production for which no cash payments are made, such as land that the farmer owns, his own labour or his family's labour, are not like cash costs, which must be paid off in


        image

        451 Brazil submits that hired and unpaid labour vary with the amount of upland cotton planted and produced. If an upland cotton farmer grows nothing at all, he or she will not hire labour. If the farmer decides to plant upland cotton, he or she will hire labour. If he plants crops, he may well have less time to dedicate to off-farm employment the more acreage he chooses to plant. Thus, both the cost of hired labour and the imputed cost of unpaid labour vary with the amount of production. So does the cost of land, whether rented or owned by a farm operator, vary with the amount of upland cotton produced. Finally, the rate of depreciation depends on the intensity of use. Machinery and equipment that is used intensively will depreciate much more rapidly than machinery and equipment that is used less intensively. Furthermore, certain equipment used in upland cotton production, especially harvesting and ginning equipment, can be rented to other producers. Thus, the cost of capital recovery of machinery and equipment is also not entirely fixed; it can vary as a function of the farmer's production decisions. A farmer will, therefore, consider the variable costs associated with capital recovery when deciding whether or not to produce upland cotton. Brazil's Rebuttal Submission, paras. 262-266.

        452 US Rebuttal Submission, para. 326.

        453 US Rebuttal Submission, para. 327.

        454 US Rebuttal Submission, para. 328.

        455 Quoting directly from the Commodity Cost and Returns Handbook, the United States notes that

        capital assets are factors of production that are not used up during a single production period, provide services over time, and retain a unique identity. The term durable asset is often used to describe physical capital because the word durable denotes not temporary or long-lived. It is common in preparing cost of production estimates to assume that durables such as machinery and buildings provide a constant quality of service over their lifetime with regular maintenance. US Rebuttal Submission, para. 331.

        WT/DS267/RW

        Page 110


        order for farmers to avoid having to "close down their businesses". It would appear that the United States is arguing that only when total cash cost exceeds market returns would farmers "close down their businesses".456 In addition, the United States points to the inherent difficulties involved in valuing items (such as land that the farmer owns, his own labour or his family's labour) for which there is no market transaction or cash outlay.


      30. Brazil has clarified that the claim that it is making is that, in the long-term, if total costs incurred in producing cotton exceed market revenues from the sales of cotton, then the cotton farmer would exit cotton farming, not necessarily farming altogether.457 Brazil considers that the argument of the United States with respect to non-cash costs squarely contradicts basic economic and accounting principles. Brazil asserts that the opportunity cost of factors of production for which no cash payments are made are nevertheless real costs.458


      31. The Panel agrees with the United States that imputing values to resources for which there are no market transactions is not as reliable as direct market valuation. It is also true, however, that the methodology used by the USDA conforms to standards recommended by the AAEA, and the application of that methodology probably produces the best estimates that can be achieved. Thus, it is our view that such difficulties in imputation do not pose a problem in using the cost data from USDA to determine whether US farmers were generating enough returns from the market to make it economically worthwhile to continue planting cotton.


      32. The Panel concurs with Brazil that the argument of the United States about total cash costs guiding cotton production decisions is not consistent with basic economic principles. The principle that resource allocation is guided by opportunity cost is a fundamental principle of economics. All resources that a farmer uses to produce cotton have a cost – which is the value that they can obtain when employed in the next best alternative. It does not matter whether the factor must be paid for in cash or not. Unless those resources used in cotton farming fetch a market return that matches their next best use, the farmer is foregoing opportunities to make better returns and would do better by moving to the next best alternative ("exiting cotton farming"). If the next best alternative is another agricultural crop like soybean or corn, exiting cotton farming will not involve moving away from farming altogether. The Panel therefore considers that opportunity costs should be included in the calculation of total costs of production.



        image

        456 See, e.g. US Comments on Brazil's Oral Statement, para. 75.

        457 Brazil's Comments on US Response to Panel Question 59, para. 139.

        458 Brazil quotes the following explanation from the Commodity Costs and Returns Estimation Handbook:


        "[c]ommodity costs and returns include estimates of both cash expenditures and noncash costs. Cash expenditures are incurred when factors of production are purchased or rented. Noncash costs occur when factors are owned. For example, if a farmer fully owns the land used to produce corn, he/she would have no expenditure for land rental or for loans to pay for the purchase of land. Yet, an economic cost arises. By owning the land and using it to grow corn, the farmer foregoes income from other uses of the land, such as renting it to another producer. These costs come about because production resources are limited and have alternative uses. If a farmer uses savings to pay for operating inputs, such as seed, fertilizer, chemicals, and fuel, and thus pays no interest on operating loans, the farmer still incurs an economic cost because the savings could have earned a return in another use. Likewise, the farmer has an opportunity cost of his/her labour used in the production of the commodity because it could have been used on another farm or in off-farm employment."


        Brazil's Comments on US Response to Panel Question 59, para. 152.

        WT/DS267/RW

        Page 111


        1. Total or variable costs of production


      33. Brazil submits that the widening gap between United States upland cotton producers' total long-term costs of production and market revenue over the period MY 2002-2005 demonstrates that much of US production would not be economically viable without the marketing loan and counter- cyclical payments.459 The United States argues that total costs do not explain short-term planting decisions. Short-term decisions on whether to plant cotton or an alternative crop are not made by reference to long-term costs of production but by reference to variable costs of production.460


      34. We note that both parties agree that in the short-term variable costs play a role in cotton producers' decision whether to plant cotton or not.461 Both parties also agree that in the long-term, market revenues should be able to cover total costs of production if farmers are to continue to plant cotton.462 It should be noted that the relevance of variable costs to production decisions in the short- term and of total costs to production decisions in the long-run is consistent with the economic literature. Given these points of agreement, the dispute between the parties is whether the question of a causal relationship between US subsidies and significant price suppression is best answered on the basis of long-run production decisions by cotton farmers or on the basis of short-run (year-to-year) planting decisions.


      35. On this question, we recall that the original panel considered that "a medium- to longer- term examination of developments in the United States upland cotton industry" was appropriate.463 It therefore relied upon "the divergence between United States producers' total costs of production and revenue from sales of upland cotton since 1997" as one factor supporting a finding of a causal relationship between certain price-contingent subsidies and significant price suppression:


        "Fourth, we find credible evidence on the record concerning the divergence between United States producers' total costs of production and revenue from sales of upland cotton since 1997. This supports the proposition that United States upland cotton producers would not have been economically capable of remaining in the production of upland cotton had it not been for the United States subsidies at issue and that the effect of the subsidies was to allow United States producers to sell upland cotton at a price lower than would otherwise have been necessary to cover their total costs.


        If market revenue alone is compared to United States upland cotton producers' total costs of production, the information before us reveals that United States upland cotton producers would on average have lost money for each planted acre in every year since MY 1998, and made a small profit in MY 1997. We do not believe the utility of the record data is fundamentally undermined by any of the criticisms levied by the United States for the purposes of this dispute, particularly as the data are calculated in accordance with a methodology which the USDA itself has deemed to be a sufficiently reliable reflection of United States upland cotton producers' costs and revenues. That the figures before us are cumulated to show a result over the six-year period from 1997-2002 also lends itself to an assessment of the medium- to longer- term examination of developments in the United States upland cotton industry. We believe that the existence of this gap between upland cotton producers' total production costs and market revenue, on the one hand, and the effect of the subsidies,



        image

        459 Brazil's First Written Submission, paras. 156-161.

        460 US First Written Submission, paras. 293-294.

        461 US First Written Submission, paras 293-294, and Brazil's Rebuttal Submission, para. 255.

        462 US First Written Submission, para. 295, and Brazil's Rebuttal Submission, para. 255.

        463 Panel Report, US – Upland Cotton, para. 7.1354.

        WT/DS267/RW

        Page 112


        on the other hand, was to sustain a higher level of output than would have occurred in the absence of the United States subsidies at issue."464 (footnotes omitted)


      36. The original panel relied on total costs of production, notwithstanding that the United States had argued that the use of an average total cost of production figure was not appropriate for constructing an annual revenue gap.465 The panel explained its approach as follows:


        "We believe that not only coverage of variable costs, but also of fixed costs, would be necessary, at least over the medium to longer term, in order for United States upland cotton producers rationally to opt to remain United States upland cotton producers. We find support for this proposition in Appellate Body Report, Canada – Dairy (Article 21.5 – New Zealand and US), paras. 87, 88. Only an analysis of the total cost of production takes account of the economic resources the producer invests in the product. Fixed and variable costs are the total amount which the producer incurs in order to produce the product and the total amount it must recoup, in the long-term, to avoid making losses. To the extent that the producer charges prices that do not recoup the total cost of production, over time, it sustains a loss which must be financed from some other source, or else the producer simply has to close down his business."466


      37. The United States appealed the panel's decision to use total costs rather than variable costs of production.467 The Appellate Body discussed the panel's approach to the consideration of total costs of production as follows:


        "We agree with the general proposition of the United States that variable costs may play a role in farmers' decision-making as to whether to plant upland cotton or some alternative crop, and how much of each crop to plant. From a short-term perspective, variable costs may be particularly important. However, from a longer-term perspective, total costs may be relevant. Based on the evidence before it regarding upland cotton production in the United States, the Panel concluded that 'the six-year period from 1997-2002 ... lends itself to an assessment of the medium- to longer-term examination of developments in the United States upland cotton industry'. The Panel found that 'the effect of the subsidies was to allow United States producers to sell upland cotton at a price lower than would otherwise have been necessary to cover their total costs'. In the circumstances of this dispute, we do not consider that the Panel's reliance on total rather than variable costs of production amounts to an error vitiating the Panel's analysis under Article 6.3(c)."468 (footnotes omitted)


      38. In comparing cotton costs and returns, the original panel considered it essential that this be done on the basis of the medium- to longer- term developments in the US upland cotton industry. We concur with this approach. Under the FSRI Act of 2002, the price-contingent marketing loan and counter-cyclical payments will be available over the lifetime of the legislation, which only lapses in MY 2007. Thus, it appears to us that the effect of the marketing loan and counter-cyclical payments on cotton farmers' production decisions could be best assessed on the basis of data covering the lifetime of the Act. Where this is not possible because a claim of "present" serious prejudice was brought before the end of the Act, the assessment of cost and return will have to be made on the basis of data covering as much of the lifetime of the FSRI Act as possible, in this instance MY 2002-2005


        image

        464 Panel Report, US – Upland Cotton, paras. 7.1353-7.1354.

        465 See e.g. Further Rebuttal Submission of the United States, (18 November 2003), paras. 116-135.

        466 Panel Report, US – Upland Cotton, footnote 1465.

        467 Appellate Body Report, US – Upland Cotton, paras. 40 and 452.

        468 Appellate Body Report, US – Upland Cotton, para. 453.

        WT/DS267/RW

        Page 113


        including projections for MY 2006. This period of time is of sufficient length to require a medium to long-term analysis. Thus, a comparison of market returns with total costs, rather than only with variable costs, is the appropriate benchmark for assessing the effects of the subsidies.


        1. The role of off-farm income


      39. While the United States accepts that total costs may be relevant to the long-term decision as to whether to continue or to exit cotton farming, it argues that this decision is not made on the basis of a cotton-only analysis of costs and returns. Rather, whole-farm decisions will be made taking into consideration whole-farm costs and returns, including off-farm revenues.469 The United States considers that the question of "off-farm income" is important because it bears on Brazil's ability to prove that without marketing loan and counter-cyclical payments, US producers would exit the business of upland cotton farming altogether.


      40. The United States points to two studies that note the increased prevalence of off-farm employment and off-farm income for farm households in the United States.470 Brazil does not dispute these studies or this trend for US agriculture as a whole. However, Brazil points out that the studies submitted by the United States suggest that off-farm income is of particular importance for small farms. Moreover, according to some studies, larger farms account for most US agricultural production.471 Finally, Brazil refers to the original panel's conclusion that off-farm income or income from growing other crops is not relevant to the dispute.


      41. In addressing the argument of the United States concerning the relevance of what it termed "off farm income", the original panel stated:


        "We disagree with the United States that an examination of United States upland cotton producers' 'off farm income' may be a legally relevant consideration in the examination before us. We are considering costs and market revenues in respect of upland cotton. Our examination is one of the upland cotton industry. We find support for this proposition in the references to 'subsidized product' and 'like product' in Article 6.3(c) of the SCM Agreement, and contextual support in such references in Article 15 of the SCM Agreement. We are inquiring into whether, in respect of upland cotton production, United States producers' revenues in respect of upland cotton were insufficient to sustain their total costs incurred. We are not looking into the possibility of cross-subsidization or cross-financing of insufficient market revenues for upland cotton that may have come from other United States industries. Indeed, the very fact that the United States relies on such cross-subsidization as a possible source of revenue would tend to support the proposition that upland cotton producers would have lost money over the longer term if they were involved in upland cotton production alone. If the infrastructure of United States upland cotton production supports the 'off farm income' then the subsidies are supportive of that income as well. If the infrastructure of United States upland cotton production does not support 'off farm income', the producer need not produce upland cotton. However, we have seen, as a factual matter, that United States producers continue to produce upland cotton, and still receive subsidies in relation to production."472



        image

        469 US First Written Submission, para. 295.

        470 Dimitri, Carolyn et. al., "The 20th Century Transformation of US Agriculture" and Hoppe, Robert

        A. and Korb, Penni, "Understanding US Farm Exits. Economic Research Report 21" June 2006, p. 20 (Exhibit US-46).

        471 Brazil's Comments on US Response to Panel Question 59, para. 141.

        472 Panel Report, US – Upland Cotton, footnote 1470.

        WT/DS267/RW

        Page 114


      42. For our part, we have turned to the two studies cited by the United States. The United States describes one of the studies (Dimitri, et al.) as an analysis of changes in US farm structure over the 20th century, and quotes a specific section of it:473


        "[A]bout a third of farm operators worked off the farm for at least 100 days in 1930.

        ... By 1970, more than half of farms had off-farm income, and by 2000, 93 per cent of farms earned off-farm income. Off-farm work has played a key role in increased farm household income; while farm household income was once below the national average, in 2002 it exceeded the national average by nearly US$8,000."


        The United States uses this quotation to substantiate its argument about the growing importance of off-farm income in US farm operators' income.


      43. The second study cited by the United States (Hoppe and Korb) looks at the link between off-farm income and the decision to exit from farming. The United States quotes the following passage from the study (to be found on page 20) and underscores the last sentence:474


        "Off-farm work has become important to farm operators. About one-third of farmers have worked off the farm at least 200 days per year—essentially fulltime— since 1978. Off-farm work could hypothetically affect exits in two ways. First, off-farm work may be the first step in an exit from farming, which would be reflected in higher exits for farms the operators of which work off-farm. Second, off-farm work might lower the probability of exit by providing farm operator households with another source of income."


      44. In the Panel's assessment, the study cited by the United States appears ambivalent about the role of off-farm income in farm exit. On the one hand, having a source of off-farm income could be a factor that reduces the likelihood of exit from farming. Alternatively, finding off-farm work may be the first step to exit from farming. In other words, the increasing importance of off-farm income in the US agricultural sector may just be reflecting the long-term exit of workers from that sector.


      45. The Panel finds stronger evidence of this ambivalence in the very next paragraph of the study in which the authors explain that the econometric model they used to estimate the impact of off- farm income on the likelihood of farm exit supports both hypotheses.475


        "The days-of-work model supports both hypotheses (table 8). On the one hand, working full-time off-farm is associated with a slightly higher exit probability than either of the alternatives (hypothesis 1). On the other hand, combining some off-farm work (1-199 days) with farm work leads to lower exits than working full-time on or off the farm (hypothesis 2). Note, however, that the range in exit probabilities is fairly narrow, about 2 percentage points, for any sales-age category." (footnote omitted)


      46. Thus, while the Panel is convinced that off-farm income has grown in importance in farm operators' income, there is ample room for debate as to the implications of this for the likelihood to exit from farming. The United States emphasizes one possible implication – off-farm work might lower the probability of exit by providing farm operator households with another source of income – for which the study finds some empirical support. But to our mind, a fair reading of the study also requires us not to neglect the alternative interpretation – off-farm work may be the first step in exiting


        image

        473 US First Written Submission, para. 295. 474 US First Written Submission, para. 296. 475 Exhibit US-46, p. 21.

        WT/DS267/RW

        Page 115


        from farming – for which the study finds equally convincing empirical support. Taking this ambivalence of the results into account, the Panel does not think that the issue of off-farm income bears as negatively as the United States believes against Brazil's ability to prove that without marketing loan and counter-cyclical payments, US producers would exit the business of upland cotton farming.


        1. US upland cotton producers' long-term costs of production and market revenues


      47. The Panel has relied on USDA cost of production surveys which combine total costs and total returns from upland cotton lint and cottonseed. This was the type of cost data relied upon by the original panel in making its findings about the US cost of production, "as the data are calculated in accordance with a methodology which the USDA itself has deemed to be a sufficiently reliable reflection of United States upland cotton producers' costs and revenues."476 In this cost of production data, ginning costs were included as part of operating costs.477


      48. In the present dispute, the United States asserts that market revenues should only include revenues from cotton lint and exclude returns from the sale of cottonseed. The reason for this, according to the United States, is that cotton lint is the product at issue in this dispute.478 As a consequence of this focus on cotton lint, the United States argues that ginning costs should be excluded from the calculation of the costs of production. It claims that "this is consistent with the practice in the United States of ginning costs being paid for out of the proceeds gained by the gin from sale of the cottonseed that is separated out in the ginning process."479


      49. Brazil disputes this statement, arguing that the United States has not presented any basis for this assertion that ginning costs are paid for out of the proceeds gained by the gin from sale of the cottonseed.480 In response to Brazil's criticism, the United States submits an exhibit containing an annual survey of eight gins in the mid-south region conducted by a certified accounting firm which includes data on revenue from cottonseed sales, warehouse rebates, gain/loss on the sale of assets, mote sales, interest and other income but no category for producer fees.481 In the view of the United States, this shows that ginning costs are not paid separately by cotton producers.482


      50. While the Panel is of the view that the cost data relied upon by the original panel is a sufficiently reliable reflection of US upland cotton producers' costs and revenues, the Panel has chosen to consider the cost and revenue data in both forms. Firstly, in the form proposed by Brazil and considered sufficiently reliable by the original panel, where revenue from cottonseed and costs from ginning are both included (see Table 6A). Secondly, in the form which the United States proposes, where only revenues from cotton lint are included and the costs of ginning are excluded (see Table 6B).


      51. Table 6A shows market revenue and total costs in the form analyzed by the original panel. Over the MY 2002-2006 period, US cotton producers' market revenue exceeded total costs of production only once, in MY 2003, when cotton prices were at their highest since MY 1997. In all other marketing years, total costs of production exceeded market revenue, sometimes by very significant margins. In MY 2002, for example, total costs exceeded market revenue by 72 per cent. Data provided by Brazil for MY 2006 suggests that total costs will continue to exceed market



        image

        476 Panel Report, US – Upland Cotton, para. 7.1354. 477 Panel Report, US – Upland Cotton, footnote 1468. 478 US First Written Submission, para. 301.

        479 US First Written Submission, para. 301.

        480 Brazil's Rebuttal Submission, para. 279.

        481 Exhibit US-94.

        482 US Rebuttal Submission, para. 340.

        WT/DS267/RW

        Page 116


        revenue. The gap is expected to be about 115.76 dollar per acre, which would be equivalent to more than a quarter of US cotton producers' market revenue.


        Table 6A: Cotton Production Costs and Returns, MY2002-2006


        (US$ per acre)

        Item

        MY 2002

        MY 2003

        MY 2004

        MY 2005

        MY 2006

        1. Market Revenue


        307.83


        561.72


        483.49


        456.69


        450.29

        2. Total Costs

        529.02

        496.74

        501.51

        544.23

        566.05

        a. Operating Costs

        278.00

        304.29

        315.35

        349.26

        362.18

        b. Overhead Costs

        251.02

        192.45

        186.16

        194.97

        203.87

        3. Revenue less Total Costs

        -221.19

        64.98

        -18.02

        -87.54

        -115.76

        Gap between revenue and total costs as share of market revenue

        -71.9%

        11.6%

        -3.7%

        -19.2%

        -25.7%

        Source: Exhibits Bra-648 and US-47.


      52. Table 6B shows market revenue from cotton lint, total costs (less ginning costs) and the difference between market revenue and total costs on a per acre basis in MY 2002-2006. Over the period examined, US cotton producers' market revenue exceeded total costs of production twice, in MY 2003, when cotton prices were at their highest since MY 1997, and also in the year immediately following that. In other marketing years, total costs of production exceeded market revenue, at times by very huge margins. In MY 2002, for example, total costs exceeded market revenue by 84 per cent. Data provided by Brazil for MY 2006 suggests that total costs will continue to exceed market revenue. The gap is expected to be about 56.25 dollar per acre, which would be equivalent to about

    14.7 percent of US cotton producers' market revenue in that marketing year. Cumulated over the five marketing years 2002-2006, US cotton farmers' costs of production have exceeded market revenues by an average of 227.49 dollars per acre.

    WT/DS267/RW

    Page 117


    Table 6B: Cotton Production Costs And Returns

    (US$ per acre)

    Item

    MY 2002

    MY 2003

    MY 2004

    MY 2005

    MY 2006

    1. Market Revenue (Cotton Lint)


    257.88


    489.72


    414.12


    392.16


    381.65

    2. Total Costs

    473.41

    408.41

    403.08

    440.22

    437.90

    a. Operating Costs (W/out

    ginning costs)


    222.39


    215.96


    216.92


    245.25


    234.03483

    b. Overhead Costs


    251.02


    192.45


    186.16


    194.97


    203.87

    3. Revenue less Total Costs


    -215.53


    81.31


    11.04


    -48.06


    -56.25

    Gap between revenue and total costs as share of market revenue


    -83.6%


    16.6%


    2.7%


    -12.3%


    -14.7%

    Source: Exhibits Bra-648 and US-47.


      1. Thus, in both tables a significant gap exists between market revenues and total costs of production. The Panel believes that the existence of this gap between upland cotton producers' total production costs and market revenue is telling. The Panel concurs with Brazil that this data provides further evidence of the essential role played by marketing loan and counter-cyclical payments in covering US upland cotton producers' long-term total costs of production and that this data provides support for Brazil's argument that, but for these subsidies, many upland cotton producers would not have been able to continue growing upland cotton and would have had to switch to alternative crops.484


        1. Other arguments


      2. The United States makes several other arguments with respect to costs and revenue in cotton. First, the United States argues that US cotton producers have covered their variable costs in MY 2002-2005. Second, the United States submits that in almost all years they have covered most if not all of their total costs of growing cotton.485 Third, the United States contends that the overwhelming majority of US production took place on farms that not only met their variable (operating) costs but also their total costs.486



        image

        483 Exhibit Bra-648 does not provide an estimate of ginning costs. Two alternative estimates were calculated for this purpose: (i) the average ginning costs during MY 2002-2005 and (ii) the result of applying the compound growth rate of ginning costs over the MY 2002-2005 period. The second method yielded a higher estimate of ginning costs for MY 2006 ($128.15 per acre) compared to the first approach ($86.60 per acre). But whichever estimate is used still yields the same result that total costs (without ginning costs) exceed market revenues (cotton lint only) in MY 2006. The operating costs of $234.03 per acre shown in Table 6B are the result of removing the higher estimate of ginning costs from variable costs.

        484 Brazil's First Written Submission, para. 163.

        485 US First Written Submission, para. 300.

        486 US Rebuttal Submission, para. 337.

        WT/DS267/RW

        Page 118


      3. The first argument of the United States is certainly borne out by the data in Tables 6A and 6B. Market revenue exceeded variable costs in all five marketing years. The second argument of the United States is somewhat imprecise because the United States offers no benchmark for judging what it means by "most". What must be clear from the outset is that, on average, cotton farming operated at a loss in four of the five marketing years shown in Table 6A and three of the five marketing years shown in Table 6B. Thus, we interpret the second argument of the United States as an attempt to describe how large these losses were in those years. Table 7 attempts to be more precise about the import of this second argument of the United States. It is based on the information in Table 6A above.487 The second row gives market revenue less operating costs, i.e. whatever is left over after paying variable costs. The third row is overhead costs. The last row gives the percentage of overhead costs that could be covered by market revenue after deducting variable costs. The second argument of the United States accurately describes what occurred in MY 2003 and MY 2004 but not what transpired in MY 2002. In MY 2005, market revenue net of operating costs is able to cover a little over half of overhead costs.


        Table 7: US Overhead Costs Covered By Market Revenue, MY 2002-2006


        (US$ per acre)

        Item

        MY 2002

        MY 2003

        MY 2004

        MY 2005

        MY 2006

        Market Revenue (less operating costs)


        29.83


        257.43


        168.14


        107.43


        88.11

        Overhead Costs

        251.02

        192.45

        186.16

        194.97

        203.87

        Part of overhead costs covered

        11.9%

        133.8%

        90.3%

        55.1%

        43.2%

        Source: Exhibits Bra-648 and US-47.


      4. To support its third argument, the United States points to data on the costs of cotton production, by cost group, which shows that only high-cost producers, accounting for approximately 8 per cent of US production, did not cover their total costs in MY 2003. Unfortunately, the United States does not have a similar breakdown by cost group for other marketing years. MY 2003 was the last year in which the USDA conducted a complete cost survey for upland cotton production. It is not clear how representative the results obtained for MY 2003 are since cotton prices were at their highest that year since MY 1997. To provide some perspective on how high cotton prices were in MY 2003, Figure 3 shows average farm and spot prices (taken from Exhibit US-83) from MY 2002 to MY 2005. The farm price in MY 2003 was 39 per cent, 49 per cent and 32 per cent higher than in MY 2002, MY 2004 and MY 2005 respectively. The spot price in MY 2003 was 27 per cent, 32 per cent and 28 per cent higher than in MY 2002, MY 2004 and MY 2005 respectively. Thus, it is reasonable to expect that the proportion of farmers who did not cover their total costs would be significantly higher in those years.


        image

        487 If we use Table 6B, the portion of overhead costs covered by market revenues (less operating costs) are 14.1%, 142.2%, 105.9%, 75.4% and 72.4% for MY 2002, 2003, 2004, 2005 and 2006 respectively. The second argument of the United States accurately describes what occurred in all marketing years except in MY 2002 where only 14.1% of overhead costs were covered. In other marketing years, overhead costs were either fully covered (as in MY 2003 and MY 2004) or about three-fourths were covered (as in MY 2005 and MY 2006).

        WT/DS267/RW

        Page 119


        Figure 3: Average Farm and Spot Prices, MY 2002-2005


        image

        70


        60


        50


        image

        Cents per Lb

        Cents per Lb

        40

        Farm price

        Spot price

        Farm price

        Spot price

        image

        30


        20


        10


        0

        2002 2003 2004 2005

        Marketing Year


        Source: Exhibit US-83.


      5. The Panel considers, however, that these additional arguments skirt around the main issue. The Panel is making an assessment involving the medium to long-term developments in the upland cotton industry. This requires the Panel to compare market revenue not only with variable costs but with total costs. And the picture that emerges from this medium- to long-term analysis is of total costs exceeding market revenue in three of the four marketing years for which we have final data if we use Table 6A and in two of the four marketing years for which we have final data if we use Table 6B. In the Panel's view, it is likely that the total costs of production will continue to exceed market revenue in MY 2006, irrespective of whether ginning costs are excluded or not.


      6. In conclusion, the Panel finds that there exists a significant gap between the total costs of production of US upland cotton producers and their market revenue. The Panel considers that this gap between costs and revenue, when analyzed in conjunction with the magnitude of the marketing loan and counter-cyclical subsidies and their importance as a share of the revenue of US cotton producers488, supports the proposition that the marketing loan and counter-cyclical payments are an important factor affecting the economic viability of US upland cotton farming. The Panel therefore also considers that without these subsidies the level of US upland cotton acreage and production would be considerably lower.489


        image

        488 Supra, para. 10.111.

        489 Thus, based on the evidence before us, we arrive at the same conclusion as the original panel, which found that "United States upland cotton producers would not have been economically capable of remaining in

        the production of upland cotton had it not been for the United States subsidies at issue" and that the subsidies at issue sustained a higher level of output than would have occurred in the absence of the subsidies. US – Upland Cotton, paras. 7.1353 and 7.1355.

        WT/DS267/RW

        Page 120


        1. The economic simulation model submitted by Brazil in this proceeding


          1. Introduction


      7. Brazil submits an economic model to simulate the effects of US subsidies on the world market for upland cotton. Brazil includes the model as part of a set of economic studies which it claims support a finding that marketing loan and counter-cyclical payments continue to cause significant price suppression. The Panel understands that the model is intended to quantify the price suppression that Brazil claims is caused by US subsidy payments. In the original dispute, an economic model was also presented by Brazil. With respect to that model, the original panel stated that:


        "we have taken [Professor Sumner's] analysis … into account where relevant to our analysis of the existence and nature of the subsidies in question, and their effects, under the relevant provision of the SCM Agreement, and have attributed to them the evidentiary weight we deemed appropriate."490


      8. In this case, we have taken the analysis of the model one step further by considering in some detail the arguments made by the parties about the model, its assumptions and results.491 To the extent possible, we have provided our assessment of these arguments, and based on that arrived at an overall conclusion about the simulation results.


      9. Brazil's model is a demand and supply model in log (natural logarithm) difference form which calculates changes from an initial equilibrium where US subsidy payments are in place. The model has just two regions, the US and the rest of the world. The model simulates a counterfactual scenario: how much would be the changes to the world price, cotton production in the US and rest of the world, and cotton exports/imports of the US and the rest of the world if US subsidy payments – principally the marketing loan and counter-cyclical payments – were eliminated?


      10. The model is simulated over the period MY 1999-2008. In order to run this simulation, assumptions have to be made about a number of parameters: the US and foreign elasticities of supply and demand and the coupling factor for marketing loan and counter-cyclical payments. Further, given that expectations about prices drive production decisions, the expected prices have to be determined as well. Three alternative assumptions are made about how farmers form price expectations: (i) the expected price is a one-year lag of actual prices and subsidy amounts, (ii) the expected price is the harvest period futures prices at the time of planting, and (iii) the expected price is equal to the FAPRI commodity outlook that is developed at the same time that growers are forming their expectations and projections.


        1. Main arguments of the parties


      11. Using the results of the simulation model, Brazil asserts that were it not for the marketing loan and counter-cyclical payments, the world price of cotton would have been higher and US upland cotton production and exports would have been substantially lower. For the period MY 2002-2005, the world price would have been higher on average by 8.2 to 8.9 per cent and US cotton production and exports would have been lower on average by 15 to 16.2 per cent and 20.8 to 22.5 per cent



        image

        490 Panel Report, US – Upland Cotton, para. 7.1209 (footnote omitted).

        491 The Panel notes the transparency of the model. The full set of equations is explicitly written out and an electronic version of the simulation model, in the form of an Excel spreadsheet, has been provided to the

        United States and to the Panel. This gives all parties to the dispute and the Panel full access to the model, its assumptions and the results. In fact, the United States has used the Excel spreadsheet provided by Brazil to conduct alternative simulations using a different set of parameter assumptions.

        WT/DS267/RW

        Page 121


        respectively, depending on the assumption made about price expectations.492 For marketing year 2005 alone, Brazil estimates that the removal of marketing loan and counter-cyclical payments increases the world price of cotton by 9.3 to 10.7 per cent, decreases US cotton production by 16.9 to 19.4 per cent and decreases US exports by 23.5 to 26.9 per cent, depending on the proxy used for deriving revenue expectations.493 For MY 2006-2008, Brazil projects a price increase of between 4.44 per cent and 6.05 per cent if both marketing loan and counter-cyclical payments were to be eliminated.494


      12. The United States strongly contests these results. Utilizing the same model provided by Brazil, the United States has carried out its own simulations using a different set of parameter values derived from the US and world cotton models developed by the FAPRI as well as another set of parameter values taken from the UNCTAD-FAO Agricultural Trade Policy Simulation Model (ATPSM). Using parameter values from the FAPRI cotton model, and assuming that the expected price is a one-year lag of actual prices, the removal of marketing loan and counter-cyclical payment programmes is projected to increase the world price by only 1.41 per cent over the period MY 2002- 2005 and 0.96 per cent over the period MY 2006-2008. Using the supply and demand elasticities from the ATPSM model, and assuming that the expected price is a one-year lag of actual prices, removal of marketing loan and counter-cyclical payment programmes is projected to lead to an increase in the world price of 2.26 per cent over the period MY 2002-2005 and 1.52 per cent over the period MY2006-2008.495 The United States argues that the results for the other price expectation models are similar to that obtained from the lagged price model.


      13. A comparison of the impact on world prices of the simulations performed by Brazil and the United States, using the one-year price lag assumption, is shown in Figure 4 below. Brazil's simulation results are shown by the top curve. The results of two US simulations are shown: the bottom curve is the outcome from using FAPRI elasticities while the middle curve shows the projected increase in the world price using the ATPSM elasticities.


        image

        492 Brazil's First Written Submission, Annex I, para. 82.

        493 Brazil's First Written Submission, Annex I, para. 81.

        494 These are the simple averages derived from Tables 2.A.1 and 2.A.3 of Brazil's First Written Submission, Annex I.

        495 US First Written Submission, Annex 1, paras. 3-4.

        WT/DS267/RW

        Page 122


        Figure 4: Simulation Results Using Different Parameter Assumptions


        16%


        14%


        12%


        Change in World Price

        Change in World Price

        10%


        8%


        6%


        4%


        2%


        0%


        image

        1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

        image

        image

        Marketing Year


        imageBrazil US FAPRI US ATPSM

        imageBrazil US FAPRI US ATPSM


      14. The disagreement between the parties relates to two issues. The first issue has to do with adequacy of the model itself. The second issue has to do with the appropriate choice of parameter values – primarily demand and supply elasticities. Although we have tried to treat each issue separately, it is important to note that the two issues are closely related.


        1. Structure of the model


      15. There appear to be several criticisms by the United States against the model submitted by Brazil. First, the model is based on a simpler and more reduced formulation than the FAPRI model and abandons critical FAPRI parameters.496 The United States notes, in this respect, that in the original proceeding, in which Brazil also submitted a model, Brazil emphasized the similarity of its model with the FAPRI model. Second, as a consequence of this more simple and reduced formulation, the model is unable to take cross-commodity impacts and cross-price elasticities into account, potentially leading to biased price effects.497 Third, Brazil's model does not allow for changes in cotton stock levels and has no stock equation, which is another problem with the model that will tend to accentuate price changes.498 Fourth, the United States also objects to the way in


        image

        496 US Rebuttal Submission, Annex I, para. 4.

        497 US Rebuttal Submission, Annex I, para. 4. See also Statement of Dr. Joseph W. Glauber, paras. 3-

        4.

        498 US Rebuttal Submission, Annex I, para. 4.

        WT/DS267/RW

        Page 123


        which different types of government payments are modelled differently, although they operate in the same way, or are directly incorporated in the revenues of the producers.499


      16. Brazil's defence is that the far more straightforward model it has used can be readily focused specifically on effects in cotton markets and the removal of cotton subsidies alone.500 Brazil explains that this approach would make the analysis more accessible to other parties and to the Panel. This single purpose use of the model – to address the question of significant price suppression of US subsidies – also explains its use of parameters that were more applicable to the case at hand rather than "all-purpose—one size fits all" parameters that are imbedded in the FAPRI system.501


        1. Parameters of the model


      17. The United States argues that many of the parameter values chosen by Brazil are "inappropriate" and "bias" the effects of the removal of US cotton subsidies on world market prices.502 The United States contends that the parameter values chosen by Brazil differ substantially from the parameter values used in the model ("Sumner I") submitted in the original dispute and with the parameters used in the FAPRI cotton models. Further, it criticizes Brazil for not providing any statistical evidence or validation for its choice of elasticities.503


      18. Brazil argues that the parameters it has chosen are more applicable to the case at hand rather than the "all-purpose—one size fits all" parameters that are imbedded in the FAPRI system. It claims that the "cumbersome nature" of the FAPRI modeling system means that it is difficult to recognize in the parameterization that different degrees of response would be expected under different policy circumstances.504


        US Supply Elasticity


      19. The parties disagree about the size of the US supply elasticity with Brazil claiming a larger value (0.8) than the United States (0.21).


      20. Brazil explains that the use of a higher supply elasticity is justified by the nature of the policy experiment being contemplated, a large and permanent reduction in per unit revenue, entailed by the elimination of the marketing loan and counter-cyclical programmes.505 Brazil cites empirical studies which suggest that by eliminating the constraints which exist under US subsidy programmes, US farmers' supply response will become substantially larger.506 Brazil claims that what it uses is a general equilibrium supply elasticity reflecting the endogenous changes in the prices of other crops.507 Brazil also expresses doubts about the accuracy and applicability of the estimates of regional supply elasticities that the United States uses to come up with a national estimate of the US supply elasticity.508


      21. The United States has used the FAPRI regional acreage equations to calculate a weighted average cotton supply elasticity for the country as a whole. This produced a considerably lower


        image

        499 US Rebuttal Submission, Annex I, para. 4.

        500 Brazil's Rebuttal Submission, Annex I, para. 4. 501 Brazil's Rebuttal Submission, Annex I, para. 6. 502 US First Written Submission, Annex 1, para. 1. 503 US Rebuttal Submission, Annex I, para. 12.

        504 Brazil's Rebuttal Submission, Annex I, para. 6. 505 Brazil's Rebuttal Submission, Annex I, para. 8. 506 Brazil's Rebuttal Submission, Annex I, para. 10. 507 Brazil's Rebuttal Submission, Annex I, para. 14. 508 Brazil's Rebuttal Submission, Annex I, para. 18.

        WT/DS267/RW

        Page 124


        estimate than that obtained by Brazil.509 The United States counters Brazil's explanation about the high US supply elasticity by pointing to equally ambitious policy scenarios that have been simulated by the FAPRI cotton model which were conducted without modifying the size of the supply elasticity.510 The United States questions the relevance of the empirical literature cited by Brazil since this literature analyzed the acreage response of producers during the early 1990s when cotton producers were allowed only limited planting flexibility under the 1990 farm bill.511 The United States also alleges that the magnitude of the elasticity chosen by Brazil suggests a neglect of cross-commodity price impacts.512


        Foreign Supply Elasticity


      22. Brazil's model assumes a foreign supply elasticity of 0.2. The United States criticizes this choice as being too low and gives an alternative value of 0.33.


      23. Brazil gives several reasons for why the foreign supply elasticity should be small. The incentive to change production faced by foreign producers is indirect, relatively small compared to the revenue loss faced by US producers, and in the opposite direction from that faced by US producers.513 Further, the transmission of world price changes to local prices occurs with a lag and, for some countries, only partially, due to trade barriers, institutional constraints, transportation costs, and costly information, among others factors.514


      24. The United States obtains its value of the foreign supply elasticity from the FAPRI international cotton model.515 It objects to Brazil's explanation about the incentive to change production faced by foreign producers being indirect and relatively small since the foreign supply elasticity measures the responsiveness of foreign supply to changes in foreign producer revenue (not US producer revenue).516 While the United States agrees that the factors mentioned by Brazil, such as trade barriers, institutional constraints, transportation costs, and costly information can have an impact on local cash prices, it believes that these impacts are captured in the FAPRI's international cotton model's estimation of detailed supply-response equations for the major cotton-producing countries.517


        US and Foreign Demand Elasticities


      25. Brazil assumes identical US and foreign demand elasticities of -0.2. The United States assumes a US demand elasticity of -0.822 and a foreign demand elasticity of -0.39.


      26. The United States criticizes the absence of a stock or inventory adjustment in Brazil's model. Shocks to supply and demand have to be absorbed only through price changes as there is no scope for absorption through stockholding behaviour.518 To account for stocks in their analysis, the United States considers a total demand elasticity for mill use and ending stocks based on MY 2003- 2005 mill use and ending stock levels and the estimated mill use and stock demand equations from the FAPRI model.519


        image

        509 US First Written Submission, Annex I, para. 19. 510 US First Written Submission, Annex I, para. 13. 511 US Rebuttal Submission, Annex I, para. 14.

        512 US First Written Submission, Annex I, para. 15. 513 Brazil's Rebuttal Submission, Annex I, para. 21. 514 Brazil's Rebuttal Submission, Annex I, para. 23. 515 US First Written Submission, Annex I, para. 23. 516 US Rebuttal Submission, Annex I, para. 20.

        517 US Rebuttal Submission, Annex I, para. 21.

        518 US First Writtten Submission, Annex 1, para. 25.

        519 US First Written Submission, Annex 1, para. 27.

        WT/DS267/RW

        Page 125


      27. Brazil explains that the lack of stock adjustment in its model is a consequence of the counterfactual scenario that is being simulated. The counterfactual involves an expected and permanent change in the overall market price level of cotton. In Brazil's view, such temporary increases or decreases in stocks in response to fluctuations in market prices are irrelevant to the demand response to the anticipated and permanent policy changes being examined in this proceeding.520


        CCP Coupling Factor


      28. Finally, Brazil and the United States disagree on the appropriate value to assign to the "coupling factor" linking counter-cyclical payments to market price. This coupling factor provides some measure of the production-enhancing effect of subsidies. Brazil uses a coupling factor of 0.4 for counter-cyclical payments while the United States argues that the correct coupling factor is 0.25, which is that assumed in the FAPRI model.


        1. Evaluation by the Panel


      29. The Panel notes the advantage of the modeling approach that Brazil has chosen to adopt. A simpler and more accessible model can be readily focused on the specific issue at hand. It should include only the relevant economic features that are crucial to a determination whether certain US Government payments led to significant price suppression.


      30. However, the Panel is also mindful of the criticism by the United States that Brazil's model "has no foundation within economic circles".521 The model has been submitted for the first time to this dispute. This means that the model needs to earn the confidence of this Panel.


      31. Furthermore, the magnitude of the impact of US subsidies on the world price depends on the parameter values assumed for Brazil's model. Each party has presented us with arguments why the parameter values it has chosen in the simulation are better than those used by the other party. The Panel does not take a position on whether the FAPRI or ATPSM parameters are better than those chosen by Brazil. Thus, the Panel is not in a position to judge the claims of the parties about the exact magnitude by which the world price would rise were marketing loan and counter-cyclical payments to cease entirely.


      32. While the Panel has refrained from making a determination about the magnitude of the impact of the US subsidies on the world price, the Panel nevertheless finds that all the simulations conducted by the parties support the view that US marketing loan and counter-cyclical payments have led to an increase in US production and exports of cotton that have then suppressed world prices. The Panel takes note of the fact that price suppression has been the outcome of all the simulation results whether one uses the parameter values proposed by Brazil or the FAPRI and ATPSM parameter values proposed by the United States.


        1. Impact of the elimination of Step 2 payments


          1. Introduction


      33. The United States eliminated the Step 2 programme effective as of 1 August 2006. The Step 2 programme was a special marketing loan provision for upland cotton that had been authorized since 1990 under successive legislation, including the FAIR Act of 1996 and the FSRI Act of 2002. It provided for the issuance of marketing certificates or cash payments to eligible domestic users and


        image

        520 Brazil's Rebuttal Submission, Annex I, para. 29.

        521 US Rebuttal Submission, Annex I, para. 7.

        WT/DS267/RW

        Page 126


        exporters of eligible upland cotton when certain market conditions existed such that US cotton pricing benchmarks were exceeded.522 As such, the programme combined elements of export subsidization, when payments were made to US exporters of eligible cotton, and import substitution, when payments were made to eligible domestic users of US upland cotton.


      34. The parties disagree on the effect of the elimination of the Step 2 programme on US production and exports of cotton.


        1. Main arguments of the parties


      35. Brazil submits that the Step 2 programme caused price suppression independent from and in addition to the significant price suppression caused by the marketing loan and counter-cyclical payments programmes.523 Although modest in comparison with the much larger effects of marketing loan and counter-cyclical payments, the world market price effects of Step 2 payments, in isolation, are clearly recognizable. Brazil submits, in this connection, that the removal of Step 2 payments will enhance the adverse effects of the counter-cyclical payments because it will result in lower US domestic prices and thereby lead to higher counter-cyclical payments.524 Brazil refers to studies by the US Congressional Budget Office and by researchers of the University of Missouri that predict that the repeal of the Step 2 programme will only have a modest downward impact on the revenues of US upland cotton producers because of the increase in counter-cyclical payments caused by the decline in domestic prices.525 Brazil also cites to studies that predict that repeal of the Step 2 programme will have no significant impact on US production and exports of upland cotton and on world market prices.526 The relatively modest effects of the removal of Step 2 payments on US production and exports mean that marketing loan payments and counter-cyclical payments continue to be the main drivers of price suppression in the world market for upland cotton. That the Step 2 programme did not affect the causal link between the marketing loan and counter-cyclical payment programmes and significant price suppression is evidenced by the fact that the planted acreage for MY 2006 is


    image

    522 Panel Report, US – Upland Cotton, para. 7.209.

    523 Brazil's First Written Submission, paras. 191-208.

    524 Brazil argues that the degree to which counter-cyclical payments will offset the loss of Step 2

    payments depends on the level of the average farm price in a particular marketing year. Counter-cyclical payments are made if the average farm price is below a "trigger price" of 65.73 cents per pound. In the rare instances when the average farm price is above the counter-cyclical trigger price, the loss of Step 2 payments will negatively impact upland cotton producers' revenue. This is because any downward movement in the average farm price that is caused by the elimination of Step 2 payments, and yet does not lower prices below

    65.73 cents per pound, will not be compensated by increased counter-cyclical payments. However, Brazil notes that in each of the last nine marketing years, the average US farm price has been below the counter-cyclical payment trigger price. It is therefore unlikely that domestic prices will rise to the level where counter-cyclical payments will not compensate US producers for a fall in domestic prices following the withdrawal of the Step 2 programme. Brazil notes that if the average farm price is below 52 cents per pound, a further decline in the price level will also not be compensated by counter-cyclical payments because the maximum counter-cyclical payment will have already been made. Thus, counter-cyclical payments will offset the loss of Step 2 payments where the average farm price is between the marketing loan rate of 52 cents per pound and the trigger price of

    65.73 cents per pound for counter-cyclical payments. Brazil's First Written Submission, paras. 198-200.

    525 Brazil's First Written Submission, paras. 201-203.

    526 According to Brazil, the study conducted by the University of Missouri researchers found that the removal of the Step 2 payments would reduce US production by about 1 per cent, US exports by 2 per cent and

    increase world market prices by 0.39 cents per pound, or less than 1 per cent. Brazil also refers to an ICAC study that concluded that elimination of the Step 2 programme would have no significant effect on US exports of upland cotton. As support for its argument concerning the limited impact of the removal of Step 2 payments, Brazil also relies on the simulation model in Annex I to its First Written Submission and the expert testimony in Annex II to its First Written Submission. Brazil's First Written Submission, paras. 203-204.

    WT/DS267/RW

    Page 127


    estimated to be 7 per cent higher than in MY 2005, notwithstanding that US cotton producers knew that the Step 2 payments would be eliminated as of August 2006.527


      1. Brazil asserts that, contrary to what the United States appears to argue, the Step 2 programme and the mandatory payments made thereunder were not the sole cause of the serious prejudice found by the original panel, and that the arguments of the United States in this regard are inconsistent with the findings of the original panel and the Appellate Body concerning the nature, magnitude and effects of the price contingent basket of measures found to cause serious prejudice.528 Brazil emphasizes that its argument that the relative effects of the Step 2 subsidy are modest compared to the combined effects of the marketing loan and counter-cyclical payments is consistent with the reality of the operation of the three subsidies. The removal of the Step 2 programme will have a relatively insignificant impact on the amount of marketing loan payments. The increase in counter-cyclical payments following the repeal of the Step 2 programme far outweighs the decrease in marketing loan payments. 529 Brazil further submits that US upland cotton producers' revenue is critically dependent upon marketing loan and counter-cyclical payments but not on Step 2 payments. Thus, the repeal of the Step 2 programme has a very limited impact on US upland cotton producers' expected or actual revenues.530 Finally, partial MY 2006 data regarding US exports does not support the argument of the United States that there are no significant production, export and world market price effects from the marketing loan and counter-cyclical subsidies. The primary cause of the decline in US exports is the world demand and supply situation. Moreover, the drop in exports in MY 2006 is not remarkable in light of the surge in exports that occurred immediately prior to the repeal of the Step 2 programme. USDA projections for MY 2006 indicate that the decline in exports will be temporary.


      2. The United States submits that Brazil's arguments about the "relatively modest" impact of removal of the Step 2 programme are without merit. First, eliminating the Step 2 programme likely does not result in an increase in the counter-cyclical payment rate but does result in lower marketing loan payments. At the season average farm price that is projected for MY 2006, there cannot be any appreciable increase in the counter-cyclical payment rate in this current marketing year.531 At the same time, the evidence suggests that elimination of the Step 2 programme will likely lead to lower marketing loan payments as a result of the increase in world market prices.532 Rather than enhancing the adverse effects of counter-cyclical payments, as alleged by Brazil, the elimination of the Step 2 programme is likely to minimize the effect of marketing loan payments. Second, Brazil's efforts to


        image

        527 In the section of Brazil's First Written Submission on the magnitude of the subsidies at issue, Brazil argues that elimination of the Step 2 subsidy from the analysis does not significantly reduce the overall magnitude of the price-contingent subsidies. According to Brazil, during MY 1999-2002 Step 2 payments represented only 13.7 per cent of the magnitude of the three price-contingent subsidies that the original panel found to stimulate production and exports and to result in lower world market prices than would prevail in their absence. Brazil's First Written Submission, para. 112.

        528 Brazil's Rebuttal Submission, para. 48.

        529 Brazil points out that a FAPRI study referred to by the United States estimates that without the Step

        2 payments the AWP would increase by an average of 0.4 cents per pound in MY 2006-2012 and that a study by the US Congressional Budget Office (CBO) estimated the reduction in marketing loan payments as a result of the elimination of the Step 2 subsidy to be US$17 million over a period of ten years. According to the CBO, counter-cyclical payments will increase over the same period by US$484 million. Brazil's Rebuttal Submission, paras. 59-61.

        530 Brazil's Rebuttal Submission, paras. 66-75.

        531 US First Written Submission, paras. 159-170. The United States submits that the projected average

        farm price for MY 2006 is only a fraction of a cent higher than the marketing loan rate. When the farm price is at or below the marketing loan rate the counter-cyclical payment rate remains fixed at its maximum level. The United States notes that none of studies cited by Brazil as support for its argument that counter-cyclical payments are likely to increase predict such an increase for the current marketing year.

        532 US First Written Submission, paras. 171-178. The United States submits that the US Congressional

        Budget Office, FAPRI and ICAC have also found that elimination of the Step 2 programme will likely have the effect of reducing the amount of marketing loan payments.

        WT/DS267/RW

        Page 128


        demonstrate the "modest" effects on the market of eliminating the Step 2 programme are undermined by its own arguments in the original panel proceeding.533 Third, data from August-October 2006 do not support Brazil's claim that elimination of the Step 2 programme has had "modest" effects. In this respect, the United States submits, in particular, that the current level of US exports of upland cotton is far below recent levels.534


        1. Evaluation by the Panel


        1. Assessing the effects on US cotton exports


      3. Brazil argues that the elimination of Step 2 payments has not diminished the significant price- suppression from marketing loan and counter-cyclical payments. While Step 2 payments by themselves have an independent price suppression effect, Brazil believes that these effects are relatively "modest" and that other US subsidies, principally counter-cyclical and marketing loan payments, will be sufficient to ensure that US cotton producers continue to produce and export cotton. The United States disputes that the effect of eliminating Step 2 payments will be "modest".


      4. In this connection, the United States refers to certain trends which it argues do not support Brazil's argument regarding the modest effects of the elimination of Step 2 payments, particularly the decline in US exports and the US share of world exports in MY 2006.535 Brazil argues that the "temporal coincidence" between the removal of Step 2 payments and a decline in US exports is not remarkable.


      5. The Panel concurs with the United States that the available evidence as of February 2007 suggests significant reductions in the volume of production and exports of US upland cotton in MY 2006. The absolute volumes of US cotton production and exports in MY 2006 are projected to fall by 9 per cent and 19.6 per cent, respectively, from MY 2005 levels. More importantly, the US share of world production is projected to fall to 18.6 per cent in MY 2006 from 20.9 per cent in MY 2005; the US share of world exports is projected to decline to 36.3 in MY 2006 per cent from 40.1 per cent in MY 2005. This means that decreases in the absolute levels of US production and exports are far sharper than for the rest of the world. The Panel also notes a document submitted by the United States that contains the results of the annual survey of planting intentions conducted by the National Cotton Council, which indicates that US upland cotton plantings are likely to be down an average of 14 per cent in MY 2007 from 2006 levels.536



        image

        533The United States submits that Brazil's argument on the relative size of the Step 2 payments as compared to the size of marketing loan and counter-cyclical payments is inconsistent with the position taken by Brazil in the original proceeding that the focus of a serious prejudice analysis should be on the effects and not on the magnitude, value or amount of the subsidy. Similarly, Brazil's argument that Step 2 payments have no impact on the level of production and exports of US upland cotton cannot be reconciled with the argument of Brazil in the original proceeding that the Step 2 programme had a significant effect on the level of US exports of upland cotton. The statement by Brazil's expert that elimination of the Step 2 programme will have little impact on the world price for cotton in the long-term is inconsistent with the statement by the same expert in the original proceeding that Step 2 was one of three factors that enabled the United States to act as the driver of world prices. US First Written Submission, paras. 181-189.

        534 US First Written Submission, para. 193.

        535 According to the United States, among these trends for MY 2006 are: (i) US exports in MY 2006

        are projected to be lower than in the previous marketing year; (ii) the US share of world exports is projected to fall to 36 per cent in MY 2006, (iii) world market prices are higher now (early in calendar year 2007) than year- ago levels, while US prices – as reflected by the New York futures market – are running well below those levels.

        536 Exhibit US-115.

        WT/DS267/RW

        Page 129


      6. The Panel considers that the data before it is certainly consistent with the argument that the elimination of Step 2 payments is having an effect on US export performance. The Panel believes that some part of the decline in the US share of the export market reflects the effect of the elimination of Step 2 payments. This is only consistent with the idea that the removal of payments which subsidize exports and which subsidize domestic users to substitute domestic production for imports will eliminate incentives to increase sales in the export market and to add to domestic production. Our view is also consistent with the results of the economic simulation model submitted by Brazil, which predicts a reduction in US production and exports. However, given that we are not yet one calendar year removed from the elimination of Step 2 payments, and that the amount of Step 2 payments was much smaller than the amounts of payments made under either the marketing loan or counter-cyclical programmes537, it is not possible for the Panel to determine with any great deal of precision or confidence just how much of the projected decline in US exports of upland cotton is due to the elimination of the Step 2 payments.


        1. Indirect effect on marketing loan and counter-cyclical payments


      7. The parties have also advanced arguments to suggest that the elimination of Step 2 payments has indirect effects, through consequential changes in the amount of counter-cyclical or marketing loan payments, which either negate or accentuate the direct impact of the elimination of the Step 2 payments.


      8. In the Panel's view, all things being equal, an export subsidy will increase the domestic price in the exporting country because a larger part of domestic production is diverted to the export market. Assuming that the exporting country is a large country, i.e. that it can affect world prices, the increased exports will also tend to depress the world price.538 The export subsidy creates a wedge between the domestic price in the exporting country and the world price. A measure that subsidizes substitution of domestic production for imports will have a similar effect as export subsidies on domestic and world prices. The additional demand for domestic production puts upward pressure on the domestic price; if the country is a large country, the reduction in its imports puts downward pressure on the world price. Thus, eliminating a programme which has both elements of export subsidization and import substitution will reverse the process, lowering the domestic price even as the world price is increased.


      9. Brazil argues that the decline in the domestic price of cotton in the United States, which comes about from the elimination of Step 2 payments, will trigger additional counter-cyclical payments.539 The United States disagrees with Brazil that the elimination of Step 2 payments will


        image

        537 In MY 2005, payments under the marketing loan and counter-cyclical programmes were about US$1.257 billion and US$838 million respectively. Table 8 below gives the amounts of payments under the Step 2 programme from MY 1999 until MY 2005. During its last marketing year, total payments under the programme amounted to US$350 million.


        Table 8: Step 2 Payments, MY 1999-2005

        (million US$)

        ITEM MY1999 MY2000 MY2001 MY2002 MY2003 MY2004 MY2005

        Step 2 payments $422 $236 $196 $429 $350 $544 $350

        Source: Brazil's First Written Submission, Table 6.

        538 We find support for this view from the analysis of export subsidies in standard international economics textbooks such as that from Paul R. Krugman and Maurice Obstfeld (2003) International Economics: Theory and Policy (6th edition) Boston, San Francisco: Addison-Wesley, 2003.

        539 Supra, para. 10.225.

        WT/DS267/RW

        Page 130


        lead to a decline in domestic cotton prices that will result in compensating increases in counter- cyclical payments.540


      10. Further, the United States argues that the elimination of the Step 2 programme is likely to result in lower marketing loan payments.541 Thus, for the United States, the elimination of Step 2 payments will have spill-over effects on marketing loan payments, which accentuate rather than negate the direct impact of eliminating the export subsidy programme. Brazil does not seem to dispute this claim although it argues that the decrease in marketing loan payments is likely to be small.542


      11. Brazil and the United States emphasize different aspects of the likely adjustment in domestic (US) and world prices resulting from the elimination of the Step 2 programme. Brazil emphasizes the reduction in the domestic price, which it argues, would tend to trigger additional counter-cyclical payments. The United States emphasizes the increase in the world price, which it argues, would tend to reduce marketing loan payments. Both parties are correct in their assessment of how domestic (US) and world cotton prices are likely to evolve, all things being the same, with the elimination of Step 2 payments. In our view the scenarios claimed by the parties – increases in counter-cyclical payments and decreases in marketing loan payments – will be played out simultaneously.


      12. There are two questions remaining. First, what are the likely magnitudes of the changes in counter-cyclical payments and marketing loan payments that will be caused by the elimination of the Step 2 programme? The second and more important question is what are the likely impacts on US cotton production and exports?


      13. On the first question, we find credible the US CBO projections, which both parties have referred to, of the likely changes in marketing loan and counter-cyclical payments that will result from the elimination of the Step 2 programme. Over the US fiscal year period 2006-2015, the CBO projects that the elimination of the Step 2 programme will reduce marketing loan payments by US$17 million and increase counter-cyclical payments by US$484 million. In this respect, one should take note of the clarification made by the United States that the projected reductions in marketing loan payments will apply only to MY 2006-2007, since no payments are expected beyond that.543


      14. In answering the second question, the Panel observes that it will not suffice to just compare the magnitude of these payments. This is because a dollar of marketing loan payments provides a greater incentive to increase production than does a dollar of counter-cyclical payments. These projected changes are also rather modest, if one takes into account the scale of the annual payments made under the marketing loan and counter-cyclical programmes, and the ten-year horizon involved. The Panel is led to conclude that the indirect impacts on US production and exports are likely to be small, given the relatively modest changes projected in the amounts of marketing loan and counter- cyclical payments, and the fact that these changes run counter to one another.


        image

        540 Supra, para. 10.227.

        541 Supra, para. 10.227.

        542 Brazil refers to the CBO report which projects a reduction of US$17 million in marketing loan payments arising from the elimination of the Step 2 programme. Brazil's Rebuttal Submission, paras. 58-60.

        543 US Rebuttal Submission, para. 184.

        WT/DS267/RW

        Page 131


        1. Other factors impacting the world market price for upland cotton


          1. Main arguments of the parties


      15. The United States submits that by failing to undertake any meaningful analysis of world market prices and the factors that influence them Brazil fails to ensure that the effects of other factors on prices are not improperly attributed to the marketing loan and counter-cyclical subsidies at issue in this proceeding. In this regard, the United States emphasizes that Brazil does not take into account the role of China's trade in cotton. Brazil ignores the fact that there exists a high correlation between China's net trade in cotton and the A-Index, that the A-Index itself has been changed to reflect the impact of China on the world cotton market, and that, as indicated by market reports submitted by Brazil, the uncertainty regarding China's trade in cotton results in downward pressures on prices.544


      16. Brazil argues that any impact of China's purchases of upland cotton on the world market price does not diminish the significant price-suppressing effects of the US subsidies at issue. Brazil emphasizes that its claim is a claim of price suppression and not of price depression and that this claim does not depend upon the actual level of world prices for upland cotton, which is influenced by numerous factors, including the purchasing behaviour of China. While China's supply and demand may explain certain price movements in the world market for upland cotton, this does not alter the fact that removal of US subsidies would result in a significant increase in the US supply to the world market, which in turn would lead to significantly higher prices. In any case, while Brazil does not disagree that China exerts a strong influence on the world market price for upland cotton, China's role is not as dominant as argued by the United States. Thus, Brazil contests that there is a significant correlation between China's exports and the A-Index and asserts that the change in the A-Index is of little importance for the overall determination of world market prices. Finally, Brazil considers that uncertainty about China's demand and supply can lead to short-term price fluctuations, but that in the long-term prices are determined by market fundamentals of demand and supply.545


        1. Evaluation by the Panel


      17. As observed by the Appellate Body and the panel in the original proceeding, in order to determine that the effect of a subsidy is significant price suppression within the meaning of Article 6.3(b) it is necessary "to ensure that the effects of other factors on prices are not improperly attributed to the challenged subsidies".546 However, the SCM Agreement prescribes no precise methodology for such an analysis of the role of other factors in the context of Article 6.3.


      18. As discussed above, the Panel has adopted a "but for" approach to the question of whether the effect of US marketing loan and counter-cyclical subsidies to upland cotton producers is significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement. The Panel considers that it is not necessary in this respect to undertake a comprehensive evaluation of factors affecting the world market price for upland cotton. Rather the question is whether the evidence before the Panel supports the conclusion that in the absence of the US marketing loan and counter-cyclical subsidies the world market price would increase significantly. The Panel considers, based on the evidence before it, that while China may play a significant role in the market for upland cotton, this does not diminish the significance of the impact of US subsidies on the world price for upland cotton as a result of their effect on US supply to the world market. Developments concerning the role of China's



        image

        544 US First Written Submission, paras. 308-318; US Rebuttal Submission, paras. 387-396.

        545 Brazil's Rebuttal Submission, paras. 324-345.

        546 Appellate Body Report, US – Upland Cotton, para. 437; Panel Report, US – Upland Cotton, para. 7.1344.

        WT/DS267/RW

        Page 132


        demand and supply do not change the fact that, with a share of world exports of around 40 per cent, the United States is capable of exerting a substantial proportionate influence on the world market.547


        1. Summary and conclusion of the Panel's analysis of Brazil's claim of serious prejudice within the meaning of Articles 5(c) and 6.3(c) of the SCM Agreement


      19. The Panel recalls that the original panel found that the effect of certain price-contingent subsidies provided by the United States in respect of upland cotton – marketing loan programme payments, user marketing (Step 2) payments, market loss assistance payments and counter-cyclical payments – was significant price suppression in the same world market for upland cotton in the period MY 1999-2002 within the meaning of Article 6.3(c) of the SCM Agreement constituting serious prejudice within the meaning of Article 5(c) of the SCM Agreement.548 In respect of this finding, the original panel recalled that under Article 7.8 of the SCM Agreement the adoption of a panel or Appellate Body Report in which it is determined that a Member has caused adverse effects within the meaning of Article 5 of the SCM Agreement requires the Member granting or maintaining the subsidy to take appropriate steps to remove the adverse effects or to withdraw the subsidy.549 550


      20. In this proceeding under Article 21.5 of the DSU, Brazil claims that the United States has failed to comply with its obligations under Article 7.8 of the SCM Agreement because two of the subsidies to US upland cotton producers that were the subject of the original panel's finding of "present" serious prejudice – marketing loan and counter-cyclical payments – cause "present" serious prejudice to the interests of Brazil within the meaning of Articles 5(c) and 6.3(c) of the SCM Agreement.


      21. The Panel has considered the arguments and evidence presented by the parties in support of their positions on the issues raised by Brazil. The Panel's evaluation can be summarized as follows.


      22. First, in view of the magnitude of the US shares of world production and exports, the Panel considers that the United States exerts a substantial proportionate influence on the world market for upland cotton, even if account is taken of the projected declines in these shares in MY 2006. This substantial proportionate influence of the United States, analyzed in light of the totality of the evidence before us, is an element that supports a finding that the marketing loan and counter-cyclical payments to US upland cotton producers cause significant price suppression in the world market for upland cotton.


      23. Second, the Panel has found that marketing loan and counter-cyclical payments affect the level of US upland cotton acreage and production as a result of their mandatory and price-contingent nature and their revenue-stabilizing effect. In this respect, the Panel has concurred with the view of the original panel that these subsidies protect or "insulate" revenues of US upland cotton producers when prices are low. We have emphasized, however, that in determining whether the structure, design and operation of these subsidies support a finding of significant price suppression under current factual conditions, we need to consider this factor in conjunction with other factors.


        image

        547 The Panel notes that Brazil has not ignored the role of China in the world market for upland cotton. The analysis of the world cotton market prepared by Brazil's experts specifically refers to the fact that China is the largest producer and consumer of cotton and is now also the largest importer of cotton. Brazil's First Written Submission, Annex II (Declaration by Andrew Macdonald).

        548 Panel Report, US – Upland Cotton, paras. 7.1416 and 8.1(g)(i).

        549 Panel Report, US – Upland Cotton, para. 8.3(d).

        550 Brazil emphasizes that in explaining its decision not to address certain claims of Brazil, the original

        panel observed that as a result of the panel's finding of present serious prejudice "the United States is obliged to take action concerning its present statutory and regulatory framework." Panel Report, US – Upland Cotton, para. 7.1501. The United States argues that this statement is not a "recommendation" and that in any event the United States has amended its statutory and regulatory framework by repealing the Step 2 programme.

        WT/DS267/RW

        Page 133


      24. Third, the Panel considers that the order of magnitude of the marketing loan and counter- cyclical subsidies is such that in conjunction with other factors, this element supports a finding that significant price suppression is the effect of the subsidy, even when account is taken of the projected decline of marketing loan payments in MY 2006.


      25. Fourth, the Panel has found that there exists a significant gap between the total costs of production of US upland cotton producers and their market revenue, suggesting that the subsidies at issue are a key factor affecting the economic viability of US cotton farming.


      26. Fifth, the Panel considers that certain arguments of Brazil are less persuasive. Thus, in arguing that there exists a temporal coincidence of subsidies and suppressed prices, Brazil in our view fails to take sufficient account of the fact that the recent trends in world market prices and US world export shares differ considerably from the trends in world market prices and US world export share in the period considered by the original panel. The Panel is also not convinced that Brazil has been able to substantiate adequately its assertion that US upland cotton acreage, production and exports do not respond to market price signals. The Panel recalls, however, its conclusion that the fact that recent years have not witnessed the sharp decline in the world market price for upland cotton that occurred during the period considered by the original panel does not necessarily mean that there is currently no price suppression within the meaning of Article 6.3(c) of the SCM Agreement. The Panel also recalls its observation that the stable US share of world production and exports does not mean that US producers are not insulated from market price signals, but that the degree of price insulation found by the original panel is now weaker, possibly because prices are not depressed as during the period examined by the original panel. Finally, the Panel recalls that all the simulations conducted by the parties support the view that US marketing loan and counter-cyclical payments have led to an increase in US production and exports of cotton that have then suppressed world prices.


      27. Sixth, the Panel has examined the role of other factors that may have an effect on the world market price for upland cotton, notably the role of China's trade in cotton. The Panel concludes that this other factor does not attenuate the link between significant price suppression and the subsidies at issue in this proceeding.


      28. Finally, with respect to the elimination by the United States of Step 2 subsidies – one of the subsidies that was the subject of the original panel's finding of "present" serious prejudice – the Panel considers that the evidence before it on the effect of the elimination of this subsidy on US exports and on the amounts of marketing loan payments and counter-cyclical payments suggests that the elimination of this subsidy does not affect the price suppressing effects of the marketing loan payments and counter-cyclical payments in the world market for upland cotton.


      29. The Panel finds, on the basis of the above summarized evaluations, that without the marketing loan and counter-cyclical payments to US upland cotton producers the level of US upland cotton acreage and production would be significantly lower and that, as a consequence, the world market price for upland cotton would be higher. The Panel recalls that to determine whether there is significant price suppression within the meaning of Article 6.3(c) as the effect of the subsidies at issue, the Panel must ascertain whether, but for the subsidies at issue, the world market price for upland cotton would have increased significantly, or would have increased by significantly more than was in fact the case. The Panel concludes that the effect of the subsidies at issue is significant price suppression in this sense.551


      30. Article 6.3(c) of the SCM Agreement provides that "serious prejudice in the sense of paragraph (c) of Article 5 may arise in any case where one or several of the following apply". (emphasis added). The Panel considers that this phrase must be interpreted to mean that "the


        image

        551 See supra, paras. 10.50 for the Panel's view on the interpretation of the term "significant".

        WT/DS267/RW

        Page 134


        situations listed in Article 6.3(a)-(d) in themselves constitute serious prejudice".552 As a consequence, a finding of significant price suppression under Article 6.3(c) of the SCM Agreement is a sufficient basis for a finding of serious prejudice within the meaning of Article 5(c) of the SCM Agreement.553


      31. In light of the foregoing considerations, the Panel finds that the United States acts inconsistently with its obligations under Articles 5(c) and 6.3(c) of the SCM Agreement in that the effect of marketing loan payments and counter-cyclical payments provided to US upland cotton producers pursuant to the FSRI Act of 2002 is significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement in the world market for upland cotton constituting "present" serious prejudice to the interests of Brazil within the meaning of Article 5(c) of the SCM Agreement.


      32. The Panel also finds, as a consequence, that by acting inconsistently with Articles 5(c) and 6.3(c) of the SCM Agreement the United States has failed to comply with the DSB recommendations and rulings. Specifically, the United States has failed to comply with its obligation under Article 7.8 of the SCM Agreement "to take appropriate steps to remove the adverse effects or ... withdraw the subsidy".


        1. CLAIM OF BRAZIL OF "PRESENT" SERIOUS PREJUDICE WITHIN THE MEANING OF ARTICLES 5(C)

          AND 6.3(D) OF THE SCM AGREEMENT


          1. Main arguments of the parties


      33. Brazil submits that the effects of the marketing loan and counter-cyclical payment subsidies in MY 2005 cause the United States to hold an increased share of the world market for upland cotton. In this respect, Brazil argues that the US share of the world market for upland cotton in MY 2005 was greater than the average US world market share during MY 2002-2004, that this increase reflects a consistent trend over the period MY 1998-2005, which Brazil considers the most representative period for purposes of Article 6.3(d), and that the evidence presented by Brazil in the context of its price suppression argument regarding the effects of the nature and magnitude of the subsidies, coupled with the commanding US production and export market share, on production, export and world market price also supports a causal link between these subsidies and the increase in the world market share of the United States.


      34. The United States argues that Brazil fails to make a prima facie case of WTO-inconsistency under Articles 5(c) and 6.3(d) of the SCM Agreement. Brazil fails to demonstrate that all of the elements of Articles 5(c) and 6.3(d) of the SCM Agreement are satisfied.554 Brazil has not demonstrated that the slight increase in MY 2005 of the US share of world production (1.53 per cent) and of the US share of world production plus beginning stocks (0.46 per cent) was not caused by the Step 2 programme. Brazil has also failed to demonstrate that this increase "follows a consistent trend over a period when subsidies have been granted". In the later regard, the United States points out that US share of world production or production plus beginning stocks is stable if one looks back to MY 1980 and has actually declined significantly if one looks back to 1960.


        image

        552 Panel Report, Korea – Commercial Vessels, para. 7.587.

        553 Panel Report, US – Upland Cotton, para. 7.1390.

        554 The United States also submits that not only are Brazil's claims against the marketing loan and

        counter-cyclical payment programmes not within the scope of this proceeding, but Brazil also fails to provide a legal basis for a finding against these programmes because it fails to demonstrate that these programmes mandate a breach of Articles 5(c) and 6.3(d). Since the legal provisions or "programmes" pursuant to which marketing loan and counter-cyclical payments are made by the US Government to US upland cotton farmers are not, as such, the subject of Brazil's claims of serious prejudice, we do not need to address these arguments of the United States.

        WT/DS267/RW

        Page 135


          1. Evaluation by the Panel


      35. Articles 5(c) and 6.3(d) of the SCM Agreement provide:


        "Serious prejudice in the sense of paragraph (c) of Article 5 may arise in any case where one or several of the following apply ...


        (d) the effect of the subsidy is an increase in the world market share of the subsidizing Member in a particular subsidized primary product or commodity17 as compared to the average share it had during the previous period of three years and this increase follows a consistent trend over a period when subsidies have been granted."


        image


        17Unless other multilaterally agreed specific rules apply to the trade in the product or commodity in question.


      36. Thus, Article 6.3(d) requires, first, a showing that there has been an increase in the world market share of the subsidizing Member in a particular primary product or commodity as compared to the average market share of that Member in the previous period of three years. Second, it must be shown that this increase follows a consistent trend over a period when subsidies have been granted.


      37. With respect to the first element that must be satisfied for a finding under Article 6.3(d), we note that the original panel interpreted the term "world market share" in Article 6.3(d) to mean the "share of the world market supplied by the subsidizing Member".555 Brazil has used this definition and has noted that this definition suggests two potential measures of "supply". First, supply can be limited to the annual level of a Member's new production of a primary product or commodity. Alternatively, supply can additionally include residual produce from previous years, which, in the context of upland cotton, is known as "beginning stocks". In this present dispute, Brazil has presented data on the US world market share using both measures of supply.


      38. Brazil argues that the world market share of the United States has increased in MY 2005, as compared to the average US share of world supply in MY 2002-2004. Brazil provides data showing that the US share of world production plus beginning stocks increased by 0.46 per cent over the average for MY 2002-2004 and the US share of world production increased by 1.53 per cent over the average for MY 2002-2004.556 The Panel notes that while the United States does not dispute these figures, the United States considers these "slight" increases to be part of ordinary fluctuations in US market share, not the effect of the marketing loan and counter-cyclical payment programmes.557


      39. The Panel agrees with the United States that Brazil has not been able to demonstrate that the small increase in the US world market share is the effect of marketing loan and counter-cyclical payments. It is not unreasonable to interpret such a small increase to be part of the ordinary fluctuations in the US market share. The sensitivity of the result, an increase in US market share compared to the average of the past three years, also becomes apparent if we change the terminal marketing year from MY 2005 to MY 2006. Brazil has been willing to admit the inclusion of MY 2006 projections provided that they are reliable. We have judged MY 2006 projections provided



        image

        555 Panel Report, US – Upland Cotton, para. 7.1454. While Brazil raised this issue in its appeal, the Appellate Body found that an interpretation of the phrase "world market share" in Article 6.3(d) was not necessary to resolve the dispute. Appellate Body Report, US – Upland Cotton, para. 511.

        556 Brazil's First Written Submission, paras. 222-226.

        557 US Rebuttal Submission, para. 402.

        WT/DS267/RW

        Page 136


        by the United States to this panel to be reliable. If one uses these projections for US and world cotton production in MY 2006, the US world market share will actually be lower in MY 2006 compared to the average of the three previous marketing years:


        Table 9: US World Market Share, MY 2002-2006


        Marketing Year

        US Share of World Production

        US Share of World Production plus Stocks

        2002

        19.5%

        17.4%

        2003

        19.2%

        17.0%

        2004

        19.3%

        16.3%

        2005

        20.9%

        17.5%

        2006

        18.6%

        16.3%

        Source: Exhibits US-27 and US-114.


      40. We note in this connection that Brazil argues that the Panel should not take into account this data for MY 2006 because Brazil's claim is a claim of "present" serious prejudice, based upon data from MY 2005, the last year for which complete data is available.558 The Panel disagrees with this argument of Brazil. The Panel must determine whether the effect of the subsidy "is" an increase in the world market share of the United States. As discussed above, the Panel considers that the use of the present tense in Article 6.3 suggests that the relevant period to consider is the present period and that where reliable data is available for MY 2006, that data must be taken into account. The Panel recalls that an important additional consideration for not limiting its analysis to MY 2005 is that the elimination of the Step 2 programme became effective on 1 August 2006.


      41. The Panel therefore finds that the evidence does not support Brazil's argument that there is an increase in the world market share of the United States in upland cotton as compared to the average US world market share during the previous period of three years. Thus, the Panel cannot find that the effect of the subsidies at issue in this proceeding is an increase in the world market share of the United States within the meaning of Article 6.3(d).559


      42. In light of this finding, the question of whether the second condition of Article 6.3(d) of the SCM Agreement has been met in this case, i.e. whether the increase in world market share follows a consistent trend over a period when subsidies have been granted, does not arise. The Panel notes that the parties disagree on how to interpret and apply the requirement of Article 6.3(d) that the increase in the subsidizing Member's world market share, as compared to the average share of the Member in the previous period of three years, be shown to follow a consistent trend over a period when subsidies have been granted.560 The Panel does not consider it necessary to address the arguments made by the parties in this regard.


      43. In light of the foregoing considerations, the Panel finds that Brazil has not made a prima facie case that the effect of marketing loan and counter-cyclical payments provided to US upland cotton producers pursuant to the FSRI Act of 2002 is an increase in the US world market share in upland cotton as compared to the average US world market share during the previous period of three years



    image

    558 Brazil's Rebuttal Submission, para. 356.

    559 Brazil argues that, the model submitted by its economic consultant shows that, but for the marketing loan and counter-cyclical payments, the US world market share would have been significantly lower in MY

    2005. The Panel considers that the terms of Article 6.3(d) require that it be demonstrated that there is an actual increase in market share of the subsidizing Member.

    560 Thus, for example, the United States argues that it must be shown under Article 6.3(d) that a

    Member's market share is consistently increasing over the previous three-year average. Brazil rejects this interpretation. US Rebuttal Submission, paras. 403-404; Brazil's Response to Panel Question 76.

    WT/DS267/RW

    Page 137


    and that this increase follows a consistent trend over a period when subsidies have been granted. Therefore, it has not been established that the United States acts inconsistently with Articles 5(c) and 6.3(d) of the SCM Agreement.


  11. BRAZIL'S CLAIM OF THREAT OF SERIOUS PREJUDICE UNDER ARTICLES 5(C) AND 6.3 OF THE SCM AGREEMENT


      1. Brazil requests the Panel to find that the United States has failed to comply with the DSB recommendations and rulings in the original proceeding in that marketing loan and counter-cyclical payments to US upland cotton producers threaten to cause serious prejudice in the form of significant price suppression in the world market for upland cotton.


      2. The United States submits that Brazil has failed to make a prima facie case of threat of serious prejudice caused by the marketing loan and counter-cyclical payments.


      3. The Panel has found that marketing loan and counter-cyclical payments under the FSRI Act of 2002 to US upland cotton producers cause "present" serious prejudice to the interests of Brazil within the meaning of Articles 5(c) and 6.3(c) of the SCM Agreement and that in this regard the United States has failed to comply with the DSB recommendations and rulings in the original proceeding. In light of these findings of "present" serious prejudice and of non-compliance by the United States with the DSB recommendations and rulings, the Panel does not consider it necessary or appropriate to make an additional finding as to whether marketing loan and counter-cyclical payments to US upland cotton producers threaten to cause serious prejudice to the interests of Brazil in the form of significant price suppression in the world market for upland cotton.561 The Panel also notes that Brazil has made clear that its claim of threat of serious prejudice is conditional. Specifically, Brazil has stated that the Panel "need only decide this claim if it is unable to find that there is present serious prejudice from marketing loan and counter-cyclical payments programmes of the FSRI Act of 2002".562


      4. The Panel therefore decides not to make a finding on Brazil's claim of threat of serious prejudice under Articles 5(c) and 6.3(c) of the SCM Agreement.


  12. BRAZIL'S CLAIM OF "PRESENT" SERIOUS PREJUDICE UNDER ARTICLES 5(C), 6.3(C) AND 6.3(D) OF THE SCM AGREEMENT AS A RESULT OF THE FAILURE OF THE UNITED STATES TO TAKE ANY ACTION BETWEEN 22 SEPTEMBER 2005 AND 31 JULY 2006


      1. Brazil requests the Panel to find that the failure of the United States to take any measure to remove the adverse effects or to withdraw the three price-contingent upland cotton subsidy programmes between 22 September 2005 and 31 July 2006 cause present serious prejudice to the interests of Brazil in the form of significant price suppression in the world market for upland cotton and an increase in the US world market share, in violation of Articles 5(c), 6.3(c) and 6.3(d) of the SCM Agreement.


        image

        561 The original panel also did not consider it necessary or appropriate to address Brazil's claims of threat of serious prejudice under Articles 5(c) and 6.3(c) of the SCM Agreement in light of its findings of present serious prejudice and of the existence of prohibited subsidies. Panel Report, US – Upland Cotton, para. 7.1503.

        562 Brazil's First Written Submission, para. 461; see also Brazil's Response to Panel Question 84. The Panel concurs with the argument in paragraph 240 of Brazil's First Written Submission that "a finding by the

        compliance Panel of either present serious prejudice or a threat thereof will suffice for the compliance Panel to find that the U.S. measures taken to comply are inconsistent with the covered agreements, within the meaning of Article 21.5 of the DSU".

        WT/DS267/RW

        Page 138


      2. In light of the Panel's ruling in paragraph 9.68 on the preliminary objection of the United States and in light of its findings above of "present" serious prejudice and of non-compliance by the United States with the DSB recommendations and rulings, the Panel does not consider it necessary or appropriate to make a finding on this claim of Brazil of "present" serious prejudice resulting from the failure of the United States to take certain action during the period 22 September 2005-31 July 2006. The Panel also notes that Brazil has presented this claim on a contingent basis.563 The Panel therefore decides not to make a finding on this claim of Brazil of "present" serious prejudice under Articles 5(c), 6.3(c) and 6.3(d) of the SCM Agreement.


  13. ATTACHMENT 1: OVERVIEW OF NEW ECONOMIC LITERATURE ON COUNTER-CYCLICAL PAYMENTS


    1. IMPACT OF COUNTER-CYCLICAL PAYMENTS


      1. Studies cited by the United States


          1. The United States submits a study "Risk Considerations in Supply Response: Implications for Counter-Cyclical Payments' Production Impact" by William Lin and Robert Dismukes.564 This study attempts to simulate the acreage effects of counter-cyclical payments for three crops – corn, soybeans and wheat. The simulation is based on an acreage model, which was estimated using state-level data during 1991-2001, drawn from farms in the Northcentral region565 of the United States. The study acknowledges the many ways in which programmes like the counter-cyclical payment programme can affect the acreage decision of farmers – through a reduction in risk, wealth effects, the increase in farm revenue, etc. Its primary focus, however, is how counter-cyclical payments reduce the price risk faced by farmers and how much this translates into increases in planting area. Essentially, counter- cyclical payments improve the probability distribution of prices faced by farmers by guaranteeing them the target price less the direct payment for that crop. The study finds that of the three crops examined (corn, soybeans and wheat), price risk reduction from counter-cyclical payments is likely to be important only in the case of soybeans. The simulation predicts that counter-cyclical payments will lead to a 3 per cent increase in soybean acreage.


          2. A second study cited by the United States, "Counter-Cyclical Payments Under the 2002 Farm Act: Production Effects Likely to be Limited" by Paul Westcott566, analyses the likely production effects of counter-cyclical payments. The author considers the case of a hypothetical corn farmer who plants all of his hundred acres of corn base with corn. He then examines how the farmer's revenues are affected by counter-cyclical payments (among other subsidies) depending on realizations of the market price. He comes to the conclusion that the impact of counter-cyclical payments differs depending upon the level of the market price. If farmers expect prices to be below loan rates, then counter-cyclical payments are at their maximum levels and become more like "fixed" payments. If farmers expect the market price to lie between the loan rate and the "effective target price", counter- cyclical payments will vary with market price and their effects may vary as well. It is in this price range that Westcott believes that, assuming the base acreage crop is planted, counter-cyclical payments have potential avenues to affect production. They do so by stabilizing farm revenues, offsetting changes in market prices with counter-cyclical payments. At the same time, Westcott points to several mitigating factors which may reduce these effects. Since commodity prices tend to move together, counter-cyclical payments on base acreage provide revenue stabilization for alternative crops. Thus this implies that the production effect need not be on cotton even though the


            image

            563 Brazil's First Written Submission, para. 461.

            564 Exhibit US-34.

            565 The region includes the states of Ohio, Indiana, Illinois, Missouri, Iowa, Minnesota, Wisconsin, and Michigan.

            566 Exhibit US-35.

            WT/DS267/RW

            Page 139


            counter-cyclical payments are made on cotton base. Further, there are other farm programmes which mitigate the revenue risk faced by farmers so that counter-cyclical payments may not have much to contribute. Finally, a large portion of output in US agriculture is produced by large producers, and to the extent that risk aversion decreases as income rises, the revenue stabilization benefits of counter- cyclical payments may be lower. Taking these mitigating factors into account, Westcott concludes that the overall production effects of counter-cyclical payments through revenue risk reduction are likely to be limited, although the study is unable to provide an estimate of this magnitude.


          3. The United States also submits a study "US Agricultural Policy: Overview and Recent Analyses" by Edwin Young, Anne Effland, Paul Westcott and Demcey Johnson.567 This paper presents an overview of recent US agricultural policies as embodied in the 1996 and 2002 Farm Acts. It also provides a survey of recent analyses regarding the effects of those policies. With respect to counter-cyclical payments under the 2002 Farm Act, it suggests that recent studies have concluded that these payments "create small but positive production incentives".


          4. Finally, the United States submits a paper entitled "Another Look at Decoupling: Additional Evidence on the Production Effects of Direct Payments" by B. Goodwin and A. Mishra in the American Journal of Agricultural Economics.568 The Goodwin and Mishra study, published in 2005, examines the extent to which production flexibility contract or Agricultural Market Transition Act (AMTA) payments under the FAIR Act affected the acreage decisions of a set of farms in the US Corn Belt who grew corn, wheat, and soybeans. In particular, the authors use farm-level data for the years 2002 and 2003 to examine the acreage decisions of Corn Belt producers. They also examine responses to a number of survey questions regarding factors affecting acreage decisions and the disposition of funds received in the form of direct payments. Farmers' responses to the survey questions suggest that direct payments, counter-cyclical payments, and the potential for base acreage updates were all relatively minor factors. Finally, their empirical results find no statistically significant relationship between direct payments and acreage decisions for corn, wheat, and soybeans.


      2. Studies cited by Brazil


          1. Brazil submits a paper "Farm-Level Production Effect From Participation in Government Commodity Programs: Did the 1996 Federal Agricultural Improvement and Reform Act Make a Difference" by Nigel Key, Ruben Lubowski and Michael Roberts.569 The paper, published in the American Journal of Agricultural Economics, looks at the production impact of participation in government commodity programmes. As such, it does not single out the influence of counter-cyclical payments from other elements of these commodity programmes. The results indicate that participants increased plantings of programme crops considerably (38-59 percentage points) more than non- participants. Estimated effects of participation on farm-size growth and total sales growth were also consistent across time, increasing 14-16 percentage points and 22-24 percentage points, respectively. Although the estimated effects do not isolate the impact of countercyclical payments, the study goes on to conjecture "that the results suggest decoupled payments affect production at the farm level".


          2. Another study submitted by Brazil, "How distorting are direct payments?" by Erik J. O'Donoghue and James Whitaker570, investigates the effects of direct payments on total harvested acres for farms growing programme crops (including upland cotton), non-programme crops and livestock. The study's main innovation is to control for the possible endogeneity of direct payments. Endogeneity can occur because, while direct payments could alter acreage changes (if the payments had wealth effects, caused expectations to change, etc.), acreage changes could also alter


            image

            567 Exhibit US-40.

            568 Exhibit US-41.

            569 Exhibit BRA-566.

            570 Exhibit BRA-568.

            WT/DS267/RW

            Page 140


            the receipt of direct payments (if base was bought or sold). The authors use the updating of base acres allowed under the 2002 Farm Act to impute causation.


          3. The authors divide farms into four categories by size of revenues. Their econometric estimates suggest that direct payments (which are "proxied" by average base acre changes for farms) affect farms differently. The smallest farms appear not to have changed their harvested acres. The next set of farms in terms of revenues increased their total cropland harvested by almost 25 per cent on average, while the third set of farms increased their total cropland harvested by nearly ten per cent. However, the largest farms decreased their total cropland harvested by approximately 20 per cent. The authors judge that altogether these magnitudes represent "substantial changes". They do not seem to have explicitly tested how direct payments affect harvested acreage for farms growing programme crops, non-programme crops or livestock. But in interpreting their econometric results, they offer the explanation that farms might be redistributing the acres from non-programme crop and livestock specialties to programme crop farms (which include upland cotton).


          4. Another paper submitted by Brazil is entitled "Are decoupled payments really decoupled?" by

    2. K. Goodwin and A. Mishra571; published in the American Journal of Agricultural Economics in

    2006.572 This study examines the production effect of Agricultural Market Transition Act (AMTA) and market loss assistance payments. It finds modest evidence that AMTA payments may lead to increased production of corn, soybeans, and wheat. However, the acreage effects are very small. Acreage elasticities for the major crops, corn and soybeans, are about 0.025 to 0.035. These estimates suggest that even a very large increase in these decoupled payments would not be expected to significantly increase acreage. However, market loss assistance payments appeared to have a more significant effect on acreage. The estimated elasticity of corn acreage with respect to the payments was 0.10, suggesting that such payments may indeed lead to more corn production.


      1. Brazil also submits a study, "Supply Response to Counter-Cyclical Payments and Base Acre Updating under Uncertainty: An Experimental Study" by Christopher R. McIntosh, Jason F. Shogren and Erik Dohlman.573 The study, which is to be published in the American Journal of Agricultural Economics, reports on a laboratory experiment which attempts to mimic the price risks faced by farmers and the incentives provided by direct and counter-cyclical payments under the 2002 Farm Act. Four scenarios are considered. The baseline scenario is where only direct payments are provided. The second scenario includes both direct and counter-cyclical payments. The third scenario allows for some uncertainty about the policy regime. The fourth scenario requires mandatory updating of base acreage. Eighty-five students recruited from Economics and Finance undergraduate classes at the University of Wyoming were the subjects of the laboratory experiment. Based on the results of the experiments, the authors conclude that counter-cyclical payments influence crop allocation decisions in the laboratory — the average subject allocated more acres toward the base crop in the presence of counter-cyclical payments. They estimate that the presence of counter-cyclical payments will increase base crop acreage by 5.43 per cent. Larger estimates (7.92 per cent increase) are obtained if apart from price risk, the subjects face uncertainty about the policy regime.


        image

        571 Exhibit BRA-618.

        572 This article appears to be based on the same study cited by the United States in Exhibit US-41. Both studies cover the same crops (corn, soybeans, wheat) and the same geographical region (Corn Belt) of the

        United States. However, there appears to be some difference in the programmes covered. Exhibit US-41 includes AMTA payments only while BRA-618 covers both AMTA and MLA payments.

        573 Exhibit BRA-565.

        WT/DS267/RW

        Page 141


        1. ARE COUNTER-CYCLICAL PAYMENTS CAPITALIZED INTO HIGHER LAND VALUES?


          1. Studies cited by the United States


      2. The studies by Hoppe and Banker574 and Young, Effland, Westcott and Johnson575 examine the extent of land rental in US agriculture. The third study by Abler and Blandford576 provides a survey of empirical studies on direct payments under the United States 1996 FAIR Act and subsequent Marketing Loss Assistance (MLA) payments.


      3. The USDA study by Hoppe and Banker shows that an estimated 44 per cent of all farmland operated in 2003 was rented.577 The study by Young, Effland, Westcott and Johnson gives even higher estimates for larger farms. They state that crop farms in which total sales exceeded US$250,000, also known as commercial crop farms, rented close to 60 per cent of their total acres operated.578 Unfortunately, neither study refers specifically to upland cotton, so it is not possible to know to what extent upland cotton farmers conformed or differed with respect to these nationwide statistics.


      4. The Abler and Blandford study was written for the OECD. The paper is a literature review of empirical studies on land allocation and production response to direct payments under the United States 1996 FAIR Act and subsequent MLA payments. Thus, it is again important to note that the study does not examine the experience with counter-cyclical payments. The main conclusion of this study on the link between such payments and land values is that PFC and MLA payments had a significant effect on land values and rental rates. Given the importance of the rental market for land in the United States, it appears that there was a relatively high "pass-through" of the additional income generated by the payments to landowners, many of whom are not the actual operators of the land. It appears that the payments primarily had the effect of increasing the value of the principal fixed asset in agriculture – land.579


      5. However, the authors also point out that the empirical studies reviewed lend support to the view that PFC and MLA payments had some impact on production relative to the counterfactual case with no PFC or MLA payments. Empirical studies indicate that the payments may have influenced planted area and possibly the use of variable inputs, particularly farm household labour. In general, the estimated impacts are modest. In the econometric studies reviewed, PFC and MLA payment variables are sometimes statistically insignificant, and when they are statistically significant they imply in most cases that each type of payment increased planted area and on-farm work hours by less than 5 per cent. In no instance do the econometric results imply an increase in planted area or on- farm work hours due to PFC or MLA payments of more than 15 per cent. Some of the synthetic studies suggest larger impacts on production, but their results are sensitive to assumptions about the degree and nature of farmers' aversion to risk. The synthetic studies also do not consider the possibility that assets owned by farmers who would exit the industry in the absence of PFC or MLA payments, could be sold or rented to other farmers, which would diminish the impact of the payments on production.580


    image

    574 Exhibit US-67.

    575 Exhibit US-40.

    576 Exhibit US-32.

    577 US First Written Submission, para. 215.

    578 See Exhibit US-40, p. 7.

    579 Exibit US-32, para. 71.

    580 Exhibit US-32, para. 69.

    WT/DS267/RW

    Page 142


  14. FINDINGS ON BRAZIL'S CLAIMS UNDER ARTICLES 10.1 AND 8 OF THE AGREEMENT ON AGRICULTURE AND ARTICLES 3.1(A), 3.2 AND 4.7 OF THE SCM AGREEMENT WITH RESPECT TO EXPORT CREDIT GUARANTEES


    1. INTRODUCTION


        1. In this section of our Report, we address Brazil's claims concerning US export credit guarantees. Brazil requests the Panel to find that US measures "taken to comply" with the recommendations and rulings of the DSB with respect to export credit guarantees either do not exist, or to the extent they exist, are not consistent with Articles 10.1 and 8 of the Agreement on Agriculture and Articles 3.1(a) and 3.2 of the SCM Agreement.


        2. Specifically, Brazil makes the following claims:


          1. First, Brazil claims that the United States has taken no action to "withdraw the subsidy" with respect to GSM 102, GSM 103, and SCGP guarantees581 outstanding on 1 July 2005.


          2. Second, Brazil claims that the measures taken by the United States to comply with the DSB recommendations and rulings are inconsistent with Articles 10.1 and 8 of the Agreement on Agriculture and, to the extent of that inconsistency, with Articles 3.1(a) and 3.2 of the SCM Agreement. More specifically, Brazil claims that the United States continues to provide prohibited export subsidies in the form of GSM 102 export credit guarantees, and has applied those export subsidies subsequent to 1 July 2005 in a manner that results in circumvention of the US export subsidy commitments under the Agreement on Agriculture.582 583


        3. Brazil requests the Panel to find that the United States has failed to comply fully with its obligation, under Article 4.7 of the SCM Agreement, to "withdraw the subsidy" and that the original Article 4.7 recommendation adopted by the DSB therefore remains in effect. Brazil also requests that the Panel find that the United States has failed to bring itself into conformity with its obligations under the Agreement on Agriculture, and that the original recommendation adopted by the DSB in this respect therefore also remains in effect.


          1. Description of the export credit guarantees at issue in this dispute


        4. We recall the main characteristics of the three CCC export credit guarantee programmes at issue in this dispute.584


        5. The GSM 102 programme guarantees credit extended by a US private bank (or, less commonly, a US exporter) to approved foreign banks using dollar-denominated, irrevocable letters of credit for purchases of US agricultural exports by foreign buyers. In practice, the foreign bank extends credit to the foreign importer and the US exporter – or more frequently, its bank (the US bank) – extends credit to the foreign bank. Of all the parties involved, only the foreign importer need


          image

          581 See above paras. 3.13 ff.

          582 The Panel notes that Brazil's claims are "as applied" claims. See infra. para. 14.41.

          583 The Panel also notes that Brazil claims that GSM 102 export credit guarantees provided to exports of three products not falling within the scope of the Agreement on Agriculture, lysine, lyocell and wood products, are inconsistent with Articles 3.1(a) and 3.2 of the SCM Agreement.

          584 While Brazil's claim with respect to outstanding guarantees concerns guarantees issued under the

          three export credit guarantee programmes that were the subject of the original panel's findings (GSM 102, GSM 103 and SCGP), Brazil's claims with respect to guarantees issued after 1 July 2005 concern guarantees issued under the GSM 102 programme, as amended.

          WT/DS267/RW

          Page 143


          not be CCC-approved. The guarantee, which the US exporter obtains from the CCC, is triggered by a default on the part of the foreign bank. The US exporter then submits a notice of default to the CCC and the CCC pays the claim consistent with the terms of the guarantee. GSM 102 guarantees generally cover 98 per cent of the principal and a portion of the interest. The GSM 103 programme, which has now been discontinued, was similar to the GSM 102 programme. The main difference between guarantees issued under the two programmes relates to the terms of the underlying credit. GSM 102 guarantees cover credit terms between 90 days and three years, GSM 103 guarantees covered credit terms between 3 years and 10 years.


        6. The SCGP programme guarantees the repayment of credit secured by promissory notes for periods not exceeding 180 days by a United States exporter to a foreign purchaser. SCGP guarantees typically cover a portion (65 per cent) of the value of principal only, the guarantee does not cover interest. The US exporter may negotiate an arrangement to be paid, in full or in part, by assigning the right to proceeds that may become payable under the CCC's guarantee to a US financial institution.585


          1. Summary of the original panel and Appellate Body's findings concerning export credit guarantees


        7. We first recall, as an introduction to this section of our Report, the main findings of the panel and Appellate Body concerning export credit guarantees in the original proceeding.


        8. As in this compliance proceeding, Brazil had, before the original panel, made claims under both the Agreement on Agriculture and, to the extent of the inconsistency with the Agreement on Agriculture, the SCM Agreement. The original panel relied on item (j) of the Illustrative List of Export Subsidies in Annex I to the SCM Agreement (hereafter "Illustrative List") as "contextual guidance" in determining that the export credit programmes at issue (GSM 102, GSM 103, SCGP) constituted export subsidies under Articles 10.1 and 8 of the Agreement on Agriculture.586 The panel declined to make additional factual findings on what it considered to be an alternative argument of Brazil that the same programmes constituted export subsidies under the general definition of "export subsidies" contained in Articles 1 and 3.1(a) of the SCM Agreement (also as context under Articles 10.1 and 8 of the Agreement on Agriculture). In the Interim Review section of its Report, the panel indicated that "Brazil's allegation invoking the elements of Articles 1 and 3.1(a) of the SCM Agreement [was] not a separate claim, but merely another argument, on a different factual basis, as to how the United States export credit guarantee programmes would meet the definition of an export subsidy in Article 3.1(a) of the SCM Agreement".587


        9. Having determined that the export credit guarantees at issue constituted export subsidies within the meaning of Article 10.1 of the Agreement on Agriculture588, the panel found that the United States had applied such subsidies in a manner which resulted in circumvention of its export subsidy reduction commitments in respect of cotton and other unscheduled products supported under the export credit guarantee programmes, as well as in respect of one scheduled product589, rice.590 To


          image

          585 See Panel Report, US – Upland Cotton, paras. 7.242-7.244 (quoted, supra, para. 3.14), Brazil's First Written Submission, paras. 351-355, Exhibit Bra-519.

          586 Panel Report, US – Upland Cotton, paras. 7.763, 7.804-7.869.

          587 Panel Report, US – Upland Cotton, para. 6.31.

          588 The panel rejected US arguments that Article 10.2 of the Agreement on Agriculture reflects the fact that export credit guarantee programmes are not subject to the export subsidy disciplines of the SCM Agreement. See Panel Report, US – Upland Cotton, paras. 7.897-7.944. The Appellate Body upheld the panel's findings in this respect, with one member of the Appellate Body Division hearing the dispute dissenting. See Appellate Body Report, US – Upland Cotton, paras. 585-627 and 631-641.

          589 I.e. a product for which the United States has undertaken export subsidy reduction commitments under the Agreement on Agriculture.

          590 Panel Report, US – Upland Cotton, paras. 7.875, 7.881 and 8.1(d)(i).

          WT/DS267/RW

          Page 144


          the extent that the programmes did not fully conform to the Agreement on Agriculture, the panel considered that they did not benefit from the exemption in Article 13 of the Agreement on Agriculture and found that they were prohibited export subsidies under Article 3.1(a) of the SCM Agreement and were also inconsistent with Article 3.2 of the same Agreement.591 The panel considered that Brazil had failed to prove its claims with respect to exports of scheduled agricultural products other than rice, including pig meat and poultry meat, and with respect to exports of unscheduled agricultural products not supported under the programmes.592


        10. The Appellate Body upheld the panel's findings with respect to unscheduled products and rice but reversed the original panel's finding that Brazil had failed to establish a violation of Article 10.1 of the Agreement on Agriculture with respect to exports of pig meat and poultry meat.593 It did not, however, "complete the analysis" in this respect. Further, the Appellate Body rejected claims by Brazil that the panel had improperly exercised judicial economy with respect to Brazil's arguments under Articles 1.1 and 3.1(a) of the SCM Agreement.594


    2. BRAZIL'S CLAIMS WITH RESPECT TO EXPORT CREDIT GUARANTEES OUTSTANDING ON

      1 JULY 2005


      1. Introduction


          1. The original panel recommended that the United States bring its export credit guarantees under the GSM 102, GSM 103 and SCGP programmes in respect of upland cotton and other "supported" unscheduled agricultural products into conformity with the Agreement on Agriculture; it also recommended, pursuant to Article 4.7 of the SCM Agreement, that the United States "withdraw" the prohibited subsidies without delay, at the latest within six months of the date of adoption of the panel report by the DSB or 1 July 2005 (whichever was earlier).595


          2. The panel's recommendations, which were upheld by the Appellate Body, were adopted by the DSB at its 21 March 2005 meeting.596 As a consequence, the United States had until 1 July 2005 to "withdraw" the prohibited export subsidies identified by the original panel.


          3. Brazil claims that the United States has taken no measures to comply with the DSB's recommendations to "withdraw the subsidy" under Article 4.7 of the SCM Agreement and otherwise bring itself into conformity with its obligations under Articles 10.1 and 8 of the Agreement on Agriculture with respect to "outstanding" guarantees, i.e. GSM 102, GSM 103 and SCGP export credit guarantees for unscheduled products and for rice issued prior to 1 July 2005 that were still outstanding on that date.597



            image

            591 Panel Report, US – Upland Cotton, paras. 7.946-7.948 and 8.1(d)(i). 592 Panel Report, US – Upland Cotton, paras. 7.882-7.896 and 8.1(d)(ii). 593 Appellate Body Report, US – Upland Cotton, paras. 763(e) and 763(f). 594 Appellate Body Report, US – Upland Cotton, paras. 763(e)(v).

            595 Panel Report, US – Upland Cotton, para. 8(3)(b). Article 4.7 of the SCM Agreement provides that: "If the measure in question is found to be a prohibited subsidy, the panel shall recommend that

            the subsidizing Member withdraw the subsidy without delay. In this regard, the panel shall specify in its recommendation the time-period within which the measure must be withdrawn." (underline added)

            596 WT/DSB/M/186.

            597 Brazil's First Written Submission, para. 341. Brazil's claims cover any guarantees that were

            outstanding on 1 July 2005, whether or not they subsequently expired. See paras. 390-393 of Brazil's Rebuttal Submission.

            WT/DS267/RW

            Page 145


      2. Main arguments of the parties


          1. Brazil argues that the United States has done nothing to bring outstanding GSM 102, GSM 103 and SCGP export credit guarantees into conformity with Articles 10.1 and 8 of the Agreement on Agriculture, or to withdraw them under Article 4.7 of the SCM Agreement. Brazil considers that the obligation to "withdraw the subsidy" applies not only to the three programmes at issue in the original dispute, but also to outstanding guarantees.598 Brazil relies on the panel and Appellate Body reports in Brazil – Aircraft (Article 21.5 – Canada), in support of its argument that the United States may not, after the "implementation deadline" of 1 July 2005 (i.e. the date by which the United States had to comply with the recommendations and rulings of the DSB), continue to perform on guarantees issued before that date and still outstanding599, either by making "payments" on claims following defaults or by rescheduling or restructuring debt covered by outstanding guarantees.


          2. Brazil also argues that the obligation to abstain from performing on financial commitments outstanding under a programme as of the deadline for implementation follows from the text of the SCM Agreement: the obligation to withdraw the subsidy under Article 4.7 of the SCM Agreement applies to both the "granting" of new subsidies and the "maintenance" of existing subsidies under Article 3.2 of the SCM Agreement.600 The United States has left the entirety of the outstanding prohibited export credit guarantees in place after the implementation date and, by continuing to perform on its commitments under the prohibited subsidies, "maintains" them unchanged. Further, continuing to perform on outstanding export credit guarantees does not satisfy the definition of "withdraw" offered by the Appellate Body in Brazil – Aircraft (Article 21.5 – Canada), in which the Appellate Body defined the term "withdraw" as to "remove" or "take away" the subsidy".601


          3. The United States considers that it has "withdrawn the subsidy" with respect to outstanding guarantees by lowering the long-term operating costs and losses of the programmes.602 To comply, the United States was required to "withdraw" the prohibited subsidies identified in paragraph 8.1(d)(i) of the original panel report, which concerned the provision of guarantees at a net cost to the government; this is what the United States had to "withdraw".603 The United States considers that the issue of rescheduling is irrelevant to Brazil's claims: any already rescheduled debt is a receivable, pursuant to which the CCC would receive money, not pay any. In response to Brazil's argument that rescheduling is a means for the United States to avoid default and the need to make a "payment" on the CCC obligation, the United States notes that the CCC does not reschedule debt prior to default, and that no obligation is owed to the CCC until it pays a claim and is subrogated to the original obligee.604


          4. Moreover, nothing in the SCM Agreement provides that withdrawing a subsidy taking the form of a programme includes an obligation to abstain from performing on commitments outstanding


            image

            598 Brazil's Rebuttal Submission, paras. 394-395 and 399; Brazil's Closing Statement, para. 30.

            599Brazil's Rebuttal Submission, paras. 396-398, citing Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 45 and Panel Report, Brazil – Aircraft (Article 21.5 – Canada), para. 6.17. See

            also Brazil's Response to Panel Question 37. Brazil also notes that the Appellate Body, in US – FSC (Article 21.5 – EC), paras. 2 and 223-231, held that a recommendation to withdraw a subsidy in the form of an overarching measure, the FSC, included the obligation to stop making pre-committed payments to existing FSCs ("foreign sales corporations"). Brazil's Closing Statement, footnote 38.

            600 Brazil's Response to Panel Question 92.

            601 Brazil's Response to Panel Question 37, para. 46.

            602 US First Written Submission, paras. 143-145. In the case of GSM 103 and SCGP, the United States

            indicates that it has completely ceased providing guarantees. Further, the United States considers that there are no "outstanding" SCGP guarantees as it is exposed to no further liability under the programme. See US Rebuttal Submission, para. 129 and US Opening Statement, para. 48.

            603 US Opening Statement, para. 43.

            604 US Rebuttal Submission, paras. 131-133 and US Response to Panel Question 109.

            WT/DS267/RW

            Page 146


            under that programme as of the deadline for implementation.605 That argument improperly equates "performing on commitments under the program" with the subsidy itself. Such an equation was appropriate in Brazil – Aircraft (21.5), where Brazil continued to issue new WTO-inconsistent bonds even after the period of implementation on the basis that it had pre-existing contractual obligations to do so. However, it is not accurate here, where the guarantees are not themselves prohibited subsidies.


      3. Main arguments of third parties


          1. Argentina submits that the obligation of the United States to withdraw the subsidy implies the termination of the export credit guarantee programmes at issue in the first proceeding, and not merely the suspension of payments or the refusal of new applications for export credit guarantees.606


      4. Evaluation by the Panel


        1. Introduction


            1. Brazil's arguments in support of its claims that the United States has failed to "withdraw" outstanding export credit guarantees rely primarily on the reasoning of the panel and Appellate Body in the Brazil – Aircraft (Article 21.5 – Canada) dispute. According to Brazil, the panel and Appellate Body reports in that case stand for the proposition that a Member's obligation to "withdraw the subsidy" pursuant to recommendations from the DSB adopted in accordance with Article 4.7 of the SCM Agreement includes an obligation to abstain from performing on financial commitments outstanding under a programme as of the implementation deadline (i.e. the date by which the Member in question had to comply with the DSB's recommendations). Brazil draws an analogy between Brazil – Aircraft (Article 21.5 – Canada) and the present dispute: it considers that the panel and Appellate Body's findings in that dispute (according to which Brazil was precluded from issuing, after the implementation deadline, PROEX bonds pursuant to letters of commitments issued before the implementation deadline) can be transposed to the factual circumstances of this case, and that as a result, the Panel should find that the United States is precluded from making "payments" in the form of the payment of claims or of debt restructuring or rescheduling.607


            2. Brazil also finds support for its position in the definition of "withdraw the subsidy" provided by the Appellate Body in Brazil – Aircraft (Article 21.5 – Canada), in which the Appellate Body defined the term "withdraw" as to "remove" or "take away" the subsidy.


            3. Given the emphasis that Brazil has put on the reasoning and the findings of the panel and Appellate Body in Brazil – Aircraft (Article 21.5 – Canada), we first proceed to examine whether, as alleged by Brazil, the Appellate Body (and panel) held in that case that the obligation to "withdraw a subsidy" under Article 4.7 of the SCM Agreement means that a Member is in all cases precluded from performing on commitments outstanding under a programme as of the deadline for implementation. In doing so, we also examine whether the analogy that Brazil seeks to draw between the two disputes is justified. We then proceed to examine whether, as Brazil argues, the Appellate Body's statement as to the meaning of "withdraw the subsidy" in that case should lead us to a finding that the United States has failed to implement the recommendation of the DSB.



              image

              605 US Opening Statement, para. 46.

              606 Argentina's Third Party Submission, paras. 9-19.

              607 In other words, from "perform[ing] on ECG's issued before adoption of the DSB's recommendation". Brazil's Response to Panel Question 37, para. 45.

              WT/DS267/RW

              Page 147


        2. Panel and Appellate Body reports in Brazil – Aircraft (Article 21.5 – Canada)


            1. The Brazil – Aircraft (Article 21.5 – Canada) dispute concerned the PROEX programme, under which the Brazilian Government granted interest rate equalization payments, notably on the export of regional aircraft. The original panel had concluded that the PROEX payments constituted prohibited export subsidies under the SCM Agreement.


            2. One of the issues before the compliance panel was whether Brazil had failed to "withdraw the subsidy" because it continued to issue bonds under the PROEX programme ("NTN-I" bonds) after the implementation deadline, pursuant to letters of commitment issued before that date.608 The Brazil – Aircraft (Article 21.5 – Canada) panel considered that it was sufficient, in order to resolve that issue, to conclude that "a Member cannot be deemed to have withdrawn prohibited subsidies if it has not ceased to act in a manner inconsistent with the WTO Agreement in respect of those subsidies".609 The panel noted that Article 3.2 of the SCM Agreement provides that a Member shall "neither grant nor maintain" export subsidies and therefore concluded that to continue to grant or to maintain prohibited export subsidies after the implementation date would be inconsistent with Brazil's obligation to withdraw these subsidies.610 The panel therefore considered that it had to determine whether the continued issuance of bonds by Brazil pursuant to letters of commitments issued under the PROEX programme before it was amended constituted the "grant" of export subsidies under Article 3.2 of the SCM Agreement. Before the panel, Brazil argued that PROEX subsidies existed under Article 1 of the SCM Agreement, (and therefore, were "granted") at the time the Brazilian Government issued a letter of commitment, not at the time the bonds were issued.611 The panel rejected Brazil's argument, recalling that in the original dispute it had found, and the Appellate Body had agreed, that for the purposes of Article 27.4 of the SCM Agreement, the export subsidies at issue were "granted" at the time of the issuance of the bonds612, and considered that a similar reasoning was appropriate under Article 3.2. The panel therefore concluded that by continuing to issue NTN-I bonds pursuant to commitments made prior to 18 November 1999, Brazil had failed to withdraw the export subsidies for regional aircraft under the PROEX programme.613


              image

              608 Commitments under the programme at issue, PROEX, were made in the form of letters of commitment issued to Brazilian manufacturers, by which the Brazilian Government undertook to provide support as specified for the transaction provided that the contract was entered into in accordance with the terms and conditions specified in the request for approval within a specified period of time (usually 90 days), and provided the aircraft was exported. Panel Report, Brazil – Aircraft (Article 21.5 – Canada), para. 2.5.

              609 Panel Report, Brazil – Aircraft (Article 21.5 – Canada), para. 6.8.

              610 Panel Report, Brazil – Aircraft (Article 21.5 – Canada), para. 6.9.

              611 See Panel Report, Brazil – Aircraft (Article 21.5 – Canada), para. 6.12.

              612 Panel Report, Brazil – Aircraft (Article 21.5 – Canada), paras. 6.10-6.11. In the original dispute, the Appellate Body had affirmed the panel's finding on this issue, though on a different reasoning. The panel

              and Appellate Body both found that the export subsidies at issue were not yet "granted" when the letter of commitment was issued, because at that point, the export sales contract had not yet been concluded and the export shipments had not yet occurred: "For the purposes of Article 27.4, we conclude that export subsidies

              ...are 'granted' when all the legal conditions have been fulfilled to entitle the beneficiary to receive the subsidies." Appellate Body Report, Brazil – Aircraft, para. 158, quoted by the 21.5 panel at para. 6.10. The Appellate Body however rejected the original panel's approach of focusing on the language of Article 1 of the SCM Agreement in order to determine when payments should be considered to have been granted for the purposes of Article 27.4 of the SCM Agreement. The panel had, notably, sought to determine whether PROEX payments constituted "direct transfers of funds" or "potential direct transfer of funds" under Article 1. The Appellate Body noted that the issue was not when or whether subsidies "existed" (Brazil had already conceded that subsidies did exist) but when the subsidies should be considered to have been "granted" for the purposes of Article 27.4. Appellate Body Report, Brazil – Aircraft, paras. 156 ff., cited in Panel Report, Brazil – Aircraft (Article 21.5 – Canada), para. 6.13.

              613 Panel Report, Brazil – Aircraft (Article 21.5 – Canada), para. 6.17. Brazil had also argued that it had a contractual obligation to issue PROEX bonds pursuant to commitments already entered into and would be

              WT/DS267/RW

              Page 148


            3. The Appellate Body affirmed the panel's findings, but considered that Brazil's arguments about when a subsidy is deemed to "exist" under Article 1.1 of the SCM Agreement, and when it is "granted" under Article 3.2 of that Agreement were not relevant to the analysis of Canada's allegation that Brazil has failed to "withdraw the subsidy".614


            4. The Appellate Body noted that Brazil did not contest that the measures in question conferred a prohibited export subsidy.615 It considered that in its ordinary meaning, the "withdrawal" of a subsidy under Article 4.7 of the SCM Agreement refers to the "removal" or "taking away" of that subsidy:


              Turning to the ordinary meaning of "withdraw", we observe first that this word has been defined as "remove" or "take away", and as "to take away what has been enjoyed; to take from." This definition suggests that "withdrawal" of a subsidy, under Article 4.7 of the SCM Agreement, refers to the "removal" or "taking away" of that subsidy. 616 (footnotes omitted)


            5. The Appellate Body then observed that Brazil conceded that it continued to issue bonds pursuant to letters of commitments issued prior to the implementation deadline, meaning that such bonds continued to be issued on "precisely the same terms and conditions as they were before".617 The Appellate Body considered that the issuance of the bonds essentially represented "disbursements" made under PROEX, given that the financing institution could choose to either sell the bonds in the market or receive payments when they became due and that Brazil was thus continuing, after the implementation deadline, to make payments under a subsidy programme found to involve prohibited export subsidies. It concluded that:


              … to continue to make payments under an export subsidy measure found to be prohibited is not consistent with the obligation to "withdraw" prohibited export subsidies, in the sense of "removing" or "taking away".618


        3. Whether the obligation to "withdraw the subsidy" includes an obligation to abstain from performing on outstanding financial commitments


            1. As is clear from the brief summary above, the issue of "prior commitments" arose, in Brazil – Aircraft (Article 21.5 – Canada), because of Brazil's arguments regarding the timing of the granting of the subsidy. The fact that the bonds were issued "pursuant to" letters of commitments that pre- dated the implementation deadline was not, in our view, an important consideration in this respect and there is no indication that either the panel or the Appellate Body would have reached a different conclusion had the bonds been issued independently of any prior commitment. In fact, it seems safe to assume that the issue of prior or outstanding "commitments" would not have been discussed at all had it not been for Brazil's arguments regarding the timing of the granting of the subsidy.


              image

              liable for breach of damages if it failed to do so. The panel rejected Brazil's arguments that such considerations should affect its conclusion: "[a]ssuming that Brazil is correct in this regard, the implication of this view would be that Members could contract to grant prohibited subsidies for years into the future and be insulated from any meaningful remedy under the WTO dispute settlement system." The panel also noted that under Article 21 of the Vienna Convention, a party to a treaty may not invoke the provisions of its internal law as justification for its failure to perform a treaty. Panel Report, Brazil – Aircraft (Article 21.5 – Canada), para. 6.16 and note 23. The Appellate Body endorsed the latter part of the panel's reasoning. Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 46.

              614 Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 44. 615 Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 44. 616 Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 45. 617 Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 45. 618 Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 45.

              WT/DS267/RW

              Page 149


            2. This reading of the reports of the panel and the Appellate Body is confirmed by the absence of any reference to prior "commitments" in the Appellate Body statement quoted above, and on which Brazil puts so much emphasis.619 The Appellate Body, unsurprisingly, found that by continuing to issue PROEX bonds as it had before – and which had been found, in the original dispute, to amount to the granting of an export subsidy – Brazil was continuing to grant the subsidies in question and had therefore not "withdrawn" them.


            3. Further, in our view, Brazil ascribes a broader meaning to the Appellate Body's use of the term "payment" than is warranted: Brazil reads that term as encompassing any disbursements made under a measure found to be a prohibited subsidy. In our view, the Appellate Body's reference to "mak[ing] payments" was clearly meant as a reference to the granting of a PROEX subsidy as the Appellate Body was not discussing the issue of "payments" in an abstract or general sense, but rather was referring to PROEX "payments" (in the form of NTN-I bonds) made on the export of regional aircraft.620 In any case, the term "payment" appears in neither of Article 4.7 or Article 3.2 of the SCM Agreement. Both provisions concern subsidies.


            4. We thus find no support, in either of the panel or the Appellate Body report in that dispute, for Brazil's contention that a Member is necessarily precluded, under Article 4.7 of the SCM Agreement, from making "payments" or otherwise performing on "commitments" pursuant to measures pre-dating the implementation deadline. Rather, as can be deducted from our analysis above, the panel and the Appellate Body's findings in that case rest on the simpler proposition that a Member has not "withdraw[n] the subsidy" where it continues to provide prohibited export subsidies after the deadline set by the DSB for implementing its recommendation. Making "payments" is inconsistent with the Member's obligation to "withdraw a subsidy" when the "payments" themselves amount to the granting of a subsidy. The fact that a subsidy is granted pursuant to a prior or outstanding commitment is, in our view, simply of no relevance.621


            5. Brazil also relies on the Appellate Body Report in US – FSC (Article 21.5 – EC). Brazil submits that the Appellate Body held in that case that a recommendation to withdraw a subsidy in the form of an overarching measure, the FSC, included the obligation to stop making "pre-committed" payments to existing FSCs (Foreign Sales Corporations).622 Again, Brazil ascribes a broader meaning to certain statements or findings of the Appellate Body than is warranted. The US – FSC (Article 21.5

              – EC) dispute concerned transitional provisions which allowed legislative measures that had been found, in the original dispute, to confer prohibited export subsidies, to continue to apply with respect to certain transactions. The "payments" in question were therefore not "pre-committed", as Brazil alleges. Rather, the legislative measure that allowed for subsidies to be provided simply continued to apply, albeit to a limited number of transactions.623 Again, we read the Appellate Body's findings in


              image

              619 That "… to continue to make payments under an export subsidy measure found to be prohibited is not consistent with the obligation to 'withdraw' prohibited export subsidies, in the sense of 'removing' or 'taking away'", Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), para. 45.

              620 See Appellate Body Report, Brazil – Aircraft (Article 21.5 – Canada), paras. 42 and 45.

              621 We also note that the panel and Appellate Body did not have to address, in Brazil – Aircraft

              (Article 21.5 – Canada), the question of whether Brazil had failed to "withdraw the subsidy" because Brazil was continuing to honour its obligations under bonds issued prior to the implementation deadline. That aspect of the dispute only concerned the issuance, by Brazil, of new bonds pursuant to letters of commitment entered into before the date of implementation. The United States notes that, in that case, Brazil presumably continued to make payments under outstanding bonds for more than a decade after the implementation date. See US Opening Statement, para. 47.

              622 Brazil refers to the Appellate Body Report, US – FSC (Article 21.5 – EC), paras. 223-231. See Brazil's Response to Panel Question 37, footnote 28 to para. 46, and Brazil's Closing Statement, footnote 38.

              623 Appellate Body Report, US – FSC (Article 21.5 – EC), para. 224. For the sake of completeness, we

              note that in the Brazil – Aircraft (Article 21.5 – Canada) case, the panel and the Appellate Body rejected arguments by Brazil that it should be allowed to continue to grant prohibited export subsidies after the end of the

              WT/DS267/RW

              Page 150


              that dispute as a reaffirmation of the principle that a Member has not "withdraw[n] the subsidy" where it continues to grant it after the end of the implementation period or where it "maintains" in place the legislative measure under which the subsidies are granted.


            6. Thus in our view, the relevant question is not whether the United States, after the end of the implementation period, makes any "payment" in a general sense under programmes that were found, in the original proceeding, to constitute prohibited export subsidies, pursuant to prior commitments. The relevant question is whether the United States continues or continued, after the end of the implementation period, to provide export subsidies.624


            7. The equivalent, in the present dispute, of the issuance of bonds in Brazil – Aircraft (Article 21.5 – Canada) is, we believe, the issuance of CCC export credit guarantees, not the payment, by the US Government, of claims under such guarantees, or the rescheduling or restructuring of debt following the payment of claims. We recall that the subsidies at issue in this dispute are not the payment of claims by the US Government (or rescheduling or restructuring of debt) but the issuance of export credit guarantees through the CCC. Brazil did not claim in the original proceeding that the United States provided export subsidies in the form of payments pursuant to existing export credit guarantees. On the contrary, Brazil challenged the very issuance of guarantees as part of programmes meeting the conditions of item (j) of the Illustrative List. The panel's recommendation to "withdraw the subsidy" concerned subsidies in the form of guarantees, not in the form of payments (of claims) under guarantees. Even before this Panel, in seeking to demonstrate that the United States circumvented its export subsidy commitments, Brazil does not refer the Panel to the volume of US agricultural exports for which claims were paid, but rather to the volume of US agricultural products covered by guarantees, whether or not claims were ultimately paid under such guarantees. Further, Brazil has not, in our view, sought to establish that the payment of a claim under a guarantee or the rescheduling/restructuring of debt are, in themselves, prohibited export subsidies, independently of the subsidy that is conferred in the form of an export credit guarantee.


            8. In sum, Brazil has not convinced us that the obligation to "withdraw the subsidy" pursuant to a recommendation adopted in accordance Article 4.7 of the SCM Agreement includes a general obligation to cease making payments or otherwise performing under outstanding or prior


              image

              implementation period because it had assumed contractual obligations to do so under municipal law (see supra, note 613) and that the Appellate Body rejected similar arguments made by the United States in US – FSC (Article 21.5 – EC). In the latter case, the United States had argued that it should be allowed to continue granting prohibited export subsidies to protect the contractual interests of private parties and to ensure an orderly transition to the new legislative regime. Appellate Body Report, US – FSC (Article 21.5 – EC), paras. 223-239. The rejection of the arguments made, respectively, by Brazil and the United States in these two disputes does not assist Brazil's argument in the present dispute. The panel and Appellate Body, in these disputes, rejected arguments by a responding Member that it should be allowed to continue granting prohibited export subsidies because it had committed to do so or because of other municipal law considerations (e.g. because of private parties' expectations). The rejection of these arguments is not equivalent to a finding that a Member is, in all cases, precluded from granting subsidies because they were pre-committed.

              624 While the Appellate Body indicated in Brazil – Aircraft (Article 21.5 – Canada) that Brazil's

              arguments regarding the timing of the existence of the granting of subsidies was not relevant to its analysis, we recall that in the original Brazil – Aircraft proceeding, both the panel and the Appellate Body reached the conclusion that PROEX subsidies were "granted" at the time of the issuance of the bonds. Were we, like the Brazil – Aircraft (Article 21.5 – Canada) panel, to consider that the "timing" of the granting of the subsidy is of relevance, we would conclude that export subsidies in the form of export credit guarantees are not "granted" at the time when the issuing agency makes a payment pursuant to a claim of default or when it reschedules or restructures debt covered by a guarantee. In our view, the recipient's "unconditional legal right" to receive the subsidy in question (i.e. export credit guarantee) arises at an earlier point in time, at the time of the issuance of the guarantee. A conclusion that the subsidies in question are granted at the time the United States makes a payment (on default) or upon the rescheduling or restructuring of debt would lead to the logically flawed result that many, if not most, of export subsidies in the form of export credit guarantees would never be granted.

              WT/DS267/RW

              Page 151


              commitments under a measure found to constitute a prohibited export subsidy. As a consequence, Brazil has not established, on that basis, that the United States has failed to "withdraw the subsidy" because it continues to pay claims pursuant to outstanding guarantees or because it reschedules or restructures debt covered by such guarantees.


        4. Meaning of the term "withdraw"


          1. Brazil also relies on the definition of the term "withdraw" provided by the Appellate Body in Brazil – Aircraft (Article 21.5 – Canada). As mentioned above, in that case, the Appellate Body interpreted that term in Article 4.7 of the SCM Agreement as requiring that a Member "remove" or "take away" the subsidy. For Brazil, "[p]aying a claim on default of a GSM 102 credit subject to the DSB's recommendation to 'withdraw the subsidy' does not 'remove' or 'take away' that subsidy".625


          2. Further, Brazil argues that the obligation to "withdraw" the subsidy in Article 4.7 of the Agreement, interpreted in the sense of "removing" or "taking away", extends to both elements of Article 3.2, the "granting" and the "maintenance" of a subsidy.626 Brazil argues that the Appellate Body, in US – FSC (Article 21.5 – EC II) held that "the obligation to 'fully' withdraw a prohibited subsidy is not achieved if the Member 'leaves the entirety or part of the original prohibited subsidy in place'".627 Thus, Brazil submits that "[b]y continuing to perform on its commitments under the prohibited subsidies, the United States 'maintains' them unchanged".628


          3. Neither of the two Appellate Body Reports cited by Brazil support its argument that the obligation to "withdraw the subsidy" includes an obligation to remove or take away individual subsidies that were granted in the past. Again, the Brazil – Aircraft (Article 21.5 – Canada) dispute concerned a situation where a Member continued to grant, after the end of the implementation period, new (individual) subsidies. The second Appellate Body Report cited by Brazil, US – FSC (Article 21.5 – EC II), addressed a situation where a Member maintained the legislative provisions that allowed the (future) granting of subsidies. In neither case did the Appellate Body indicate that the Member concerned was required to "withdraw" (in the sense of "removing" or "taking away") individual subsidies that had already been granted. Thus, Brazil has pointed to no precedent for the proposition that the obligation to "withdraw" in the sense of "removing" or "taking away", whether viewed in terms of the granting of a subsidy or in terms of its maintenance, extends to individual subsidies that were granted under a programme found to confer prohibited export subsidies, prior to the end of the implementation period.


          4. We need not, in this dispute, develop a comprehensive analysis of the extent of a Member's obligation to "withdraw the subsidy" pursuant to a recommendation adopted under Article 4.7 of the SCM Agreement. In particular, we need not decide whether the obligation to "withdraw the subsidy" pursuant to a panel recommendation adopted under Article 4.7 of the SCM Agreement imposes an obligation on a Member to "remove" or "take away", retrospectively, individual subsidies that were provided in the past. Brazil has emphasized that the remedy it is seeking is not a retroactive remedy and that its claims are entirely prospective in nature.629 It is therefore the Panel's understanding that Brazil considers that the WTO Agreements (in particular, the SCM Agreement) do not provide for retroactive remedies. Brazil has not explained how its interpretation of the obligation set out under Article 4.7 would not amount to a retroactive remedy, even though it would require the United States



            image

            625 Brazil's Response to Panel Question 37, para. 46.

            626 Article 3.2 of the SCM Agreement provides that "A Member shall neither grant nor maintain" prohibited subsidies.

            627 Brazil's Response to Panel Question 92, para. 199, referring to Appellate Body Report, US – FSC

            (Article 21.5 – EC II), para. 84 (emphasis added by Brazil).

            628 Brazil's Response to Panel Question 92, para. 200.

            629 Brazil's Response to Panel Question 92, para. 201.

            WT/DS267/RW

            Page 152


            to, in effect, eliminate subsidies previously granted and therefore take action with respect to acts which took place before the end of the time-period within which the United States had to "withdraw the subsidy". A finding on our part that the United States is precluded from making payments (or rescheduling or restructuring debt) pursuant to guarantees issued before 1 July 2005 would, in our view, amount to a retroactive remedy, similar to a finding that a Member must demand repayment, from the recipient, of a subsidy in the form of a direct transfer of funds.


          5. In light of the foregoing considerations, the Panel finds that Brazil has not established that the United States has failed to comply with the panel recommendations, adopted by the DSB, to "withdraw the subsidy" with respect to export credit guarantees issued prior to the end of the implementation period on 1 July 2005630, and to bring its measures into conformity with the Agreement on Agriculture.


    3. BRAZIL'S CLAIMS CONCERNING GSM 102 EXPORT CREDIT GUARANTEES ISSUED AFTER

      1 JULY 2005


      1. Introduction


          1. Brazil claims that the US measures taken to comply with the DSB recommendations and rulings are inconsistent with Articles 10.1 and 8 of the Agreement on Agriculture and, to the extent of that inconsistency, with Articles 3.1(a) and 3.2 of the SCM Agreement. This claim concerns GSM

            102 export credit guarantees provided to exports of unscheduled products and exports of three scheduled products – rice, pig meat and poultry meat631 – since the end of the implementation period,

            i.e. 1 July 2005. In addition, Brazil claims that GSM 102 export credit guarantees provided to exports of three products not falling within the scope of the Agreement on Agriculture, lysine, lyocell and wood products, are inconsistent with Articles 3.1(a) and 3.2 of the SCM Agreement.


          2. Brazil indicates that its claims of inconsistency under the Agreement on Agriculture and the SCM Agreement are "product-specific": Brazil claims that the United States has applied the GSM 102 programme in a manner that circumvents its export subsidy commitments with respect to unscheduled products and three scheduled products. Thus, we understand Brazil to be challenging GSM 102 export credit guarantees "as applied".


          3. We examine each of these claims in turn, starting with Brazil's claims under Articles 10.1 and 8 of the Agreement on Agriculture.


      2. Brazil's claims under Articles 8 and 10.1 of the Agreement on Agriculture


          1. Article 8 and 10.1 of the Agreement on Agriculture read as follows: Article 8: Export Competition Commitments

            Each Member undertakes not to provide export subsidies otherwise than in conformity with this Agreement and with the commitments as specified in that Member's Schedule.


            [...]



            image

            630 Given this conclusion, we need not determine whether the fact that the original panel's findings were reached on the basis of item (j) means, as the United States submits, that only the export credit programmes, and not individual export credit guarantees, needed to be withdrawn.

            631 The Panel has rejected the US request for a preliminary ruling concerning Brazil's claims relating to exports of pig meat and poultry meat. See supra, para. 9.27.

            WT/DS267/RW

            Page 153


            Article 10: Prevention of Circumvention of Export Subsidy Commitments


            1. Export subsidies not listed in paragraph 1 of Article 9 shall not be applied in a manner which results in, or which threatens to lead to, circumvention of export subsidy commitments; nor shall noncommercial transactions be used to circumvent such commitments.


          2. A violation of Article 10.1 requires a demonstration of two elements: first, that "export subsidies not listed" in Article 9.1 are "applied", and second, "in a manner which results in ... circumvention of export subsidy commitments." Brazil argues that GSM 102 are export subsidies under the terms of Article 10.1. It was not in dispute before the original panel that export credit guarantees are not included in the list of export subsidies contained in Article 9.1 of the Agreement on Agriculture, and that Article 10.1 was therefore the applicable provision.632 Neither of the parties in this compliance proceeding suggests otherwise.


          3. The Agreement on Agriculture does not generally define the meaning of the term "export subsidies". Article 1(e) of the Agreement provides that export subsidies are "subsidies contingent upon export performance, including the export subsidies listed in Article 9 of this Agreement". Panels and the Appellate Body have, in interpreting the meaning of "export subsidies" in Article 10.1, relied, inter alia, on the relevant provisions of the SCM Agreement (including Articles 1 and 3 and items of the Illustrative List) as "context".633 These provisions define what measures constitute, respectively, "subsidies", and "export subsidies", for the purposes of the SCM Agreement.


          4. Regarding the "circumvention of export subsidy commitments" in Article 10.1 of the Agreement on Agriculture, the disciplines to which Members are subject differ as between products for which a Member has entered "export subsidy reduction commitments" in its Schedule of commitments (so called "scheduled" products) and those for which it has made no such commitments ("unscheduled" products). For the latter, Members are subject to a general prohibition on the provision of export subsidies. This results from the language of the ending clause of Article 3.3 of the Agreement on Agriculture ("... shall not provide such subsidies [i.e. export subsidies under Article 9.1] in respect of any agricultural product not specified in [Section II of Part IV ] of its Schedule") and Article 10.1, which precludes a Member from "applying" subsidies not listed in Article 9.1 in a manner that results (or threatens to result) in circumvention of its export subsidy commitments.634 With respect to scheduled products, a Member is allowed to grant export subsidies up to the level of its reduction commitments in its Schedule. Reduction commitments are expressed on a budgetary outlay (total amount of subsidy provided) basis as well as on a quantity (volume) basis for each agricultural product concerned. Any export subsidy provided beyond the specific reduction commitments applicable to the agricultural product in question is prohibited.


          5. Given that only some of Brazil's claims of circumvention under Article 10.1 of the Agreement on Agriculture are subject to the special burden of proof rules of Article 10.3 of the Agreement 635


            image

            632 Panel Report, US – Upland Cotton, para. 7.788.

            633 See Appellate Body Report, US – FSC, paras. 136-141; Panel Report, US – FSC (Article 21.5 - EC), para. 8.116; Appellate Body Report, US – FSC (Article 21.5 – EC), paras. 187-196; Panel Report, US – Upland

            Cotton, paras. 7.797-7.803; Panel Report, Canada – Dairy (Article 21.5 – New Zealand and US), paras. 5.152- 5.162, Panel Report, Canada – Dairy, paras. 7.125 ff. This last panel considered that Article 9.1 of the Agreement on Agriculture can also serve as "context" to interpret the meaning of "export subsidy" under Article 10.1 of the Agreement.

            634 The term "export subsidy commitments" includes both reduction commitments undertaken with respect to scheduled products, and the commitment not to provide any subsidy on exports of unscheduled

            products. See Appellate Body Report, US – FSC, paras. 145-147.

            635 Under the usual rules on the burden of proof, the complaining Member would bear the burden of proving both aspects of its claim: that (1) export subsidies are provided; (2) the total amount of subsidies

            WT/DS267/RW

            Page 154


            the partial reversal of the burden of proof operated by Article 10.3 applies only to claims under the Agreement on Agriculture, and only with respect to scheduled products636 – we first conduct our analysis as if Brazil bore the burden of proof with respect to both the quantitative and the subsidization elements of its claims, i.e. as if Article 10.3 did not apply. Thus, we first proceed to determine whether GSM 102 export credit guarantees are "export subsidies" for the purpose of Article 10.1 of the Agreement on Agriculture. We conduct that analysis by relying on the provisions of the SCM Agreement as contextual guidance. Only if Brazil fails to demonstrate that GSM 102 export credit guarantees are export subsidies will we proceed – for those claims of Brazil that are subject to Article 10.3 of the Agreement on Agriculture (scheduled products) – with the analysis mandated under Article 10.3, first determining whether Brazil establishes that the United States exported the products concerned in quantities exceeding its commitments and then shifting the burden on the United States to prove that no subsidies were provided with respect to the quantities exported in excess of its commitments.


      3. Are GSM 102 export credit guarantees "export subsidies" – legal bases for Brazil's arguments


          1. We proceed to the first step of our analysis, whether GSM 102 export credit guarantees constitute "export subsidies" under Article 10.1 of the Agreement on Agriculture. As suggested above, given that the Agreement on Agriculture does not contain a comprehensive definition of the term and that the Agreement does not provide other guidance as to whether and under what conditions export credit guarantees may constitute export subsidies, we consider that it is appropriate to refer to the SCM Agreement as contextual guidance to determine whether GSM 102 guarantees are "export subsidies" under Article 10.1 of the Agreement on Agriculture. The parties have not suggested that we proceed otherwise.


          2. Brazil relies on two legal bases to demonstrate that GSM 102 export credit guarantees are "export subsidies" in the sense of Article 3.1(a) of the SCM Agreement. First, Brazil asks the Panel to find that GSM 102 export credit guarantees meet the definition of an export subsidy under the terms of Articles 1.1 and 3.1(a) of the SCM Agreement, i.e. that they are "financial contributions", that confer a "benefit" and that are contingent upon export performance. Alternatively, Brazil asks the Panel to conclude that GSM 102 export credit guarantees fall within the scope of item (j) of the


            image

            provided exceeds the Member's commitments (expressed in terms of budgetary outlays) for the product in question or the total quantity of exports supported by the subsidy(ies) in question exceeds the Member's commitments (expressed in volume of exports) for the product in question. Article 10.3 of the Agreement on Agriculture operates a partial reversal of the burden of proof in certain disputes under Articles 3, 8, 9 and 10 of the Agreement on Agriculture where the complaining Member alleges that volume commitments were exceeded. See, for a discussion of the effects of Article 10.3, Appellate Body Report, US – Upland Cotton, paras. 642 ff., Appellate Body Report, Canada – Dairy (Article 21.5 – New Zealand and US II), paras. 67 ff. Article 10.3 "'cleaves the complaining Member's claim' into two parts: a quantitative aspect, and an export subsididization aspect, 'allocating to different parties the burden of proof with respect to the two parts." Appellate Body Report, US – Upland Cotton, para. 644, quoting from Appellate Body Report, Canada – Dairy (Article 21.5 – New Zealand and US II), para. 69. For those claims subject to Article 10.3, consistent with the usual rules of burden of proof, it is for the complaining Member (in this dispute, Brazil) to prove that the responding Member (here the United States) has exported an agricultural product in quantities exceeding its commitment level. If the complaining Member succeeds in establishing this quantitative part of the claim, then under Article 10.3, the burden of proof shifts to the responding Member to prove that the quantities exported in excess of its commitments are not subsidized. Appellate Body Report, Canada – Dairy (Article 21.5 – New Zealand and US II), para. 74, quoted in part in Appellate Body Report, US – Upland Cotton, para. 645. The reversal of the burden of proof is total and complete – the complaining Member does not have to make a prima facie case that subsidisation exists. See Appellate Body Report, Canada – Dairy (Article 21.5 – New Zealand and US II), para. 75.

            636 Appellate Body Report, US – Upland Cotton, paras. 649, 652.

            WT/DS267/RW

            Page 155


            Illustrative List of Export Subsidies. Item (j) (in conjunction with Article 3.1(a)) indicates that the following are prohibited export subsidies:637


            (j) The provision by governments (or special institutions controlled by governments) of export credit guarantee or insurance programmes, of insurance or guarantee programmes against increases in the cost of exported products or of exchange risk programmes, at premium rates which are inadequate to cover the long-term operating costs and losses of the programmes.


          3. Brazil regards these two alternative "bases" as separate claims, rather than arguments in support of a single claim.638 As we have already mentioned, Brazil first asks the Panel to consider its "claim" that GSM 102 guarantees meet the definitional elements of an "export subsidy" under Articles 1 and 3 of the SCM Agreement. Only if the Panel rejects Brazil's "claim" in this respect does Brazil ask the Panel to also consider its "claim" under item (j) of the Illustrative List of Export Subsidies.639


          4. In our view, the two legal bases on which Brazil relies in support of its argument that GSM 102 export credit guarantees are export subsidies under the terms of the SCM Agreement and the Agreement on Agriculture are not separate claims, but are rather alternative arguments as to why these guarantees are export subsidies under Articles 3.1(a) and 3.2 of the SCM Agreement.640 A Panel finding under either of the legal bases relied upon by Brazil would, in both cases, be that the United States is providing export subsidies under the terms of Article 3.1(a) of the SCM Agreement and under Article 10.1 of the Agreement on Agriculture. Item (j) of the Illustrative List does not, in itself, provide that certain measures are prohibited, it merely indicates that certain measures are export subsidies under Article 3.1(a). In any case, we recall that we are, at this point, only relying on Article 3.1(a) of the SCM Agreement as contextual guidance to determine whether GSM 102 export credit guarantees are export subsidies under Article 10.1 of the Agreement on Agriculture. There can be no doubt that, in this context, the two legal bases relied upon by Brazil are indeed mere alternative arguments, and not separate claims.


          5. We first proceed to examine Brazil's arguments that GSM 102 export credit guarantees are export subsidies because they fall under the scope of item (j) of the Illustrative List.641 Our decision first to proceed with Brazil's arguments under item (j) is motivated by the fact that there is no dispute between the parties that measures that fall within the scope of item (j) are, under Article 3.1(a) of the SCM Agreement, per se export subsidies.642



            image

            637 This is by virtue of Article 3.1(a) of the SCM Agreement, which provides that:

            "Except as provided in the Agreement on Agriculture, the following subsidies, within the meaning of Article 1, shall be prohibited:

            (a) subsidies contingent, in law or in fact , whether solely or as one of several other conditions, upon export performance, including those illustrated in Annex I;" (footnotes omitted, emphasis added).

            638 Brazil's Comments on US Response to Panel Question 38, para. 112.

            639 Brazil indicates that its request for review of its item (j) claim solely in the alternative stands even if the Panel considers item (j) to be the source of an alternative argument, rather than an alternative claim. See Brazil's First Written Submission, footnote 593.

            640 The original panel reached the same conclusion. See Panel Report, US – Upland Cotton, para. 6.31.

            641 This is similar to the approach followed by the original panel. See Panel Report, US – Cotton, para.

            7.803.

            642 The United States considers that item (j) of the Illustrative List provides the only test to determine

            whether export credit guarantees are export subsidies. See, inter alia, US First Written Submission, paras. 62-

            70. Brazil considers that export credit guarantees may be found to constitute export subsidies within the meaning of Article 3.1(a) of the SCM Agreement either because they meet the terms of item (j) or because they

            WT/DS267/RW

            Page 156


          6. As a result of our conclusion that the two legal bases invoked by Brazil are alternative "arguments", not "claims", were we to conclude that GSM 102 export credit guarantees are export subsidies by virtue of the fact that they fall within the scope of item (j), we do not consider that we would necessarily need to determine whether they also, independently, meet the definitional elements of an export subsidy under Articles 1 and 3 of the SCM Agreement.


            4. Brazil's arguments under item (j) of the Illustrative List of Export Subsidies


            1. Original panel's analysis under item (j) of the Illustrative List of Export Subsidies


          7. Before proceeding with our analysis, we recall the main findings of the original panel in its analysis under item (j) of the Illustrative List, as context to determine whether the programmes at issue conferred export subsidies under Article 10.1 of the Agreement on Agriculture.


          8. The original panel considered that item (j) does not require the use of any particular methodological approach or accounting philosophy to determine whether a programme is run at a net cost to the government.643 The panel considered that it was reasonable, on the basis of the evidence and arguments submitted to it, to take into account the period since 1992.644 But it also considered that the analysis under item (j) needed not be purely retrospective and could take into account elements of the structure, design and operation of the measure which may affect the long-term operating costs and losses of the GSM 102, GSM 103 and SCGP programmes.645


            1. Past performance of the programmes


          9. The panel reviewed data submitted by the parties concerning the profitability of the programmes. Data under different methodologies was presented by the parties and relied upon by the panel. The first approach concerned the "net present value" approach to budget accounting used by the US Government with respect to export credit guarantees. The US Government projects, in each year's budget, the "net present value" of the costs associated with its export credit guarantees issued in that year.646 The United States budget data for 1992 to 2002 (and estimates for 2003 and 2004) showed a positive "guaranteed loan subsidy" (i.e. a loss). Although the panel recognized that these amounts were only initial estimates of the long-term costs of the programme, representing neither "random guesses nor historically verifiable real amounts", it considered that the consistently positive subsidy numbers indicated that the US Government "believes, based upon its own assessment, that it may not, even over the long term, be able to operate the export credit guarantee programmes without



            image

            meet the definitional elements of an export subsidy under Article 3.1(a) of the SCM Agreement (and, by reference, Article 1 of the SCM Agreement), i.e. that they are "financial contributions", that confer a "benefit" (thereby constituting a "subsidy" under Article 1), which is contingent upon export contingency under Article 3.1(a). See Brazil's Rebuttal, paras. 454-481. A number of third parties (Australia, Canada, Chad, and China) consider, like Brazil, that export credit guarantees can be prohibited export subsidies under either Articles 1.1 and 3.1(a) of the SCM Agreement or item (j) of the Illustrative List. The Panel understands the European Communities to agree with the United States' position. See Australia's Third Party Submission, para. 24, Canada's Oral Statement Submission, paras. 8-13, Chad's Third Party Submission, paras. 50-56, China's Third Party Submission, paras. 4-8 and China's Oral Statement, EC Oral Statement, para. 33.

            643 Panel Report, US – Upland Cotton, para. 7.804. 644 Panel Report, US – Upland Cotton, para. 7.831. 645 Panel Report, US – Upland Cotton, para. 7.835.

            646 A positive net present value means the US Government is granting a "subsidy" to borrowers and therefore expecting a loss, a negative present value means that the programme is generating a "profit." This

            value does not include administrative costs. See Panel Report, US – Upland Cotton, para. 7.842. Administrative costs are usually relatively modest in comparison to the value itself (US$5 million in the most recent years).

            WT/DS267/RW

            Page 157


            some net cost to government".647 The panel also took into account the fact that "actual historical experience was a primary factor" on which the estimates are based.648


          10. These initial estimates are subject to annual re-estimations over the lifetime of the guarantees involved. The United States argued in this respect that the re-estimates, over time, erased the initial positive "estimate" figure, in other words, that taking the re-estimates into account, the programmes would not give rise to a net cost to the US Government. The United States provided the panel with cumulative re-estimates on a cohort-specific basis (a cohort is comprised of all guarantees issued in a given year). These figures showed a positive subsidy of US$230 million. The United States noted an overall downward re-estimate of US$1.9 billion (i.e. overall, re-estimates brought the subsidy down), that, with the exception of 2002 (for which only very recent data was available), the trend for all cohorts was favourable as compared to the original subsidy amount, and that the data for the 1992- 1996 and 1999 cohorts indicated profitability. According to the United States, when data for the – as yet uncompleted 2001 and 2002 cohorts were removed, the US$230 million was almost entirely eliminated. While the panel took into account the fact that the process of estimates and re-estimates by cohort is necessarily a fluid one that cannot be definitively analyzed until a cohort is "closed", it disagreed with the United States that data for certain of the more recent cohorts should be disregarded; the panel was also not persuaded that cohort re-estimates over time, would necessarily not give rise to a net cost to the US Government. It therefore rejected the arguments of the United States in this respect.649


          11. The second "methodology" under which the original panel examined the US programmes was a constructed cost formula submitted by Brazil, based on data from the US budget.650 Brazil submitted a table which showed a net cost to government of US$1.08 billion from FY 1993 to FY 2002 (using data taken in the "prior year" column in the US budget).651 The panel considered it an "an additional informed appraisal, based on information generated by the US Government itself, of the performance of export credit guarantee programmes" which, the panel noted, takes into account the actual level of guarantees issued by the CCC in a given year, as opposed to an estimate of this level.652 The panel compared the result of this formula with evidence submitted by the United States, which it asserted, reflected a profit of US$630 million to the US Government. The difference between the results of Brazil's and the United States' calculations corresponded to the amount of rescheduled claims (US$1.6 billion), which the United States did not treat as "outstanding claims": rather, the standard accounting treatment of reschedulings by the CCC was to treat them as new direct loans; this, in the panel's view, represented an instant reduction of the amount in outstanding claims in the year in which the rescheduling was finalized. The panel thus shared Brazil's concerns that the United States' treatment of rescheduled debt understated the net cost to the US Government.653


          12. The panel also took into account the fact that the CCC financial statements for the years 2002 and 2003 indicated a "credit guarantee liability" of US$411 million and US$22 million, respectively. The panel observed that these "credit guarantee liability" amounts "are not actual losses. They are but another indicator, used and relied upon by the US government, to assess the estimated long-term cost


        image

        647 Panel Report, US – Upland Cotton, para. 7.843.

        648 Panel Report, US – Upland Cotton, para. 7.843.

        649 Panel Report, US – Upland Cotton, paras. 7.852-7.854.

        650 The formula calculated the following: (Premiums collected + Recovered principal and interest (line

        88.40 of the US budget) + Interest revenue (line 88.25)) – (Administrative expenses (line 00.09) + Default claims (line 00.01) + Interest expense (line 00.02)). See Panel Report, US – Upland Cotton, para. 7.844.

        651 See the table at para. 7.845 of the Panel Report, US – Upland Cotton.

        652 Panel Report, US – Upland Cotton, para. 7.846.

        653 Panel Report, US – Upland Cotton, para. 7.851.

        WT/DS267/RW

        Page 158


        to the United States government of export credit guarantees", and that these amounts were consistently positive.654


          1. Overall, the panel found that these considerations relating to the past performance of the programmes supported a view that the programmes were run at a net cost to the government.655


            1. Structure, design and operation of the programme


          2. The panel also took into consideration the structure, design and operation of the export credit guarantee programmes. Several elements in this respect led the panel to believe that the programmes were "not designed to avoid a net cost to government".656 The panel took into consideration the fact that the CCC borrows from the US Treasury and that access to Treasury borrowings facilitates the functioning of the programmes and that, for a creditor, the guarantees provide the backing of the full faith and credit of the US Government. Further, the panel was of the view that the premiums were not geared toward ensuring adequacy to cover long-term operating costs and losses for the purposes of item (j) of the Illustrative List. First, it noted in this respect the 1 per cent fee cap imposed by US legislation on the fees.657 Second, it considered that the premiums were not risk-based, either with respect to country risk or the creditworthiness of the borrower in an individual transaction. Country risk and foreign obligor risk only impacted the US Government subsidy estimates and re-estimates, and not the fees charged. 658 Moreover, the statute imposed no limitation on the amount of guarantees that could be provided annually to a high-risk country or to high risks transactions. The CCC was, therefore, able to provide large amounts of guarantees to high-risk countries with a resulting high rate of default. The panel also noted the lack of substantial or regular review of the premiums by the US Government to ensure that they were adequate to cover long-term operating costs and losses. Finally, the panel noted evidence (an annual review of fees by the USDA) that indicated that the US Government did not consider that the fees were risk-based and that they did not cover the "subsidy cost of the programme"; the document also stated that the 1 per cent fee cap likely prevented the adoption of risk-based fees.659 Thus, for the panel, the evidence indicated that the premiums were not proportionate to, reflective of, nor geared towards, covering long-term operating costs and losses.660


            1. Overall conclusion by the original panel


          3. Thus, on the basis of the totality of the evidence, the panel found that Brazil had established that the three export credit guarantee programmes were provided "at premium rates which are inadequate to cover the long-term operating costs and losses of the programmes" within the meaning of item (j) of the Illustrative List. This view was "based on a careful consideration of the evidence, taken as a whole, and no one element, in isolation, is determinative".661 The panel therefore found that the GSM 102, GSM 103 and SCGP export credit guarantee programmes constituted a per se export subsidy within the meaning of item (j) of the Illustrative List and were therefore export subsidies under Article 10.1 of the Agreement on Agriculture.


            image

            654 Panel Report, US – Upland Cotton, para. 7.855. 655 Panel Report, US – Upland Cotton, para. 7.856. 656 Panel Report, US – Upland Cotton, para. 7.857. 657 Panel Report, US – Upland Cotton, para. 7.860.

            658 Country risk is the risk – represented by a credit rating – associated with a given country, and

            usually relates to the risk of political or economic crises, etc; foreign obligor risk is the risk specific to an individual foreign obligor.

            659 Panel Report, US – Upland Cotton, para. 7.865. 660 Panel Report, US – Upland Cotton, para. 7.867. 661 Panel Report, US – Upland Cotton, para. 7.867.

            WT/DS267/RW

            Page 159


            1. Appellate Body


          4. The Appellate Body agreed with the panel's approach to assessing whether the programmes at issue met the conditions set forth in item (j) of the Illustrative List662 and that "[t]he text of item (j) does not suggest that this provision requires a Panel to choose one particular basis for the calculation and then to make a precise quantification of the difference between premiums and long-term operating costs and losses on that basis".663


          5. Like the original panel, we conduct a two-step analysis. First, we review the evidence of a quantitative nature submitted by the parties to determine whether the GSM 102 programme is provided at premium rates that are inadequate to cover its long term operating costs and losses; second, we examine the evidence pertaining to elements of the "structure, design and operation" of the GSM 102 programme submitted by Brazil.


            1. "Quantitative" analysis under item (j)


              1. Main arguments of the parties


          6. Brazil submits that the US Government continues to predict that GSM 102 export credit guarantees will be provided at a loss, even under the new fee schedule. Brazil relies on the "subsidy" estimates for FY 2006, 2007 and 2008, which continue to be positive. Brazil refers to the findings of the original panel, which took similar projections as an indication that the US Government considers that the programmes would be provided at a net cost.664


          7. The United States submits that current budget financial data shows that the fee structure in place during the original panel proceeding provided premia more than adequate to cover the long term costs and losses of the programmes at issue in that proceeding. The data – which concerns subsequent reestimates of initial subsidy estimates – became available as more cohorts of export credit guarantees closed. The data demonstrates that the programmes were profitable even before the suspension of the GSM 103 and SCGP programmes and the significant modifications of the fee structure and eligibility under the programmes. Further, the United States argues that the initial subsidy estimates in the US budget are not derived from the actual experience of the CCC. The original estimates are calculated before any use is made of the export credit guarantee programme in the year for which the estimate is made and are based on an historically overly-optimistic projection of the actual use of the programme. In addition, the CCC is, in establishing these estimates, required to use government-wide estimation rules, without regard to the actual experience specific to the CCC export credit guarantee programmes.665


          8. Brazil counters that the result of the US Government "re-estimates" is elsewhere tracked in the CCC's Financial Statements. The Financial Statements report a positive "credit guarantee liability" of US$220 million for guarantees outstanding as of 30 September 2006. Brazil also submits "cash basis" calculations which, in its view, confirm that the programmes have not been profitable since 1992.666


          9. The United States rejects Brazil's reading of the "credit guarantee liability" figure in the CCC's Financial Statements. In addition, the United States questions the significance of Brazil's cash



            image

            662 Appellate Body Report, US – Upland Cotton, paras. 665 ff.

            663 Appellate Body Report, US – Upland Cotton, para. 665.

            664 Brazil's First Written Submission, paras. 433-437; Brazil's Opening Statement, paras. 252-253.

            665 US First Written Submission, paras. 100-104; US Rebuttal Submission, paras. 108-126

            666 Brazil's Rebuttal Submission, paras. 482 ff.

            WT/DS267/RW

            Page 160


            basis accounting methodology and submits that Brazil omits to include recoveries in respect of guarantees issued before 1992 in its calculation.667


            1. Evaluation by the Panel


          10. Before this Panel, Brazil again relies on the initial or original "subsidy estimates" reported in the US Budget. The 2007 US budget projected a "positive subsidy" (or loss) of US$125 million for the 2006 cohort and US$114 million for the 2007 cohort.668 The 2008 budget revised these numbers to, respectively, US$71 and US$60 million, and in addition, projected a US$53 million subsidy estimate for the guarantees issued in the 2008 year.669 None of these figures include administrative costs (which amount to US$5 million a year).670


          11. We recall that the original panel found that the initial estimates were one of the "considerations relating to the past performance of the programmes" that supported its conclusion that the programmes were provided "at premium rates which are inadequate to cover the long-term operating costs and losses of the programmes" within the meaning of item (j) of the Illustrative List.671 The original panel found as follows with respect to these initial estimates:


            "We realize that these amounts are initial estimates of the long-term costs to the United States government. They are not, however, mere random guesses as to the amount of possible, but highly unlikely, costs to the government. Nor, at the other extreme, are they historically verifiable real amounts that have been, or actually will be, disbursed by the United States government. Rather, this is a methodology used and relied upon by the United States government to assess the estimated long-term net cost to the United States government of export credit guarantees. Actual historical experience is a "primary factor" on which estimates are based.672 The consistently positive numbers in the United States budget guaranteed loan subsidy line indicate to us that the United States government believes, based upon its own assessment, that it may not, even over the long term, be able to operate the export credit guarantee programmes without some net cost to government." 673 (emphasis original)


          12. We agree with the original panel that consistently positive initial subsidy estimates provide an indication by the US Government that it expects the export credit guarantee programmes in question to be run at a net cost. It is in our view highly significant that the US Government continues to


            image

            667 US Rebuttal Submission, paras. 84 ff.

            668 The CCC is required to make annual subsidy estimates as a result of the Federal Credit Reform Act of 1990 (the "FCRA"). The purpose of the reform implemented by the FCRA is to "measure more accurately

            the costs of Federal credit programs". See 2 U.S.C. § 661(a), Exhibit Bra-545. The net present value analysis attempts to calculate the value today of a future stream of income or cost; under its budget accounting, the US Government identifies an annual "cost" in terms of the "net present value" associated with its export credit guarantee programmes. See Panel Report, US – Upland Cotton, para. 7.842.

            669 See line 233001 of the 2007 (for years 2006 and 2007) and 2008 budgets (respectively, Exhibits Bra-544 and US-71). The revision occurs because the subsidy estimate that first appears is based on an estimate

            of the use of the programme. Budgets for the two subsequent years following that budget progressively take into account the actual use of the programme. See discussion, infra, para. 14.76.

            670 See line 00.09 for each of the 2006, 2007, and 2008 budgets (respectively, Exhibits Bra-544 and

            US-71).


            671 Panel Report, US – Upland Cotton, paras. 7.856, 7.867.

            672[original footnote] Federal Accounting Standards Advisory Board, 'Statement of Federal Financial

            Accounting Standards, No. 19, Technical Amendments to accounting Standards for Direct Loans and Loan Guarantees' in Statement of Federal Financial Accounting Standards No. 2" (March 2001), p. 16, para. 36, reproduced in Exhibit BRA-118 states: "Actual historical experience of the performance of a risk category is a primary factor upon which an estimation of default cost is based".

            673 Panel Report, US – Upland Cotton, para. 7.843.

            WT/DS267/RW

            Page 161


            project, at the time of their issuance, that new GSM 102 guarantees issued under the revised programme will be provided at a net cost.


          13. We now address the arguments the United States makes against the use of the initial estimates to assess the long term net cost of the GSM 102 programme to the US Government. The United States raises two arguments in this respect.674 First, the United States argues that the initial subsidy estimates are inflated because they are issued before any use is made of the programme and therefore do not reflect the actual use of the programme. Second, the United States argues that the initial subsidy estimate is unreliable because the CCC is required, in preparing the estimates, to use government- wide estimation rules, including mandated risk assessment country grades, without regard to CCC's actual default experience. The United States submits in support of this second argument a Report from the General Accounting Office which explains that the principal factors used in the calculation of the original subsidy rate are provided by the Office of Management and Budget (OMB).675 For the United States, this "one-size-fits-all approach ... largely explains the continuing presence of an original subsidy estimate for CCC export credit guarantee programmes in fiscal years 2006 and 2007".676


          14. In both cases, however, the United States is reasserting arguments that it made in the original proceeding and that were, at least implicitly, rejected by the original panel.677


            image

            674 We address below, in paras. 14.78 ff. the evidence submitted by the United States with respect to subsequent reestimates of the initial subsidy estimates.

            675 US Rebuttal, paras. 108 ff. and Exhibit US-73. The document indicates that the OMB provides such agencies with "expected loss rates", a key component of their subsidy costs. OMB also provides the

            discount rates that are used to calculate subsidy estimates. The United States also explains that, starting with fiscal year 2003, OMB changed its method for determining expected loss rates because emerging finance literature indicated its former approach might overstate losses to the Government. OMB's methodology, however, continues to make certain assumptions that are not consistent with CCC's experience nor with its programme. More specifically, the United States indicates that the OMB uses "the market price of credits with the lowest ICRAS rating (category 11) [as] the predominant basis for recovery rates." OMB's use of the market price of the "lowest-rated credits is based on the assumption that this value represents the most the US government would recover in the event of default." Each sovereign borrower or guarantor is rated on an 11- category scale, ranging from A through F– (or their numerical counterparts, categories 1-11). For CCC, however, no country below a certain ICRAS rating is eligible for the programme at all. Consequently, OMB's imposed assumption about recovery rates as part of the calculation of expected loss rates is much more conservative than would be warranted with respect to the CCC". The United States also argues that, to date, the OMB has not incorporated any data specific to CCC experience in its calculations of expected loss rates that it requires CCC to employ in the preparation of its budget at the beginning of each fiscal year.

            676 US Rebuttal, para 116.

            677 For instance, the United States made the following statements in the original proceeding:


            "The original subsidy estimate, in contrast [to the "actual" estimate figure two years thereafter] begins with what is an historically overly optimistic projection of actual use of the program and then is required to use the government-wide estimation rules, including mandated risk-assessment country grades without regard to the actual experience specific to the CCC export credit guarantee programs. These two factors are the answer to Brazil's question, 'why does [CCC] continue to offer original estimates that are so high?' (footnotes omitted).

            (US Further Submission, 30 September 2003, para. 147.)

            ...

            "The United States has previously noted that the original subsidy estimate in the US budget begins with an historically overly optimistic projection of the actual use of the programme and then the use of government-wide estimation rules is required, including mandated risk assessment country grades without regard to the actual experience specific to the CCC export credit guarantee programmes. The original subsidy estimate for these cohorts, like the

            WT/DS267/RW

            Page 162


          15. Notwithstanding these arguments of the United States, the original panel concluded that the original subsidy estimate could be relied upon as an indicator of the long term net cost of the programmes to the US Government – or at least of an indicator of the US Government's assessment of such costs.678


          16. As Brazil points out "[t]he process by which the CCC arrives at initial estimates of long-term programme performance has not changed since the original proceedings; nor have the criteria used by the CCC to make this assessment".679 There are, in sum, no changes in the underlying circumstances that would justify us revisiting the original panel's reasoning in this respect. 680


          17. In any case, we would note that while the United States is factually correct that the initial estimates are established before any use is made of the programme and that the initial estimates are based on the projected use of the programme, these initial estimates are subsequently re-adjusted in the US Budget two years after the budget year in which the initial estimate appears.681 In no case has



            image

            original subsidy estimate for all cohorts, necessarily reflects no actual operating experience with respect to that cohort." (footnotes omitted).

            (US Response to original panel's Question 221(g), Panel Report, US – Upland Cotton, Annex I, page I-371)

            ...

            "The United States has noted that such [original] estimates are compelled by government-wide accounting rules for credit programs..." (US Further Rebuttal Submission, 18 November 2003, para. 196)

            ...

            "Brazil also mysteriously alleges that 'according to CCC's 2002 financial statements, CCC has been relieved of what the United States argues are onerous government-wide accounting rules that 'compel' projection of enormous losses.' CCC, however, has never been so 'relieved.' It remains compelled to adhere to the requirements of the federal Credit Reform Act of 1990, and relevant provisions of the Office of Management and Budget Circular A-11, implementing that legislation. CCC remains subject to government-wide requirements for subsidy estimates and the risk categories mandated by OMB with respect to exposure to debt from different countries. The government-wide rules continue to dictate the methodology for calculation of estimates, and reestimates..." (US Closing Statement at the Second Panel Meeting, 3 December 2003, para. 10).

            678 Panel Report, US – Upland Cotton, para. 7.843 (cited above, para. 14.70):


            "We realize that these amounts are initial estimates of the long-term costs to the United States government. They are not, however, mere random guesses as to the amount of possible, but highly unlikely, costs to the government. Nor, at the other extreme, are they historically verifiable real amounts that have been, or actually will be, disbursed by the United States government. Rather, this is a methodology used and relied upon by the United States government to assess the estimated long-term net cost to the United States government of export credit guarantees. Actual historical experience is a 'primary factor' on which estimates are based. The consistently positive numbers in the United States budget guaranteed loan subsidy line indicate to us that the United States government believes, based upon its own assessment, that it may not, even over the long term, be able to operate the export credit guarantee programmes without some net cost to government." (footnotes omitted, emphasis added).

            679 Brazil's Response to Panel Question 105, para. 300.

            680 See Appellate Body Report, US – Softwood Lumber VI (Article 21.5 – Canada), para. 103, citing Appellate Body Report, Mexico – Corn Syrup (Article 21.5 – US), para. 121.

            681 Each annual budget provides long-term subsidy estimates for three cohorts of export credit guarantees: the cohort of guarantees to be issued in the fiscal year that has not yet begun (for instance, in the

            2008 budget, the 2008 cohort), the cohort of guarantees issued in the fiscal year that has partly elapsed at the time the budget is released (in the 2008 budget, the 2007 cohort), and the cohort of guarantees issued in the fiscal year that has most recently ended (in the 2008 budget, the 2006 cohort). In subsequent years, re-estimates

            WT/DS267/RW

            Page 163


            such an adjustment turned a positive subsidy estimate into a negative one.682 We recall that, in this case, we are only interested in the presence of a net cost to the US Government resulting from the provision of GSM 102 export credit guarantees. We need not quantify such a cost. Therefore, the fact that the initial subsidy estimate may be overstated does not, in our view, mean that it is not a reliable predictor of a net cost to the US Government.


          18. Nor do we see a reason to depart from the original panel's conclusion that the subsidy estimates are, at least in part, based on historical experience and can be read as the US Government's own assessment of the long term costs of the programme(s). Brazil also directs us to the USDA's "Agriculture Financial Standards Manual", which states that "[i]n estimating default costs, the following risk factors are considered: (1) loan performance experience...".683 and to a "self- assessment" of the programmes by the USDA Foreign Agricultural Service (FAS) in 2004.684 The self-assessment states that the credit models used to calculate the subsidy estimate were revised in FY2001 and 2003 and "currently provide reliable estimates".685 While we take note of the United States' arguments that the OMB continues to impose expected loss rates or estimated default costs, in our view, this does not bring into question the fact that this is the method which is relied upon by the US Government itself to assess the long-term net cost of its export credit guarantee programmes at the time of issuance of the guarantees.


          19. The United States also submits that new US budget data that became available after the original panel conducted its analysis demonstrates that the programmes were in fact not provided at a net cost to the US Government even before it took measures to comply with the DSB recommendations. The data submitted by the United States concerns cumulative subsequent "re- estimates" of the initial subsidy estimates on a cohort-specific basis: initial "subsidy" estimates in the US budget are subject to re-estimates over the lifetime of the guarantees involved.686 As mentioned above, the United States had made similar arguments before the original panel, but the panel was not convinced that the evidence presented by the United States showed that the cohort estimates (which showed an overall positive subsidy estimate, i.e. a "loss") would eventually turn into a negative figure. We understand the United States to argue that the Panel should take into consideration the reestimates data in addition to the initial subsidy estimates. The United States is therefore also challenging the fact that initial estimates may (by themselves, without taking the re-estimates into account) be relied


            image

            are tracked in Table 8 of the Federal Credit Supplement accompanying the budget. Thus, for each cohort, subsequent budgets will provide three estimates: the estimate reported in the budget for the first fiscal year takes no account of the actual use of the programme, the estimate reported in the budget for the second fiscal year partly takes into account the actual use of the programme, and the estimate reported in the budget for the third fiscal year (and which is released after the cohort has closed) is based on the actual use of the programme. The amounts reflecting the actual use of the programme usually decline in absolute amounts. See Brazil's Rebuttal Submission, para. 508 and accompanying footnote. We do not believe there was any misunderstanding between the parties and the Panel on this issue in the original dispute. See for instance, US Further Submission (30 September 2003), para. 149:


            "Actual guarantee issuance can first be reflected only in the budget two fiscal years after the original subsidy estimate. Once the actual use of the program is determined all subsequent reestimates are based on that figure, not on the original subsidy estimate. Other than with respect to interest (because of independent market forces), a downward reestimate never occurs based on the original subsidy estimate."

            682 See Exhibits Bra-617 and US-8.

            683 Exhibit Bra-616, US Department of Agriculture, Office of the Chief Financial Officer, Credit, Travel, and Accounting Division, Agriculture Financial Standards Manual (May 2004), pages 120-121.

            684 See Brazil's Opening Statement, para. 253.

            685 Exhibit Bra-588 (Agricultural Export Credit Guarantee Programs Assessment, ExpectMore.gov, Section 3.CR2).

            686 US First Written Submission, para. 83, also referring to the original panel report, para. 7.843.

            WT/DS267/RW

            Page 164


            upon as a reliable indicator of the long-term net cost of the programmes or, at least, as representing the US Government's assessment of the projected long-term net cost of the programme.


          20. We note that the new data submitted by the United States concerns the three programmes examined by the original panel (GSM 102, GSM 103 and SCGP). The issue before us is whether the GSM 102 programme, as revised, is provided at premium rates inadequate to cover its long term operating costs and losses. The original panel found that the United States provided the three programmes at a net cost to its government. That issue has been definitively disposed of in the original proceeding and is not before us. However, we must consider the arguments and evidence submitted by the United States to the extent that they are relevant to determine whether the revised GSM 102 programme is run at a net cost to the US Government or, as we understand the United States' argument, that we should rely on the reestimates rather than the initial subsidy estimates as an indicator of the long-term cost of the programme. The United States submits to the Panel a table687 that includes reestimates for the 1992 to 2006 cohorts. The table shows, for all three programmes, an aggregate overall negative "subsidy estimate net of reestimates" (i.e. anticipated "profit") of US$926 million for the 1992-2002 cohorts (i.e. the cohorts examined by the original panel) and of US$403 million for the 1992-2006 cohorts.


          21. The original panel was not persuaded that cohort reestimates, over time, would necessarily not give rise to a net cost to the United States Government.688 Similarly, we are not convinced that the re-estimates data submitted by the United States establishes that all positive subsidy estimates – and in particular, those for cohorts 2006-2008 – will eventually turn into negative ones. We note in this respect that the US Government continues to project, for the new cohorts of GSM 102 guarantees issued after 1 July 2005 (i.e. after the CCC revised the programme's fee structure and discontinued its use of the GSM 103 and SCGP programmes), that the guarantees will be provided at a long-term net cost.


          22. Nor are we convinced that the subsidy reestimates data demonstrates that the GSM 102, GSM 103 and SCGP programmes were provided at no net cost to the US Government in the concerned periods. The reestimates data provided by the United States (except those related to the 1994 and 1995 cohorts, which have closed) are revised estimates. They indicate that the US Government now projects the cohorts of guarantees at issue to be profitable. But, because they are revised estimates, they do not establish that the programmes were provided at no net cost to the US Government.689 We further recall that the original panel's conclusion that the three programmes at issue fell within the scope of item (j) was based on a number of indicators of the past performance of the programmes, in addition to the subsidy estimates.690 The original panel took into consideration "cash accounting" evidence submitted by Brazil and the "credit guarantee liability" figure reported in the CCC's Financial Statements.691 The evidence considered by the Panel under both of these methodologies pointed to the programmes being provided at a net cost to the US Government. The CCC continues to report, in its Financial statements, a "credit guarantee liability": the CCC's 2005-2006's Financial


            image

            687 The table was submitted by the United States as part of its Response to Panel Question 111. It includes data contained in the 2008 budget. The United States had initially submitted a table containing re- estimates data contained in up to the 2007 budget. See paras. 87-89 of the US First Written Submission.

            688 Panel Report, US – Upland Cotton, para. 7.853.

            689 The Panel notes that, according to the re-estimates data submitted by the United States, certain

            cohorts (the 1997, 1998 and 2002-2006 cohorts) continue to present a positive "subsidy estimate net of re- estimates".

            690 The United States recognizes this. See US Response to Panel Question 105: "... in the original

            panel's conclusions, the original subsidy estimates were not dispositive on the question of whether the three export credit guarantee programs met the test of item (j). It was but one factor in the original panel's analysis."

            691 The panel's conclusion also rested on its findings that a number of elements of the structure, design

            and operation of the programme indicated that the programme was not designed to cover its long term operating costs and losses. See supra, para. 14.61.

            WT/DS267/RW

            Page 165


            Statement692 report a "credit guarantee liability" of US$220 million.693 The original Panel noted the following with respect to the "credit guarantee liability" figure reported in the CCC's Financial Statements:


            "The CCC defines the term "credit guarantee liability" as the estimated cash outflows of the guarantees on a net present value basis. "Liability" is defined as "...a probable future outflow or other sacrifice of resources as a result of past transactions or events." We observe that these amounts are not actual losses. They are but another indicator, used and relied upon by the United States government, to assess the estimated long-term cost to the United States government of export credit guarantees. They are consistently positive, indicating to us that the CCC believes, based upon its own assessment, that it may not, even over the long term, be able to operate the export credit guarantee programmes without some net cost to government." 694 (footnotes omitted)


          23. Notwithstanding that the original panel relied in part on the "credit guarantee liability" figure reported in the CCC's Financial Statements, the parties disagree as to the meaning of this figure.695 The United States argues that the credit guarantee liability is merely a balance sheet entry (corresponding to a future disbursement under the programme) of a liability for which an offsetting entry of an asset exists and that the CCC has an expectation of a recovery associated with the payment for default on the guaranteed loan, such that such receivables may not properly be considered "losses".696 Brazil rejects the United States' reading as being inconsistent with the definition of the term "credit guarantee liability" provided in the Financial Statements themselves. Brazil draws attention to the use of the word "net": this, in Brazil's view, belies the US assertion that the credit guarantee liability figure in the CCC financial statements includes only a liability for which an offsetting entry of an asset exists:697


            "Credit guarantee liabilities represent the estimated net cash outflows (loss) of the guarantees on a net present value basis. To this effect, CCC records a liability and an expense to the extent, in management's estimate, CCC will be unable to recover claim payments under the Credit Reform Export Credit Guarantee programs." (emphasis by Brazil). 698


          24. We see no reason to depart from the line of reasoning adopted by the original panel. It is clear, in our view, that the "estimated net cash outflow" of outstanding guarantees means that with respect to guarantees that were outstanding as of 30 September 2006, the US Government estimates


            image

            692 Brazil's Rebuttal Submission, paras. 484-486.

            693 See Exhibit Bra-585 (Audit Report, Commodity Credit Corporation's Financial Statements for Fiscal Years 2006 and 2005, November 2006, p. 11 of Notes to the Financial Statements). The financial

            statements define the term "credit guarantee liability" in the following terms:

            "Credit guarantee liabilities represent the estimated net cash outflows (loss) of the guarantees on a net present value basis. To this effect, CCC records a liability and an expense to the extent, in management's estimate, CCC will be unable to recover claim payments under the Credit Reform Export Credit Guarantee programs."

            694 Panel Report, US – Upland Cotton, para. 7.855.

            695 The parties have, in their Responses to Panel Questions 107-108 and Comments thereon, provided an extensive discussion of their arguments on this point.

            696 US Responses to Panel Questions 107 and 108. See also the US Rebuttal Submission, paras. 87-88.

            697 Brazil's Comments on US Response to Panel Questions 107 and 108, para. 298.

            698 See Brazil's Comments on US Response to Panel Questions 107 and 108, paras. 297-298 and Exhibit Bra-585, p. 11.

            WT/DS267/RW

            Page 166


            future disbursements of US$220 million.699 Thus, in our view, the "credit guarantee liability" reported in the CCC's financial statements provides another indication of the US Government's view that further costs will be incurred with respect to guarantees outstanding as of 30 September 2006.


          25. As mentioned above, Brazil also submits evidence concerning the cumulative receipts and disbursements for the three programmes pursuant to a "cash basis accounting" methodology. The chart submitted by Brazil in this respect is similar to the results of the constructed "cost" formula it submitted to the original panel700, but covers the period from FY1993 to FY2005 (the table submitted to the original panel covered the period up to 2002). The table (Exhibit Bra-613) is as follows:


            US$


            FY


            Premiums collected (88.40)

            Recovered principle and interest (88.40)


            Interest revenue (88.25)


            Sub Total


            Administrative expenses (00.09)


            Default claims (Line 00.01)


            Interest expense (Line 00.02)


            Sub-Total

            1993

            27,608,000

            12,793,000

            15,672,000

            56,073,000

            3,320,000

            570,000,000

            0

            573,320,000

            1994

            20,893,000

            458,954,000

            0

            479,847,000

            3,381,000

            422,363,000

            0

            425,744,000

            1995

            18,000,000

            62,000,000

            0

            80,000,000

            3,000,000

            551,000,000

            10,000,000

            564,000,000

            1996

            20,000,000

            68,000,000

            26,000,000

            114,000,000

            3,000,000

            202,000,000

            61,000,000

            266,000,000

            1997

            14,000,000

            104,000,000

            26,000,000

            144,000,000

            4,000,000

            11,000,000

            62,000,000

            77,000,000

            1998

            17,000,000

            81,000,000

            54,000,000

            152,000,000

            4,000,000

            72,000,000

            62,000,000

            138,000,000

            1999

            14,000,000

            58,000,000

            0

            72,000,000

            4,000,000

            244,000,000

            62,000,000

            310,000,000

            2000

            16,000,000

            100,000,000

            99,000,000

            215,000,000

            4,000,000

            208,000,000

            62,000,000

            274,000,000

            2001

            18,000,000

            149,000,000

            125,000,000

            292,000,000

            4,000,000

            52,000,000

            104,000,000

            160,000,000

            2002

            21,000,000

            155,000,000

            61,000,000

            237,000,000

            4,000,000

            40,000,000

            93,000,000

            137,000,000

            2003

            20,000,000

            187,000,000

            56,000,000

            263,000,000

            4,000,000

            102,000,000

            84,000,000

            190,000,000

            2004

            22,000,000

            219,000,000

            61,000,000

            302,000,000

            4,000,000

            130,000,000

            81,000,000

            215,000,000

            2005

            508,000,000

            0

            0

            508,000,000

            4,000,000

            190,000,000

            80,000,000

            274,000,000

            2,914,920,000

            3,604,064,000


          26. Over the period FY 1993-2005, the table shows programme receipts of US$2.9 billion, against programme disbursements of US$3.6 billion, for a net loss of over US$689 million.701


            image

            699 We do not understand the United States to fundamentally disagree with the original panel's interpretation of the meaning of the "credit guarantee liability" reported in the CCC's Financial Statements. See US Response to Panel Questions 107 and 108, para. 271:


            "Turning to the Panel’s question as to whether the liability figure may be considered a loss, the United States notes the original panel’s explanation that a 'liability' is defined as "... a probable future outflow or other sacrifice of resources as a result of past transactions or events." The United States considers that this is an accurate and proper definition of liability. The original panel specifically noted that such a 'probable future outflow' does not equate with 'loss.' Citing the Federal Accounting Standards Advisory Board Statements of Federal Financial Accounting Concepts and Standards, the panel noted that the definition of 'loss,' in contrast, is 'any expense or irrecoverable cost, often referred to as a nonrecurring charge, an expenditure from which no present or future benefit may be expected.' CCC clearly has an expectation of recovery on the receivable it has obtained in exchange for its payment for a default on a guaranteed loan. Therefore, the receivables are not properly considered to be 'losses.'" (footnotes omitted)

            700 Exhibit Bra-613. See Panel Report, US – Upland Cotton, paras. 7.844-7.845 for a discussion of the data presented to the original panel.

            701 Exhibit Bra-613 (CCC Cost Formula). See also Exhibit Bra-614 (Budgets of the United States Government, Fiscal Year 1993-2006). The United States notes two minor omissions with respect to the 2005

            WT/DS267/RW

            Page 167


          27. The United States considers that Brazil's table is deficient because it does not include amounts recovered under guarantees issued prior to 1992 (the United States indicates that such recoveries, in 1993-2005, total over US$5.6 billion).702 We disagree. The Panel does not believe that it would be appropriate to include recoveries for pre-1992 guarantees unless all costs and all other revenues associated with the same guarantees are also included. In addition, the Panel considers that including these amounts would not allow a comparison with the re-estimates data submitted by the United States, which only covers guarantees issued after 1992.


          28. This "cash basis accounting" data leads to results which stand in sharp contrast to the "reestimates" data submitted by the United States. This is not surprising. The difference between the two methods primarily has to do with the fact that the data submitted by the United States includes not only amounts for a past period, but also projected disbursements and receipts. Brazil's cash basis formula however only captures receipts and disbursements to date. It does not, therefore, record projected receipts and disbursements, including (as the United States indicates) projected recoveries under rescheduled amounts.703


          29. The "credit guarantee liability" figure reported in the CCC's Financial Statements and the cash accounting basis data submitted by Brazil at the very least raise questions as to whether subsequent subsidy reestimates indeed establish that the programmes at issue in the original proceeding were not provided at a net cost to the US Government.


          30. In conclusion, notwithstanding the reestimates data submitted by the United States, we find it highly significant that the US Government continues, at the time of their issuance, to project that new GSM 102 export credit guarantees will, in the long-term, be provided at a net cost to the Government. In light of this, we consider that the initial subsidy estimates provide a strong indication that GSM 102 export credit guarantees are provided against premia which are inadequate to cover the long-term operating costs and losses of the GSM 102 programme.


          31. Brazil also adduces other evidence which further convinces us that the GSM 102 programmes fall within the scope of item (j) of the Illustrative List. We analyze this evidence below.


            1. Comparison with MPRs under the OECD Arrangement


              1. Main arguments of the parties


          32. Brazil asks us to take into consideration the minimum premium rates ("MPRs") provided in the OECD Arrangement on Officially Supported Export Credits (hereafter "the Arrangement").704



            image

            fiscal year: First, Brazil overlooked US$60 million in additional revenue in line 88.25 ("interest on univested funds"), second, Brazil (understandably) reproduced a printing error in the US budget: premiums collected in that year amounted to just above US$21 million; the balance of US$487 million for FY 2005 should have been printed in the recovery line 88.40 for the financing account. This does not affect the sub-total of revenues for the 2005 year. US Rebuttal Submission, footnote. 145.

            702 The United States refers to Exhibit US-68, budget line 88.40 of the "liquidating account" (relative to pre-1992 guarantees); Brazil only uses line 88.40, which pertains to the "financing account" (relative to post-

            1991 guarantees). See US Rebuttal Submission, para. 95: the US otherwise agrees, that with the two small exceptions noted in the preceding footnote, Brazil reports the correct figures.

            703 US Response to Panel Question 102.

            704 See Brazil's First Written Submission, paras. 438-443, Exhibits Bra-546 (OECD Arrangement –

            2005 revision) and Bra-547 OECD document TD/PG(2004)10/FINAL, 6 July 2004, Premium and Related Conditions: Explanation of the Premium Rules of the Arrangement on Officially Supported Export Credits (the Knaepen Package). The Panel notes that Brazil, in its Rebuttal Submission, includes its arguments with respect to the OECD MPRs in its discussion of the design, structure and operation of the GSM 102 programme. See Brazil's Rebuttal Submission, paras. 499-500.

            WT/DS267/RW

            Page 168


            Brazil argues that the MPRs represent the assessment by the Participants in the OECD Arrangement – including the United States – of the fees that prudently should be charged in order to meet long-term costs and losses. Brazil recognizes that the Arrangement – and therefore, the MPRs – does not apply to export credits in the agricultural field. Yet, Brazil asks us to take into consideration its comparison of GSM 102 fees with the MPRs as a "qualitative" demonstration that GSM 102 fees are well below the level at which they should be to be consistent with item (j) of the Illustrative List.705

          33. The United States rejects the relevance of Brazil's arguments based on the OECD MPRs.706 The United States submits that item (j) of the Illustrative List provides that the proper comparison is between the premium rates charged under the programme at issue and the long-term operating costs and losses of that programme. Item (j) of the Illustrative List does not refer to the Arrangement as the standard under which to assess whether a programme covers its long term operating costs and losses under item (j) of the Illustrative List.


            1. Main arguments of third parties


          34. Japan707 argues that Brazil's reliance on the OECD Arrangement in this proceeding is inappropriate. Japan notes that the Arrangement does not cover agricultural products and that the risks of financing arrangements for bulk agricultural commodity products, such as cotton, are not only substantial, but also unique. Moreover, the scope of the Arrangement is limited to export credits with terms of two years or more.


            1. Evaluation by the Panel


          35. It is clear to the Panel that the MPRs do not directly apply in determining whether an export credit guarantee programme falls within the scope of item (j) of the Illustrative List. While the MPRs were developed at least in part to ensure that Participants would comply with item (j), item (j) itself does not refer to the OECD Arrangement or to any other international agreement or benchmark in determining whether the premia are inadequate to cover the long term operating costs and losses of export credit guarantee programmes and other programmes covered by item (j) of the Illustrative List. As the United States points out, this contrasts with the second paragraph of item (k) of the Illustrative List, where such a reference is made. In consequence, there is in our view no basis to treat the OECD MPRs as providing a legally binding benchmark to determine whether an export credit guarantee programme falls within the scope of item (j) of the Illustrative List. In any event, GSM 102 fees are not subject to the Arrangement – and therefore to the MPRs – because the Arrangement does not apply to exports of agricultural commodities. Moreover, the Arrangement only applies to official support for terms of two years or more, while GSM 102 guarantees, which are provided for tenors up to three years, are also provided for tenors of less than two years.708


          36. That being said, in our view, while the MPRs have no legal status in the context of an analysis under item (j) of the Illustrative List, they may, from an evidentiary point of view, provide an indication as to whether GSM 102 fees are sufficient to cover the long-term operating costs and losses of the programme. Indeed, recognizing that the MPRs do not directly apply to the GSM 102 programme, Brazil asks us to take into consideration its comparison of GSM 102 fees with the MPRs as a "qualitative" demonstration that GSM 102 fees are well below the level at which they should be to be consistent with item (j) of the Illustrative List.709 We note that the OECD explains that the Arrangement's benchmark rates, including MPRs, are set "to ensure that Participants to the



            image

            705 Brazil's Response to Panel Question 43, para. 69.

            706 US Response to Panel Question 42.

            707 Japan's Third Party Submission, paras. 24-26.

            708 Arrangement, para. 5(c).

            709 Brazil's Response to Panel Question 43, para. 69.

            WT/DS267/RW

            Page 169


            Arrangement charge premium rates in addition to interest charges that ... are not inadequate to cover long-term operating costs and losses associated with the provision of export credits". Thus, the MPRs were developed, within the context of the OECD, with a view to ensure that Participants would respect the disciplines of item (j):


            "The objectives of the Knaepen Package, as reflected in Article 22 of the Arrangement on Officially Supported Export Credits (the "Arrangement"), are to ensure that Participants to the Arrangement charge premium rates in addition to interest charges that:

            1. cover the risk of non-repayment of export credits (i.e. credit risk); and

            2. are not inadequate to cover long-term operating costs and losses2 associated with the provision of export credits."


              image

              2 This obligation is in relation to the list of prohibited subsidies, i.e. Item (j) of Annex I of the WTO Agreement on Subsidies and Countervailing Measures (ASCM), which prohibits the provision by governments (or special institutions controlled by governments) of export credit guarantee or insurance programmes, of insurance or guarantee programmes against increases in the cost of exported products or of exchange risk programmes, at premium rates which are inadequate to cover the long-term operating costs and losses of the programmes. 710


          37. The MPRs may thus be regarded as representing an assessment, developed by and agreed upon by the export credit experts of the Participants to the Arrangement, of the premia levels that are necessary to ensure that export credit guarantee programmes cover their long-term operating costs and losses. We note, in this respect, that the United States itself, as a Member of the OECD and a Participant to the Arrangement, participated in the elaboration of the MPRs.


          38. We consider that, in this particular case, because of the magnitude of the difference between the MPRs and GSM 102 fees, the MPRs may provide an indication, on an informed basis, of the fact that GSM 102 fees are set at a level which is insufficient to cover the long term operating costs and losses of the programme. We emphasise that we take into account Brazil's comparison of GSM 102 fees with the MPRs because of the importance of the difference between them; we do not suggest that any difference in this respect could be relied upon as an indication that an export credit guarantee programme meets the criteria of item (j) of the Illustrative List.


          39. Using the formula for the calculation of MPRs contained in Annex VI to the Arrangement, Brazil finds that, for 1063 of 1064 comparison points, GSM 102 fees fall below the Arrangement's MPRs.711 On average, the MPRs are 106 per cent above GSM 102 fees.712


            image

            710 Exhibit Bra-547, Introduction, para. 1.

            711 Exhibit Bra-548 compares GSM 102 fees for transactions involving annual repayment of principal to MPRs, Exhibit Bra-549 performs the same comparison for GSM 102 fees involving semi-annual repayment

            of principal. The charts submitted by Brazil include both country and tenor-specific comparisons for any country (i) for which GSM 102 provides a risk rating of 1-6; and, (ii) for which any GSM 102 allocation has been made available in any year during the period FY 2001 through FY 2006 (even if those allocations were not used). Brazil uses the list of countries assigned a GSM 102 risk rating published on 28 July 2006. Confirmation of GSM 102 allocations is drawn from the fiscal year-end editions of "Monthly Summaries of Export Credit Guarantee Program Activity" See Brazil's First Written Submission, paras. 439-440. The sole "comparison point" for which GSM 102 fees are below the MPRs relates to China. Brazil notes that the CCC ranks China in category 6, while the OECD ranks China in category 2. Brazil also notes that Standard & Poor's ranks China as A/Stable/A1, making China's sovereign risk comparable to that of Botswana, Chile, Estonia, Lithuania and South Korea, all countries that are ranked by both the OECD and the CCC in risk category 2. See Brazil's First Written Submission, para. 441, footnote 615.

            712 Brazil's First Written Submission, para. 440 and Exhibits Bra-548 and Bra-549.

            WT/DS267/RW

            Page 170


          40. In addition, Brazil highlights the fact that, for longer tenors and higher country risk, GSM 102 fees are well below the OECD MPRs and that the difference between the two increases significantly as country risk increases and tenors lengthen.713


          41. The fact that GSM 102 fees are not subject to the Arrangement does not, in our view, fundamentally undermine the value of Brazil's comparison of GSM 102 fees with the MPRs as a qualitative indication that the fees are inadequate to cover the long term operating costs and losses of the programme.


          42. With respect to product coverage, Brazil submits that, if anything, the MPRs would increase if agricultural products were made subject to it. That is because, in Brazil's view, in an agricultural transaction, the value of the product sold holds little value to the guarantor (agricultural products are normally consumed shortly after sale or may not preserve). In contrast, industrial products are more durable and offer a better security to the guarantor, should he wish to obtain one.714 We agree. The fact that agricultural products are perishable in our view suggests that agricultural transactions, if anything, contain an added element of risk to the guarantor, as the products sold would hold little value on which security may be taken. As a consequence, there is in our view no reason why the MPRs should be lowered were agricultural products to fall under its product coverage.


          43. With respect to the fact that the Arrangement only applies to transactions of two years or more, we note Brazil's indication that the MPR formula allows the determination of fees for a shorter period of time, which has allowed Brazil to calculate the MPRs for tenors of less than two years. But in any case, we note that Brazil has included in its comparison the fees for tenors that would be covered by the Arrangement (tenors of 24, 30 and 36 months). Our conclusion above that GSM 102 fees are well below the MPRs holds true even with respect to only these three tenors, as does our conclusion that the difference between the two increases significantly as risk in the form of higher country risk and longer tenors increases.


          44. Thus, in our view, the factual differences between the product coverage and the coverage in term of tenors of GSM 102 guarantees and the OECD MPRs do not fundamentally undermine the value of the comparison performed by Brazil. We recall that we consider that the MPRs provide an indication – on an informed basis – that GSM 102 fees are insufficient to cover the long term costs and losses of the programme because of the significant difference between the GSM 102 fees and the MPRs. Any adjustment to take into consideration the differences pointed out by the United States would not, in our view, lead to a markedly different result.


            image

            713 See the following table submitted by Brazil in Exhibit Bra-547, p. 15 (the table concerns guarantees with annual repayment of principal, Brazil submits a similar table with respect to guarantees with semi-annual repayment of principal in Exhibit Bra-548):


            Average Percentage by which MPR Fees Exceed GSM-102 Rates (Annual Repayment of Principal)

            USDA Risk Level

            9

            Months

            12

            Months

            15

            Months

            18

            Months

            24

            Months

            30

            Months

            36

            Months

            All Tenors

            0

            19%

            23%

            22%

            23%

            27%

            27%

            30%

            24%

            1

            20%

            26%

            25%

            27%

            33%

            35%

            40%

            29%

            2

            24%

            32%

            31%

            34%

            43%

            46%

            54%

            38%

            3

            65%

            68%

            69%

            76%

            99%

            110%

            130%

            88%

            4

            79%

            93%

            95%

            104%

            135%

            150%

            177%

            119%

            5

            101%

            122%

            124%

            136%

            173%

            192%

            225%

            153%

            6

            138%

            175%

            179%

            194%

            240%

            264%

            306%

            216%

            All risk levels

            81%

            98%

            100%

            109%

            138%

            152%

            178%

            122%

            714 Brazil's First Written Submission, paras. 442-443.

            WT/DS267/RW

            Page 171


            1. Structure, design and operation of the GSM 102 programme


              1. Main arguments of the parties


          45. Brazil recalls that the original panel took into consideration elements of the structure, design and operation of the three programmes at issue in the original proceeding to conclude that they fell within the scope of item (j) of the Illustrative List. For Brazil, a programme that breaks even by chance would still not be structured or designed to do so in the long-term. Brazil considers that the GSM 102 programme is not designed to cover its long-term operating costs and losses. Brazil identifies five elements in this respect. First, the original panel found it relevant that the CCC borrows from the US Treasury and that access to US Treasury borrowings facilitate the functioning of the CCC programmes. This still holds true. Second, Brazil considers that the United States has not adopted a truly risk-based fee schedule. It recalls that the original panel relied on a USDA statement acknowledging that moving to a truly risk-based system would require raising fees to a level that "would most likely exceed" the one per cent fee cap imposed by US law on GSM 102 fees. The United States has not removed the cap. Third, although the CCC establishes individual risk ratings for each approved foreign obligor, those risk ratings do not affect the GSM 102 fee that is charged. Brazil considers that prudent fiscal management compels commercial banks to take varying borrower risk into account not only via exposure limits, but also through fees. Fourth, Brazil recalls that GSM 102 fees are well below the OECD MPRs. Fifth, the US Government continues to predict, year after year, that the programmes will be loss making – that, on one view of the data, the programmes have beaten the odds and met their costs and losses does not mean that they are structured and designed to do so.715


          46. The United States submits that the GSM 102 programme is structured and designed to cover its long term operating costs and losses.716 The United States notes that higher premia are now assessed in respect of obligors in higher-risk countries. Fees increase with each of the 6 eligible country risk categories, and all countries outside those 6 categories are wholly ineligible. The United States reclassified into an ineligible risk category a number of countries previously eligible under the programmes. Fees also increase with tenor. Also, the United States increased fees, on average, by 46 per cent across all programmes and 23 per cent, on a trade-weighted basis, for the GSM 102 programme alone over fiscal year 2004, the last year in which the prior fee schedule exclusively applied. In addition, the CCC ceased issuing guarantees under the GSM 103 and SCGP programmes.717 Of the three programmes originally examined, the GSM 102 programme exhibited the lowest risk profile. These changes made to implement the DSB recommendations and ruling bolster other disciplines already in place. The United States notes among those the need for foreign banks to be approved before a guarantee can be issued in respect of a transaction (this means in effect, that some countries that are listed as potentially eligible for GSM 102 export credit guarantees are de facto ineligible because no banking institution in the country is CCC-approved). Also, the CCC establishes internal bank limits to govern exposure to potential defaults by obligor banks in individual transactions.718 In any case, the budget data reflects the profitability of the programme over the last 14 years, even before the adoption of the recent changes in 2005. The profitability of the programmes

            is neither "against the odds", nor "avoiding losses by good fortune", as Brazil argues.719



            image

            715 For Brazil's arguments, see Brazil's First Written Submission, paras. 444-446; Brazil's Rebuttal Submission, paras. 487-512; Brazil's Opening Statement, paras. 240-255.

            716 See US First Written Submission, paras. 73-80; US Rebuttal Submission, paras 101-107, US Opening Statement, paras. 9.

            717 US First Written Submission, paras. 73-80. 718 US First Written Submission, paras. 77-78. 719 US Rebuttal Submission, paras. 101 ff.

            WT/DS267/RW

            Page 172


          47. The United States considers that Brazil's argument that the CCC is part of the US Government and can borrow from the US Treasury would lead to the result that a government agency could never satisfy the item (j) test.720 With respect to the one per cent statutory fee cap, the United States submits that the operating assumption of the USDA statement on which Brazil relies was that all of the then-eligible countries would remain eligible. Instead, in light of the fact that the fee cap has not been repealed, the CCC eliminated from eligibility 22 countries with respect to which participation in the programme might require fees in excess of the statutory maximum.721 As to foreign obligor risk, the United States submits that no WTO rule governs how a government programme shall ensure that its premia are adequate to cover its long-term operating costs and losses.


            1. Evaluation by the Panel


          48. This dispute concerns a new fee structure for which, by definition, no past performance data exists and therefore for which only projections of the long term net cost of the programme to the US Government are available. Given this, evidence concerning the "structure, operation and design" of the programme would seem to be particularly relevant. This is consistent with the approach of the original panel, which also took into consideration such prospective elements to determine that the programmes at issue in the original dispute were not designed to meet their long term costs and losses.722


          49. The original panel considered that:


            "In this dispute, in order objectively to assess premiums in relation to long-term operating costs and losses, we believe it is also appropriate for us to take into account aspects of the structure, design and operation of the measure before us. We are entitled to inquire whether the programme, including in terms of the premiums charged, was set up in such a way that the total of all premiums would be likely to cover the total of all operating costs and losses under the programme. For example, where a programme does not provide for premium rates that are fully reflective of the risks of a particular transaction, this might be one indicator that the programme was set up in such a way that its long-term operating costs and losses have to be borne, in total or in part, by the government."723 (footnote omitted)


          50. We therefore proceed to examine whether, as Brazil argues, there are elements of the structure, design and operation of the GSM 102 programme that indicate that it is not designed to cover its long term costs and losses.


          51. First, we recall that the original panel noted that the CCC has access to funds from the US Treasury and benefits from the full faith and credit of the US Government.724 This is still the case, and were the GSM 102 programme to incur massive losses, it would have access to additional funds from the US Treasury.


          52. Second, the original panel found that the fees charged by the CCC were not "risk based". This was one of the elements that led to the panel's conclusion that the programmes were not designed


            image

            720 US Rebuttal Submission, para. 104.

            721 US Rebuttal Submission, para. 106.

            722 The panel considered that "[i]n terms of the structure, design and operation of the export credit guarantee programmes before us, there are several elements which lead us to believe that the programmes are

            not designed to avoid a net cost to government." Panel Report, US – Upland Cotton, para. 7.857.

            723 Panel Report, US – Upland Cotton, para. 7.805.

            724 Panel Report, US – Upland Cotton, para. 7.858.

            WT/DS267/RW

            Page 173


            to cover their long term costs and losses.725 The original panel noted the following with respect to the fees not being affected by either country risk or foreign obligor risk:


            "... the premiums are not risk-based726, either with respect to country risk or the creditworthiness of the borrower in an individual transaction.727 All countries eligible for any of the CCC export credit guarantee programmes are categorized according to risk. CCC categorization of countries is based on a United States government internal risk classification system, which, according to the United States, 'is administratively controlled and may not be released outside of the US Government'.728 However, the United States has confirmed that "a country's risk classification has no impact on the premiums payable under the United States export credit guarantee programmes".729 Rather than the premiums, it is the United States government subsidy estimates and re-estimates for the export credit guarantee programmes that are determined in large part by the obligor's sovereign or non- sovereign country risk grade."730 731


          53. The panel also took into consideration the fact that the statute governing the CCC export credit guarantee programmes did not impose any limitation on the amount of guarantees that could be provided annually to any high risk country (in the aggregate or individually) or to high risks transactions. It concluded that the "CCC is, therefore, able to provide large amounts of guarantees to high-risk countries with a resulting high rate of default".732


          54. Brazil notes that GSM 102 fees still do not take into consideration foreign obligor risk at all.733 Although the CCC establishes individual risk ratings for each foreign obligor approved for participation in the GSM 102 programme, those risk ratings do not affect the GSM 102 fees. Brazil notes that none of the CCC-approved foreign obligors banks enjoy a credit rating superior to that of its sovereign.734 In addition, Brazil argues that prudent fiscal management compels commercial banks to take varying borrower risk into account not only via exposure limits, but also through fees, and that


            image

            725 See Panel Report, US – Upland Cotton, paras 7.861-7.866.

            726 [Original footnote:] The US Government acknowledges this. See Exhibit US-150 and United States' Response to Panel Question No. 86.

            727 [Original footnote:] e.g. United States' response to Panel Question No. 85. The United States has

            confirmed to us that under the SCGP "CCC does not determine the creditworthiness of participating importers[...]". Rather, according to the United States, the risk sharing between the CCC and the exporter is intended to ensure that due diligence is performed.

            728 [Original footnote:] United States' response to Panel Question No. 86. According to 7 USC 5622, the CCC shall not make credit guarantees available in connection with sales of agricultural commodities to any country that the Secretary determines cannot adequately service the debt associated with such sale.

            729 [Original footnote:] See e.g. Exhibit US-150 and United States' response to Panel Question No. 86.

            See also, "US General Accounting Office, Report to the Ranking Minority Member, Committee on Agriculture, Nutrition and Forestry, US Senate, Former Soviet Union, Creditworthiness of Successor States and US Export Credit Guarantees, GAO/GGD-95-60", February 1995, pp. 135-136, reproduced in Exhibit BRA-181: "Although GSM 102 recipient countries vary significantly from one another in terms of their risk of defaulting on GSM 102 loans, CCC does not adjust the fee that it charges for credit guarantees to take account of country risk. [...] CCC fees that included a risk-based component might not cover all of the country risk, but they could help to offset the cost of loan defaults." In Exhibit BRA-153, Audit Report of CCC for FY 2000, June 2001, USDA's Office of the Inspector General stated that fees for the GSM 102 and GSM 103 programmes had not been changed in 7 years, and as a result "may not be reflecting current costs." Without accounting for country risk in the fees charged, the CCC foregoes an opportunity to cover the cost of loan defaults.

            730 [Original footnote:] e.g. Exhibit US-150.

            731 Panel Report, US – Upland Cotton, para. 7.861.

            732 Panel Report, US – Upland Cotton, paras. 7.862-7.863.

            733 Brazil's Rebuttal Submission, paras. 496-498.

            734 Brazil's First Written Submission, para. 416.

            WT/DS267/RW

            Page 174


            the GSM 102 programme's failure to adjust its guarantee fees to take account of material risks is not cured by diversifying its portfolio of inaccurately-priced guarantees.735


          55. The United States argues that the new fee structure is risk-based. With respect to foreign obligor risk, the United States considers that the changes made to the GSM 102 programme bolster other disciplines already in place. The CCC conducts an independent analysis of each foreign bank in order to establish the risk rating of each foreign bank obligor. The United States argues that Brazil's assertion that because GSM 102 fees do not vary according to the particular foreign bank obligor, the CCC "ignores" foreign obligor risk is inaccurate and that it is similarly inaccurate to assert that the CCC assumes that foreign banks have a credit rating that is identical to that of the sovereign: to the contrary, the CCC establishes an exposure limit with respect to each bank. Once such limit is attained, no further guarantees of obligations of that bank are issued.736 The United States submits that a Member may choose to control exposure to defaults by foreign banks (and therefore, costs) by either establishing fees that vary with foreign obligor risk or, as the CCC does, by establishing internal bank limits.


          56. We see no reason to depart from the reasoning adopted by the original panel, and find that GSM 102 fees are not risk-based insofar as they do not take into account one of the main elements of risk associated with the guarantees, the risk of default that is specific to an individual foreign obligor. The impact of this is not insignificant – Brazil points out that over 90 per cent of GSM 102 guarantees are issued with respect to non-sovereign foreign obligors.737 The fact that the United States sets bank limits with respect to each foreign bank obligor does not, in our view, alter the fact that foreign obligor risk is not reflected in GSM 102 fees.


          57. Third, Brazil notes that the one per cent fee cap has not been repealed. Brazil refers the Panel to a statement of the USDA, in which the latter acknowledged that moving to a risk-based system would raise fees to a level that "would most likely exceed 1 per cent...".738 The statement was taken into consideration by the original panel. The one per cent fee cap applied to GSM 102 and SCGP fees.


          58. The United States refutes Brazil's argument by noting that the USDA statement in question was made before measures were taken to implement the DSB recommendations and rulings. Under the new fee structure, fees increase with country risk and with tenor (length of the loan), and all countries outside the eligible risk categories are wholly ineligible under the programme, a step aimed at reducing the risk exposure of the programmes.739 740 The United States also brings to our attention



            image

            735 Brazil's Rebuttal Submission, para. 498.

            736 US First Written Submission, para. 79.

            737 Brazil's First Written Submission, para. 416, Exhibits Bra-541 and Bra-542.

            738 Brazil's First Written Submission, para. 418, quoting Panel Report, US – Upland Cotton, para. 7.865

            and Exhibit Bra-543 (Annual Review of Fees for USDA Credit Programs, 25 March 2003, p. 4 (Exhibit US-150 in the original proceeding)).

            739 Fees continue to vary according to the duration of the underlying loan (30 days to 3 years) and the frequency of the repayment of principal (annual or semi-annual) as they did under the previous fee structure.

            740 The Panel notes that, in their comments on the interim report of the panel, both parties indicated that

            there are 6, rather than 7 eligible country risk categories under the GSM 102 programme. The parties have, in their submissions, been inconsistent in their designation of the number of eligible country risk categories. Compare, for instance, US First Written Submission, para. 74 to US Response to Panel Question 113; see also Brazil's First Written Submission, para. 334. The Panel conducts its analysis of GSM 102 fees on the basis of its understanding, supported by the evidence submitted by the parties, that countries in categories 0-6 are eligible under the GSM 102 programme.

            WT/DS267/RW

            Page 175


            the fact that a number of high-risk countries have been reclassified into an ineligible category (in addition, some countries are de facto ineligible as no bank in the country is approved).741


          59. Brazil argues in response that the changes the United States made had no significant impact on the profitability of the programme. First, Brazil argues that discontinuing applications for GSM 103 and SCGP could not have had a significant impact on the riskiness of the CCC's guarantees portfolio as GSM 102 export credit guarantees made up 93 per cent of the CCC guarantees portfolio;742 second, the fact that 19 countries are now ineligible does not significantly impact the riskiness of the portfolio. Brazil points out that, as of the implementation deadline, no GSM 102 guarantees were outstanding for 14 of those 19 countries,743 and that, for 4 of the 5 remaining countries, GSM 102 guarantees outstanding on the same date amounted to less than one half of one per cent of total outstanding GSM 102 guarantees. Of the countries that had GSM 102 guarantees outstanding, only Algeria accounted for outstanding guarantees of any significance.744 745


          60. We note that, of relevance to this issue, the United States has indicated that the fees were increased, on average, by 46 per cent for all export guarantee programmes, and by 23 per cent for GSM 102 export credit guarantees (on a weighted-average basis) compared to the last year under the prior fee schedule. Further, the United States explains that the GSM 102 programme was the most profitable of the three programmes examined by the original panel.


          61. The original panel did not quantify the margin by which the fees for the programmes at issue fell below the level that would have been sufficient to cover the long term operating costs and losses of the programmes. We are therefore unable to assess whether a 46 per cent average fee increase (23 per cent trade-weighted increase) is sufficient to ensure that the GSM 102 programme is not provided at a net cost to the US Government.746 We cannot rule out that changes made by the US Government to the programmes may affect the USDA's evaluation of whether the one per cent fee cap still remains a constraint on the adoption of risk-based fees. Yet, we consider that the evidence submitted by Brazil concerning outstanding guarantees with respect to countries that have now been made ineligible suggests that the factual context in which the USDA's statement was made has not been altered fundamentally. In this context, we regard the non-repeal of the one per cent fee cap as an indication that the fees charged for GSM 102 export credit guarantees are still not "risk-based".


        image

        741 US First Written Submission, paras. 75, 77.

        742 Brazil's Rebuttal Submission, para. 15, Exhibit Bra-586 (Commodity Credit Corporation, Guarantee Loan Program Summary for FAS as of 30 June 2005).

        743 See Brazil's Opening Statement, para. 171 and Exhibit Bra-586 (Commodity Credit Corporation, Guarantee Loan Program Summary for FAS as of 30 June 2005). See also Brazil's Rebuttal Submission,

        para 17.

        744 See Brazil's Opening Statement, para. 17. Exhibit Bra-586 indicates outstanding guarantees for Algeria totalling US$584 million over a total of US$7.9 billion in outstanding guarantees (7.38 per cent of that

        amount).

        745 The Panel notes that, of the countries that were ineligible on 1 July 2005, only Algeria (7.4 per cent) and Pakistan (3.9 per cent) had significant shares of GSM 102 guarantees outstanding on 1 July 2005; all others

        had less than 0.3 per cent of the total outstanding GSM 102 guarantees and together amounted to just over

        0.8 per cent of total outstanding guarantees. Exhibit Bra-586. Pakistan was later returned to eligibility. See US First Written Submission, footnote 15. The divergence in the number of countries referred to by the parties is explained by the fact that some countries became ineligible after 1 July 2005 and others were returned to eligibility after that date.

        746 We note however that Brazil submits that multiplying the figures in the "premiums collected" column in Exhibit Bra-613 by 1.46 per cent to reflect the average fee increase for all programmes would still

        result in a net loss of US$574 million (US$632 million if multiplied by 1.23 per cent to reflect the 23 per cent trade-weighted increase in GSM 102 fees calculated by the United States). Brazil's exercise of course assumes that the Panel would consider that the cash-basis accounting provides the relevant assessment of the long-term cost of the programme to the US Government in the 1992-2005 period. In any case our analysis only concerns GSM 102 fees, and not the fees applicable under the three programmes at issue in the original proceeding.

        WT/DS267/RW

        Page 176


          1. Our assessment of the effects of the remaining presence of the one per cent fee cap is confirmed by the evidence presented by Brazil with respect to the insufficient "scaling" of GSM 102 fees with the risks that it does take into consideration – country risk and tenor.


          2. Brazil argues that the one per cent fee cap has prevented the United States from implementing a fee structure that is truly risk-based: for Brazil, "where no fee cap stands in the way of taking relevant risks more fully into account, the upward slope of the fee curve varies dramatically from the insignificant scaling that can be observed with respect to GSM 102 fees".747 Brazil refers the Panel to the rate of increase, with risk, of fees charged by the ExIm Bank (the US export credit agency) for two of its products, the Letter of Credit Insurance (LCI) and the Medium-Term Export Credit Insurance (MTI).748


          3. The United States rejects Brazil's arguments in this respect.749 It considers that the fees are sufficiently scaled to take account of country risk, that the SCM Agreement imposes no standards in this respect, and that, in any case, the only issue under item (j) of the Illustrative List is whether the programme's "structure, design and operation" – including any scaling – when taken as a whole indicate that the premia are inadequate to cover the long-term costs and losses of the program".750


          4. In addition, the United States objects to Brazil's comparison of GSM 102 fees with those charged by the ExIm Bank. The United States notes that the MTI is not available for agricultural


            image

            747 Brazil's First Written Submission, paras. 419 and 446 and Brazil's Rebuttal Submission, para. 495:


            "...despite all the pomp attached by the United States to the adoption of 'a risk-based fee schedule,' calling something 'risk-based' does not make it truly reflective of the risks involved. Brazil recalls that the statutorily-imposed one-per cent fee cap imposed on GSM 102 does not permit the program to set fees that take account of all the risks faced by the CCC in issuing GSM 102 ECGs. USDA has acknowledged that moving to a truly risk-based system would require raising fees to a level that 'would most likely exceed 1 per cent ….' Despite the original panel's conclusion that the fee cap was significant in preventing the United States from respecting its WTO obligations, the United States has not removed the cap..."


            In the footnotes of this paragraph of its Rebuttal Submission, Brazil refers to its argument that the scaling of GSM 102 fees in terms of country risk and tenor is insufficient when compared to the rate of increase of ExIm Bank fees under the LCI and MTI. See also Brazil's Response to Panel Question 43, para. 73, where Brazil makes a similar argument in the context of its discussion of the OECD MPRs. We therefore examine the parties' arguments concerning Brazil's comparison of GSM 102 and ExIm Bank LCI and MTI fees in this section of our findings, but note that they were made primarily in the context of addressing Brazil's arguments that GSM 102 fees confer a benefit because they are provided for a below-market fee.

            748 Brazil compares GSM 102 fees to those of the LCI for tenors of up to 360 days and to those of the MTI for tenors exceeding 360 days. Unlike GSM 102, ExIm Bank has no single product with tenors from

            30 days to three years. See Brazil's First Written Submission, para. 384 and Exhibits Bra-530, pp. 54 and 59. The LCI provides protection against default on irrevocable letters of credit issued by foreign financial institutions. For agricultural commodities, cover is normally limited to 180 days but may be extended to 360 days on a case-by-case basis. 98 per cent of the principal and a portion of the interest are covered by the insurance for exports of agricultural commodities. The MTI covers transactions involving obligations from a foreign bank, and protects the insured against nonpayment due to the foreign bank’s failure to pay because of unforeseen commercial or political reasons. The MTI does not cover transactions for exports of agricultural products. It provides cover for 100 per cent of the transaction amount. Brazil indicates that MTI cover is available for repayment terms between 1 and 5 years, exceptionally 7 years. See Brazil's Comments on US Reponses to Panel Question 98 and Exhibit Bra-534.

            749 Again, the United States makes these specific arguments in the context of answering Brazil's arguments that GSM 102 export credit guarantees confer a benefit because they are provided for a below-market

            fee; but it is clear from the US Response to Panel Question 99 that the United States considers these arguments to be relevant to Brazil's item (j) arguments.

            750 US Response to Panel Question 99, para. 234.

            WT/DS267/RW

            Page 177


            products and that MTI fees are subject to the OECD Arrangement's MPRs, which cover transactions of two years or more, whereas hundreds of GSM 102 transactions totalling hundreds of millions of dollars are for terms of less than 2 years.751 The United States also identifies a number of other differences between GSM 102 guarantees and the ExIm Bank products, notably in terms of interest cover752 and principal coverage. Principal coverage is 100 per cent under the MTI, compared to 98 per cent for a GSM 102 guarantee. Further, the ExIm Bank allows, under the LCI product, the insured to arrange recourse or "pass back" to a third party of all or any part of any uninsured amount. The United States submits that this is not permitted under the GSM 102 programme:


            "Upon default the claimant must subrogate to CCC the claim to the entire amount in default, not just the guaranteed portion. CCC then has the right to recover from the obligor all moneys in default. Unlike the LCI product, the holder of the GSM-102 guarantee is therefore unable to sell the unguaranteed portion of the obligation to another party to insulate himself from any exposure to principal default. Under the MTI product, the holder is similarly insulated from any risk because of the full 100 percent coverage. Inferior coverage of principal relative to the Ex-Im products renders the GSM-102 guarantee less desirable, which is reflected in the price parties are willing to pay. Brazil has also not accounted for the difference in 'pass back' policies of the programs in its comparisons."753


          5. We agree with Brazil that a significant difference in the rate of increase of fees between ExIm Bank's products and the CCC's GSM 102 may indicate that the latter do not sufficiently increase with risk and, as a consequence, are not truly "risk-based". The fact that ExIm Bank's MTI is not available for agricultural products does not, in our view, fundamentally undermine the value of the comparison, nor does the fact that the financial products are not identical, provided that adjustments are made to render them comparable. Taking into account the concerns of the United States that the interest coverage of the products differs, Brazil estimates that GSM 102 fees should be adjusted by approximately 5 per cent to take this difference into account.754 We consider that Brazil's adjustment is appropriate.755 With respect to principal coverage, we note that Brazil was able to use a variable on


        image

        751 US Response to Panel Question 98, para. 223.

        752 Under ExIm Bank's MTI product, the obligor is required to use a form of promissory note promulgated by the Bank. The MTI covers "100 per cent of the rate provided in the note". See Exhibit Bra-

        535. Under the LCI product, "the insured interest rate is generally prime rate minus 0.5 per cent or the rate stated in the credit agreement, whichever is less". The "prime rate" refers to the prime rate published in the Wall Street Journal, under the table "Money Rates". See Exhibit Bra-531. In contrast, GSM 102 guarantees cover "eligible interest", which is the lesser of the interest rate specified between the US bank (or exporter) and the foreign bank or 55 per cent of the average investment rate of the most recent Treasury 26-week bill auction. Exhibit US-153. According to the United States, the interest cover of both LCI and MTI greatly exceeds that of the GSM 102 guarantee. The United States provides a comparison of the GSM 102 interest cover (55 per cent of the applicable 26-week T-bill rate with the LCI interest cover (Wall Street prime rate minus 0.5 per cent), and concludes that LCI interest cover was, on 1 March 2007, 491 basis points higher than CCC interest coverage (7.75 per cent vs. 2.839 per cent) (the difference varied between 374 and 499 basis points since 1 July 2005). The United States also notes that interest cover for the LCI is less than under the MTI). US Response to Panel Question 98, paras. 226-231 and Exhibit US-155.

        753 US Response to Panel Question 98, para. 232.

        754 See Brazil's Comments on the US Response to Panel Question 98, paras. 243-267.

        755 Brazil relies on the maximum possible difference in interest coverage between GSM 102 and ExIm Bank guarantees – i.e. as alleged by the United States, a difference of 5 per cent (7.8 per cent minus 2.8 per

        cent). Brazil however submits evidence which satisfies us that proceeding on this basis is conservative and favourable to the United States as the 7.8 per cent – which represents the LCI interest coverage cap, and which Brazil applies for both LCI and MTI fees – in fact overestimates the actual interest coverage in recent years. Brazil assumes that in case of default, the CCC would pay 100.8 per cent of the remaining principal (98 per cent towards principal and 2.8 per cent towards interest) and that the ExIm Bank would, in case of default, pay

        105.8 per cent of the remaining principal (98 per cent towards principal, and 7.8 per cent towards interest).

        WT/DS267/RW

        Page 178


        the ExIm Bank's website fee calculator to calculate MTI fees on the basis of a principal coverage of 98 per cent. With respect to the possibility to arrange recourse, or "pass-back" under the ExIm Bank LCI product, Brazil submits that the United States' contention that the holder of a GSM 102 export credit guarantee is not "permitted" to insure the remaining 2 per cent of principal not covered by the guarantee is misleading and incorrect insofar as "the United States implies that although a GSM 102 ECG only covers 98% of principal, in the event of default and payment by CCC to the holder of the ECG of 98% of principal, the CCC enjoys a claim on 100 per cent of any monies recovered from the foreign obligor".756 Brazil submits that, under section 1493.130(c) of the GSM 102 Regulations, the guarantee holder's claim to the 2 per cent of debt not covered by the guarantee ranks pari passu with the CCC's own claim to recover the 98 per cent when the CCC pursues recovery. In light of this fact, the Panel does not consider that the difference in principal coverage between GSM 102 export credit guarantees and the LCI would materially affect the fee comparison performed by Brazil.


          1. In any case, the Panel recalls that it is not comparing the fees charged, on the one hand, by the CCC for the issuance of GSM 102 guarantees and, on the other hand, by the ExIm Bank for the issuance of LCIs and MTIs. Rather, the Panel is considering the rate of increase (or scaling) of fees with increased risk, as an element to assess whether the one per cent fee cap constrains the adoption of a risk-based fee schedule. The fact that the financial products being compared are not identical (for instance, in terms of product coverage) does not fundamentally undermine the comparison of the "scaling" of their fees. 757


          2. Brazil has in our view convincingly demonstrated that GSM 102 fees only minimally respond to increased country risk and increased risk in the form of longer tenors, as compared to the fees charged by the ExIm Bank for its LCI and MTI products.758 The Panel notes, in particular, that while the fees for the lowest country risk categories are relatively similar between GSM 102 and the LCI


            image

            Brazil notes that this assumes that interest payments by the borrower on the loan are due annually – an assumption that is favourable to the United States, since if interest is due on a shorter interval, the dollar magnitude of the difference in interest coverage will decrease (less interest will be due at the time of default). See Brazil's Comments on the US Response to Panel Question 98, in particular footnote 409 thereof. The amended results of Brazil's fee comparison are reported in Exhibits Bra-693 and Bra-694.

            756 Brazil's Comments on US Response to Panel Question 98, para. 272.

            757 Further, we recall our finding above (see the Panel's discussion of Brazil's arguments concerning the

            OECD MPRs, supra, para. 14.101) that, if anything, fees should increase where risk in respect of agricultural products is covered. Likewise, the Panel does not regard differences in terms of the tenors of the financial products compared as material. Brazil has established that the "scaling" of GSM 102 fees is greatly inferior to that under either LCI or MTI fees.

            758 Tables in Figure 32, para. 400 of Brazil's First Written Submission. With respect to the rate of increase with country risk, see also the following table submitted by Brazil and which reports the average percentage by which ExIm Bank fees exceed GSM 102 fees:


            GSM 102 annual repayment of the principal

            GSM 102 semi-annual repayment of the principal

            Risk Level 6

            206%

            175%

            Risk Level 5

            145%

            125%

            Risk Level 4

            109%

            94%

            Risk Level 3

            76%

            67%

            Risk Level 2

            30%

            29%

            Risk Level 1

            28%

            27%

            Risk Level 0

            19%

            20%

            Table 31 of Brazil's First Written Submission, para. 396. The table does not adjust GSM 102 fees for the difference in interest coverage but, as indicated above, we do not consider that an adjustment of approximately 5 per cent would materially affect the comparison of the rates of increase of the fees.

            WT/DS267/RW

            Page 179


            and MTI, there is a sharp difference between the fees charged by the ExIm Bank and the CCC for the highest risk categories.759


          3. The significant difference between the rates of increase for, on the one hand, GSM 102 fees and, on the other hand, the ExIm Bank's LCI and MTI products suggests that the 1 per cent fee cap imposed by US law on GSM 102 fees does indeed prevent the adoption of a truly risk-based fee structure by the CCC.760


          4. Further, we note that, in all cases, GSM 102 fees vary by, at most, 100 per cent between the lowest and highest country risk category, for identical tenors and repayment frequencies.761 In our view, an increase in fees of, at most, 100 per cent suggests that fees insufficiently take account of this factor. We recall that the countries placed in risk category 0 will by definition present a very low (even virtually inexistent) risk of default and that countries in the highest risk categories present a high risk of default. Finally, we recall our finding above concerning the increasing difference between GSM 102 fees and the OECD MPRs as risk increases.762 These additional elements lend further support to our conclusion that the rate of increase of GSM 102 fees is insufficient to meaningfully account for increased risks in the form of higher country risk and longer tenors and that the one per cent fee cap prevents the adoption, by the CCC, of risk-based fees for the GSM 102 programme.


            image

            759 For instance, according to Exhibit Bra-548, the respective fees for GSM 102 and for ExIm Bank's LCI (with annual repayment of principal) for a country falling within each agency's risk category 6 are, for a tenor of 12 months: respectively, 0.612 per cent (GSM 102) and 2.410 per cent (ExIm Bank LCI); for a tenor of 36 months: 1.0 per cent (GSM 102), 4.325 per cent (ExIm Bank MTI). In contrast, the fees for a country falling under each agency's risk category 2 are: 12 months tenor: 0.373 per cent (GSM 102), 0.570 per cent (ExIm Bank LCI); for 36 months: 0.637 per cent (GSM 102), 1.060 per cent (ExIm Bank MTI). The Panel notes that the ExIm Bank and the CCC's risk categories do not perfectly correspond. See Brazil's First Written Submission, footnote 508:

            "GSM 102 risk categories correspond with ExIm Bank risk categories as follows:

            • Countries rated by CCC in GSM 102 risk category 6 have an average ExIm Bank risk category of 5.64;

            • Countries rated by CCC in GSM 102 risk category 5 have an average ExIm Bank risk category of 4.78;

            • Countries rated by CCC in GSM 102 risk category 4 have an average ExIm Bank risk category of 4.13;

            • Countries rated by CCC in GSM 102 risk category 3 have an average ExIm Bank risk category of 3.33;

            • Countries rated by CCC in GSM 102 risk category 2 have an average ExIm Bank risk category of 2.00;

            • Countries rated by CCC in GSM 102 risk category 1 have an average ExIm Bank risk category of 1.83;

            • Countries rated by CCC in GSM 102 risk category 0 have an average ExIm Bank risk category of 1.33.

        Thus, GSM 102 fees for GSM 102 risk categories 2, 3, 4, 5 and 6 are compared, respectively,

        to ExIm Bank fees for ExIm Bank risk categories 2, 3, 4, 5 and 6. GSM 102 fees for GSM 102 risk categories 0 and 1 are compared, respectively, to ExIm Bank fees for ExIm Bank risk categories 1 and 2."

        760 Again, applying an adjustment of approximately 5 per cent to account for the differences between the two products would not, in our view, affect meaningfully the overwhelming difference in the manner in

        which the increased riskiness of guarantees associated with higher country risk and longer tenors is taken into consideration by LCI/MTI and GSM 102 fees.

        761 US Response to Panel Question 99, para. 235: "under the current fee schedule, the difference in premia for the same tenor between the least risky countries and the most risky countries is in all cases between

        81.5 per cent and 100.6 per cent"; see also the GSM 102 Fee Rate Schedule, Exhibit Bra-505.

        762 See supra, para. 14.99.

        WT/DS267/RW

        Page 180


          1. Brazil also relies, in support of its arguments that the GSM 102 programme is not designed to cover its long term operating costs and losses, on the fact that the CCC continues to project that the programme will be run at a net cost to the US Government in the long term and on the fact that GSM 102 fees are well below the OECD MPRs. The Panel notes that it has already found these elements to provide an indication that GSM 102 export credit guarantees are provided at premium rates which are inadequate to cover the long-term operating costs and losses of the GSM 102 programme.


          2. In light of these considerations – the CCC's access to funds from the US Treasury, which facilitates the functioning of the programme, the fact that GSM 102 fees do not vary with foreign obligor risk, the fact that the one per cent fee cap has not been repealed and in our view prevents the adoption of risk-based fees (notably due to the insufficient "scaling" of GSM 102 fees) – we conclude that the GSM 102 programme is not designed to cover its long term operating costs and losses.


            1. Overall conclusion of the Panel's analysis under item (j) of the Illustrative List


          3. We recall that we have conducted our analysis under the general rules of the burden of proof,

            i.e. we have assumed that it was for Brazil to establish that GSM 102 export credit guarantees constitute export subsidies because they fall within the terms of item (j) of the Illustrative List.763


          4. We find that Brazil has met its burden of proof in this respect and that it has established that the GSM 102 export credit guarantee programme constitutes an "export subsidy" because it is provided against premiums which are inadequate to cover its long term operating costs and losses under the terms of item (j) of the Illustrative List. We reach this conclusion on the basis of the totality of the evidence submitted by Brazil in this respect. We recall that we have taken into consideration the following elements:


            1. The US Government continues to project that new GSM 102 export credit guarantees issued under the new fee schedule will be provided at a net cost to the Government;


            2. GSM 102 fees are well below the OECD MPRs, which we consider to provide an informed appraisal of the level of fees necessary to cover the long-term operating costs and losses of an export credit guarantee programme;


            3. Elements of the structure, design and operation of the GSM 102 programme indicate that the programme is not designed to cover the long term operating costs and losses of that programme.


          5. On the basis of this evidence, taken as a whole, we find that Brazil has established that GSM 102 export credit guarantees constitute per se export subsidies within the meaning of Article 3.1(a) of the SCM Agreement because they are provided as part of a programme that falls within the scope of item (j) of the Illustrative List. Relying on the definition of "export subsidy" in Article 3.1(a) of the SCM Agreement as context, we conclude that GSM 102 export credit guarantees are "export subsidies" within the meaning of Article 10.1 of the Agreement on Agriculture.


          6. We recall that we consider that the two legal bases relied upon by Brazil in support of its argument that GSM 102 export credit guarantees are "export subsidies" under the terms of Article 3.1(a) of the SCM Agreement are not separate claims, but mere separate arguments in support of that contention. We do not consider it necessary to examine whether, in addition to constituting export subsidies because they fall within the scope of item (j) of the Illustrative List, these guarantees


            image

            763 See supra, para. 14.47.

            WT/DS267/RW

            Page 181


            also meet the definitional elements of an export subsidy under Articles 1.1 and 3.1(a) of the SCM Agreement. In any case, we consider that our findings with respect to item (j) of the Illustrative List suffice to resolve the matter between the parties.764


            1. Whether the United States has applied GSM 102 guarantees in a manner that results in circumvention of its export subsidy commitments under the Agreement on Agriculture


              1. Introduction


          7. Having concluded that GSM 102 export credit guarantees provide export subsidies under the terms of Article 10.1 of the Agreement on Agriculture, we must now determine whether the United States has, subsequent to 1 July 2005, applied such export subsidies in a manner which has resulted in circumvention of its export subsidy commitments.765


          8. The elements necessary to establish that the export subsidies resulted in the circumvention of the United States' export subsidy commitments vary depending on whether scheduled or unscheduled products are at issue. For unscheduled products, "circumvention" of the United States' export subsidy commitments will occur if any export subsidies (in the form of GSM 102 export credit guarantees) are provided in respect of any quantity of exports of the product in question.766 For scheduled products, "circumvention" will occur if the United States provides export subsidies to volumes of exports of the product at issue in excess of its "quantity" reduction commitments or of its "budgetary outlay" reduction commitments.767


          9. Brazil submits evidence concerning the quantity of exports of certain products that benefited from export subsidies in the form of GSM 102 export credit guarantees since 1 July 2005. We examine whether this evidence suffices to establish Brazil's claims of circumvention below, first with respect to unscheduled products, and in a second time, with respect to scheduled products.


            1. Unscheduled products


          10. Brazil submits evidence to the effect that the CCC has, subsequent to 1 July 2005, provided GSM 102 guarantees to support the export of the following unscheduled products:768


            1. between 1 July and 30 September 2005:769 cotton, oilseeds (including soybeans/soybean meal), protein meals, fresh vegetables, hides/skins and tallow.770


              image

              764 We note that in the original proceeding the Appellate Body rejected a claim by Brazil that the original panel had improperly exercised judicial economy in refusing to make additional factual findings under Articles 1 and 3 of the SCM Agreement. See Appellate Body Report, US – Upland Cotton, paras. 726-733.

              765 Given our finding that Brazil has established that GSM 102 export credit guarantees constitute export subsidies under Article 10.1 of the Agreement on Agriculture, we need not conduct the analysis that

              would normally follow from the partial reversal of the burden of proof under Article 10.3 of the Agreement on Agriculture. See supra, para. 14.47.

              766 See Appellate Body Report, US – FSC, para. 150. See also the ending clause of Article 3.3 of the

              Agreement on Agriculture: "... a Member shall not provide export subsidies ... in respect of any agricultural product not specified in [Section II of Part IV] of its Schedule."

              767 This follows from the fact that a Member's export subsidy reduction commitments include both a

              volume ("quantity") and a financial ("budgetary outlay") component. See Appellate Body Report, EC – Export Subsidies on Sugar, para. 200 and the text of Article 3.3, which provides that "... a Member shall not provide export subsidies ... in respect of the agricultural products or groups of products specified in Section II of Part IV of its Schedule in excess of the budgetary outlay and quantity commitment levels specified therein."

              768 See Brazil's First Written Submission, paras. 450 ff.

              WT/DS267/RW

              Page 182


            2. in FY 2006 (1 October 2005-30 September 2006): cotton, oilseeds, soybeans/soybean meal, protein meals, hides/skins, tallow and corn products.771


          11. Given that the United has made no specific export subsidy reduction commitments for these products, the United States has committed itself to providing no export subsidies on the export of these products (Article 3.3. of the Agreement on Agriculture). Thus, we find, on the basis of the evidence submitted by Brazil, that the United States has after 1 July 2005, applied export subsidies in the form of GSM 102 export credit guarantees in a manner which resulted in circumvention of its export subsidy commitments under the Agreement on Agriculture.


            1. Scheduled products


          12. With respect to scheduled products, Brazil's claims under Article 10.1 of the Agreement on Agriculture concern exports of three products – rice, pig meat and poultry meat. The subsidy reduction commitments of the United States for these products are as follows:


            Agricultural product

            Final Commitments

            Outlays (US$)

            Quantity (MT)

            Rice

            2,368,925

            38,544

            Pig meat

            497,055

            395

            Poultry meat

            14,554,827

            27,994

            (Source: United States Schedule XX, Part IV, Section II)


          13. We recall that we have rejected the US arguments that Brazil's claims in respect of pig meat and poultry meat are not properly before us.772


          14. Brazil determines the quantity of export of each of these three products as follows: it takes the value of GSM 102 exporter applications received for each product in the period concerned (July- September 2005 and FY 2006). It then divides that amount by the average US export price for the commodity in question for FY 2006 (US$/metric ton)773 and thereby obtains the volume, in metric tons, of exports of the product in question in the period concerned.



            image

            769 I.e. the portion of FY 2005 running from 1 July 2005 – GSM 102 export credit guarantee activity is tracked on the basis of a fiscal year running from 1 October to 30 September (FY 2005 ran from 1 October 2004 to 30 September 2005; FY 2006 ran from 1 October 2005 to 30 September 2006).

            770 Compare the summaries of export credit activity as of July 2005 (Exhibit Bra-551) and for the whole of FY 2005 (Exhibit Bra-511). We note that Brazil also submitted evidence to the effect that the

            United States also provided GSM 102 export credit guarantees to exports of lysine and wood products in the same period. We address Brazil's claim under the SCM Agreement in respect of these products below in Section XIV.D.

            771 Exhibit Bra-513. Exports of lyocell and wood products (wood pulp) also benefited from GSM 102 export credit guarantees in the same period. We address Brazil's claims under the SCM Agreement in respect of

            these products below in Section XIV.D.

            772 See supra, section IX.B.2.

            773 Brazil obtains the US export price from the following sources: FY 2006, rice: October to September average of monthly prices of US long grain rough rice, bulk, FOB, New Orleans from the 13 October 2006

            ERS/USDA "Rice Outlook." See Exhibit Bra-552. July-September 2005, rice: average of July to September 2005 monthly prices of US long grain rough rice, bulk, FOB, New Orleans from the

            WT/DS267/RW

            Page 183


            Agricultural product

            Quantity commitment

            Exports under GSM 102 (July-

            September 2005) (MT)

            Exports under GSM 102 (FY

            2006) (MT)

            Rice

            38,544

            54,430774

            197,030775

            Pig meat

            395

            -

            22,562776

            Poultry meat

            27,994

            86,572777

            381,906778


          15. The United States does not challenge Brazil's calculations, nor the manner in which Brazil estimates the relevant price for each agricultural product. We would, on the basis of the evidence provided by Brazil, be ready to conclude that Brazil has made a prima facie case that the United States has applied export subsidies in the form of GSM 102 in a manner which has resulted in the circumvention of the US export subsidy reduction commitments for these three products.


          16. Before doing so, however, we address one issue which has not been discussed by the parties in this dispute, but which the Panel must nevertheless resolve. This issue arises from the fact that the US Schedule of commitments indicates, with respect to the quantitative aspect of the US commitments, that "[a]ll quantity commitments are for the period 1 July 2002 through 30 June 2003 while budgetary outlay commitments are implemented on a fiscal year basis beginning 1 October 2002".779


          17. Thus, while the evidence submitted by Brazil in support of its claims concerns the periods from 1 July to 30 September 2005 (the part of FY 2005 subsequent to the end of the implementation period) and from 1 October 2005 to 30 September 2006 (FY2006), the United States' subsidy commitments, expressed in quantity terms, run from the period 1 July to 30 June of each year.780


          18. The evidence presented by Brazil concerning the period from 1 July to 30 September 2005 does not, in our view, raise any specific difficulty: the totality of the period covered by the evidence presented by Brazil falls within only one of the US "quantity commitment year" (Brazil's evidence


            image

            13 October 2006 ERS/USDA "Rice Outlook." Exhibit Bra-552. For pig meat, FY 2006: average of October 2005 to August 2006 monthly unit values (total value/total quantity) of US pigmeat (HS 0203) from the "USITC Interactive Tariff and Trade DataWeb). Exhibit Bra-553. For poultry meat, July-September 2005: average of October 2004 to September 2005 monthly unit values (total value/total quantity) of US poultry meat (HS 0207), FOB, from the "USITC Interactive Tariff and Trade DataWeb". Exhibit Bra-554. FY 2006, poultry meat: average of October 2005 to August 2006 monthly unit values (total value/total quantity) of US poultry meat (HS 0207), FOB, from the "USITC Interactive Tariff and Trade DataWeb". Exhibit Bra-554.

            774 Exhibit Bra-552.

            775 Exhibit Bra-552.

            776 Exhibit Bra-553.

            777 Exhibit Bra-554.

            778 Exhibit Bra-554.

            779 United States Schedule XX, note 1.

            780 A somewhat similar issue arose in the original proceedings: before the original panel, Brazil made claims with respect to the period 30 June-1 July 2002 (the period corresponding to the United States' quantity

            reduction commitments), whereas US export credit guarantees activity was reported on the basis of the period running from 1 October to 30 September. See Panel Report, US – Upland Cotton, para. 7.878. The Panel did not address the issue of the partial overlap of the time periods, but the Appellate Body commented on the issue and indicated that the Panel should have sought to address it. See Appellate Body Report, US – Upland Cotton, para. 691.

            WT/DS267/RW

            Page 184


            covers the first three months of the US "quantity commitment year" running from 1 July 2005 to 30 June 2006)". But the evidence presented by Brazil with respect to FY 2006 (i.e. 1 October 2005 to 30 September 2006) only partially overlaps with the US "quantity commitment year", which runs from 1 July 2005 to 30 June 2006.781 Explained differently, US exports of scheduled products during the period running from 1 July 2006 to 1 October 2006 should be computed as part of a different year for the purpose of determining whether the United States has respected its export subsidy reduction commitments (expressed in quantity terms).782


          19. That being said, in the present case, we are satisfied that Brazil has, at the very least, made a prima facie case that the United States has circumvented its quantitative export subsidy commitments. This is because, for all three products, the quantity of exports having benefited from GSM 102 export credit guarantees exceeds the US quantity commitments by a large margin. The quantity of exports exceeds the United States' quantitative commitments by a factor of 5 in the case of rice, 57 in the case of pig meat and 14 in the case of poultry meat.783 We could only conclude that Brazil has not proved (or at least made a prima facie case) that the United States has exceeded its quantity commitments if it were assumed, in the "best" case (rice), that 80 per cent of the exported quantity having benefited from subsidies was exported in the last three months of the year (from 1 July to 1 October 2006). Such an assumption would not be reasonable, nor has the United States presented any arguments that Brazil's demonstration with respect to "actual circumvention" under the Agreement on Agriculture is incorrect. We therefore conclude that Brazil has – at the very least – made a prima facie case that the United States exceeded its export subsidy reduction commitments for each of rice, poultry meat, and pig meat, during FY 2006.


          20. As a result, the Panel finds that the United States has applied export subsidies in the form of GSM 102 export credit guarantees in a manner which has resulted in the circumvention of the US export subsidy reduction commitments for rice, poultry meat and pig meat during FY 2006 (1 October 2005 to 30 September 2006), and for rice and poultry meat in the period from 1 July to 30 September 2005.


          21. The Panel also finds, as a consequence, that by acting inconsistently with Articles 10.1 and 8 of the Agreement on Agriculture the United States has failed to comply with the DSB recommendations and rulings. Specifically, the United States has failed to bring its measures into conformity with the Agreement on Agriculture.



            image

            781 We note in passing that the two periods for which Brazil submits evidence fall within the same US "quantity commitment year".

            782 We also note that the issue does not arise with respect to unscheduled products: this is because Article 3.3 of the Agreement on Agriculture prohibits the United States from granting any subsidy whatsoever to

            exports of unscheduled products. Even if the United States' compliance with its obligation not to provide any export subsidies to these products was computed on the basis of the "period" for which it makes commitments in its Schedule, there would be "circumvention." This is because the period within which the United States' "budgetary outlay commitments" are to be assessed matches the period for which Brazil provides evidence (in both cases, the "year" runs from 1 October to 30 September). Because any quantity of exports benefiting from export subsidies necessarily corresponds to a budgetary outlay, no additional enquiry would be necessary to conclude that the United States has "circumvented" its export subsidy commitments with respect to unscheduled products.

            783 Expressed differently, the quantities of exports having benefited from exports subsidies exceed the

            United States' commitments by, respectively, 511 per cent, 5712 per cent, and 1379 per cent (compare the second and fourth columns of the table in para. 14.143, supra).

            WT/DS267/RW

            Page 185


            1. Claims that GSM 102 export credit guarantees are inconsistent with Articles 3.1(a) and

              3.2 of the SCM Agreement


          22. Brazil claims that to the extent that export subsidies in the form of GSM 102 export credit guarantees are provided in a manner inconsistent with the Agreement on Agriculture, they are not shielded from challenge under Articles 3.1(a) and 3.2 of the SCM Agreement.784 Brazil notes that the opening clause of Article 3.1 of the SCM Agreement provides that "subsidies contingent... upon export performance" are prohibited, "[e]xcept as provided in the Agreement on Agriculture...," meaning that Members are prohibited from granting or maintaining export subsidies to products covered by the Agreement on Agriculture except where the Agreement on Agriculture allows such subsidies. Brazil argues that for unscheduled products, the United States is not due the "safe harbor" afforded by the opening clause of Article 3.1 of the SCM Agreement at all, and that as a result, the prohibition on export subsidies in Articles 3.1(a) and 3.2 applies fully to the "GSM 102 export subsidy". With respect to rice, pig meat and poultry meat, Brazil argues that the United States' entitlement to the "safe harbor" is limited and that Articles 3.1(a) and 3.2 of the SCM Agreement apply to the extent that the quantities of products having benefited from GSM 102 export subsidies exceed the United States' quantitative export subsidy reduction commitments for these three products.


          23. The United States has not specifically responded to Brazil's arguments in this respect.


          24. Article 3.1 of the SCM Agreement reads as follows:


            "Except as provided in the Agreement on Agriculture, the following subsidies, within the meaning of Article 1, shall be prohibited:


            1. subsidies contingent, in law or in fact, whether solely or as one of several other conditions, upon export performance, including those illustrated in Annex I;" (footnotes omitted; emphasis added)


          25. There is, in our view, no question that, to the extent that the GSM 102 export credit guarantees do not conform fully to the provisions of the Agreement on Agriculture, they are subject to the prohibition on export subsidies under Articles 3.1(a) of the SCM Agreement.785 We have already concluded that GSM 102 export credit guarantees confer export subsidies under Article 10.1 of the Agreement on Agriculture, relying on item (j) of the Illustrative List as "context" to reach that conclusion. We see no reason to reach a different conclusion when considering whether GSM 102 export credit guarantees constitute export subsidies under the terms of Article 3.1(a) in itself, and not as "context" in an analysis under Article 10.1 of the Agreement on Agriculture.


          26. As indicated in our findings in paras. 14.140 and 14.149, above, the United States has, since 1 July 2005, provided export subsidies with respect to unscheduled products (between 1 July and 30 September 2005, on exports of cotton, oilseeds – including soybeans/soybean meal, protein meals, fresh vegetables, hides/skins and tallow and in FY 2006, to exports of cotton, oilseeds, soybeans/soybean meal, protein meals, hides/skins, tallow and corn products) and has provided export subsidies in excess of its export subsidy reduction commitments to exports of scheduled products – rice, poultry meat and pig meat during FY 2006, and rice and poultry meat in the period from 1 July to 30 September 2005.



            image

            784 Brazil's First Written Submission, paras. 457-460.

            785 We need not, in this case, address the question of whether, now that Article 13 of the Agreement on Agriculture (the Peace Clause) has lapsed, there may be a violation of Articles 3.1(a) and 3.2 of the SCM

            Agreement in respect of all exports (i.e. even those that conform to the disciplines of the Agreement on Agriculture). Brazil makes no claims to that effect.

            WT/DS267/RW

            Page 186


          27. The Panel therefore finds that the United States has acted inconsistently with Articles 3.1(a) and 3.2 of the SCM Agreement by providing export subsidies to unscheduled products since 1 July 2005 and by providing export subsidies to scheduled products in excess of its commitments under the Agreement of Agriculture since 1 July 2005.


          28. The Panel also finds, as a consequence, that by acting inconsistently with Articles 3.1(a) and

        3.2 of the SCM Agreement the United States has failed to comply with the DSB recommendations and rulings. Specifically, the United States has failed to comply with the DSB recommendation, adopted pursuant to Article 4.7 of the SCM Agreement, to "withdraw the prohibited subsidies" identified in paragraph 8.1(d)(i) of the report of the original panel.


    4. CLAIMS WITH RESPECT TO LYSINE, LYOCELL AND WOOD PRODUCTS


      1. As indicated above, Brazil initially included three products – lysine, lyocell and wood products – in the list of products concerned by its claims under Article 10.1 and 8 of the Agreement on Agriculture, and to the extent of the inconsistency with the Agreement on Agriculture, under Articles 3.1(a) and 3.2 of the SCM Agreement. After the United States pointed out that the products are not covered by the scope of the Agreement on Agriculture, Brazil indicated that it is not pursuing its claims under the Agreement on Agriculture with respect to these products.786 We therefore need not decide whether these three products in fact fall outside the scope of the Agreement on Agriculture.


      2. The United States contends that Brazil's claims under Articles 3.1 and 3.2 of the SCM Agreement are not properly before us. The United States submits that in its request for the establishment of a panel, Brazil presented claims under Articles 3.1 and 3.2 of the SCM Agreement "as a result and to the extent of the violation of Article 10.1 (and, as a consequence, Article 8) of the Agreement on Agriculture".787 We therefore need to decide whether Brazil's claims concerning these three products under the SCM Agreement are properly before us.


        image

        786 See Panel Question 50 and Brazil's Response thereon:


        "Does Brazil maintain its claims with respect to the three unscheduled products (lyocell, lysine, wood products) identified by the United States as falling outside the scope of the Agreement on Agriculture? (see paragraph 83 of the United States' Rebuttal)

        Brazil maintains its claims under Articles 3.1(a) and 3.2 of the SCM Agreement with respect to GSM 102 ECGs for lyocell, lysine and wood products. To the extent that these products fall outside the scope of the Agreement on Agriculture, the compliance Panel need not, before addressing Brazil's prohibited subsidy claims under the SCM Agreement, address whether the provision of GSM 102 ECGs for these products circumvents US agricultural export subsidy commitments."

        787 US Comments on Brazil's Response to Panel Question 50:

        "In its response, Brazil appears to acknowledge that lyocell, lysine, and wood products fall outside the scope of the Agreement on Agriculture. By definition, therefore, the provision of GSM 102 export credit guarantees in respect of lyocell, lysine, and wood products cannot circumvent US agricultural export subsidy commitments under that agreement. In its request for establishment of a panel, Brazil presented claims under Articles 3.1 and 3.2 of the SCM Agreement "as a result and to the extent of [the alleged] violation of Article 10.1 (and, as a consequence, Article 8) of the Agreement on Agriculture." As there can be no "violation of Article 10.1 (and, as a consequence, Article 8) of the Agreement on Agriculture," there can be no claim under the SCM Agreement with respect to the GSM 102 export credit guarantees in respect of lyocell, lysine, and wood products under the express terms of Brazil's request for panel establishment. Where a claim is not set out in the request for panel establishment – as here – it is not part of the matter referred to the panel and falls outside the scope of a panel proceeding." (footnotes omitted)

        WT/DS267/RW

        Page 187


      3. The relevant parts of Brazil's request for the establishment of a panel read as follows: "(b) Inconsistency of measures taken to comply with the covered agreements

          1. Brazil believes that the United States provides export subsidies (under Articles 1, 3.1(a) and 3.2 of the SCM Agreement, as well as under item (j) to the Illustrative List of Export subsidies included as Annex I to the SCM Agreement (the "Illustrative List")) through the GSM 102 and SCGP programs, and ECGs provided under these programs, and has applied those export subsidies subsequent to 1 July 2005 in a manner that results in circumvention of the United States' export subsidy commitments. In this respect, the United States' measures taken to comply are not consistent with Articles 10.1 and 8 of the Agreement on Agriculture and Articles 3.1(a) and 3.2 of the SCM Agreement.


          2. Specifically, and notwithstanding the US measures taken to comply cited as items (ii) and (iii) in Section B above, the GSM 102 and SCGP programs, and ECGs provided thereunder, are export subsidies under the Agreement on Agriculture and the SCM Agreement, based on the following two grounds:


            • The programs and ECGs provided thereunder provide and constitute financial contributions (loan guarantees) that confer benefits on recipients, within the meaning of Article 1 of the SCM Agreement, and that are contingent upon export performance, within the meaning of Article 3.1(a) of the SCM Agreement; and, separately and independently,


            • The programs do not levy premium rates adequate to cover the long- term operating costs and losses of the GSM 102 and SCGP programs, within the meaning of item (j) of the Illustrative List.


          3. Further, subsequent to 1 July 2005, ECGs under the GSM 102 and SCGP programs have been applied to circumvent US export subsidy commitments, within the meaning of Article 10.1 of the Agreement on Agriculture. Brazil is concerned that ECGs under the GSM 102 and SCGP programs have been provided subsequent to 1 July 2005 to support the export of: upland cotton and other unscheduled products; and, rice, pigmeat and poultry meat, in excess of US reduction commitment levels for those products.788


          4. As a result and to the extent of this violation of Article 10.1 (and, as a consequence, Article 8) of the Agreement on Agriculture, the GSM 102 and SCGP programs, and ECGs provided thereunder, are subject to the prohibition against export subsidies included in Articles 3.1 and 3.2 of the SCM Agreement. On the grounds cited in paragraph 26 above, Brazil believes that the GSM 102 and SCGP programs, and ECGs provided thereunder, are prohibited export subsidies, within the meaning of Articles 3.1(a) and 3.2 of the SCM Agreement."789



        image

        788 Exports that benefited from GSM 102 and SCGP ECGs subsequent to 1 July 2005 are currently reflected in the July 2005 and September 2005 (fiscal year end), as well as July 2006 "Monthly Summary of Export Credit Guarantee Activity," USDA FAS Online, available at http://www.fas.usda.gov/excredits/Monthly/ecg.html.

        789 WT/DS267/30, paras. 25-28.

        WT/DS267/RW

        Page 188


      4. The Panel agrees with the United States that Brazil's claims under Articles 3.1(a) and 3.2 of the SCM Agreement are contingent upon a finding by the Panel that the United States has applied export subsidies inconsistently with its obligations under the Agreement on Agriculture, and only to the extent of that inconsistency.


      5. We note in particular that Brazil's request for the establishment of a panel uses the expression "in this respect" in the introductory paragraph of the section concerning Brazil's export credit guarantees claims. Read in conjunction with the immediately preceding sentence ("and has applied those export subsidies ... in a manner that results in circumvention of the United States' export subsidy commitments") and taking into consideration the fact that Brazil uses "circumvention of the United States' export subsidy commitments" to refer to a violation, by the United States, of its obligation under Article 10.1 of the Agreement on Agriculture790, it is clear to the Panel that Brazil's claims of inconsistency with 3.1(a) and 3.2 of the SCM Agreement depend on a finding by the Panel that the export subsidies in question have resulted in the circumvention of the United States' export subsidy commitments.


      6. In addition, paragraph 28 of Brazil's panel request, which explains in more details Brazil's claims, indicates that it is "[a]s a result" and "to the extent" of the violation of Articles 10.1 and 8 of the Agreement on Agriculture that the US export credit guarantee programmes in question are subject to the prohibition against export subsidies included in Articles 3.1 and 3.2 of the SCM Agreement.


      7. As a result, the Panel considers that Brazil's claims under Articles 3.1(a) and 3.2 of the SCM Agreement with respect to lyocell, lysine and wood products are not covered by Brazil's request for the establishment of a panel and are therefore not part of its terms of reference.


  15. CONCLUSIONS AND RECOMMENDATION


    1. In accordance with its mandate under Article 21.5 of the DSU, the Panel has examined the "existence or consistency with a covered agreement of measures taken" by the United States "to comply with the recommendations and rulings" adopted by the DSB in the original proceeding. The Panel concludes that:


      With respect to the measure taken by the United States to comply with the DSB recommendations and rulings relating to the original panel's finding of inconsistency with Articles 5 and 6 of the SCM Agreement:


      1. The United States acts inconsistently with its obligations under Articles 5(c) and 6.3(c) of the SCM Agreement in that the effect of marketing loan and counter- cyclical payments provided to US upland cotton producers pursuant to the FSRI Act of 2002 is significant price suppression within the meaning of Article 6.3(c) of the SCM Agreement in the world market for upland cotton constituting "present" serious prejudice to the interests of Brazil within the meaning of Article 5(c) of the SCM Agreement. By acting inconsistently with Articles 5(c) and 6.3(c) of the SCM Agreement the United States has failed to comply with the DSB recommendations and rulings. Specifically, the United States has failed to comply with its obligation under Article 7.8 of the SCM Agreement "to take appropriate steps to remove the adverse effects or. ..withdraw the subsidy".



        image

        790 See the first sentence of para. 27 of Brazil's request:

        "Further, subsequent to 1 July 2005, ECGs under the GSM 102 and SCGP programs have been applied to circumvent US export subsidy commitments, within the meaning of Article 10.1 of the Agreement on Agriculture."

        WT/DS267/RW

        Page 189


      2. Brazil has not made a prima facie case that the effect of marketing loan and counter- cyclical payments provided to US upland cotton producers pursuant to the FSRI Act of 2002 is an increase in the US world market share in upland cotton as compared to the average US world market share during the previous period of three years and that this increase follows a consistent trend over a period when subsidies have been granted. Therefore, it has not been established that the United States acts inconsistently with Articles 5(c) and 6.3(d) of the SCM Agreement.


        With respect to the measure taken by the United States to comply with the DSB recommendations and rulings relating to the original panel's findings of inconsistency with Articles 10.1 and 8 of the Agreement on Agriculture and Articles 3.1(a) and 3.2 of the SCM Agreement:


      3. Regarding GSM 102 export credit guarantees issued after 1 July 2005 the United States acts inconsistently with Article 10.1 of the Agreement on Agriculture by applying export subsidies in a manner which results in the circumvention of US export subsidy commitments with respect to certain unscheduled products791 and certain scheduled products792, and as a result acts inconsistently with Article 8 of the Agreement on Agriculture. Regarding GSM 102 export credit guarantees issued after 1 July 2005 the United States also acts inconsistently with Articles 3.1(a) and 3.2 of the SCM Agreement by providing export subsidies to unscheduled products793 and by providing export subsidies to scheduled products794 in excess of the commitments of the United States under the Agreement on Agriculture. By acting inconsistently with Articles 10.1 and 8 of the Agreement on Agriculture and Articles 3.1(a) and 3.2 of the SCM Agreement the United States has failed to comply with the DSB recommendations and rulings. Specifically, the United States has failed to bring its measures into conformity with the Agreement on Agriculture and has failed "to withdraw the subsidy without delay".


      4. With respect to certain export credit guarantees issued prior to 1 July 2005, Brazil has not established that the United States has failed to "withdraw the subsidy without delay".


    2. The Panel considers that to the extent that the measures taken by the United States to comply with the recommendations and rulings adopted by the DSB in the original proceeding are inconsistent with the obligations of the United States under the covered agreements, these recommendations and rulings remain operative. In this regard, the Panel recalls that the recommendations and rulings that were adopted by the DSB in the original proceeding were as follows:


      "(a) we recommend pursuant to Article 19.1 of the DSU that the United States bring its measures listed in paragraphs 8.1(d)(i) and 8.1(e) above into conformity with the Agreement on Agriculture;


        1. as required by Article 4.7 of the SCM Agreement, we recommend that the United States withdraw the prohibited subsidies in paragraphs 8.1(d)(i) and 8.1(e) above without delay. The time-period we specify must be consistent with the


          image

          791 The unscheduled products at issue are (i) in the period 1 July-30 September 2005: cotton, oilseeds (including soybeans/soybean meal), protein meals, fresh vegetables, hides/skins and tallow; and (ii) in the period 1 October 2005–30 September 2006: cotton, oilseeds, soybeans/soybean meal, protein meals, hides/skins, tallow and corn products.

          792 The scheduled products at issue are: (i) in the period 1 July-30 September 2005: rice and poultry meat; and (ii) in the period 1 October 2005-30 September 2006: rice, poultry meat and pig meat.

          793 Supra, footnote 791.

          794 Supra, footnote 792.

          WT/DS267/RW

          Page 190


          requirement that the subsidy be withdrawn 'without delay'. In any event, this is at the latest within six months of the date of adoption of the Panel report by the Dispute Settlement Body or 1 July 2005 (whichever is earlier);


        2. pursuant to Article 4.7 of the SCM Agreement, we recommend that the United States withdraw the prohibited subsidy in paragraph 8.1(f) above without delay and, in any event, at the latest within six months of the date of adoption of the Panel report by the Dispute Settlement Body or 1 July 2005 (whichever is earlier); and


        3. we recall that, in respect of the subsidies subject to our conclusion in paragraph 8.1(g)(i) above, pursuant to Article 7.8 of the SCM Agreement:

      '7.8 Where a panel report or an Appellate Body report is adopted in which it is determined that any subsidy has resulted in adverse effects to the interests of another Member within the meaning of Article 5, the Member granting or maintaining such subsidy shall take appropriate steps to remove the adverse effects or shall withdraw the subsidy'.

      Accordingly, upon adoption of this report, the United States is under an obligation to 'take appropriate steps to remove the adverse effects or ... withdraw the subsidy'".795


    3. The Panel has set out above in para. 9.79 its views on the interpretation of the obligation under Article 7.8 of the SCM Agreement to "take appropriate steps to remove the adverse effects or ... withdraw the subsidy".


image


image

795 Panel Report, US – Upland Cotton, para. 8.3. As stated by the panel in US – FSC (Article 21.5 – EC II), "once the DSB adopts a dispute settlement report, the findings and recommendations in that report become collective, operative DSB rulings and recommendations." US – FSC (Article 21.5 – EC II), para. 7.35.