Source(s) of the information:

Lawyers, other representatives, expert(s), tribunal’s secretary

Recourse to Arbitration by the United States under Article 22.6 of the DSU - Decision by the Arbitrator

TABLE OF CASES CITED IN THIS REPORT

Short TitleFull Case Title and Citation
Argentina – Hides and Leather Award of the Arbitrator, Argentina – Measures Affecting the Export of Bovine Hides and Import of Finished Leather – Arbitration under Article 21.3(c) of the DSU, WT/DS155/10, 31 August 2001, DSR 2001:XII, 6013
Australia – Salmon Award of the Arbitrator, Australia – Measures Affecting Importation of Salmon – Arbitration under Article 21.3(c) of the DSU, WT/DS18/9, 23 February 1999, DSR 1999:I, 267
Brazil – Aircraft (Article 22.6 – Brazil) Decision by the Arbitrators, Brazil – Export Financing Programme for Aircraft – Recourse to Arbitration by Brazil under Article 22.6 of the DSU and Article 4.11 of the SCM Agreement, WT/DS46/ARB, 28 August 2000, DSR 2002:I, 19
Canada – Aircraft Credits and Guarantees (Article 22.6 – Canada) Decision by the Arbitrator, Canada – Export Credits and Loan Guarantees for Regional Aircraft – Recourse to Arbitration by Canada under Article 22.6 of the DSU and Article 4.11 of the SCM Agreement, WT/DS222/ARB, 17 February 2003, DSR 2003:III, 1187
EC – Bananas III (Ecuador) (Article 22.6 – EC) Decision by the Arbitrators, European Communities – Regime for the Importation, Sale and Distribution of Bananas – Recourse to Arbitration by the European Communities under Article 22.6 of the DSU, WT/DS27/ARB/ECU, 24 March 2000, DSR 2000:V, 2237
EC – Bananas III (US) (Article 22.6 – EC) Decision by the Arbitrators, European Communities – Regime for the Importation, Sale and Distribution of Bananas – Recourse to Arbitration by the European Communities under Article 22.6 of the DSU, WT/DS27/ARB, 9 April 1999, DSR 1999:II, 725
EC – Hormones Appellate Body Report, EC Measures Concerning Meat and Meat Products (Hormones), WT/DS26/AB/R, WT/DS48/AB/R, adopted 13 February 1998, DSR 1998:I, 135
EC – Hormones (Canada) (Article 22.6 – EC) Decision by the Arbitrators, European Communities – Measures Concerning Meat and Meat Products (Hormones), Original Complaint by Canada – Recourse to Arbitration by the European Communities under Article 22.6 of the DSU, WT/DS48/ARB, 12 July 1999, DSR 1999:III, 1135
EC – Hormones (US) (Article 22.6 – EC) Decision by the Arbitrators, European Communities – Measures Concerning Meat and Meat Products (Hormones), Original Complaint by the United States – Recourse to Arbitration by the European Communities under Article 22.6 of the DSU, WT/DS26/ARB, 12 July 1999, DSR 1999:III, 1105
Korea – Alcoholic Beverages Award of the Arbitrator, Korea – Taxes on Alcoholic Beverages – Arbitration under Article 21.3(c) of the DSU, WT/DS75/16, WT/DS84/14, 4 June 1999, DSR 1999:II, 937
US – 1916 Act (EC) (Article 22.6 – US) Decision by the Arbitrators, United States – Anti-Dumping Act of 1916, Original Complaint by the European Communities – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS136/ARB, 24 February 2004, DSR 2004:IX, 4269
US – FSC (Article 22.6 – US) Decision by the Arbitrator, United States – Tax Treatment for "Foreign Sales Corporations" – Recourse to Arbitration by the United States under Article 22.6 of the DSU and Article 4.11 of the SCM Agreement, WT/DS108/ARB, 30 August 2002, DSR 2002:VI, 2517
US – Gambling Appellate Body Report, United States – Measures Affecting the Cross-Border Supply of Gambling and Betting Services, WT/DS285/AB/R and Corr.1, adopted 20 April 2005, DSR 2005:XII, 5663
US – Gambling Panel Report, United States – Measures Affecting the Cross-Border Supply of Gambling and Betting Services, WT/DS285/R, adopted 20 April 2005, as modified by Appellate Body Report, WT/DS285/AB/R, DSR 2005:XII, 5797
US – Gambling (Article 21.5 – Antigua and Barbuda) Panel Report, United States – Measures Affecting the Cross-Border Supply of Gambling and Betting Services – Recourse to Article 21.5 of the DSU by Antigua and Barbuda, WT/DS285/RW, adopted 22 May 2007
US – Offset Act (Byrd Amendment) (Brazil) (Article 22.6 – US) Decision by the Arbitrator, United States – Continued Dumping and Subsidy Offset Act of 2000, Original Complaint by Brazil – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS217/ARB/BRA, 31 August 2004, DSR 2004:IX, 4341
US – Offset Act (Byrd Amendment) (Canada) (Article 22.6 – US) Decision by the Arbitrator, United States – Continued Dumping and Subsidy Offset Act of 2000, Original Complaint by Canada – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS234/ARB/CAN, 31 August 2004, DSR 2004:IX, 4425
US – Offset Act (Byrd Amendment) (Chile) (Article 22.6 – US) Decision by the Arbitrator, United States – Continued Dumping and Subsidy Offset Act of 2000, Original Complaint by Chile – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS217/ARB/CHL, 31 August 2004, DSR 2004:IX, 4511
US – Offset Act (Byrd Amendment) (EC) (Article 22.6 – US) Decision by the Arbitrator, United States – Continued Dumping and Subsidy Offset Act of 2000, Original Complaint by the European Communities – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS217/ARB/EEC, 31 August 2004, DSR 2004:IX, 4591
US – Offset Act (Byrd Amendment) (India) (Article 22.6 – US) Decision by the Arbitrator, United States – Continued Dumping and Subsidy Offset Act of 2000, Original Complaint by India – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS217/ARB/IND, 31 August 2004, DSR 2004:X, 4691
US – Offset Act (Byrd Amendment) (Japan) (Article 22.6 – US) Decision by the Arbitrator, United States – Continued Dumping and Subsidy Offset Act of 2000, Original Complaint by Japan – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS217/ARB/JPN, 31 August 2004, DSR 2004:X, 4771
US – Offset Act (Byrd Amendment) (Korea) (Article 22.6 – US) Decision by the Arbitrator, United States – Continued Dumping and Subsidy Offset Act of 2000, Original Complaint by Korea – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS217/ARB/KOR, 31 August 2004, DSR 2004:X,4851
US – Offset Act (Byrd Amendment) (Mexico) (Article 22.6 – US) Decision by the Arbitrator, United States – Continued Dumping and Subsidy Offset Act of 2000, Original Complaint by Mexico – Recourse to Arbitration by the United States under Article 22.6 of the DSU, WT/DS234/ARB/MEX, 31 August 2004, DSR 2004:X,4931
US – Section 110(5) Copyright Act (Article 25.3) Award of the Arbitrators, United States – Section 110(5) of the US Copyright Act – Recourse to Arbitration under Article 25 of the DSU, WT/DS160/ARB25/1, 9 November 2001, DSR 2001:II, 667
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

I. INTRODUCTION

A. INITIAL PROCEEDINGS

1.1.
On 20 April 2005, the Dispute Settlement Body (DSB) adopted the report of the Panel in this dispute, as modified by the report of the Appellate Body.1
1.2.
The Appellate Body inter alia upheld the original panel's finding that the United States' Schedule includes a commitment to grant full market access in gambling and betting services and upheld the Panel's finding that the United States acts inconsistently with Article XVI:1 and sub-paragraphs (a) and (c) of Article XVI:2 by maintaining certain limitations on market access not specified in its Schedule. The Appellate Body reversed the Panel's finding that the United States had not shown that the three federal statutes are "necessary to protect public morals or to maintain public order", within the meaning of Article XIV(a). The Appellate Body modified the conclusion of the panel with respect to Article XIV as a whole and found, instead, "that the United States has demonstrated that the Wire Act, the Travel Act, and the Illegal Gambling Business Act are measures 'necessary to protect public morals or maintain public order', in accordance with paragraph (a) of Article XIV, but that the United States has not shown, in the light of the Interstate Horseracing Act, that the prohibitions embodied in those measures are applied to both foreign and domestic service suppliers of remote betting services for horse racing and, therefore, has not established that these measures satisfy the requirements of the chapeau".2
1.3.
On 19 August 2005, an arbitrator established under Article 21.3(c) of the DSU determined that the "reasonable period of time" for the United States to implement the recommendations and rulings of the DSB in this dispute was 11 months and 2 weeks from the date of adoption of the Panel and Appellate Body Reports by the DSB. The United States was consequently awarded until 3 April 2006 to bring its measures into conformity with its obligations under the GATS.3
1.4.
On 8 June 2006, Antigua and Barbuda (hereafter "Antigua") requested consultations with the United States under Article 21.5 of the DSU.4 On 6 July 2006, Antigua requested the establishment of a panel under Article 21.5 of the DSU.5 At its meeting on 19 July 2006, the DSB referred the matter to the original panel, if possible. The report of the Article 21.5 compliance panel was adopted on 25 May 2007. The compliance panel found that the United States had failed to comply with the recommendations and rulings of the DSB in this dispute.
1.5.
On 21 June 2007, Antigua requested authorization from the DSB6, under Article 22.2 of the DSU, to suspend the application to the United States of concessions and related obligations of Antigua under the General Agreement on Trade in Services (the "GATS" and the Agreement on Trade-Related Aspects of Intellectual Property Rights (the "TRIPS Agreement") amounting to an annual value of US$3,443 billion.

B. REQUEST FOR ARBITRATION AND PROCEEDINGS

1.6.
On 23 July 2007, the United States objected to the level of suspension proposed by Antigua and Barbuda pursuant to Article 22.6 of the DSU and claimed that Antigua had not followed the principles and procedures of Article 22.3 of the DSU in its request.7 At its meeting of 24 July 2007, the DSB agreed that the matter raised by the United States had been referred to arbitration.
1.7.
The arbitration was undertaken by the original panelists of the Article 21.5 compliance panel as follows:

Chairperson: Mr Lars Anell

Members: Mr Mathias Francke

Mr Virachai Plasai

1.8.
An organizational meeting was held on 7 August 2007 to discuss proposed working procedures and the timetable for the proceedings. At that meeting, both parties requested, inter alia, additional time for the preparation of their communications to the Arbitrator. The final timetable and working procedures adopted by the Arbitrator were transmitted to the parties on 22 August 2007 (see the text of the Working Procedures in Annex 1).
1.9.
In accordance with the timetable adopted by the Arbitrator, Antigua presented a communication concerning the methodology for calculating the proposed level of suspension ("Methodology Paper") on 31 August 2007, and the United States and Antigua submitted written submissions to the Arbitrator on 19 September and 4 October 2007 respectively. The Arbitrator met with the parties on 18 October 2007. After the meeting, the Arbitrator posed questions to the parties, to which the parties provided written responses on 2 November 2007. Both parties were further provided with an opportunity to comment in writing on the responses of the other party to the questions of the Arbitrator. The decision of the Arbitrator was circulated on 21 December 2007.

II. PRELIMINARY ISSUES

A. MANDATE OF THE ARBITRATOR

2.1.
Article 22.7 of the DSU provides that:

"The arbitrator[1] acting pursuant to paragraph 6 shall not examine the nature of the concessions or other obligations to be suspended but shall determine whether the level of such suspension is equivalent to the level of nullification or impairment. The arbitrator may also determine if the proposed suspension of concessions or other obligations is allowed under the covered agreement. However, if the matter referred to arbitration includes a claim that the principles and procedures set forth in paragraph 3 have not been followed, the arbitrator shall examine that claim. In the event the arbitrator determines that those principles and procedures have not been followed, the complaining party shall apply them consistent with paragraph 3."

[1] The expression "arbitrator" shall be interpreted as referring either to an individual or a group or to the members of the original panel when serving in the capacity of arbitrator.

2.2.
In this proceeding, the United States challenges two distinct aspects of the request by Antigua for the suspension of certain obligations under the covered agreements. First, it challenges the level of suspension that Antigua seeks an authorization for (i.e. the figure of US$3,443 billion as the level of nullification or impairment of benefits accruing to Antigua). Secondly, the United States also challenges the choice of the sectors and agreement in which Antigua is proposing to carry out the suspension (i.e. certain obligations under the GATS and under the TRIPS Agreement). The Arbitrator therefore has two distinct determinations to make in these proceedings.

1. Mandate in relation to the proposed level of suspension

2.3.
Antigua has requested an authorization to suspend the application to the United States of concessions and related obligations of Antigua under the GATS and the TRIPS Agreement, in an amount of an "annual value of US$3,443 billion", which it considers to "match the level of nullification or impairment of benefits accruing to Antigua and Barbuda". The United States, however, has objected to the level of suspension of concessions and other obligations proposed.
2.4.
Article 22.4 of the DSU provides that "[t]he level of the suspension of concessions or other obligations authorized by the DSB shall be equivalent to the level of nullification or impairment". Article 22.6 further provides that "if the Member concerned objects to the level of suspension proposed..., the matter shall be referred to arbitration".
2.5.
As noted above, Article 22.7 provides that "[t]he arbitrator acting pursuant to paragraph 6 (…) shall determine whether the level of such suspension is equivalent to the level of nullification or impairment (..)".8 The DSU provides no further detail how exactly such equivalence might be established. Some guidance is provided, however, by previous arbitral decisions in which such assessments were carried out further to a recourse to arbitration under Article 22.6 of he DSU.9
2.7.
In approaching this task, we note, as other arbitrators have, that, while the purpose of suspension of concessions or other obligations under the covered agreements as foreseen in Article 22.1 of the DSU is to "induce compliance" by the Member concerned with its obligations under the covered agreements, this does not mean that such suspension may be authorized beyond what is "equivalent" to the level of nullification of impairment.11 Rather, in setting out the requirement that suspension be "equivalent" to the level of nullification or impairment, Article 22.4 of the DSU requires a degree of "correspondence or identity"12 between the level of the suspension to be authorized and the level of the nullification or impairment of benefits.
2.8.
This means that it is necessary to determine what this level of nullification or impairment of benefits is, in order to compare it to the requested level of suspension.13 Further, past arbitrators have also considered that, if they determined that the proposed level is not equivalent to the level of nullification or impairment as required by the DSU, then it was also their duty to estimate the level of suspension that they considered to be equivalent to the impairment suffered, with a view to contributing to the objective of prompt and positive settlement of disputes embodied in the DSU.14 This is also what the United States is asking the Arbitrator to do in this dispute.
2.9.
In light of these elements, we understand our mandate in relation to Antigua's proposed level of suspension to require us to determine whether the annual amount of US$3,443 billion proposed by Antigua is equivalent to the level of nullification or impairment of benefits accruing to Antigua under the GATS as a result of the failure of the United States to bring its GATS-inconsistent measures into conformity with its obligations. If we find that it is not, then we will need to determine what the level of such nullification or impairment is.
2.10.
We are also mindful that Article 22.7 of the DSU provides that the Arbitrator "shall not examine the nature of the concessions or other obligations to be suspended". Our analysis will, accordingly, be limited to an examination of the level of the proposed suspension, and will not extend to a consideration of the nature of the concessions or other obligations proposed for suspension.
2.11.
We note that the United States has also requested us to require Antigua to indicate how it will ensure that any suspension of concessions does not exceed the level of nullification and impairment found by the Arbitrator. In the view of the United States, without this information it would be impossible for the Arbitrator to determine the equivalence of the level of suspension of concessions with the level of nullification and impairment, as is required by DSU Article 22.715 and the Arbitrator should thus not find that Antigua is allowed to suspend TRIPS concessions.16 In Antigua's view, however, the imposition of a requirement for it to specify how it will ensure that the level of suspension of concessions does not exceed the level of nullification determined by the Arbitrators would not be within the terms of reference of the Arbitrators under Article 22.7 of the DSU.
2.12.
Without prejudice to our views as to whether our mandate under Article 22.7 of the DSU allows us to impose such a requirement on Antigua, we find it more appropriate to consider this matter once we have made the relevant determinations on the level of suspension and the principles and procedures of Article 22.3 of the DSU, especially in light of the fact that the United States concern appears to relate primarily to the suspension of obligations under the TRIPS Agreement, a matter which will only become pertinent if and when we determine that Antigua may seek suspension of obligations under that Agreement. We therefore leave the discussion of our mandate in relation to this question for a later stage of our award.17

2. Mandate in relation to the sector(s) and agreement(s) in which suspension is sought

2.13.
Antigua's request for authorization to suspend concessions or other obligations refers to both subparagraph (b) and subparagraph (c) of Article 22.3, and it explains that "[b]ecause the withdrawal of concessions solely under the GATS is at present not practicable or effective, and the circumstances are sufficiently serious to justify Antigua and Barbuda exercising its rights under Article 22, Antigua and Barbuda requests authorization to suspend concessions and other obligations under the TRIPS".18 The United States, in its request for arbitration under Article 22.6, claimed that Antigua's proposal "does not follow the principles and procedures set forth in paragraph 3 of Article 22 of the DSU".19
2.14.
Article 22.3 sets out, in subparagraphs (a) to (c), certain principles and procedures to be followed by a complaining party seeking to suspend concessions or other obligations, as to the sector(s) and/or covered agreement in which the suspension can be sought. Subsequent subparagraphs of Article 22.3 further elaborate on the factors that must be taken into account in applying these principles and procedures as well as the definition of "sectors" for the purposes of this provision and the procedural requirements to be followed in requesting authorization to suspend concessions or other obligations in another sector or under another covered agreement.
2.15.
Article 22.7, as noted above, provides that "if the matter referred to arbitration includes a claim that the principles and procedures set forth in paragraph 3 have not been followed, the arbitrator shall examine that claim" and that "[i]n the event the arbitrator determines that those principles and procedures have not been followed, the complaining party shall apply them consistent with paragraph 3".
2.16.
There has only been one dispute to date in which claims relating to the observance of the principles and procedures of Article 22.3 of the DSU have been considered (EC – Bananas III).20 In that dispute, in determining the scope of their authority to review the principles and procedures of subparagraphs (b) and/or (c) of Article 22.3, the arbitrators considered that "the fact that the powers of Arbitrators under subparagraphs (b)-(c) are explicitly provided for in Article 22.6 implies a fortiori that the authority of Arbitrators includes the power to review whether the principles and procedures set forth in these subparagraphs have been followed by the Member seeking authorization for suspension".21
2.17.
In this instance, Antigua's request, as noted above, refers both to subparagraph (b) and to subparagraph (c) of Article 22.3. Antigua requested an authorization to suspend obligations both under the GATS and under the TRIPS Agreement. To the extent that it sought to retaliate under the GATS, it sought to do so in a different sector from that in which the violation was found. In addition, Antigua sought an authorization to suspend certain obligations under the TRIPS Agreement. In a subsequent communication, Antigua clarified that it was now seeking an authorization to suspend obligations only under the TRIPS Agreement.22
2.18.
Our mandate with respect to Antigua's request to suspend concessions or other obligations under the TRIPS Agreement is therefore to review whether the principles and procedures set forth in the various subparagraphs of Article 22.3 of the DSU have been followed by Antigua, as the Member seeking authorization for suspension in this dispute.

3. Order of analysis

2.19.
As noted above, the DSU provides for specific mandates in respect of the two aspects of the matter before us. To the extent, however, that our determinations in respect of one aspect may have a bearing in our determinations in respect of the other, we consider it appropriate to examine these two aspects in an order that will allow us to fully take into consideration the relevant elements for each determination, and to then make overall conclusions. Specifically, we consider it appropriate, in the circumstances of this dispute, to examine the issues before us in the following order:

(a) first, the level of nullification or impairment;

(b) second, the principles and procedures of Article 22.3 of the DSU, taking into account to the extent relevant the level of nullification or impairment as previously determined;

(c) third, the US request in relation to equivalence, taking in to account our determination, in the preceding section, as to the sectors and/or covered agreements under which Antigua may seek to suspend obligations;

(d) finally, our conclusions.

B. BURDEN OF PROOF AND PRESENTATION OF EVIDENCE

2.20.
Antigua has argued that the burden of proof in Article 22.6 arbitrations is well settled, and that it is not for the complaining party to demonstrate that its request for suspension of concessions or other obligations conforms to the requirements of Article 22, rather that it is for the responding party to establish a prima facie case that the complaining party has not done so. If the responding party is able to establish a prima facie case that the complaining party has not conformed to the requirements of Article 22, then, Antigua argues, the burden shifts to the complaining party to rebut that presumption.23
2.21.
In response to a question by the Arbitrator, the United States indicated that it agreed that the party referring the matter to arbitration has the initial burden of showing that the proposed level of suspension is not equivalent to the level of nullification or impairment.24 At the same time, the United States considers that this burden does not mean that the allegations and factual assertions of Antigua enjoy any presumption of correctness or any special weight simply because Antigua has put them forward.25
2.26.
With respect to the US claims that Antigua has not followed the principles and procedures of Article 22.3 of the DSU, as noted above, only one relevant precedent exists (EC – Bananas III). In that dispute, the arbitrator found that the same rules applied for the allocation of burden of proof, namely that it was for the Member challenging the proposal to demonstrate that the principles and procedures of Article 22.3 had not been followed. Specifically, the Arbitrator in that dispute considered that it was "for the European Communities to challenge Ecuador's considerations of the principles and procedures set forth in Articles 22.3 (b)-(d)" and that "[o]nce the European Communities has shown prima facie that these principles and procedures have not been followed, and that the factors listed in subparagraph (d) were not taken into account, however, it [was] for Ecuador to rebut such a presumption".29
2.28.
We also note that the Working Procedures adopted by the Arbitrator in these proceedings provide that "each party shall submit all factual evidence to the Arbitrator no later than in its written submission to the Arbitrator, except with respect to evidence necessary for the purposes of rebuttal or for answers to questions".

C. REQUEST BY THE UNITED STATES FOR AN OPEN HEARING

D. REQUEST BY THE EUROPEAN COMMUNITIES FOR THIRD PARTY STATUS

2.30.
On 3 August 2007, the Arbitrator received a communication by the European Communities, requesting to be accorded third-party status in these proceedings. The Arbitrator sought the views of both parties on this request, and received these views in writing on 9 August 2007.

E. CONSULTATION OF THE INTERNATIONAL MONETARY FUND (IMF) AND THE EASTERN CARIBBEAN CENTRAL BANK (ECCB)

2.32.
On 19 November 2007, the Arbitrator informed the parties that, in the light of the parties' responses to its questions and because of the need for further clarification of certain factual issues, it considered that it would be useful for it to seek technical information from the IMF and the ECCB concerning the exact source and contents of the data collected by them, which had been referred to in these proceedings. The Arbitrator therefore proposed to address a request for information to both organizations and requested the parties to comment on the text of draft communications to be sent to these two organizations.
2.33.
The United States provided some specific drafting suggestions. Antigua, for its part, admitted to "a degree of ambivalence regarding the proposed enquiries".34 On the one hand, Antigua was certain that whatever responses the Arbitrators would get as a result of these enquiries, they would support the evidence and claims made by Antigua during the course of this arbitration, but on the other hand, in the view of Antigua, the enquiries could also be viewed as reflecting adversely upon the veracity or accuracy of fundamentally uncontested evidence submitted by Antigua in this arbitration, and as an effort by the Arbitrators to engage in fact finding on behalf of a litigant that has essentially failed to do so itself, despite having many opportunities.
2.34.
The Arbitrator considered Antigua's arguments, in particular its concern that the Arbitrator may be engaging in fact finding "on behalf of " the United States. While the Arbitrator agreed that the purpose of an expert consultation should not be to substitute for a party's failure to make its case, the Arbitrator considers that this is not what it was proposing to do in seeking information from the IMF and the ECCB. Rather, the aim of the Arbitrator was to seek additional clarification of some of the factual elements that had been put before it by the parties, to make its determination on the basis of the best possible information.
2.35.
In light of these considerations, the Arbitrator decided to proceed with the consultation. Accordingly, it revised the text of the draft communications in light of the comments of the parties, and addressed requests for information to both the IMF and the ECCB on 22 November 2007 (see text reproduced in Annexes C and B). A response was received from the IMF on 14 December 2007 (see text reproduced in Annex D). No response had been received from the ECCB by the time the Arbitrator finalized its work.

III. ASSESSMENT OF THE PROPOSED LEVEL OF SUSPENSION

3.1.
As noted above in paragraph 1.5, Antigua has requested an authorization to suspend the application to the United States of "concessions and related obligations" of Antigua under the GATS and the TRIPS Agreement, in an amount of an "annual value of US$3,443 billion", which it considers to "match the level of nullification or impairment of benefits accruing to Antigua and Barbuda". The United States, however, has objected to this proposed level of suspension, pursuant to Article 22.6 of the DSU.

A. MAIN ARGUMENTS OF THE PARTIES

3.2.
In its Methodology Paper, Antigua presented three calculations in support of its proposed level of suspension of concessions or other obligations, and explained that its calculations were based on a counterfactual that assumed that to comply with the recommendations and rulings of the DSB, the United States would allow Antiguan operators to provide unrestricted access to cross-border remote gambling and betting services to United States consumers.35 Specifically, Antigua indicated that it made the following assumptions in relation to the counterfactual:

1. The United States adheres to its GATS commitments for remote gambling and betting services as established in the dispute;

2. The United States recognizes that Antiguan law governs Antiguan-based remote gaming operators serving customers located in the United States;

3. The United States does not interfere with the electronic transfer of funds between customers and Antiguan remote gaming operators;

4. The United States does not interfere with advertising by Antiguan remote gaming operators;

5. Antiguan remote gaming operators are not compelled to invest significant resources to counteract United States measures to restrict gaming operators from providing remote wagering services to United States consumers.36

3.4.
The United States, however, disputes Antigua's choice of counterfactual. Specifically, it disputes the assumption that the compliant US measure would be a measure that allowed for "all types of cross-border gambling services", for which it considers that Antigua has no legal or factual basis.38 In the view of the United States, the Arbitrator needs to take into account "the possibility of compliance under different approaches, not just the approach preferred by Antigua".39 The United States submits that the DSB recommendations and rulings are addressed to discriminatory treatment of gambling on horseracing, so that the real question in the scenario where the United States achieves compliance by allowing market access for foreign suppliers is as follows: "what would be the economic effect if the United States were to adopt measures that allow Antiguan operators (and operators of all other WTO Members) to provide cross-border remote gambling services on horseracing?"
3.5.
The United States thus considers that a measure that would open, without discrimination, access to remote gambling services in respect of horseracing only would constitute compliance with the rulings. The United States identifies this as the pertinent scenario, if the assumption is to be that it would allow some access to remote gambling. In the United States view, this is the "most likely" scenario.
3.6.
The United States also identifies, as a scenario that would in its view be consistent with the DSB rulings and recommendations, a situation in which it would ensure that domestic service suppliers are also prohibited from supplying remote betting services on horseracing.40 However, it does not suggest that this be the scenario retained for the purposes of the counterfactual in these proceedings.
3.7.
Antigua considers that the United States misses the key starting point when it "takes the fundamental position in its submission that the nullification and impairment to Antigua only extends to hypothetical trade in betting on horse racing that might be available to Antiguan operators from customers in the United States".41 Antigua analyses the rulings, in particular the role of the affirmative defence under Article XIV of GATS and its relationship to market access commitments under Article XVI, to conclude that the United States should be barred from presenting a completely new argument (distinguishing between horseracing and other forms of remote gambling) in the hopes of minimizing its exposure to Antigua for having failed to come into compliance with the DSB rulings".42
3.8.
Antigua also recalls that the United States bears the burden of proof "to overturn the presumptions established by Antigua's Request and the Methodology Paper".43 Antigua further considers it "unconscionable" for the United States to have done nothing to come into compliance in the time that it should have, and now ask the arbitrators to determine what may have constituted compliance.44 It considers that, in the circumstances of this dispute, where there has been no attempt to comply, the only assumption that may be used by the Arbitrator in assessing the level of nullification and impairment is "withdrawal of the measure" as set out in Article 3.7 of the DSU.45
3.9.
The United States also takes issue with other aspects of Antigua's methodology, in particular the underlying data, which it considers to be "inherently unreliable" and as providing "no useful information in this arbitration".46

B. APPROACH OF THE ARBITRATOR

3.10.
As noted above, Article 22.7 provides that "[t]he arbitrator acting pursuant to paragraph 6 (…) shall determine whether the level of such suspension is equivalent to the level of nullification or impairment (..)".47
3.11.
In this instance, we are therefore required to determine whether the amount of US$3,443 billion proposed by Antigua is equivalent to the level of nullification or impairment of benefits accruing to Antigua. As has been noted by previous arbitrators, for that purpose, we must first determine what the level of nullification or impairment of benefits accruing to Antigua is.48
3.12.
Antigua's proposed level of suspension is based on an assessment of the annual amount of trade that it considers that it has lost, as a result of the maintenance of the inconsistent US measures beyond the end of the reasonable period of time for implementation. To calculate the level of such lost trade, Antigua relies on a counterfactual scenario intended to reflect what the situation would have been, if the United States had complied with the DSB recommemdations and rulings.
3.13.
The United States does not disagree with the basic approach adopted by Antigua in determining its level of nullification or impairment of benefits. Specifically, the United States agrees that the level of nullification and impairment may be calculated based on a "counterfactual", under which the Member concerned is assumed to have adopted measures in compliance with the DSB recommendations and rulings".49
3.14.
We note that this approach has been applied in prior proceedings under Article 22.6 of the DSU.50 We agree that the use of a counterfactual to assess the level of exports that would have accrued to Antigua, had the United States complied with the rulings, constitutes an appropriate basis for assessing the level of nullification or impairment of benefits accruing to Antigua in this dispute and, in light of the parties' common view in this respect, we accept it as the basis for our determination.
3.15.
Although the parties are in agreement on the suitability of this general approach, the United States disagrees both with the counterfactual upon which Antigua has based its calculations and with the data used by Antigua to then estimate the level of its lost exports under that counterfactual. The threshold issue of the choice of counterfactual has a crucial bearing on subsequent steps in the determination of the level of nullification or impairment of benefits, as it effectively determines the basis on which subsequent calculations are to be made. We will therefore consider this issue first, before turning to the actual calculation of the level of nullification.

C. CHOICE OF COUNTERFACTUAL

1. Assessment of Antigua's proposed counterfactual

3.20.
The US challenge to Antigua's counterfactual arises from the fact that it assumes a means of achieving compliance with the DSB rulings that is, in the view of the United States, "extraordinarily unrealistic" in this dispute. In the view of the United States, there is no factual or legal basis for Antigua to assume that the United States would comply with the DSB's recommendations and rulings in this dispute by providing unrestricted access to its remote gambling sector to Antiguan operators.
3.21.
Antigua's counterfactual is based on an assumption that the United States, by virtue of its specific commitments under the GATS, is required to provide unrestricted access to its remote gambling market to Antiguan operators. Antigua has explained that it is entitled to assume that the United States would comply with its obligations by allowing such access to the United States remote gambling market for Antiguan operators. Antigua finds support for this assumption in particular in the nature of the rulings (a violation of a market access commitment under Article XVI of GATS), as well as in the terms of Article 3.7 of the DSU, which provide that the first objective of the dispute settlement system is usually to secure the withdrawal of the measures concerned if these are found to be inconsistent with the provisions of any of the covered agreements.
3.22.
The United States acknowledges that the scenario proposed by Antigua would constitute a means of compliance with the recommendations and rulings of the DSB.56 However, it argues that this scenario is "extremely unrealistic". The United States considers that only benefits that can reasonably be expected to accrue to the complaining party can form the basis for the calculation of nullification or impairment of benefits. The United States further considers that the Arbitrator's enquiry "is simply what counterfactual – based on the specific facts or circumstances in the dispute and the specific DSB recommendations and rulings – would be the most likely form of compliance".57
3.23.
In considering this issue, we first recall that the burden rests upon the United States to demonstrate that the level of suspension proposed by Antigua is not equivalent to the level of nullification or impairment of benefits it has suffered as a result of the continued application of the inconsistent US measures. For that purpose, it would not be sufficient, in our view, for the United States to simply identify the existence of alternative possible means of compliance, other than that envisaged by Antigua in its counterfactual. Rather, for the United States to succeed in its challenge, it must persuade us that the particular scenario identified by Antigua as the basis for its counterfactual is not such as to accurately reflect the level of nullification or impairment of benefits accruing to Antigua under the GATS as a result of the continued application by the United States of inconsistent measures.
3.28.
We find support for our position in past determinations. We thus note that the arbitrator in US – Bananas III case selected what it determined to be a "reasonable counterfactual" for the purposes of its calculation, taking into account the circumstances of the case. We also note that in the US – Section 110(5) Copyright Act (Article 25) dispute59, the arbitrator based its determination of the level of nullification or impairment on historical figures on royalty income prior to the introduction of the TRIPS-incompatible limitation to copyright, rather than on a counterfactual that would have assumed no limitations in respect of the right in question.60 We further note the finding that "the level of royalty income which the European Communities could reasonably expect EC right holders to receive is, in the view of the Arbitrators, limited to licensing revenue from the numbers of users that would be licensed"61 (rather than the entirety of fees that could by law have been collected). We further note the Arbitrator's observation in that dispute that:

It should be recalled, in this context, that the inquiry into the level of benefits which the European Communities could expect to accrue to it if Section 110(5)(B) were brought into conformity with the TRIPS Agreement is hypothetical in nature. The Arbitrators consider that, in such a situation, it is necessary to proceed with caution, such that only those benefits which the European Communities could, in good faith and taking into account all relevant circumstances, expect to derive from Articles 11bis(1)(iii) and 11(1)(ii) are found to be nullified or impaired".62

3.29.
In response to a question by the Arbitrator, Antigua commented that, in that case, the parties were not in disagreement over the nature of the benefits that should accrue to the European Communities, but rather over the level of nullification or impairment, and that the "reasonable expectations" at issue did not refer to what was expected of the United States government but to what the European Communities could reasonably expect of non-governmental third-party participants in extracting the full value out of the impaired rights.63
3.31.
With these considerations in mind, we now turn to an examination of whether the specific counterfactual upon which Antigua bases its calculations accurately reflects the benefits accruing to it in this dispute. As noted above, it is for the United States to demonstrate to us that it does not.
3.32.
Antigua considers that the nature of the benefits accruing to Antigua is reflected in Article XVI of the GATS and in the US schedule, and that for the purposes of this arbitration, it means that the United States is under an obligation to provide unrestricted market access for Antiguan providers of all cross-border remote gambling and betting services to consumers in the United States. In Antigua's view, whether it should "reasonably" expect such benefit is not an issue, since it has a legal right to enjoy it.64 Antigua also considers that it is "entitled to ignore the failed Article XIV defences of the United States in making Antigua's assumptions in this proceeding".65 The United States, however, notes that, in this dispute, the DSB found that the US measures restricting remote gambling were provisionally justified under the public morals exception set out in Article XIV(a) of the GATS, so that Antigua is wrong in asserting that the United States has an unconditional obligation to allow access to each and every type of service covered by the service sectors included in the Member's schedule of GATS commitments.66 In the US view, the specific problem found with the US measures at issue was with respect to the limited issue of the regulation of remote gambling on horseracing.67
3.33.
As we understand it, the question before us therefore essentially turns around whether and how we should take into account, in assessing the proposed counterfactual, the fact that the Appellate Body found that the US measures at issue were "necessary to protect public morals or maintain public order" in accordance with paragraph (a) of Article XIV, even though it ultimately found these measures not to be justified under Article XIV.
3.34.
Both parties have presented extensive arguments relating to the rulings in the underlying dispute to explain their proposed counterfactuals and justify opposing views on this question. We find it useful, in these circumstances, to consider first the terms of the Appellate Body's rulings in this dispute.
3.35.
The Appellate Body made, inter alia, the following findings:

- first, with respect to Article XVI of the GATS, the Appellate Body upheld the Panel's finding that:

"by maintaining the Wire Act, the Travel Act, and the Illegal Gambling Business Act, the United States acts inconsistently with its obligations under Article XVI:1 and sub-paragraphs (a) and (c) of Article XVI:2";68

- further, with respect to Article XIV of the GATS, the Appellate Body found that:

"the United States has demonstrated that the Wire Act, the Travel Act, and the Illegal Gambling Business Act are measures 'necessary to protect public morals or maintain public order', in accordance with paragraph (a) of Article XIV, but that the United States has not shown, in the light of the Interstate Horseracing Act, that the prohibitions embodied in those measures are applied to both foreign and domestic service suppliers of remote betting services for horse racing and, therefore, has not established that these measures satisfy the requirements of the chapeau."69

3.36.
We note that the findings of the Appellate Body contain an initial finding that the measures at issue (three federal laws) are inconsistent with Article XVI of the GATS, followed by findings that these same measures are "necessary to protect public morals or maintain public order" within the meaning of paragraph (a) of Article XIV, but that the United States ultimately "has not established that these measures satisfy the requirements of the chapeau" of that provision.
3.38.
The Appellate Body's finding of a violation of Article XVI of the GATS in the underlying proceedings was based on a finding that the United States had a commitment, under sub-sector 10.D of the GATS, to provide unrestricted access to the cross-border supply of remote gambling and betting services. In light of this, we agree that the nature of the benefits accruing to Antigua in this dispute is reflected in Article XVI of the GATS and in the US Schedule, and that these benefits relate to Antigua's access to the supply of cross-border gambling and betting services in the United States.
3.39.
The question we must consider, however, is whether Antigua's assumption that such access would come in the form of unrestricted access to the US market with respect to all forms of remote gambling and betting services was reasonable, in the circumstances of this dispute.
3.40.
Antigua's proposed counterfactual assumes that, as of 3 April 2006 (the end of the reasonable period of time for implementation), the United States would have granted it unrestricted access to its remote gambling market. As Antigua itself notes, it is based on an assumption, by Antigua, that it was entitled to "ignore" the United States invocation of Article XIV of the GATS in the underlying proceedings. The United States objects that this is an "extraordinarily unrealistic" scenario, in view in particular of the fact that the United States bans remote gambling for strong policy reasons of protecting public morality and public order, and that it is entitled under the GATS to maintain such a ban as long as it is not applied in a manner that arbitrarily or unjustifiably discriminates between operators in different jurisdictions71. The United States also argues that the specific problem found with the US measures at issue was the "limited issue of the regulation of remote gambling on horse racing".72
3.42.
We further note that the Appellate Body found that US measures restricting access to the remote gambling market were "necessary" within the meaning of Article XIV of the GATS for the protection of public morals and public order, and found that the United States had not shown, "in the light of the Interstate Horseracing Act, that the prohibitions embodied in those measures are applied to both foreign and domestic service suppliers of remote betting services for horse racing and, therefore, had not established that these measures satisfy the requirements of the chapeau".73
3.43.
In light of these elements, we consider that Antigua's assumption that it could simply "ignore" the failed US defence under Article XIV of the GATS, and that the United States would comply with the rulings by providing unrestricted access to all sectors of its remote gambling market for Antiguan operators, was not reasonable, taking into account the particular circumstances of this dispute.
3.44.
In making this determination, we have considered carefully Antigua's arguments that it was entitled by right to expect unrestricted market access, because the United States was obligated to provide such treatment in accordance with Article XVI of the GATS and its specific commitments, and that Antigua was also entitled to expect "withdrawal" of the inconsistent measure (i.e. in this dispute, in Antigua's view, the withdrawal of the US restrictions on access to the remote gambling market), in accordance with Article 3.7 of the DSU.
3.45.
We agree with Antigua that the United States has an obligation, under Article XVI of the GATS and the US schedule, to provide market access on the cross-border supply of remote gambling services. At the same time, however, the GATS also recognizes the right of Member to regulate in order to meet their national policy objectives74 and Article XIV specifically recognizes the possibility of adopting measures necessary to protect public order and public morals. In this dispute, such legitimate policy objectives were consistently invoked by the United States in order to restrict access to its remote gambling market despite the existence of a specific commitment, and the Appellate Body has found that the three federal laws at issue in the proceedings were "measures... necessary to protect public morals or to maintain public order".75 In these circumstances, a reasonable counterfactual must take into account these US policy objectives.
3.47.
In this respect, we note the observations of the compliance panel in this dispute:

"The possible form of measures taken to comply with a recommendation under Article 19.1 of the DSU will depend on the rulings of the DSB in a particular dispute. For example, if a measure has been found inconsistent with a covered agreement, or unjustified under an otherwise available exception, due to the way in which the measure is applied, compliance with the recommendation could presumably be achieved by a change in the application of the measure, without necessarily a change to the text of the measure itself or that of any written implementing measures. The present dispute illustrates this point."78

3.48.
We also note the observation made by the arbitrator appointed under Article 21.3(c) to determine the reasonable period of time for implementation in this dispute that "I am … conscious of the fact that any legislation adopted by the United States will inevitably, as the Appellate Body Report demonstrates, bear on questions of public moral and public order".79
3.49.
In sum, our conclusion that Antigua's counterfactual is not reasonable, taking into account the circumstances of this dispute, is not modified by our consideration of Antigua's arguments in respect of the nature of its rights under Article XVI of the GATS and of Article 3.7 of the DSU. Rather, our analysis of these elements confirms to us that, as a matter of law, the United States may have had a range of WTO-consistent means at its disposal in order to implement the recommendations and rulings in this dispute, not limited to a complete opening of its remote gambling and betting services market.

2. Alternative counterfactual

3.50.
Having determined that the counterfactual used by Antigua to estimate the benefits accruing to it in this dispute did not accurately reflect such benefits, we must now determine what would constitute a reasonable counterfactual, in the circumstances of this dispute.
3.51.
We first recall our determination above that a reasonable counterfactual for the purposes of this dispute would have to take into account the US public policy objective of protecting public morals and public order.
3.52.
The United States identified, as a potential means of implementation of the DSB's recommendations and rulings in this dispute, a complete prohibition on the provision of remote gambling services in the United States.80 Antigua appears to agree that this could constitute a form of compliance.81 However, neither party has suggested that this scenario should form the basis of the calculation of the level of nullification or impairment of Antigua's benefits for the purposes of these proceedings. We therefore do not consider it further.
3.53.
The other alternative scenario that has been presented to us is that identified by the United States, which assumes that the United States would provide unrestricted market access for remote gambling and betting services only in respect of horseracing gambling and betting.
3.54.
The United States describes this scenario as the "most likely", and explains that the specific problem found with the US measure at issue was with respect to the limited issue of the regulation of remote gambling on horseracing, and in particular on the United States inability to demonstrate an absence of discrimination under the chapeau of GATS Article XIV with respect to remote gambling services on horseracing. In the view of the United States, nothing in the DSB recommendations required it to clarify or change its measures with respect to other forms of remote gambling.82 Moreover, according to the United States, nothing in the DSB recommendations and rulings, nor in the chapeau of GATS Article XIV, would require it to treat all types of remote gambling identically, and a counterfactual that allowed remote gambling on horseracing, but disallowed other types of remote gambling, is entirely plausible.83
3.55.
Antigua, however, considers that such a scenario would not bring the United States into compliance. Antigua argues in particular that in the underlying proceedings, the United States alleged that it was entitled to maintain the offending measures by application of Article XIV of the GATS and that it prohibited all remote gambling because, by its very nature, remote gambling presented risks and other pernicious features that were not subject to amelioration through regulation or any other means.84 In Antigua's view, the United States can therefore not now assert that it is "necessary" to prohibit all remote gambling to protect its citizens if it expressly allows remote gambling in any context. Antigua also notes that the United States never argued that remote gambling on horseracing was somehow "safer" than other types of remote gambling, or that Antiguan service providers should be allowed to do only what domestic providers can lawfully do.85 Antigua further considers that it is not for the Arbitrator in this proceeding to assess whether such hypothetical measures can bring the United States in compliance with its obligations under the GATS.
3.56.
In approaching this part of our determination, we are mindful that our mandate in these proceedings is not to determine the consistency with the WTO covered agreements of hypothetical compliance measures. We also note that, as we have stated above, whether the scenario at issue is the "most likely", as described by the united States, is not pertinent as such in our determination. Rather, as determined above, we must assess whether the proposed scenario could constitute a "plausible" or "reasonable" compliance scenario, in the circumstances of the dispute, for the purposes of calculating the level of nullification or impairment of benefits accruing to Antigua in the dispute.
3.57.
We first observe, in this respect, that, as the United States noted, the specific aspect of the United States' measures that was found to lead to an arbitrary discrimination within the meaning of the chapeau of Article XIV of the GATS was the treatment of remote gambling and betting in respect of horseracing. This was the sole basis upon which the Appellate Body determined that the US measures were not justified under Article XIV of the GATS. Specifically, the Appellate Body found that:

"We find, instead, that the United States has demonstrated that the Wire Act, the Travel Act, and the IGBA fall within the scope of paragraph (a) of Article XIV, but that it has not shown, in the light of the IHA, that the prohibitions embodied in these measures are applied to both foreign and domestic service suppliers of remote betting services for horse racing. For this reason alone, we find that the United States has not established that these measures satisfy the requirements of the chapeau."86 (underlined emphasis added)

3. Separate opinion

3.62.
One of the arbitrators is unable to agree with the analysis and conclusions reflected in paragraphs 3.41 to 3.61 above. In the view of this arbitrator, it was not unreasonable for Antigua to assume, in the circumstances of this case, a counterfactual scenario under which the United States would provide unrestricted access to its remote gambling and betting market.
3.63.
In the view of this arbitrator, it is appropriate to refer, as a starting point, to the findings and conclusions of the panel and the Appellate Body in this case. Specifically, the Appellate Body has determined that, although the three federal laws at issue are measures "necessary to protect public morals or public order", the United States had not demonstrated that they were applied in accordance with the requirements of the chapeau of Article XIV of the GATS.
3.64.
The Appellate Body made this determination "in light of" only one specific discrimination that it identified in the application of the measures. However, this does not necessarily imply, as the United States suggests, that the specific problem found with the US measure at issue was restricted to the "limited issue of the regulation of remote gambling on horse racing"88. Rather, the overall conclusion of the Appellate Body was that the measures at issue (rather than simply the discriminatory treatment provided in respect of horseracing) were, as a result, not justified under Article XIV of the GATS. As the compliance panel in this case noted:

"It is true that the Appellate Body found that the United States had demonstrated that the measures at issue were 'justified' under paragraph (a) of Article XIV of the GATS. However, this was not a finding on Article XIV in its entirety. The Appellate Body expressly confirmed that Article XIV contemplates a 'two-tier analysis' – first, under one of the paragraphs of Article XIV, and then under the chapeau. There was no finding that the measures were consistent with the chapeau or with Article XIV in its entirety nor, hence, with the United States' obligations under the GATS, and there is no concept recognized under the DSU of provisional or transitional consistency with a recommendation of the DSB."89

3.65.
Article XIV of the GATS allows Members to maintain certain trade-restrictive measures that would otherwise be inconsistent with their obligations under this Agreement, in order to fulfil certain legitimate objectives. However, these measures are required to be applied in such a manner that they do not result in unjustifiable discrimination or disguised restrictions to trade. In other words, the protection afforded by the exceptions under the various paragraphs of Article XIV of GATS is conditional upon an application of the measures at issue that does not turn them into instruments of undue discrimination or protection. In this case, the United States was not able to justify that it applied the measures consistently with these requirements. Furthermore, the compliance panel report also identified further aspects in the application of the measures that could reflect a discriminatory application of the measures at issue, although it made no findings in respect of such aspects.90
3.66.
This arbitrator agrees with the determination in paragraph 3.41 above that it is appropriate to take into account the specific circumstances of the case in order to assess whether the counterfactual proposed by Antigua accurately reflects the benefits accruing to Antigua under the GATS that have been nullified or impaired in this case. However, this arbitrator considers that, in the circumstances of this case, it was not unreasonable for Antigua to base its counterfactual on an assumption that the United States would open its market to the cross-border provision of remote gambling services. Such an assumption is fully compatible with the nature of the Appellate Body's findings in this case, as analysed above. In fact, as noted above, the United States itself acknowledges that this scenario would constitute compliance with the DSB recommendations and rulings in this case.
3.67.
By contrast, it is not clear that the alternative counterfactual scenario envisaged by the United States, which involves a partial opening of a limited segment of the market, constitutes a more reasonable assumption in the circumstances of the case. In particular, assuming, for the sake of argument, that such a scenario would constitute compliance with the recommendations and rulings of the DSB in this case (something that is not within our mandate to determine), it is not clear how the United States proposes to reconcile the protection of public morals or public order with the opening of one segment of the market (horseracing).
3.68.
In light of the above, this arbitrator is not persuaded that the United States has demonstrated that Antigua's proposed counterfactual was unreasonable in the circumstances of the case.
3.69.
This does not imply, however, that this scenario would constitute the only way in which the United States could have complied with these recommendations and rulings. As noted above, Members are free to choose the WTO-consistent means by which they will comply with DSB rulings, and there may be a number of ways in which the United States could have complied, and still could comply, with the rulings in this case. Nor does this determination imply that US policy objectives should not be taken into account. On the contrary, this arbitrator agrees with the determination of the majority that a reasonable counterfactual must take into account the US policy objectives, as stated in para. 3.45 above. In that sense, nothing would require the US to abandon its objective of protecting public morals or public order, under Article XIV(a) of the GATS, in implementing, in the future, the recommendations and rulings of the DSB. Indeed, it is quite conceivable that the United States could have found ways in which to address these concerns while protecting such interests. Other WTO Members have chosen to open their market to remote gambling, and it would not be reasonable to assume that such Members do not also have similar policy objectives.
3.70.
This arbitrator also notes Antigua's argument that the objective of inducing compliance with the rulings of the DSB is to be taken into account in the Arbitrator's assessment of Antigua's counterfactual. In the view of this arbitrator, the objective of inducing compliance cannot lead to suspension being authorized in excess of equivalence with the level of nullification or impairment of benefits, as foreseen in Article 22.6 of the DSU, and this must remain the benchmark against which the proposed level of suspension is assessed. At the same time, however, in a situation where different means of compliance might form the basis of a counterfactual in order to determine the level of nullification or impairment of benefits, the complaining party would not be prevented from selecting a counterfactual that may lead to a higher level of nullification or impairment than others, provided that such counterfactual is reasonable.
3.71.
In should be borne in mind, in this respect, that the Member concerned has had the opportunity to comply with the rulings at issue, and that the very reason for the existence of countermeasures under the DSU is to induce compliance with the covered agreement that has not taken place within the period foreseen in the DSU. In these circumstances, the complainant is by necessity obliged to make certain assumptions as to how compliance might have taken place. This has no implications, however, as to the means through which the Member concerned might actually choose to comply in the future, possibly including, in this case, through the adoption of measures that involve a restriction to trade fully consistent with the provisions of Article XIV of GATS.
3.72.
Finally, although this arbitrator disagrees with the conclusions reached by the majority of the Arbitrator in respect of the choice of counterfactual, it nonetheless sees no added value in pursuing the calculations on the basis of Antigua's counterfactual, which the majority of the Arbitrator does not consider to be reasonable. Accordingly, the assessments and calculations in the following section reflect the common view of the Arbitrator as to the calculation of the level of nullification or impairment, on the basis of the counterfactual proposed by the United States.
3.73.
This arbitrator also wishes to highlight that, even if calculations had been pursued on the basis of Antigua's proposed counterfactual, a number of parameters would have had to be taken into account. In particular, due account would have had to be taken of the fact that, in a scenario under which the United States would open the entirety of its remote gambling market, Antigua would face competition not only from other foreign operators but also from domestic US operators that currently also have no access to the market.

D. CALCULATION OF THE LEVEL OF NULLIFICATION

3.74.
In light of our determinations above, we now proceed with an assessment of the exports "lost" by Antigua on the basis of a counterfactual under which the United States would have provided unrestricted market access to Antiguan operators only in respect of horseracing gambling and betting.
3.75.
For this purpose, we first review the approach of Antigua in respect of the calculations and methodologies, and then, as necessary, the approach of the United States, before turning to the approach that we have adopted, taking into account our review of the parties' approaches.

1. Review of the approach proposed by Antigua

(a) Description of the approach proposed by Antigua

3.76.
In its calculations, Antigua assumes complete access for all of its remote gambling services to the United States market as the appropriate counterfactual.91 The core data used by Antigua comes from the Quarterly eGaming Statistics Report of the private gambling consulting group "Global Betting and Gaming Consultants" (GBGC) and from its underlying database. In the absence of bilateral data, Antigua uses data from the GBGC database on global remote gaming revenues for Antigua for the years 1999-2006 and projections for Antigua for the years 2007-2012 as a starting point.92 Antigua considers its global remote gambling revenues to be a good proxy of its revenues from the US, claiming that at least 80 per cent of the revenue of Antiguan remote gaming operators has been from customers in the United States.93 The second set of key data in Antigua's approach are historical data on global remote gaming revenues of the whole industry (not only the Antiguan industry) for the years 1999-2006 and projections for the years 2007-2012 gathered from the GBGC's Quarterly eGaming Statistics Report. The two data set together allow Antigua to calculate its global market share.94
3.77.
Antigua is of the view that its remote gaming industry drove the growth of the remote global gaming markets until 2002. Antigua assumes that both the historical data and the projections by GBGC reflect the adverse impact on Antigua of the US measures as of a certain date and that the difference between actual revenues at that date and actual revenues after that date is principally due to the US measures.95 Hence, Antigua seeks to determine how its remote gaming revenues would have developed in the absence of the US measures, in order to compare those counterfactual revenues to its actual revenues beginning 4 April 2006, i.e. after the expiry of the reasonable period of time for the United States to come into compliance.
3.78.
Antigua presents three "models" to calculate the counterfactual Antiguan remote gaming revenues per annum starting from the historical time series and following the same basic framework. For its Constant Market Share model, Antigua takes its global market share in 2003 (21 per cent) as a basis for constructing the counterfactual. It assumes that absent the US measures, it would have retained that market share. Hence, it multiplies the global remote gaming revenues of the whole industry for each year 2004-2012 with the 21 per cent share in order to arrive at annual counterfactual Antiguan remote gaming revenues. For the years 2006-2012, it then subtracts the actual/projected global remote gaming revenues for Antigua from the counterfactual numbers to obtain the Antiguan annual global remote gaming revenues lost beginning 4 April 2006 (in order to take into account that the time period after the reasonable period of time is less than a year it reduces the calculated annual 2006 amount by about one quarter). It then multiplies the average annual losses by an assumed output multiplier ("small island economy multiplier"). In the absence of better estimates, it proposes to use a multiplier of 1.41, established for the tourism industry in Barbados.96 Finally, it discounts the "trade loss" so obtained for each year 2006-2012 to the year 2007, utilizing a 9.38 per cent discount rate. This rate is the median weighted average cost of capital for SIC Code 7999, which includes the gaming industry, according to the Ibbotson Cost of Capital 2006 Yearbook. Antigua, finally, takes the simple average over those 6¾ years yielding an average annual global remote gaming revenue loss over the 2006-2012 period of US$3.45 billion.97
3.79.
The other two models proposed by Antigua only differ in the way in which the annual counterfactual Antiguan remote gaming revenues are calculated. Rather than assuming a constant share of the global market, Antigua proposes, in the Constant Growth Rate Model, to take its actual global remote gaming revenues of 2001 (US$2,392 billion as per GBGC) and proceed in two steps to determine the counterfactual amounts for each following year. For the years 2002-2006, Antigua assumes that this amount would have been unchanged. For the years, 2007-2012 it assumes that it would grow at an annual rate of 8.7 per cent. The latter growth rate constitutes the compound annual growth rate of global remote gaming revenues for the whole industry excluding Antigua. Antigua then proceeds in the same manner subtracting the actual/projected global remote gaming revenues for Antigua from the counterfactual numbers to obtain the Antiguan annual global remote gaming revenues lost beginning 4 April 2006. After applying the multiplier, discounting and averaging, it obtains an average annual global remote gaming revenue loss of US$2,403 billion.98
3.80.
Finally, under the fixed revenue model Antigua simply presumes that it would have retained its actual global remote gaming revenues of 2001 (US$ 2,392 billion as per GBGC) in each following year until 2012. Again, these counterfactual amounts are reduced by the actual/projected global remote gaming revenues for Antigua to obtain the Antiguan annual global remote gaming revenues lost beginning 4 April 2006. After applying the multiplier, discounting and averaging, it obtains an average annual global remote gaming revenue loss of US$1,614 billion.99
3.81.
Apart from the counterfactual, the United States takes issue with other aspects of Antigua's methodology, in particular its underlying data, which it considers as "inherently unreliable" and as providing "no useful information in this arbitration".100 In the following, we will review the arguments by the parties on each of the major concerns raised in relation to Antigua's proposed approach, followed by an assessment by the Arbitrator.

(b) Nature of GBGC data, data gathering methodology and data inconsistencies

(i) Arguments by the parties

3.82.
The United States notes that only insufficient explanations of GBGC's data gathering and processing methodology have been given and that the data contains internal inconsistencies and has been substantially revised from one quarterly report to the next. To begin with, the United States states that it has been unable to evaluate the basis for the data and the methodologies used by GBGC. The United States notes that the "Quarterly eGaming Statistics Report" by GBGC was not submitted by Antigua to the Arbitrator with its methodology paper. The United States was only able to identify a report entitled "Quarterly Data Report", which was on sale on the GBGC website.101 The United States claims that it is not clear if GBGC produced this report at the behest of Antigua for the purposes of the dispute and, hence, whether it may contain self-serving allegations.102
3.83.
The United States continues to consider the GBGC data unreliable on methodological grounds even after submission by Antigua of first an extract from the May 2007 release of the GBGC Quarterly eGaming Statistics Report along with an explanation by GBGC of some of its contents103 and, thereafter, of the full May and October 2007 releases of the Report as well as a letter by GBGC explaining its methodology in gathering the data.104 The United States observes that GBGC in its cover letter admits that during the "early years" (presumably 2000, when it started selling gambling data, and following years) "the information available was more limited." The United States holds that such information must have been very "limited", since, even for current figures, the only source of data GBGC uses for non-public operators is guidance provided by unspecified contacts in the industry and information from private companies with which GBGC has consultancy arrangements.105 In that regard, the United States believes further explanation is required on how the information is requested, whether sampling is used, and if so, how.106
3.84.
The United States also sees the extent of data revisions by GBGC as a sign of methodological problems. It points to GBGC's statement that it constantly revises its numbers and its admission that "previous editions" of its estimates have "little relevance." According to the United States, this suggests that the figures, notably the ones for Antigua in the early years (i.e. around the year 2000), could be substantially different in any future revision of the data, and that the relevance of the available GBGC in that regard is therefore questionable. As an indication of the severity of such revisions, the United States notes the 35 per cent downward revision of North American online gambling revenues from the May 2007 to October 2007 releases of GBGC's eGaming Report.107
3.85.
The United States also points to what it perceives to be data inconsistencies. It notes that Antigua's asserted revenues for 2001 (based on the May 2007 release of GBGCs eGaming Report) exceeds the figure for all revenues from players in North America for the same year (according to the October 2007 release of the eGaming Report).108 It also notes the inconsistency between regional statistics in the May 2007 release of the GBGC database and the revenue figures of Antigua, Curaçao and Cost Rica, which taken together surpass the total for the entire Central American and Caribbean region in certain years.109
3.86.
The United States notes that the breakout for Antigua is not contained in GBGC's regular eGaming Statistics Reports. Since it appears to have been prepared especially for Antigua for the purposes of the arbitration, the United States concludes that the Antiguan data could not have been commonly relied upon by others in the online gambling industry110 (and, hence indirectly be confirmed in the market place). The United States also observes that the more detailed analysis GBGC purports to have undertaken as of 2005 (reviewing the information base from 2003 onwards) seems to focus on poker gambling sites rather than sportsbetting, as the main gambling activity by Antiguan operators.111
3.87.
Antigua replies that the GBGC data is publicly available in the sense that any one who is willing to pay for the information may have access to it.112 It recalls that GBGC and its data existed before Antigua's use of its data in the estimation of the level of nullification or impairment. It also explains that the Antiguan data has not been established for Antigua, but provided by GBGC to Antigua upon its request from the underlying database, but that it is not a standing feature of GBGC's Quarterly eGaming Statistics Report, unlike some of the other regional and global statistics.113
3.88.
Antigua is of the opinion that, in addition to the letter from GBGC explaining GBGC's data gathering methodology114, the best proof of the reliability of the GBGC data is its widespread use in the global marketplace. Antigua notes that as an independent company GBGC relies upon industry participants and other parties for its revenues. It has every reason to be as accurate as it can possibly be in order for its data to be considered reliable and useful by its customers. Antigua doubts that GBGC would be long in business otherwise.115 Antigua provides a list of institutions that have used GBGC data in support of its good credentials. It makes specific reference to GBGC data used in a report commissioned by the European Commission.116
3.89.
Concerning data revisions and inconsistencies, Antigua notes that GBGC refines its data on an ongoing basis (including the allocation of revenues to particular jurisdictions).117 It observes that adjustments for the changing value of the United States currency can lead to relatively minor inconsistencies among data sets, particularly those compiled at different points in time.118

(ii) Analysis by the Arbitrator

3.90.
At the outset, we note that questions of data availability and reliability present themselves as a particular challenge in this proceeding. In considering data quality, we have to make an evaluation in both absolute and relative terms. In other words, in analysing the strengths and weaknesses of the GBGC data, we have to keep in mind the question whether alternative sources of data put forward by parties really constitute an improvement, given that these may suffer from certain deficiencies themselves.
3.91.
We agree with the United States that the information concerning data gathering and processing methodology provided to Antigua by GBGC leaves much to be desired. At the same time, we think that, in the absence of more detailed background information from GBGC, the wide use of its data in the market place can be seen as giving some indirect support to its figures, since it implies a certain degree of acceptance amongst observers familiar with the industry. It may also show that alternative data sources are somewhat hard to come by. We are also not convinced why GBGC would have an interest in providing wrong numbers to its clients who may base some of their business decisions on that data and would quickly discard GBGC as a reliable provider in case of persistent flaws.
3.92.
We acknowledge that some of the aggregate figures that are standing features of GBGC's Quarterly eGaming Statistics Reports are perhaps more reliable than the country data and can certainly be useful to our analysis. The former may be assumed to be periodically subjected to the judgement of a broader range of users, to rely to a larger extent on publicly reported information and to obviate the need to allocate revenue on the basis of additional assumptions. We take it that the revenue data for Antigua may be affected by such problems, especially in the early years. We note in this regard, that the revisions for Antigua undertaken by GBGC between the May and October 2007 releases of the Quarterly eGaming Statistics Report are quite substantial for the years 2000-2002 (downward corrections of between one third and one fourth), less so for 2003, while the years 1999 and 2004-2006 remain virtually unchanged. We also note that the data inconsistencies identified by the United States between regional aggregates and country data, including on Antigua, in the May 2007 release of the Quarterly eGaming Statistics Report appear to have been removed in the October release.119 While these revisions are severe, they are perhaps not enough to discredit the GBGC data, especially since the changes of the Antiguan data for more recent years have been less dramatic. However, if the GBGC data were to be used, we would take some comfort in obtaining additional confirmation of the order of magnitude of Antigua's revenues from remote gambling over those years.

(c) Relationship between GBGC remote gambling revenue data and Antigua's exports of remote gambling services to the United States

(i) Arguments by the parties

3.93.
The United States questions to what extent Antigua's remote gambling revenues according to GBGC constitute exports of remote gambling services from Antigua to the United States.120 The United States notes that the GBGC data at most shows gambling revenue collected from persons in North America by operators located and/or licensed somewhere in the Western Hemisphere. The data do not purport to show what revenues actually flow back to any jurisdiction.121
3.94.
Three issues appear to be involved: Firstly, the GBGC data does not provide for Antigua's gambling revenues from the United States, but only for its supposed remote gambling revenues from the world or for the world's remote gambling revenues from North America.122 Secondly, the United States considers the GBGC methodology not to be sufficiently clear in relation to how revenue estimates were calculated and how country shares of supply and demand or specific gambling products were determined. As a result, it would be unclear whether revenues were allocated on the basis of where licensing existed, operations were located or company headquarters were located.123 The United States holds that the fact that a website is licensed by Antigua does not necessarily indicate where the website's operations are located.124 Websites may obtain licenses from multiple jurisdictions for marketing or other purposes, while operating out of a different jurisdiction altogether.125 Consequently, the United States considers it is unclear how revenues from companies with licenses in several jurisdiction are allocated to a specific country. The United States notes that GBGC allocates revenue based on the operation of the server, but that no explanation is given of underlying assumptions in relation to GBGC's statement that "given the complexity of the structures of some of the organizations involved in the online gambling industry some assumptions have to be made by GBGC in this regard."126 Thirdly, the United States submits that for gambling revenue actually to be considered an export of Antigua, that revenue must be generated from internet operations actually located in Antigua and must be returned to those same operations in Antigua. Otherwise, such revenue would not be associated with an Antiguan service export, and any reduction in that level of revenue cannot be considered nullification and impairment suffered by Antigua.127 It notes that Antigua has conceded that many of the gambling revenues associated with websites licensed by Antigua are held in foreign banks by foreign nationals and are never returned to Antigua.128 The United States goes on to explain that Antigua may act as a point of sale for business owners in third countries. Its exports would be measured by the stream of payments it receives from foreign firms for its selling services, which would be a modest fraction of the claimed revenues.129
3.95.
On the first issue, Antigua provides a letter from GBGC stating that it estimates the historical share of Antigua's remote gambling revenues from the United States to amount to between 80 and 90 per cent of its total revenues. Antigua also notes that certain Antiguan operators in conversations have estimated their US revenues to lie between 90 and 95 per cent and that Antigua's Directorate of Gaming of Antigua uses an 80 per cent figure. Antigua also refers to some publicly listed companies reporting estimates in a similar range.130 On the second aspect, Antigua notes that remote gambling revenues are allocated among countries by the location of the server that processes the wagers, a methodology used not only by GBGC131, but also in the 2006 Report commissioned by the European Commission. As far as the third point is concerned, Antigua holds that the United States is wrong in arguing that revenues generated by an Antiguan operation do not count as Antigua's exports if they accrue to "foreign nationals" or if the revenues are held in "foreign banks". It notes that for the purposes of measuring GDP (as the more relevant concept in measuring revenues generated in a particular jurisdiction compared to GNP), it does not matter who owns Antiguan companies or where the profits go, as long as the remote gaming revenues are earned through the operation of computer services in Antigua.132 Antigua also notes that all remote gambling revenues should be considered exports, as there is no domestic market for these services.133

(ii) Analysis by the Arbitrator

3.96.
Concerning the question to what extent Antigua's global revenues from remote gambling constitute exports by Antigua to the United States of such services, we, first of all, regret that the GBGC revenue data for Antigua is only given at the global level and not by destination, specifically in regard to revenues from the United States. We observe, in that context, that the services trade statistics referred to by the United States are not available on a bilateral basis either. The anecdotal evidence provided by Antigua, while unsatisfying, provides at least some indication of the United States share in Antigua's remote gambling revenues.
3.97.
As far as multiple licences in several jurisdictions are concerned, it seems reasonable to us that revenues be allocated across countries on the basis of where the service is produced (i.e. wagers are processed, etc.). In the absence of any evidence to the contrary, the location of the server seems to us to be an acceptable proxy for determining where the service is produced. This is what GBGC, as a general rule, purports to be doing, although we are unable to verify from the information provided to us how thoroughly this is done and cross-checked for consistency. However, the evolution of revenues for Antigua and some competitors in the region over the years gives a vague indication that GBGC does account for shifts in revenues following possible shifts in server location.
3.98.
Finally, we agree with the United States that for gambling revenue to be considered an export by Antigua, it must be generated from internet operations located there. However, we fail to understand why the United States contends that such revenue ceases to be an Antiguan export if it is not returned to those same operations, for instance because payments are received by a foreign owner or on a foreign bank account. In regard to conceptual issues regarding exports/imports, we refer to the IMF's Fifth Edition of the Balance of Payments Manual published in 1993, noting also that the United States itself uses IMF balance of payments data for its calculations. The Manual clarifies that despite the connotation, the balance of payments is not concerned with payments, as that term is generally understood, but with transactions (para. 26). A transaction is defined as an "economic flow that reflects the creation, transformation, exchange, transfer, or extinction of economic value and involves changes in ownership of goods and/or financial assets, the provision of services, or the provision of labour and capital" (para. 13). The Manual states that the balance of payments records transactions between residents and non-residents (and not between nationals and foreign nationals). The concept of residence is based on a transactor's centre of economic interest. In particular, we note that an enterprise is said to have a centre of economic interest and to be a resident unit of a country when the enterprise is engaged in a significant amount of production of goods and/or services there (para. 73). From this we conclude, as suggested by Antigua, that a service "produced" by a resident and provided to a non-resident should be considered an export by that country even if the resident is a foreign national or payment is received on a non-resident bank account. Hence, we believe that such elements cannot constitute conclusive evidence that the remote gambling services that Antigua sells to the United States are in fact "produced" elsewhere, and we have not received any evidence as to where such a production would take place, if not in Antigua.

(d) Relationship between GBGC remote gambling revenue data and Antigua's reported exports and GDP

(i) Arguments by the parties

3.99.
The United States argues that the GBGC data does not comport with data on Antigua's exports and GDP.134 The United States holds that the GBGC data are inconsistent with official economic statistics prepared by international institutions, such as the Eastern Caribbean Central Bank, the International Monetary Fund and the WTO.135 The United States supplies data from all of these sources showing that exports by Antigua of "other services" (ECCB), "other business services" (IMF) and "other commercial services" (WTO), which should include gambling services, are not more than around $50 million per annum in recent years, and hence, of a completely different order of magnitude than the $1 to 2 billion of remote gaming revenue cited by Antigua on the basis of the GBGC data.136 The United States also provides ECCB data on the development of Antigua's GDP over time, again noting inconsistencies, with alleged remote gaming revenue exceeding total GDP per annum and with changes in alleged remote gaming revenues not being reflected in GDP.137
3.100.
The United States acknowledges that exports may exceed the value of a country's GDP, as is the case for e.g. Singapore, where in 2005 goods and services exports were 143 per cent greater than GDP. However, the United States holds that Antigua's claimed net revenues from exported internet gambling services in 2001 in addition to its reported exports of goods and services would exceed the value of its GDP by over 300 per cent or be twice in excess of Singapore, the country in the world with the highest ratio of exports to GDP on the basis of official data. Hence, it concludes that even accepting that exports can and do exceed GDP, the Antiguan claims are so out of bounds relative to the experience of other Members that the GDP comparison may serve to question Antigua's export claims.138
3.101.
The United States further notes that even if Antiguan export revenues from gambling services were unrecorded (claimed to be around $2.4 billion in 2001), Antigua has made no claim that its $483 million of imports in 2001 are underrecorded that may have offset the claimed surges in gambling export revenues.139 Hence, the surplus could only have derived from additional unrecorded value added within the Antiguan economy itself, corresponding to a GDP that would have been almost 360 per cent larger than Antigua's officially reported GDP. The United States considers it to be implausible that such an enormous understatement of the size of the Antiguan economy exists.140
3.102.
The United States adds that the footprint of the alleged massive inflow and outflow of revenues during the 1999-2005 period should have been felt in the rest of the Antiguan economy, and would have been measured by the ECCB and the IMF.141 In particular, the United States observes that according to footnote 46 to Antigua's answer to question 39 by the Arbitrator, the data considered by the ECCB in its balance of payments reporting include "financial statements and foreign exchange records reported by the commercial banks to the ECCB, and ECCB's financial statement." The United States claims that gambling revenues and the corresponding monetary flows as well as uses of these revenues, notably for domestic consumption and imports, should be reflected in the balance of payment accounting based on bank transactions, even if operators systematically withhold information on the level of their gambling revenues. The ECCB balance of payments statistics submitted by Antigua, however, do not show monetary inflows on the order claimed by Antigua.142 The United States also observes that in certain years gambling revenues reportedly declined, e.g. in 2003 by $693 million or an equivalent of 92 per cent of GDP, while GDP was up $40 million.143
3.103.
Lastly, the United States submits that the points pertaining to Antigua's explanation of its prior use of a 10 per cent of GDP estimate are some of the most plausible, and relevant, statements made by Antigua regarding its level of nullification and impairment: In particular, the United States refers to Antigua's statement that this figure was based on estimates of the direct impact of the industry on the country in the nature of salaries paid, domestic rents and purchases and similar direct expenditures; and correlations made between estimated employment and the size of the overall domestic workforce. The United States believes this figure to be a more realistic assessment, since any government would notice economic inflows that dwarfed the rest of its economy, because the spillover effects on consumption and employment would be the main factors in overall economic growth or decline.144
3.104.
Antigua dismisses the comparisons by the United States between its claimed gaming revenues and its officially reported export and GDP data.145 Antigua holds that the ECCB data appears to be the source of both the IMF and WTO statistics cited by the US146 and that the ECCB confirmed that its data did not include revenues earned by operators who were licensed to engage in interactive wagering and gaming, since operators did not report.147 Antigua also disputes the relevance of historical comparisons of trade-GDP ratios, comparisons with jurisdictions, where no remote gambling industries exist, or comparisons with the non-remote gambling sector. It holds that remote gaming is an unprecedented form of trade and that the only relevant comparisons in this arbitration should be between Antigua and similarly situated small economies with a remote gaming industry, such as the Khanawake Territory, for which the GBGC data shows revenues from remote gaming of similar size to those of Antigua.148
3.105.
Antigua also finds fault with the way in which the United States calculates export to GDP ratios. It notes that the United States increases the numerator (Antigua's goods and services exports) by the claimed level of remote gambling revenues, but leaves the denominator constant (ECCB reported GDP). Antigua claims that, in order to do this calculation correctly, the denominator would need to be increased by the value-added of remote gaming to Antigua's GDP. Making some assumptions about profits, as suggested by the US, Antigua claims that, in this manner, its export to GDP ratio would be in line with comparator economies.149
3.106.
Antigua adds that besides unreported revenues from remote gambling, expenditures would also have to be taken into account to determine the true impact on GDP of the remote gambling industry. In particular, Antigua notes that, while some domestic spending by operators, on employee wages, facilities expenses and the like, might find their way into GDP – even if not actually reported by the operators – many would not. In particular, Antigua states that the greatest expenditures of remote gambling operators, namely, financial services, advertising, computers, servers and other hardware and software development, take place primarily if not exclusively outside of the jurisdiction, i.e. large part of "inputs" are imported and a significant portion of profits are expatriated.150 It notes that both revenues and expenditures of Antiguan operators are not taken into account by the ECCB in establishing Antigua's GDP.151
3.107.
Finally, Antigua provides some additional reasons why revenues from remote gambling do not show up in GDP figures. In particular, it states that a substantial portion of the "knock-on" effect of remote gaming revenues in the economy may itself find its way back into the remote gaming sector or into other sectors of the Antiguan economy that are under- or incorrectly reported or under-represented (a not uncommon feature of developing economies). Specifically in regard to the question why the decline of the remote gambling sector does not appear to affect GDP, Antigua states that to some extent activities within the reported sectors of the Antiguan economy influenced by remote gaming revenues have decreased with the decline of the industry; hence, it would not be usual to see a substitution effect within the economy, in essence working to mitigate losses as much as possible.152

(ii) Analysis by the Arbitrator

3.108.
Concerning the relationship between the GBGC data for Antigua and Antigua's reported exports and GDP, e.g. pursuant to IMF statistics, we agree with the United States that (i) Antigua's alleged remote gambling export revenue appear large relative to other exports by Antigua and as a percentage of GDP, and that (ii) even if revenues are not reported, there should be a trace of related transactions in Antigua's balance of payments. On the first aspect, we take it that remote gambling is a relatively recent phenomenon and that that revenue margins, especially for early market entrants, may have been extraordinary initially and led to revenue levels that appear large in comparison to traditional export revenues. We note that in other economies mentioned by the parties, revenues from remote gambling may also be large in comparison to other export revenues and may lead to substantial increases in the trade to GDP ratio. This may be even more so in smaller and less diversified economies traditionally running substantial current account deficits.153 Nevertheless, from the limited information we have, we tend to agree that Antigua's initial revenue figures based on the May 2007 release of the GBGC database appear out of line by all historical standards. The substantial revisions carried out by GBGC for its October 2007 release lends support to this assessment. Even assuming that the October 2007 figures for Antigua in the early years had been correct (although those still appear generous), we have reason to believe that such extraordinary revenues must be a temporary phenomenon and are quickly competed away. We will discuss the question of competition further below, but note already that adjustments even to the October 2007 numbers appear necessary to bring them more in line with other economic statistics.
3.109.
On the second aspect raised in the previous paragraph, we agree with the United States that sizeable export revenues would reflect a considerable value added in the Antiguan economy, and hence corresponding increases in GDP, unless counterbalanced by other positions in the balance of payments, notably imports. In fact, it may well be that the net contribution of the remote gambling industry, despite substantial amounts of revenues, to the external accounts and GDP is small, if, as Antigua notes, important gambling-related imports (notably banking, advertising services etc.) exist and profits are transferred abroad with the industry being mostly foreign-owned. While Antigua claims that both revenues and expenditures are unreported, the United States notes that the ECCB and IMF compile Antigua's balance of payments not only on the basis of company surveys (where remote gambling operators may not respond and, hence, neither revenues nor spending may be reported), but also uses financial statements and foreign exchange records reported by commercial banks, where some of those transaction should show. We sought clarification in this regard from both the ECCB and IMF. In its letter dated 13 December 2007 the IMF confirms that the balance of payments statistics for Antigua and Barbuda, which are compiled by the ECCB and supplied to the IMF, do not separately identify remote gambling services and that the aggregate of personal, cultural and recreational services has a zero entry for all year shown in the IMF's 2006 Balance of Payments Yearbook. Moreover, the IMF notes that the survey report forms used by the ECCB do not indicate where gambling services are separately identified or covered and that the description of balance of payments data compiled by the ECCB and set out in Part III of the Balance of Payments Yearbook does not refer to gambling services. However, the IMF also confirms that transactions related to gambling may be captured in the financial account, because financial transactions are, to some extent, reported by commercial banks, and the transactions of gambling site operators would be an undifferentiated component of higher level financial account aggregates reported by these institutions.154 Likewise, transactions related to gambling, such as imports of business services and repatriation of profits by gambling site operators, should be covered in the Yearbook data, but would be undifferentiated components of higher level aggregates as well. The IMF concludes that it cannot determine how comprehensively the data are captured in practice. Hence, we are confident that it would be equally difficult, if not impossible, for us to make an educated guess on balance-of-payments entries of Antigua, which, in addition, over recent years, do not show any unusual magnitudes or movements.

(e) Absence of direct evidence of revenues of remote gambling operators licensed by Antigua and comparison of GBGC remote gambling revenue data with revenue figures from other sources

(i) Arguments by the parties

3.110.
The United States notes that that Antigua claims to have exempted operators licensed by Antigua from any form of official reporting requirements155, but then admits that financial statements for gambling companies with operations located in Antigua, which could give support or refute the GBGC data, are actually available to the government.156 The United State s holds that Antigua should not be rewarded for failing to disclose such information, which could have been provided with redactions or in summary tables for privacy concerns or fears of prosecution. It notes that Antigua does not even explain whether it has reviewed such financial statements, or whether the data in those statements is in accord with the GBGC estimates. The United States draws the conclusion that the actual data does not support Antigua's claims pursuant to the GBGC data, or else it would have been used, and that adverse inferences may be drawn, when a party to a dispute declines to provide relevant information.157
3.111.
Antigua confirms that Antiguan remote gaming operators are subject to a number of reporting obligations. This includes the obligation to provide the Directorate of Gaming of Antigua, at least once per year, access to and examination of the books, accounts and financial statements of each licence holder.158 However, Antigua declines to disclose financial information from Antiguan operators, principally for the reason that the United States has used filings of public companies as a basis for criminal complaints in the past.159 It only cites gross profits of US$895.7 million in 2005 that five Antiguan remote gaming operators had confidentially reported to the Antiguan Directorate of Offshore Gaming (Division of Gaming, Financial Services Regulatory Commission). This information is contained in a letter by the Director of Gaming to Mr. Mark Mendel, dated 1 October 2007, written for the purposes of this proceeding.160
3.112.
Instead, Antigua provides alternative sources that provide data for the remote gambling market of a similar order of magnitude as the GBGC data, including the United States General Accounting Office Report on "Internet Gambling: An Overview of the Issues".161 Antigua also notes that historical revenue estimates for the global remote gaming industry by Christiansen/Cummings Associates, Inc. (apparently, the name was later changed into Christiansen Capital Advisers, CCA), a US-based consultancy, were similar to GBGC data.162 Upon our request concerning the evolution of the size of the US market, Antigua specifically has provided data on US gross gambling revenues and, in addition, US internet gambling revenues from CCA published for the years 1995-1997 (gross revenues only) and 1999-2006 in the gaming publication "International Gaming & Wagering".163 Data on the US market is also included in GBGC's Quarterly eGaming Statistics Report (May and October 2007 releases) on annual remote gambling revenues broken down by certain regions and countries, including the US, and by four types of gambling activities. Antigua notes that the GBGC and CCA data on the US remote gaming market for 2006 are also on the same order of magnitude.164
3.113.
Antigua also provides revenue and employment data for the years 2001 (for some companies 2003) to 2006 from publicly listed companies which are active in the online gambling business and some of which are licensed in Antigua.165 Antigua observes that there are wide disparities between the companies as far as revenues per employee are concerned. This may be explained by the varying degree of customer service in an internet-based business, such as support centres for customer enquiries or wager acceptance over the telephone. However, Antigua concludes that the reported revenues per employee multiplied by the number of employees are broadly supportive of the GBGC revenues reported for Antiguan operators.166 In order to verify this, Antigua provides information from the Directorate of Gambling, inter alia, on the number of employees directly employed by licence holders for the 1998 (2000 for employment) to 2007 time period (which replaces Antigua's earlier estimate of 3000 employees in 1999 that had taken into account both direct and indirect employment).167 Antigua also refers to a report commissioned by the European Commission in 2006, which provides estimates of total employment and revenues in the EC remote gaming industry resulting in approximately US$1 million of revenues per employee in 2006.168 It submits that such figures as well as comparisons to other e-commerce companies show how much revenue can be generated by a relatively small number of employees in this highly automated, technical industry.169
3.114.
The United States claims that the CCA data should not be given any weight, since Antigua did not provide copies of the cited CCA documents nor information as to CCA's sources. Moreover, it claims that figures provided on the CCA website apparently do not match Antigua's description of CCA figures.170 Upon our request for data on the size of the US remote gambling market, if not at the country level, but at least for individual states, the United States notes that, since the gambling services Antigua wishes to provide are currently criminal in the US, there are no official US statistics on such activities. It also states that it does not have information on any specific state's statistics on remote gambling.171 However, it provides consumption data (from statistics of the Bureau of Economic Analysis (BEA) of the US Department of Commerce on personal consumption expenditures by type of product) for the non-remote gambling market, including annual time series data from 1995 to 2006 on pari-mutual net receipts, 90 per cent of which are believed to be accounted for by betting on horseracing.172
3.115.
The United States contests that the revenues per employee data for various companies provided by Antigua can be used to estimate Antigua's remote gambling revenues, since the sources of this data were not included with Antigua's answers173 and since the range of values presented is large and refers to "leading remote gaming companies", supposedly at the exclusion of less successful online and gaming companies that would have revenues per employee at much lower values.174
3.116.
The United States concedes that the data provided by Antigua on the number of employees engaged in offering online gambling services appear less unrealistic than its revenue predictions on the basis of GBGC data. It also notes that employment figures are significantly lower than Antigua's previously stated number of 3000 employees. The United States submits that this employment data is more useful than Antigua's revenue estimates in assessing the level of nullification and impairment. The United States notes that according to Antigua, in the last five full calendar years (2002-2006), Antigua had an average of about 450 employees, although it is not known how many worked on gambling tied to United States citizens, or on a particular gambling activity, such as horse racing.175 However, the United States also observes that the numbers of employees increased between 2002 and 2005, while during the same time period the GBGC breakout for Antigua's remote gambling revenues shows a decline, calling further into question the accuracy of the GBGC figures.176

(ii) Analysis by the Arbitrator

3.117.
We think that the provision of information from Antigua's remote gambling operators directly, in particular financial statements, would certainly have been helpful. However, it is not entirely clear whether the financial reporting requirements by Antiguan remote gambling operators allow for this information to be retained within the government. Antigua only states that these companies have "to provide the Directorate, at least once per year, access to and examination of the books, accounts and financial statements" (emphasis added).177 We consider that we are not in a position to pass judgement on the use of that sort of information in criminal proceedings in the United States in the past, nor do we need to. Nevertheless, we reckon that it should be possible to produce some kind of summary information even from a limited or informal information base on those companies that also takes account of the sensitivities involved in relation to company-specific data. While Antigua has not provided this type of information, we recognize that it has made efforts to substantiate the order of magnitude of the GBGC data in other ways. In particular, we appreciate the collection of data on revenues per employment from publicly listed companies and the provision of historical employment data in the Antiguan remote gambling sector directly from Antigua's Directorate of Gaming. We note quite some variation in the data between companies and years. Although this information may only allow for a rough estimation of average revenues, it provides at least for another reality check of the GBGC data. We also note that the United States considers the employment data by Antigua to be "less unrealistic" than the GBGC revenue data and "more useful" in assessing the level of nullification and impairment.178
3.118.
We also note that both Antigua and the United States have provided figures that can be used to model the development of the US market, which we deem to be an important element in the construction of a more realistic counterfactual.

(f) Application of a multiplier to the remote gambling revenue data

(i) Arguments by the parties

3.119.
The United States holds that there is no basis under the DSU for "multiplying" trade effects by a factor to reflect other economic effects. Moreover, it observes that the purpose of a multiplier is to reflect effects on internal transactions within the Antigua economy – such as effects on the purchase of Antiguan goods or services by Antiguan gambling operators. As such, the transactions which would serve as the basis for Antigua's suggested multiplier are not lost Antiguan exports, and thus are not properly included in a measurement of Antigua's nullification or impairment of GATS benefits. Moreover, the United States notes that Antigua cannot reconcile (1) its argument that gambling revenue has a 40 per cent carryover effect on the Antiguan economy (as suggested by the multiplier it proposes), (2) its argument that gambling revenue dwarfs the rest of its economy, and (3) the fact that Antigua's official economic statistics (leaving aside the reporting or non-reporting of direct gambling revenue) show no significant "multiplier" effect from the allegedly massive influx of gambling revenue. Even if direct gambling revenue were under-reported, the United States argues that there is no basis for believing that all other economic figures on the Antiguan economy are not simply wrong, but are wrong in the exact way necessary to mask the effects of the allegedly massive increases and decreases of gambling revenue over the 1998 to 2006 period.179
3.120.
The United States also notes that prior arbitrations under Article 22.6 of the DSU calculated the level of nullification and impairment of the benefit under the covered agreement and not some broader measure of overall economic impact of non-compliance.180 It observes that the equivalence standard under Article 22.6 of the DSU would not be respected if just the level of nullification and impairment was increased by a multiplier. It points to Article 22.3 of the DSU (which sets out the principles to be followed in determining the sector or agreement under which retaliation is to be sought) as the appropriate place for a consideration of broader economic elements, and sees no legal basis for this being done in the context of calculating the level of nullification or impairment under Article 22.6 of the DSU.181
3.121.
Antigua assumes that revenue losses by the Antiguan remote gambling industry led to additional losses in other sectors of the economy, lower income and government revenues. It holds that these "indirect" effects need to be taken into account on top of the direct trade effects in calculating the total level of nullification and impairment and that multipliers are frequently used in economic modelling in that regard.182
3.122.
Antigua agrees that there is no precedent in which arbitrators considered a multiplier, but declines that it was not allowable under the DSU.183 It also notes that the Arbitrator in EC – Bananas III (US) contemplated indirect consequences, even if these ultimately were not taken into account, and that multiplier effects were an accepted feature of economic modelling.184 Antigua adds that just because multipliers have not been used in past arbitrations, this should not mean that multipliers cannot or should not be used. It also states that Antigua's status as a developing country Member should be relevant in this regard.185

(ii) Analysis by the Arbitrator

(g) Point in time, when US measures began to affect Antigua's exports of remote gambling services to the United States and loss of market share to competitors

(i) Arguments by the parties

3.124.
The United States notes that Antigua bases its estimates, at least in part, on gambling exports for the years 2001-2002, which it argues were the high point of Antigua's exports of gambling services to the US. The United States agrees with Antigua that historical levels of internet gambling services are indicative of the level of nullification or impairment. It explains that although remote gambling has at all relevant times been unlawful under US Federal criminal statues, enforcement of those federal statutes against foreign operators offering services over the internet has been difficult. It concludes that historical levels of gambling services exports are instructive as to the levels that might exist if remote gambling had never been outlawed in the United States.186 More precisely, the United States "can accept that exports in 2001-2002 are somewhat instructive regarding the (unlawful) market in the United States for remote gambling services".187 However, the United States believes that any other effects that have influenced Antigua's exports of remote gambling services to the United States would need to be factored out from the calculation of the level of nullification and impairment. In particular, the United States believes that "any decline from 2001 levels would also be due to increasing competition from other gambling operators".188
3.125.
Concerning the "turning point" in time, in its answers to questions by the Arbitrator, the United States further elaborates that it agrees that that historical levels are somewhat instructive as to the level of nullification and impairment resulting from United States non-compliance with the DSB's recommendations and rulings, but that it disagrees that the US measures first began to affect Antigua's exports in 2001-2002 and that this is reflected in the GBGC data.189 The United States notes that its measures have remained unchanged throughout the relevant time period.190 The United States notes in this regard that the Cohen prosecution and conviction took place in 1998 and 2000 respectively and that operators from other jurisdictions increased their operations in the US market after 2001, showing that the criminal laws cannot be the cause of any absolute or relative loss of Antiguan market share in the provision of gambling services to US consumers. The United States notes that Antigua's own data shows a growth in exports to the United States by operators in other jurisdictions, showing the historical trend that Antigua is losing market share, and submits that Antigua's contentions about the changing United States enforcement environment were developed after the fact, in order to match the GBGC figures.191
3.126.
Concerning other factors, notably competitors from other countries, the United States is of the opinion that US measures cannot be held responsible for Antigua's loss of market share (especially on a worldwide scale), since US enforcement measures would have affected all providers in the United States market and not only Antigua.192 In addition, the United States refers to the observation in Antigua's methodology paper that the Antiguan Government itself enacted stricter regulations in mid-2001193, although Antigua claims that these regulations were enacted as a result of United States pressure.194
3.127.
The United States then moves on to discuss other factors that may be responsible for Antigua's drop in global market share and should not be attributed to US measures, notably low market entry barriers to internet gambling and the ensuing growth in competition from operators in other locations.195 According to the US, Antigua has not provided any evidence that its remote gaming industry possesses particular attributes that might place its operators at an advantage as compared to competitors in other countries.196 It concludes that if internet gambling is a growing trend, as asserted by Antigua, Antigua would continue to lose market share to other nations.197 More specifically regarding the United States market, Antigua would face competition also from domestic providers, were the United States to legalize internet gambling. In contrast, up to the present, US criminal laws banning remote gambling have been applied rigidly to operators located in the United States while foreign operators have remained outside the reach of United States criminal laws.198
3.128.
The United States observes that Antigua's own data indicates that is has lost market share to gambling service suppliers from other countries.199 Using the GBGC data (May 2007 release)200 the United States points to two time series showing Antigua's position relative to suppliers from other countries: First, the United States observes that Antigua's gambling revenues increased until 2001 and then declined, while world revenues from North America ("online gross gambling yield by player location") increased continually. According to the United States this implies that non-Antiguan suppliers have been entering the United States market, and doing quite well.201 Second, the United States notes that the country-specific GBGC data provided by Antigua, including the data on Costa Rica, Curaçao and Malta, show increasing world revenues for all of these countries with the exception of Antigua.202 The same is true for the South/Central America and Caribbean group in the regional breakdown of remote gambling revenues by operator location, which shows a smooth rising trend. The United States interprets this trend as evidence that the Antigua-specific allocations for 2001-2002 are wrong, that gambling operators diversified their operations by moving servers to other locations in the region and/or that new gambling operators started up in the region. Since the US measures apply equally to operators from Antigua and from other countries, including operators elsewhere in the South/Central America and Caribbean region, the United States concludes that either Antigua never had as large a market share as it asserts, or (if Antigua did have a high, early market share), that share was lost as Antiguan operators moved servers to other countries in the region, or as a result of new competitors located in other countries in the region. The United States also observes that, according to GBGC, both the sports and non-sports betting have been growing in the United States market; hence, the data does not support Antigua's only explanation of how United States enforcement measures supposedly harmed Antiguan operators more than other operators, namely that Antigua focuses on sports betting, as opposed to Internet poker. The United States also notes that there is no reason to believe that the 2002 United States enforcement actions involving attempts to block credit card and PayPal payments to overseas gambling operators would have any greater effects on sports gambling than poker gambling, or any greater effects on Antiguan operators than operators in other jurisdictions.203
3.129.
The United States concludes that Antigua has presented no convincing reason to believe that United States actions could have such a disproportionate and unique effect on gambling service providers in Antigua as opposed to other gambling service providers located in other markets. The United States cites Antigua as admitting that its initially high market share was because it was an early entrant in the market,204 and that its initial market share eroded over time.205 The United States concludes that there is no information to support the conclusion that United States actions had any effect on Antigua's market share, nor is there reason to believe that Antigua would benefit uniquely from any future change in United States regulations.206
3.130.
Concerning timing, Antigua maintains that historical levels of remote gaming services exports by Antigua to the United States prior to 2002 are instructive as to the levels that might exist in the absence of the inconsistent measures.207 According to Antigua, 2002 is the year when measurable damages began to accrue, which is why it does not propose any "lost revenues" for 2001 and earlier.208 Antigua also explains that for its constant market share model it did not use the 2001 value of 59 per cent (but the 2003 share of 21 per cent), since it did not believe that this was a realistic market share number, given the growth in global remote gaming generally and, in particular, the fact that much of that growth was occurring in areas such as poker that were not the major market segment served by Antigua.209
3.131.
Concerning other factors causing the decline in Antigua's exports of remote gambling services to the United States, Antigua dismisses the United States arguments. While it admits that it was inevitable that it lost some market share globally as the result of the growth of poker on the internet and the growth of remote gaming in other parts of the world, Antigua is primarily a United States facing, sports betting jurisdiction and therefore more severely affected by the US measures than operators in other locations.210 Antigua claims that the three highest profile prosecutions of foreign remote gaming brought by United States authorities in the past few years have been brought against operators with Antiguan licensed subsidiaries and substantial Antiguan operations, whereas to date the United States has not prosecuted any remote gaming operators licensed and located in European jurisdictions, such as Gibraltar, Malta and Guernsey, nor has it prosecuted any operators licensed by the Khanawake Mohawk Tribe of Canada. It goes on to state that the 2001-2003 time frame saw a large number of operators move from Antigua to European jurisdictions that were perceived as being less likely targets of United States efforts than Antigua.211
3.132.
Antigua provides a time line of actions by the United States and other major events212 that are each said to have an adverse impact on the Antiguan remote gaming industry.213 As an early entrant in the remote gambling market, Antigua professes its confidence that in the absence of the US measures, it would have stood a good chance of successfully competing against potential new entrants. Antigua attributes its competitive edge to innovative product offerings, experience and particular knowledge of the American sports betting market, location in a time zone convenient to US consumers and tax advantages allowing for larger operating margins. Antigua is convinced that its competitive advantage would also play out against potential US competitors.214 It recommends abstaining from speculations about new entrants and rely on historical information.215
3.133.
Antigua concludes that the decline in the Antiguan industry is solely attributable to the United States actions and hat no adjustments should be made to the counterfactual level of exports for any other possible factors, including the introduction of "stricter regulations" by Antigua itself in 2001, which it claims came as a result of direct pressure from the United States.216 It holds that its fixed revenue and constant growth models produce conservative estimates of counterfactual exports and that its constant market share model already reflects the loss of overall market share that Antigua experienced due to factors perhaps unrelated ‑ or less related ‑ to the activities of the United States government.217

(ii) Analysis by the Arbitrator

3.134.
We note that both Antigua (in two out of three models) and the United States in its alternative approach (described in section III.D.2(a) below) use the year 2001 as the point in time after which the US measures began to affect Antiguan export of remote gambling services to the United States. Also, we do not see any concrete proposal by the parties for an alternative "turning point". While the United States emphasizes that its measures have been in place for a long time before that date, it also acknowledges that enforcement has been difficult against foreign remote gambling providers and that it "can accept that exports in 2001-2002 are somewhat instructive regarding the (unlawful) market in the United States for remote gambling services".218
3.135.
Nevertheless, as foreshadowed above in section III.D.1(d), we must take into account that other factors, over time, in particular an increase of competition, may have affected Antigua's 2001 levels of remote gambling revenues even in the absence of the US measures. In order to not to falsely attribute such declines in revenue to the US measures, we are mindful of the fact that the US measures, in particular the enforcement actions involving the prohibition for credit card companies and systems such as PayPal to make payments to overseas gambling operators, are capable of affecting all providers, and not only those from a particular country of origin. As far as the risk of criminal proceedings is concerned, all remote gambling operators a priori seem to be an equally likely target, even though a number of Antiguan operations may have been subject to prosecution in the past.
3.136.
In particular, we recognize that the GBGC data shows that world revenue from North America has continued to rise and that the individual countries from South/Central America and the Caribbean as well as the region as a whole have also experienced further increases in their overall revenues. It seems fair to us assume that countries in the region (as opposed to say European providers) are not unlikely to be actual competitors of Antigua in the United States market, since they benefit from some of the same advantages (possibly more similar cost structures, time zones and the like). We see no reason to believe that these countries would be affected differently by the US measures than Antigua. Hence, part of Antigua's loss may be a reflection of its loss in competitiveness vis-à-vis its competitors and their increase in revenues at Antigua's expense. In the light of this, we do not think it would be appropriate for us to follow Antigua's proposition that its revenue loss is entirely due to the US measures and that no adjustments to reflect relative shifts in competitiveness need to be made.
3.137.
But we do not believe either that Antigua's loss in market share has nothing to do with the US measures and is entirely due to new entrants and competitors. While Antigua might have lost market share to competition in any event, it may have seen its revenues decline by more than it would have in the absence of the US measure. In trying to disentangle these impacts, we find it less helpful to look at Antigua's global market share, as the United States does in its alternative approach summarized below in section III.D.2(a), which may be affected by developments in other markets and gambling activities, in which Antigua is not present, and in relation to providers, against whom Antigua is not actually competing. Instead, as suggested above, we do believe that it is important to consider Antigua's relative position vis-à-vis competitors that are competing in the US market and are likely to be equally affected by the US measure. If this is so, we may not be able to follow either one of the two extreme (all or nothing) approaches proposed by the parties.
3.138.
An additional consideration regarding competition needs to be made as a function of the counterfactual scenario chosen. As discussed above in section III.C, Antigua's approach presumes that it would continue to provide remote gambling services under the same conditions it faced in 2001, i.e. in the absence of competition from domestic providers in the US. However, if the cross-border provision of remote gambling services were to be legalized, this would likely apply to both foreign and domestic providers. In the absence of any historical experience, any assumption about how Antigua would fare in such a situation belongs to the realm of pure speculation. As a consequence, Antigua's scenario, which completely disregards the possible entry of domestic competitors, seems clearly deficient to us.219
3.139.
For the moment, the only evidence we have for a competitive situation to exist that includes domestic providers is in the remote betting market on horseracing. Antigua's observation that its operators do very little business in horse race betting220, certainly for a range of commercial considerations, could, at least in part, be also related to the intensity of competition in a segment of the market, where domestic and foreign providers already compete with one another.

(h) Comparison of actual and counterfactual exports of Antigua's remote gambling services to the United States

(i) Arguments by the parties

3.140.
The United States agrees with Antigua on the basic concept that the difference between the actual amount of exports that are affected by the WTO-inconsistent measure and the amount of exports in a "counterfactual" scenario, whereby the offending party has brought the WTO inconsistent measure into conformity within the reasonable period of time, typically represents the level of nullification or impairment.221 From parties' original written submissions, hence, it appears that there is agreement that the appropriate point in time for making the comparison between the actual level of exports and the calculated counterfactual is the end of the reasonable period of time, in this case 3 April 2006.222 However, the United States later states that the precise point in time in that regard should be the date of referral of the matter to arbitration, as this was the date when the terms of reference of the arbitrator were established.223
3.141.
Concerning Antigua's projection of industry revenues into the future (2012, until which date GBGC industry projections are available), averaging and discounting to the year 2007224, the United States does not directly address this assumption, but refers to the counterfactual as a hypothetical situation in which the responding party brought the WTO-inconsistent measure into conformity within the reasonable period of time, which appears to preclude consideration of developments beyond that date.225
3.142.
Antigua further elaborates in this regard that in order to determine the level of nullification or impairment it needs to be determined what the annual prospective Antiguan exports of remote gambling and betting services to the United States would be if the United States had come into compliance on 3 April 2006. Antigua is of the opinion that this is not to say that in developing the counterfactual and coming to a determination of the level of those "annual prospective exports" only information as of that date is to be taken into consideration.226 At the same time Antigua argues that while the end of the reasonable period of time marks the point at which Antigua's trade loss begins to accrue, this does not mean that the level of Antigua's nullification or impairment should be determined by reference to a "snapshot" of sorts on that date. It argues that a 3 April 2006 "snapshot" would miss the enactment of the UIGEA in October of that year, which Antigua claims has had a massive adverse impact upon the actual amount of revenues that Antiguan operators are able to earn. This is why Antigua proposes to use projections up to the end of 2012. Antigua argues that using historical data combined with reasoned, professional estimates of developments in the market in the foreseeable future and averaging these figures is a reasonable and generally accepted method for addressing anomalies that might arise in one year as opposed to another. It also notes that averaging has played a role in several previous arbitrations.227

(ii) Analysis by the Arbitrator

(i) Concluding remarks

2. Review of the approach proposed by the United States

(a) Description of the approach proposed by the United States

3.146.
As discussed above, the United States contests the counterfactual proposed by Antigua, namely a measure that would allow for all types of cross-border gambling services.228 It holds that one means to come into compliance would be for the United States to ensure that domestic service suppliers are also prohibited from supplying remote betting services on horseracing as the area where a discriminatory treatment between domestic and foreign service suppliers was found by the compliance panel.229 The United States observes that in the latter case the level of nullification and impairment would be zero, but does not pursue this approach as an alternative to Antigua's calculation.230 Instead, it recalls that the DSB recommendations and rulings were addressed to the discriminatory treatment of gambling on horseracing. From that it concludes that a counterfactual scenario could be taken as a basis for the calculation of the level of suspension of concessions where "the United States were to adopt measures that allow Antiguan operators (and operators in all other WTO Members) to provide cross-border remote gambling services on horseracing".231 On the basis of its proposed counterfactual, the United States seeks to establish what the level of exports from Antigua to the United States of remote betting services on horseracing would be absent any discriminatory treatment.
3.147.
For export data by Antigua, the United States relies on publicly available services trade statistics by the ECCB, IMF and WTO. Finding that the respective statistics from these three sources are of a similar magnitude, the United States takes as a starting point for its own calculations the WTO figure of Antigua's "other commercial services" exports in 2001 ($47 million), which it sees as "a hard cap on the absolute highest figure of Antigua's gambling service exports in 2001"232, since it both comprises exports to the world (not just the United States) and more than just gambling services exports.233
3.148.
The United States then seeks to determine the share of Antigua's remote gambling services exports on horseracing. In the absence of government statistics on remote gambling in the United States (both overall and on breakdowns by type of gambling activity), the United States supplies publicly available government statistics from the website of the Bureau of Economic Analysis (US Department of Commerce) on non-remote gambling in the United States showing that horserace gambling accounts for no more than 7 per cent of total gambling in each of the years 2003 to 2006.234 It then applies the 7 per cent share to the US$47 million of Antigua's other commercial services exports to arrive at US$3.3 million worth of exports. It then corrects this figure for the impact of competition on Antigua's export revenues by reducing it in the same proportion as the fall in Antigua's share of global remote gaming revenue according to the GBGC report235, i.e. by multiplying it with the term 7 per cent (in 2006) over 50 per cent (in 2001), to arrive at US$462,000 or approximately US$500,000. The United States concludes that the latter number constitutes the annual level of nullification or impairment, with the US$3.3 million figure serving as an upper limit.236
3.149.
Besides its earlier arguments that the United States has failed to overturn the presumptions established by Antigua's proposed methodology on any point237, Antigua provides specific arguments dismissing the methodology proposed by the US, in particular on the grounds of the underlying counterfactual and data used. In the following, we will review the arguments by the parties on the major concerns raised in relation to the approach proposed by the US.

(b) Antigua's exports of remote gambling services on horseracing

(i) Arguments by the parties

3.150.
As discussed above, Antigua challenges the counterfactual presumed by the United States on a number of grounds.238 Therefore, it initially has not argued for a different estimate of the share of betting on horseracing in Antigua's overall exports of remote gambling services, since it considered that share to be of no relevance to this Arbitration. In addition, it notes that if only horseracing was important, the United States would not seek to withdraw its relevant GATS commitments.239 Antigua also believes that its operators do very little business in horse race betting, but says it does not possess precise statistics for Antigua operators.240 It notes that GBGC does not provide a breakdown of Antigua's remote gambling revenues by type of gambling activity.241
3.151.
Nevertheless, Antigua seeks to provide alternative statistics to the ones proposed by the United States that may give some indication of the share of betting on horseracing in overall gambling activities. For one, Antigua makes reference to a 2002 report on remote gambling by the journal "International Gaming & Wagering Business" mentioning a 22 per cent share of horse racing in the American market.242 Antigua also provides CCA's annual breakdown by type of gambling activity for the years 1995-1997 and 1999-2006 on United States Gross Annual Gaming Revenues excluding remote gambling revenues, published in various volumes of "International Gaming & Wagering."243 If pari-mutuel betting is supposed to mostly relate to betting on horse races, as conjectured by the US244, revenues in that category almost stay constant over those years with the overall market more than doubling. This results in a declining share from about 8 per cent in 1995 to 3 per cent in 2006. Antigua also provides data from GBGC's Quarterly eGaming Statistics Report (May and October 2007 releases) on annual remote gambling revenues broken down by countries, including the US, and by four types of gambling activities, including sportsbetting which may be presumed to comprise betting on horse racing.245 GBGC data also includes time series data for the world and several geographical regions providing a breakdown of gambling revenues from online and offline betting on horseracing combined.246
3.152.
The United States defends the use of its proposed share considering it to be reasonable to assume that there is a correspondence between the levels of remote and non-remote versions of the same types of gambling. It also notes that Antigua, in its written submission, has not argued for a different estimate of the ratio between horserace remote gambling and all remote gambling.247 According to the US, the 7 per cent ratio can then be used as a reasonable estimate of Antigua's ratio of horserace remote gambling exports to total remote gambling exports.248
3.153.
The United States also states that absent information by Antigua based on consultation with its own operators, an allocation based on non-remote gambling in the United States provides the best basis for an estimate.249 It also states that the allocation of 22 per cent of the total amount of remote wagers placed by customers in the United States on horse races quoted by Antigua from a 2002 report published by "International Gaming & Wagering Business"250 should not be relied upon in this proceeding, since a copy of this report was not provided by Antigua for comments as to its contents.251 It also dismisses the GBGC data, which only provides global figures of remote and non-remote gambling revenues from horseracing.252

(ii) Analysis by the Arbitrator

3.154.
We note that all of the proposed figures for the share of Antigua's exports of betting services on horseracing to the United States are quite approximate, albeit in different respects. One weakness of all of the proxies is that they refer to consumption as opposed to production shares. Hence, they denote the distribution of shares among different gambling activities demanded, which may be different from the composition of Antigua's remote gambling services supplied to the US.
3.155.
The 7 per cent share from the US Bureau of Economic Analysis has the advantage of referring to the US market (and not to a larger geographical region), but covers the non-remote market only. It is not unreasonable to assume that certain gambling activities are more or less enjoyable than others depending on whether they are carried out in situ or remotely. Hence, the shares of different gambling activities may vary in the remote and non-remote market. The shares from CCA (referring also to non-remote pari-mutuel betting, i.e. principally betting on horseracing, in the United States) fall in the same range of figures, at least for the earlier years.
3.156.
The GBGC data contains figures on the remote market in the US, but only provides rougher aggregates of gambling activities, notably sportsbetting, which supposedly includes betting on horseracing. Alternatively, GBGC provides data on betting on horseracing combining the remote and non-remote market, but this data is only available at the regional level. Finally, Antigua cites a 22 per cent share of remote betting on horseracing according to a journal article. However, that number seems somewhat out of line with the GBGC data and we are unaware of any particular popularity of horseracing in the United States compared to other countries that would justify a substantial difference in these shares.
3.157.
It seems that the best approximations under the circumstances are the Bureau of Economic Analysis share for the non-remote market on betting in horseracing in the United States or the GBGC figures, which have the advantage of referring to the remote (or at least the combined remote and non-remote) market, but employ larger aggregates in terms geographical regions or gambling activities.

(c) ECCB/IMF/WTO data on Antigua's services exports and alternative revenue estimates

(i) Arguments by the parties

3.158.
Antigua notes that IMF and WTO data are based upon information reported by the ECCB.253 It states that the ECCB, by its own admission, does not capture the revenues or expenditures of the Antiguan remote gambling industry in its entirety. In that regard, it recalls the ECCB statement in connection with its overall data gathering that a major problem has been weak survey response.254 Antigua further explains that there is no legal obligation on behalf of the remote gaming operators (or any other private component of the Antiguan economy for that matter) to answer ECCB surveys or respond to ECCB enquiries.255 Antigua expresses the view that trying to extrapolate remote gaming revenues from flawed GDP data cannot be done in any meaningful manner.256
3.159.
Antigua also notes that the alternative data proposed by the United States is only supported by ECCB figures that does not include the relevant data (remote gambling revenues), while the data it uses in its proposed approach (GBGC) data is supported by several sources.257
3.160.
The United States maintains that the IMF and WTO remain the best available source of data on the level of Antigua's services exports and claims that Antigua has not established that IMF and WTO data do not include export revenues associated with Antiguan gambling operators. It observes that, at most, Antigua has shown that the ECCB was unable to include in its figures data on Antiguan gambling operators, because Antigua has chosen to shield those companies from any disclosure requirements.258 The United States also observes that the ECCB in its letter to Antigua259 only states that the GDP figures for Antigua do not include revenues earned by operators who are licensed to engage in interactive wagering and gaming, but that the ECCB does not comment on whether these revenues were included or not in its balance of payments information.260 Finally, the United States holds that if no direct data is available on Antigua's exports of gambling services, then this would also include the GBGC data. The United States also notes that in none of the IMF Annual Article IV consultations with Antigua, nor in the WTO Trade Policy Reviews has the service trade data in the balance of payments been questioned.261
3.161.
To provide further support for the order of magnitude of its calculations, the United States proposes three other ways to estimate Antigua's exports of remote gambling services to the United States on the basis of figures that Antigua has mentioned itself.
3.162.
First, the United States takes the statement by Antigua in its request pursuant to Article 22.2, namely that its "gambling and betting sector accounted for more than ten per cent of the gross domestic product of Antigua and Barbuda", to be about the relative sizes of Antigua's exports to its GDP, concluding that Antigua's gambling services exports to the entire world must have been on the order of $68 million, if the GDP of 2001 is taken as a basis for that calculation.262
3.163.
Second, the United States proposes to take the US$15.8 million figure for wages and salaries earned in the Antiguan internet gambling industry in 2001, as per Antigua's statement in para. 66 of its written submission, as a basis for estimating Antigua's net revenues from remote gambling. Making some assumptions about labour and profit shares, the United States estimates total net revenues to be between US$136 million and US$190 million depending on the precise shares chosen, which it calls a reasonable estimate of net revenues based on Antigua's own allegation of salaries of gambling operators.263 The United States concludes that this type of approach is the only means of determining historical Antiguan gambling revenues that would comport with economic reality.264
3.164.
Third, at the oral hearing, the United States observed that the 2007 Trade Policy Review on Antigua and Barbuda, Report by the Secretariat265 stated that interactive wagering and/or gaming operators in Antigua enjoyed the same tax exemptions as offshore financial services companies, with the exception of an income tax of 3 per cent introduced in 2001 by the International Business Corporation (Amendment) Act, 2001. Tax records could then be used to calculate the operators' income, i.e. gross take-in minus payouts to clients on bets. Since the Antigua Government did not provide any comments on the WTO Secretariat report in that regard, this information would need to be assumed to be correct.266
3.165.
On the first point, Antigua concedes that it cannot trace the source of that statement, which appear to relate to the year 1999. Antigua believes that the statement could be attributable in whole or in part to (i) basic ignorance within the government as to the true size of the remote gaming industry; (ii) estimates on the direct impact of the industry on the country in the nature of salaries paid, domestic rents and purchases and similar direct expenditures; and (iii) correlations made between estimated employment and the size of the overall domestic workforce.267
3.166.
Antigua rejects also rejects the second approach as being full of conjecture. It asserts that none of the figures are supported by any data, nor is there any support for the assumptions made and methodology used by the United States.268
3.167.
Finally, in regard to income taxation, Antigua counters that the three per cent tax introduced in 2001 was repealed in 2003 and has never been imposed, implemented or collected. It clarifies that the only direct fees paid by remote gaming operators to the Antiguan government are in respect of licensing and related fees in the nature of "flat" fees.269

(ii) Analysis by the Arbitrator

3.168.
We note that the WTO services trade data on Antigua used by the United States is based on information from the IMF, which in turn has been sourced from the ECCB. The ECCB in its letter confirms that revenues by Antiguan remote gambling operators are not included in its GDP accounting for Antigua. In addition, the IMF, in its letter dated 13 December 2007, confirms that the balance of payments statistics for Antigua and Barbuda are compiled by the ECCB and supplied to the IMF and that the IMF does not conduct any independent data collection of importers or exporters of gambling services. As mentioned above in paragraph 3,109, the IMF further explains that the data supplied by the ECCB does not separately identify remote gambling services, that the aggregate of personal, cultural and recreational services has a zero entry for all year shown in the IMF's 2006 Balance of Payments Yearbook, that survey report forms used by the ECCB do not indicate where gambling services are separately identified or covered, and that the description of balance of payments data compiled by the ECCB and set out in Part III of the Balance of Payments Yearbook does not refer to gambling services. Hence, if we must assume that the available balance-of-payments information on Antigua does not include revenues from remote gambling, it does not seem justified to us to use aggregates, such as "other commercial services", as an upper limit for the calculation of the level of counterfactual exports knowing that this aggregate maybe should, but does not contain the required information. Also, as we mentioned before in paragraph 3,109, in regard to other transactions related to gambling, the IMF concludes that it cannot determine how comprehensively the data are captured in practice, and we have noted that there are no unusual movements in other BOP entries, such as the debit position of various business services in the current account or of investment income, that could, (even vaguely) be related to changes in the use of remote gambling revenues and allow for an educated guess of their size.
3.169.
Concerning the three alternative "back-of-the-envelope" calculations mentioned by the United States, we note that the one approach using reported wages and salaries in the Antiguan industry as a starting point could have been promising. However, the United States does not provide any evidence on reasonable estimates of the wage share in total earnings. At a minimum, one could have provided information from a range of companies in the remote gambling business on their total revenues from internet gambling as well as on the number of employees and wage costs in that segment. Regarding the other two approaches, the necessary tax information does not appear to exist at all (the relevant legislation on income taxation was repealed shortly after its promulgation and has never been implemented) or, as far as the Antiguan remark on the contribution of gambling to GDP is concerned, it is not clear what is being referred to (apparently there is no documentation underlying the rather vague statement made, nor even a reference to GDP in a particular year).

(d) Loss of market share to competitors

(i) Arguments by the parties

3.170.
As described above in section III.D.2(a), the United States approach factors in an alleged loss of competitiveness by Antigua using a 7 over 50 per cent factor representing Antigua's global market share in 2006 and 2001 respectively to adjust the calculated level of nullification or impairment downwards. Referring to the arguments by parties portrayed already in section III.D.1(g)(i), we recall that Antigua opposes any adjustment to the counterfactual level of exports for other factors arguing that the revenues losses of its remote gambling industry are entirely attributable to the US actions270, while the United States maintains that Antigua's loss of global market share cannot be attributed to the US measures at all.271

(ii) Analysis by the Arbitrator

3.171.
As explained above in section III.D.1(g)(ii), we consider the issue of competition to be important, but we do not think that it is appropriate to use Antigua's global market share in an attempt to quantify Antigua's loss of competitiveness. Instead, Antigua's competitive position should be gauged in relation to its competitors in the United States market that are likely to be equally affected by the US measures. As stated above, being unable to rely on the way parties have modelled (or not) the issue of competition, we are required to develop our own approach in that regard.

3. Approach of the Arbitrator

(a) Antigua's exports of remote gambling services to the United States

3.175.
In view of the non-reporting of remote gambling services data by Antigua and the consequent non-inclusion of such information in the relevant categories of the services trade statistics compiled by the ECCB/IMF/WTO, we are left with no real alternative to the GBGC data for a time series of Antigua's revenues from remote gambling. We have to assume that such revenues indeed constitute exports, i.e. that the domestic market in Antigua is negligible and that the allocation of revenues to various jurisdiction has been done according to the actual location of operations. On the basis of some of the additional information provided, we can check whether the order of magnitude of the GBGC data is not out of line with other statistics. Notably, we will calculate an average of revenues per employee using the public company data provided by Antigua. Another interesting point of comparison would have been an estimation of total revenues on the basis of the 2001 wage bill in Antigua's remote gambling sector, but, unfortunately, we were not provided with any evidence as to the average share of wages in total revenues.

(i) GBGC data

3.176.
We note that the GBGC data on Antigua's annual remote gambling revenues for the years 1999-2006 from the October 2007 release of the Quarterly eGaming Statistics Report has been significantly revised downwards, especially for the years before 2002, compared to the data of the May release, which was used by Antigua in its methodology paper.272 Given GBGC's own corrections, we feel, of course, compelled to use the most recent update of the data. We also recall that these figures constitute Antigua's global revenues and that no information on bilateral exports of gambling services from Antigua to the United States has been provided. However, Antigua provides some anecdotal evidence that the US share in Antigua's total revenues from remote gambling should be around 80 per cent. In the absence of better information and of any indication about how this share has developed over time, we feel that the most we can do is to reduce the GBGC time series of Antiguan remote gambling revenues by 20 per cent to obtain an estimate of its revenues from the United States and to do at least some justice to our impression that the United States indeed represents the dominant market for Antigua.
3.177.
Like Antigua (in two of its three models) and the United States, we take the 2001-2002 period as the turning point when the US measures began to affect Antiguan exports of remote gambling services to the United States. Following the basic idea of Antigua's constant revenue methodology, we calculate the difference between Antigua's 2001 revenues and its actual revenues in each year 2002 to 2006. From these numbers, we determine the average annual revenue loss in the US market for Antigua to be about $304 million, according to the GBGC data.273
3.178.
However, only part of this loss is due to the US measures. By the same token, the assumption that Antigua's revenues would have remained constant represents a conservative starting point, since Antigua may well have benefited from an increase in US demand (that may have increased Antigua's revenues or cushioned its loss to competitors compared to the 2001 level). To take account of market developments, Antigua, in its methodology paper, proposes to use either a constant market share or constant growth methodology. We think that the constant market share methodology is not appropriate for a number of reasons. As was said before, Antigua's market share may be influenced by a number of events totally unrelated to the question at hand. Also, it is unlikely that an early mover advantage in a market with seemingly low barriers to entry guarantees a constant market share; to the contrary, as long as "supernormal" profits can be earned in a particular market, one would expect new entry and declining market shares of early entrants, even if revenues continue to grow. Most importantly, the GBGC data itself shows signs of increased competition that do not justify such an assumption. Concerning Antigua's constant growth assumption, we think that its 2001 revenues are the product of extraordinary growth in the early years and, if anything, would grow at a decelerating rather than linear rate, more in line with an S-curve pattern commonly observed in competitive markets. Despite the very limited number of observations, a logistic curve can be fit to the pre-2002 Antiguan revenue data, using the Levenberg-Marquardt algorithm. Its saturation point k converges to US$1,411 million. Taking this value as a benchmark, the average annual loss would have been US$411 million. Yet, we are not going to pursue this approach further preferring instead to model the impact of competition as well as possible increases in United States demand directly.

(ii) Revenues per employee

3.179.
As stated above, we use the information on revenues per employee from public listed companies as well as the data on employment in the remote gambling sector provided by Antigua to conduct a rough test of the order of magnitude of Antigua's remote gambling revenues.274 The data for seven publicly listed companies (three of which are licensed in Antigua) was taken from annual reports and is available for the years 2001 (for some companies 2002, 2003 or 2004) to 2006. Taking the average over those years gives an annual value of about $446,000/employee, and $447,000/employee if extreme values are removed.275 Multiplying the former value by the difference between the number of employees in 2001 and the number of employees in each year 2002 to 2006 (adjusted for the assumed 80 per cent US market share) and taking the average over those years results in an average annual revenue loss of about $196 million.276
3.180.
In the following, we will make further adjustments to both of these numbers.

(b) Competing suppliers

3.181.
Analysing the arguments by parties and the available evidence, we have found that we cannot agree with either extreme position taken by parties that either none or all of Antigua's loss in market share experienced since 2002 is due to its loss of competitiveness vis-à-vis providers from other locations. Using the October release of the GBGC data provided by Antigua, we find it particularly revealing that world revenues increase for both individual countries/territories in the region, namely Costa Rica, Curaçao and the Khanawake Territory, as well as for the South/Central America and Caribbean region as a whole, while the opposite is true for Antigua. As discussed above, it seems reasonable to us to assume that countries in the region compete with Antigua for the US market and a priori should be affected similarly by the US measures than Antigua. While these countries also lose global market share between 2002 and 2006, this decline is less pronounced than that for Antigua. In order to operationalize the notion of competition, we interpret the underperformance of Antigua since 2002 in terms of its contribution to the regional share in global remote gambling revenues as a loss in competitiveness by Antigua compared to its main competitors that it would have suffered even in the absence of the US measures. If this is so, only part of the revenue loss calculated in III.D.3(a)(i) and III.D.3(a)(ii) above should be attributed to the US measures.
3.182.
To make the necessary adjustments, we proceed as follows: Assuming that Antigua would have continued to do as well as its Central/South American and Caribbean competitors, its contribution to the region's global market share would have remained constant at the 2001 value. However, its contribution declined each year to reach about half of its 2001 value in 2006.277 Hence, the revenue loss by Antigua should be corrected for by the decline in its relative contribution to the region's global market share for each year 2002 to 2006. If this is done the average annual revenue loss calculated in III.D.3(a)(i) and III.D.3(a)(ii) above is reduced to $164 million and $128 million respectively.278
3.183.
If, in addition, we include the Khanawake Territory in the group of regional competitors, the average annual revenue loss calculated in III.D.3(a)(i) and III.D.3(a)(ii) above would only amount to $118 million and $108 million respectively. However, while inclusion of the Khanawake Territory makes sense conceptually, we prefer to avoid the use of country-specific information, besides Antigua, as far as possible, since we have more confidence in GBGC's global and regional estimates. Specifically, for the Khanawake Territory, we note some small inconsistencies in certain years with the North America total of revenues by operator location that, unlike for Antigua, have not been corrected between the May and October 2007 releases of the GBGC database.

(c) Gambling on horseracing

3.184.
In light of the chosen counterfactual and our analysis above in III.D.2(b)(ii), we have identified two strategies concerning an approximation of the unknown share of gambling on horseracing in Antigua's exports of remote gambling services to the US, which seem preferable over the other possible options identified by parties. The first option is to use the GBGC data combining remote and non-remote revenues from gambling on horseracing in various regions of the world. It is not clear whether the regional numbers refer to player or operator location. Since the main markets are separated out, it may refer to the former; however, the figure for sportsbetting in North America is rather small. Hence, it is probably more prudent to take the world total. Globally, the share of horseracing in all gambling activities declines from 13 per cent in 2001 to 9 per cent in 2006, while the total amount stays about constant. Applying the average share of 11 per cent to the figures in boldface calculated in III.D.3(b) above, one obtains an estimated annual revenue loss by Antigua from gambling on horseracing of about $18 million and $14 million respectively.
3.185.
If the 7 per cent share of gambling on horseracing in the United States non-remote market is used (based on the average for the years 2003 to 2006), the amounts calculated III.D.3(b) above would be reduced to about $11 million and $9 million respectively.

(d) Growth in US demand

4. Conclusion

IV. PRINCIPLES AND PROCEDURES OF ARTICLE 22.3 OF THE DSU

4.1.
As noted in Section I.B, Antigua requested to be authorized to suspend concessions and other obligations under the GATS and the TRIPS Agreement, in accordance with the principles and procedures of Article 22.3 of the DSU.

A. MAIN ARGUMENTS OF THE PARTIES

4.2.
In its request, Antigua states that, in considering what concessions and obligations to suspend, it applied the principles and procedures set forth in Article 22.3 of the DSU, and makes its request pursuant to Articles 22.3(b) and (c). Antigua notes that is a developing country with a population of approximately 80,000, and that with a combined landmass of only 442 square kilometres, Antigua is by far the smallest WTO Member to have made a request for the suspension of concessions under Article 22 of the DSU and realises the difficulty of providing effective countermeasures against the world's dominant economy. Antigua argues that its natural resources are negligible and as a result not only are the country's exports limited (approximately US$4.4 million annually to the United States) but Antigua is required to import a substantial amount of the goods and services needed and used by the people of the country. Antigua submits that on an annual basis, approximately 48.9 per cent of these imported goods and services come from single source providers located in the United States. The imposition of additional import duties on products imported from the United States or restrictions imposed on the provision of services from the United States by Antigua will, in Antigua's view, have a disproportionate adverse impact on Antigua by making these products and services materially more expensive to the citizens of the country. Given the vast difference between the economies of the United States and Antigua, additional duties or restrictions on imports of goods and services from the United States would have a much greater negative impact on Antigua than it would on the United States. In fact, ceasing all trade whatsoever with the United States (approximately US$180 million annually, or less than 0.02 per cent of all exports from the United States) would have virtually no impact on the economy of the United States, which could easily shift such a relatively small volume of trade elsewhere.
4.3.
Antigua further considers that the circumstances are serious enough to justify the suspension of concessions or obligations under other covered agreements in addition to the GATS. As a result of Antigua's lack of natural resources, the bulk of the economy is dependent upon tourism and the provision of banking and other financial services. Antigua explains that, initially with the cooperation of the United States Government, it looked to the provision of gambling and betting services as a way to diversify its economy and create the growth needed to assist the country in moving from a developing to a more advanced status, and that recent efforts by the United States Government to further prohibit and impede the provision of these services to consumers in the United States have greatly harmed the Antiguan service providers, while domestic providers in the United States continue to prosper in the absence of prohibition and criminal prosecution. Under the circumstances, the United States' prohibition of these services and its non-compliance with the recommendations and rulings of the DSB in this dispute forces Antigua to proceed in the manner requested despite the difference in level of development with the United States, the large disparity in their trade relations and despite the harsh economic reality affecting Antigua which affects its ability to exercise its rights under Article 22.
4.4.
Antigua also considers that the suspension of concessions and other obligations corresponding to a value of US$3,443 billion and wholly applied to the importation of services from the United States is neither practicable nor effective for various reasons. First, Antigua and Barbuda made no commitments under the sector at issue in this dispute, GATS Sector 10.D., "Sporting and Other Recreational Services," in its Schedule of Specific Commitments under the GATS (GATS/SC/2) ("Antigua's Schedule"). Second, with respect to most of the other services covered by Antigua's Schedule, as noted above suspension of concessions in the form of higher duties, tariffs, fees or other restrictions would have a disproportionate impact on the economy of Antigua and little or no impact on the United States. Third, even if Antigua were to rely exclusively on a suspension of concessions under the GATS, it would clearly not be able to recover the full amount of nullification and impairment caused by the inconsistent measures.
4.5.
Additionally, in Antigua's view, the United States' continued non-compliance renders the circumstances serious enough, within the meaning of Article 22.3(c) of the DSU, to justify the imposition of appropriate countermeasures under other covered agreements, given that Antigua's gaming industry will continue to suffer serious losses, the government of Antigua will be deprived of critical revenue, the people of Antigua will be enjoined from participating in much needed employment and the overall economy of the nation will continue to suffer adverse effects for such time as the United States does not withdraw the measures at issue in this dispute or remove their adverse effects.
4.6.
The United States, however, considers that Antigua has not followed the principles of Article 22.3 of the DSU, which required it to first seek to suspend concessions in the same sector as that covered by gambling services (namely, the United States states, "10. Recreational, Cultural and Sporting Services"), and that if it is not practicable to do so, Antigua must then seek to suspend concessions or obligations under the same agreement, namely the GATS.279
4.7.
The United States considers that Antigua has not provided an adequate explanation of why it cannot seek to suspend concessions under the GATS, as opposed to the GATS and also the TRIPS Agreement. The United States notes that Antigua makes conclusory statements that it is a developing country and that it generally needs to import goods and services, but provides no specific explanation of why it would not be practicable or effective to suspend concessions on specific services. The United States also considers that the argument that the level of Antigua's request is greater than all of its trade with the United States, and thus it must suspend concessions under other agreements, is flawed because the amount of the request is out of line with any economic reality. Once the level is recalculated, it is in line with Antigua's services imports and suspension of services concessions is possible (and required). In the United States view, the argument that no suspension in either goods or services could have an impact on the United States because of the disparity in the relative size of both economies cannot provide a carte blanche permission to deviate from the general principles set out in Article 22.3.
4.8.
Antigua considers that with the explanations it has provided in its request, it has clearly and unambiguously set forth its claims in accordance with the principles of Article 22.3 of the DSU280, so that it is now for the United States to establish "that it is both reasonable and effective for Antigua to seek [to suspend] concessions either in the same sector under the GATS or under one or more other sectors of the GATS in which Antigua has made commitments."
4.9.
Antigua also indicates in its written submission that it has no remedy under the GATS, most fundamentally because its volume of imports in the services sector is nowhere near sufficient to absorb the level of concessions that it is entitled to.281 Specifically, Antigua then further explains why suspension under GATS is practicable or effective neither in the same sector nor in other sectors.

B. OVERALL APPROACH BY THE ARBITRATOR

4.10.
Article 22.3 sets out certain principles and procedures to be followed by a complaining party seeking to suspend concessions, as to the sector(s) and/or covered agreement in which the suspension can take place, which the United States claims that Antigua did not follow in its request.
4.11.
Article 22.7 of the DSU provides that "if the matter referred to arbitration includes a claim that the principles and procedures set forth in paragraph 3 have not been followed, the arbitrator shall examine that claim" and that "[i]n the event the arbitrator determines that those principles and procedures have not been followed, the complaining party shall apply them consistent with paragraph 3". We are therefore required to examine the US claim that Antigua has not followed the principles and procedures of Article 22.3 of the DSU.
4.12.
As a general principle, Article 22.3(a) of the DSU provides that suspension of concessions or other obligations should first be sought in the same sector as that in which a violation was found. Article 22.3 provides in relevant part that:

"In considering what concessions or other obligations to suspend, the complaining party shall apply the following principles and procedures:

(a) the general principle is that the complaining party should first seek to suspend concessions or other obligations with respect to the same sector(s) as that in which the panel or Appellate Body has found a violation or other nullification or impairment."

4.13.
Subparagraphs (b), (c) and (d) of Article 22.3 further specify the principles and procedures to be followed by a complaining party wishing to seek suspension in another sector, or another agreement, than that in which a violation was found:

"(b) if that party considers that it is not practicable or effective for it to suspend concessions or other obligations with respect to the same sector(s), it may seek to suspend concessions or other obligations in other sectors under the same agreement;

(c) if that party considers that it is not practicable or effective to suspend concessions or other obligations with respect to other sectors under the same agreement, and that the circumstances are serious enough, it may seek to suspend concessions or other obligations under another covered agreement."

4.14.
In addition, subparagraph (d) provides that:

"(d) in applying the above principles, that party shall take into account:

(i) the trade in the sector or under the agreement under which the panel or the Appellate Body has found a violation or other nullification or impairment, and the importance of such trade to that party;

(ii) the broader economic elements related to the nullification or impairment and the broader economic consequences of the suspension of concessions or other obligations."

4.15.
In considering our mandate with respect to the principles and procedures set forth in Article 22.3 of the DSU, we find it useful to refer to the analysis of the arbitrators in the EC – Bananas III (Ecuador) case. As noted in Section I.A above, in determining the scope of their authority to review the principles and procedures relating to requests for suspension of concessions or other obligations under subparagraphs (b) and/or (c), the arbitrators in that case considered that "the fact that the powers of arbitrators under subparagraphs (b)-(c) are explicitly provided for in Article 22.6 implies a fortiori that the authority of Arbitrators includes the power to review whether the principles and procedures set forth in these subparagraphs have been followed by the Member seeking authorization for suspension".282
4.16.
The arbitrators then considered the terms of Article 22.3, including the fact that a certain margin of appreciation is left to the complaining party in arriving at conclusions in respect of certain factual elements (i.e. "if that party considers" in subparagraphs (b) and (c)) as well as the fact that the party concerned is required to apply the principles of Article 22.3 (i.e. "shall apply the following principles and procedures" in the introductory clause of Article 22.3). On the basis of that textual analysis, the arbitrators determined that:

"[T]he margin of review by the Arbitrators implies the authority to broadly judge whether the complaining party in question has considered the necessary facts objectively and whether, on the basis of these facts, it could plausibly arrive at the conclusion that it was not practicable or effective to seek suspension within the same sector under the same agreements, or only under another agreement provided that the circumstances were serious enough."283

4.17.
The arbitrators in the EC – Bananas III (Ecuador) case also considered more broadly the terms of Article 22.3 and noted that "these provisions imply a sequence of steps towards WTO-consistent suspension of concessions or other obligations which respects both a margin of appreciation for the complaining party as well as a margin of review by Arbitrators, if a request for suspension under Article 22.2 is challenged under Article 22.6".284
4.18.
We agree with these determinations. Specifically, we agree that the principles and procedures set forth in Article 22.3 of the DSU, which require the complaining party to make certain determinations, imply "a margin of appreciation" for the complaining party in making these determinations. At the same time, Article 22.3 sets out specific principles and procedures, that the complaining party must follow, and we understand the role of the arbitrator acting pursuant to Article 22.7 of the DSU to involve a review of whether those principles and procedures have been followed. We agree with the arbitrators in EC – Bananas III (Ecuador) that this includes a determination "whether the complaining party in question has considered the necessary facts objectively" and also "whether, on the basis of these facts, it could plausibly arrive at the conclusion that it was not practicable or effective to seek suspension within the same sector under the same agreements, or only under another agreement provided that the circumstances were serious enough".
4.20.
In this instance, Antigua first indicated in its request that it "may" suspend "horizontal and/or sectoral concessions and obligations for the following sector contained in the Antigua schedule: 2. Communication services", as well as under the TRIPS Agreement. However, it now considers that suspension of concessions or other obligations under the GATS, whether in the sector in which a violation was found or in other sectors, is not practicable or effective, and that the circumstances were serious enough, so that it may seek to suspend concessions or other obligations under the TRIPS Agreement (under which no violation was found).287
4.21.
Since Antigua seeks to suspend obligations under another covered agreement than that in which a violation was found (i.e. the TRIPS Agreement rather than the GATS), this implies that it made two successive determinations, under the "sequence of steps" foreseen in Article 22.3 of the DSU: first, a determination that it was not practicable or effective to suspend concessions or other obligations in the sector in which a violation was found, in accordance with Article 22.3 (b), and secondly, a determination that it was also not practicable or effective to suspend concessions or other obligations under another sector within the agreement in which a violation was found, and that the circumstances were serious enough, so that it was entitled to seek suspension of concessions or other obligations under another agreement as foreseen in subparagraph (c). We will consider these two aspects in turn.

C. PRELIMINARY OBSERVATIONS

4.22.
Before we start our assessment of Antigua's determinations, we find it useful to make some preliminary observations to take into account the fact that we have found, in Section III above, that the annual level of nullification or impairment of benefits accruing to Antigua is in the amount of US$ 21 million, rather than US$3,443 million as estimated by Antigua.
4.23.
Antigua's arguments in relation to the availability of suspension of concessions or other obligations under the GATS rely in part on assumptions concerning the amount of suspension that it is entitled to. Specifically, Antigua has explained that the "first and most fundamental reason" for which it is not practicable or effective to suspend concessions or other obligations under the GATS is that the volume of its imports from the United States in services is nowhere near sufficient to absorb the level of suspension of concessions that it is entitled to.288 The United States, for its part, considers that Antigua's argument is flawed because the amount of the request is out of line with any economic reality. Once the level is recalculated, the United States argues, it is in line with Antigua's services imports and suspension of services concessions is possible (and required).
4.24.
We agree that, to the extent that the annual level of nullification or impairment of benefits accruing to Antigua exceeds the total annual level of imports of services by Antigua from the United States, it would not be possible for Antigua to suspend obligations only in the services sector to cover the entire level of nullification or impairment. However, we have determined in paragraph 3,188 above that the level of nullification or impairment of benefits accruing to Antigua is significantly lower than Antigua had estimated, at US$21 million. We will therefore take this finding into account, in reviewing Antigua's determinations.
4.25.
With these preliminary considerations in mind, we now turn to Antigua's determination that suspension of concessions or other obligations is not practicable or effective, in the same sector as that in which a violation was found, or in any other sector under the GATS.

D. REVIEW OF ANTIGUA'S DETERMINATION IN RESPECT OF "SAME-SECTOR" SUSPENSION

4.26.
As noted above, it is only if the complaining party considers that it is "not practicable or effective" to suspend obligations with respect to the same sector that it may seek to suspend obligations in other sectors (or, ultimately, under another agreement). In addition, Article 22.3 subparagraph (d) requires the complaining party to take into account two specific aspects in applying the principles of paragraphs (a) to (c).
4.27.
We have determined in the previous section that our task is to examine whether, in making a determination in this case, Antigua, as the complaining party, has considered the necessary facts objectively and whether, on the basis of these facts, it could plausibly arrive at the conclusion that it was not practicable or effective to seek suspension with respect to the same sector within the same agreement.
4.28.
In order to conduct this assessment, we must clarify the elements that are relevant to such determination, and specifically the meaning of the terms "not practicable or effective" within the terms of Article 22.3(b), as well as the role of the elements identified in subparagraph (d) in this assessment, before turning to Antigua's determination in this case.

1. The principles and procedures of Article 22.3(b)

4.30.
We consider that these interpretations provide a useful starting point for our assessment of Antigua's determination that it was not practicable or effective for it to suspend obligations with respect to the same sector under the GATS.
4.31.
With respect to the elements referred to in subparagraph (d) of Article 22.3, which the complaining party is required to "take into account" in making its determination, we also find it useful to refer to the interpretations of the arbitrators in EC – Bananas III (Ecuador).
4.32.
With respect to the reference in subparagraph (d)(i) to "the trade in the sector or under the agreement under which the panel or Appellate Body has found a violation" and the "importance of such trade" to the complaining party, that arbitrators did not "exclude the possibility" that trade in the entirety of the relevant sector or agreement may be pertinent, but considered that these criteria related primarily to trade nullified or impaired by the WTO-inconsistent measure at issue.
4.33.
In our view, the ordinary meaning of the terms of subparagraph (d)(i) suggests that a consideration of the entirety of "trade in the sector" under which a violation was found is pertinent, rather than, as the EC – Bananas III (Ecuador) arbitration suggests, primarily the "trade nullified or impaired by the WTO-inconsistent measure at issue". This appears to us also to be consistent with the context of this provision as well as with its object and purpose.
4.36.
We agree with these determinations.
4.37.
In conclusion, we understand subparagraph (d) of Article 22.3 to require a consideration, by the complaining party, of the following elements:

(a) the trade in the relevant sector (as defined in subparagraph (f)) and the importance of such trade to the complaining party (in this case Antigua), including considerations relating to the relative importance of that trade to the complaining party and to the Member to which the requested suspension would apply (the United States);

(b) "broader economic elements" relating to the Member suffering the nullification or impairment (in this case Antigua);

(c) "broader economic consequences" of the suspension of concessions or other obligations, both for the party which was found not to be in compliance with WTO law (the United States) and for the complaining party (Antigua).

4.38.
With these general considerations in mind, we now review Antigua's determination that it was not practicable or effective to suspend concessions or other obligations with respect to the same sector as that in which a violation was found under the GATS in this case.

2. Review of Antigua's determination that is not practicable or effective to suspend concessions or other obligations in the "same sector"

4.39.
In order to review Antigua's determination, we find it useful to first clarify what the relevant "sector" is, as well as the scope of the "concessions or other obligations" to be considered. We will then turn to Antigua's determination on the practicability and effectiveness of suspension of these obligations, including its consideration of the factors identified in subparagraph (d).

(a) Identification of the relevant sector and scope of obligations to be considered

4.40.
Subparagraph (a) of Article 22.3 provides that :

"[T]the general principle is that the complaining party should first seek to suspend concessions or other obligations with respect to the same sector(s) as that in which the panel or Appellate Body has found a violation or other nullification or impairment" (emphasis added).

4.41.
Article 22.3 (f) of the DSU further defines the term "sector" for the purposes of Article 22.3. Specifically, it defines the term "sector" as meaning:

"[W]ith respect to trade in services, a principal sector as identified in the current "Services Sectoral Classification List" which identifies such sectors".

4.42.
As is clarified in footnote 14 of the DSU, this list is contained in document MTN.GNS/W/120, which identifies 11 principal sectors.
4.43.
In this case, a violation was found in sub-sector "10.D" (Sporting and Other recreational services). The "principal sector" in which the violation was found, as defined in Article 22.3(f), is therefore Sector 10 ("Recreational, Cultural and Sporting Services"). Antigua has confirmed, in response to a question from the Arbitrator, that it agrees that the entirety of Sector 10 is relevant for the purposes of this determination.294
4.44.
Article 22.3 refers to "concessions or other obligations" (emphasis added) with respect to the relevant sector. This indicates, in our view, that the entire range of obligations incurred under the relevant agreement with respect to the sector concerned is relevant in considering whether suspension is practicable or effective in that sector. This would include, as expressly noted in the text, scheduled concessions, as well as, as is also expressly stated in the text, any "other obligations" incurred with respect to the sector at issue.
4.45.
In this case, this means that the relevant range of obligations that should be considered for the purposes of assessing whether suspension is practicable or effective with respect to the same sector includes not only any specific commitments made by Antigua with respect to Sector 10 (Recreational, Cultural and Sporting Services), but also other obligations under the GATS that apply to this sector, such as the MFN obligation contained in Article II of the GATS.

(b) Antigua's determination that it is "not practicable or effective" to suspend concessions or other obligations with respect to the same sector

(i) Arguments of the parties

4.46.
In its request for authorization to suspend concessions or other obligations, Antigua noted that it had made no commitments under the sector at issue in this dispute, "namely GATS Sector 10.D, 'Sporting and Other Recreational Services,'"295 in its Schedule of Specific Commitments, and did not elaborate further on the potential suspension of concessions or other obligations with respect to Sector 10 as a whole ("Recreational, Cultural and Sporting Services").
4.47.
The United States argues that Antigua's explanation of its rationale for seeking suspension of obligations under the TRIPS Agreement in its request under Article 22.3 is "unconvincing" and does not follow the principles of Article 22.2. The United States further notes that Antigua's identification of the relevant sector under the GATS as "10.D" (Sporting and Other Recreational Services) is incorrect, as the relevant sector is all of Sector 10 (Recreational, Cultural and Sporting Services). The United States notes that Antigua has in fact made commitments under 10.A (Entertainment Services)296, and that it has provided no explanations of why it is not seeking to suspend concessions for Entertainment Services, where it has specific market access and national treatment commitments.297
4.48.
In its written submission, Antigua accepts that the relevant sector is Sector 10, and further elaborates on the possibility of same-sector suspension within the entirety of Sector 10 under the GATS. Antigua considers that it is clear from "prior arbitrations" that a complaining party does not have to take into consideration "trade in a sector where it has not made any commitments"298 and notes that the only trade within the same sector with respect to which it has made commitments in its GATS Schedule is sub-sector 10.A ("Entertainment Services"). It states that it has been unable to "locate any statistical sources" on trade in entertainment services and believes that such trade is negligible in its overall volume. Antigua further considers that suspension of concessions or other obligations in this sector "would most likely impair the already limited entertainment options available to Antiguan citizens while having virtually no impact on the United States at all"299.
4.49.
In response to a question from the Arbitrator concerning the possible suspension of obligations other than specific commitments, such as the MFN obligation under the relevant agreement, Antigua responded that this was most likely an academic question in the context of this dispute and that "even if this enquiry were carried out to its farthest extent to where Antigua had no obligations to the United States under the GATS at all", "the trade disparity is so great [between Antigua and the United States] that United States service providers would suffer little harm at all, if any, while Antiguan consumers would be forced to scramble for replacement services at uncertain cost".300
4.50.
The United States also generally argues that Antigua has not provided an adequate explanation of why it cannot seek to suspend concessions under the GATS, as opposed to under the GATS and the TRIPS Agreement. The United States notes that Antigua's request includes "conclusory statements along the lines that Antigua is a developing country and that it generally needs to import goods and services, but Antigua provides no specific explanation of why it would not be practicable or effective to suspend concessions on specific services".301
4.51.
More generally, Antigua argues that the "first and most fundamental reason" why it is neither practicable nor effective for it to seek suspension under the GATS is that the volume of its imports from the United States in services is nowhere near sufficient to absorb the level of suspension of concessions that it is entitled to.302 Antigua argues that, given that its total imports of services in 2005 amounted to US$208.11 million, it would not be possible for Antigua to reach a level of suspension of concessions or other obligations equivalent to the level of nullification or impairment "either under the same sector of the GATS in which the violation was found [Sector 10] or under the entire GATS".303 The United States, on the other hand, considers that the argument that the level of Antigua's request is greater than all of its trade with the United States, and thus it must suspend concessions under other agreements, is flawed because the amount of the request is out of line with any economic reality. Once the level is recalculated, the United States argues, it is in line with Antigua's services imports and suspension of services concessions is possible (and required).304

(ii) Assessment by the Arbitrator

4.52.
As noted above, Antigua considers that it was only required to take into consideration, for the purposes of determining whether suspension is practicable or effective, trade in sectors where it has made a commitment.305 It notes that the only trade within Sector 10 under the GATS with respect to which it has made commitments in its GATS Schedule is sub-sector 10.A ("Entertainment services"). It states that it has been unable to "locate any statistical sources" on trade in that sub-sector and believes that such trade is negligible in its overall volume. Antigua further considers that suspension of concessions or other obligations in this sector "would most likely impair the already limited entertainment options available to Antiguan citizens while having virtually no impact on the United States at all"306.
4.53.
The United States also appears to have considered, at least initially, that the only trade that was relevant in this determination was that in respect of which Antigua has made a specific commitment, namely sub-sector 10.A.307 The United States did not, however, specifically respond to any of the arguments provided by Antigua in its written submission in respect of the volumes of trade concerned and the potential adverse impact of a suspension under this sub-sector.
4.54.
As we have explained above, we believe that the range of obligations to be considered for the purposes of a determination of whether suspension is practicable or effective in the same sector is not limited to those sub-sectors in which specific commitments have been made. Nonetheless, we consider first Antigua's determination in respect of sub-sector 10.A ("Entertainment Services"), where it has made such commitments.308
4.55.
Antigua first notes that it has been unable to locate statistical sources in relation to trade in entertainment services and thus concludes that the volume of such trade must be negligible. We note that the United States does not dispute this assertion. Nor has it suggested, in fact, that there exists any amount of imports from the United States falling under sub-sector 10.A, in respect of which Antigua could suspend concessions or other obligations. We also note that the IMF Balance of Payments Statistics country tables submitted by the United States provide a breakdown of Antigua's services imports by category, including a category for "personal, cultural and recreational services" under which no amount is reported for the years 1998 to 2005.
4.56.
In these circumstances, we consider that Antigua could plausibly arrive at the conclusion that it was not practicable or effective for it to suspend concessions or other obligations under the GATS in respect of sub-sector 10.A, under which it has made a specific commitment.
4.57.
However, as observed above, we consider that, in addition to market access and national treatment specific commitments, other obligations under the GATS, including general GATS obligations, should also have been taken into account by Antigua in making its determination as to whether suspension was practicable or effective in Sector 10. In this respect, we note that Antigua has only made a very general observation, in response to a question from the arbitrator, that "even if this enquiry were carried out to its farthest extent to where Antigua had no obligations to the United States under the GATS at all", "[t]he trade disparity is so great that United States service providers would suffer little harm at all, if any, while Antiguan consumers would be forced to scramble for replacement services at uncertain cost".309 Antigua also observes that the United States has not argued that Antigua has a practical and effective remedy in respect of "other obligations" under the GATS.310
4.58.
In considering this matter, we first note that there is only a limited number of such "other obligations" under the GATS, that Antigua would be able to suspend under the whole of Sector 10, i.e. including sub-sector "10.B, News agency services"; "10.C, Libraries, archives, museums and other cultural services"; "10.D, Sporting and other recreational services"; and "10.E, Other". In our view, the main relevant obligation in this respect is the MFN obligation, contained GATS Article II, which obliges Antigua to accord immediately and unconditionally to US services and service suppliers treatment no less favourable than that it accords to like services and service suppliers of any other country.
4.59.
We recall our observation above that the IMF Balance of Payments statistics do not reflect any volumes of imports by Antigua in relation to "Personal, Cultural and Recreational Services" between 1998 and 2005. This would tend to suggest that no significant volumes of imports of such services takes place. We also note that the United States has not in fact suggested that any amount of imports takes place with respect to Sector 10 from the United States, that could form the basis for suspension of obligations under the GATS. We also note Antigua's argument that "[t]he trade disparity is so great that United States service providers would suffer little harm at all, if any, while Antiguan consumers would be forced to scramble for replacement services at uncertain cost"311, and the fact that the United States has not given any specific indication that Antigua's assessment in respect of the practicability or effectiveness of suspension of other obligations under Sector 10 is incorrect.
4.60.
In light of these elements, we find that Antigua could plausibly arrive at the conclusion that it was not practicable or effective for it to suspend concessions or other obligations under the GATS in respect of Sector 10.

(c) Whether Antigua has taken into account the factors identified in subparagraph (d) of Article 22.3

4.61.
We must now consider whether Antigua has taken due account, in making its determination, of the factors identified in subparagraph (d) of Article 22.3.
4.62.
Antigua's arguments to explain that same-sector suspension is not practicable or effective, as examined above, relate to the volume of trade in the relevant sector and to the potential impact of suspending concessions or other obligations in that sector both for Antigua (which it considers would be negative) and for the United States (which it considers would be virtually non-existent).
4.63.
These elements fall squarely within the range of factors that are pertinent for the purposes of paragraphs (i) and (ii) of subparagraph (d). Specifically, we find that these considerations relate to "the trade in the sector … under which the panel or the Appellate Body has found a violation" within the meaning of subparagraph (d)(i), as well as to the "broader economic elements" and "broader economic consequences" within the meaning of subparagraph (d)(ii).
4.64.
We find, therefore, that Antigua has taken into account the relevant elements in subparagraph (d) of Article 22.3, in determining that it was not practicable or effective for it to seek suspension in the GATS sector in which the violation was found in accordance with subparagraph (b) of Article 22.3.

(d) Conclusion

4.65.
In light of our determinations in sections (b) and (c) above, we find that Antigua could plausibly arrive at the conclusion that it was not practicable or effective to suspend concessions or other obligations under the GATS in respect of Sector 10, and determine that suspension of concessions or other obligations is not practicable or effective, with respect to the same sector as that in which a violation was found. We also find that the United States has not demonstrated that Antigua had not followed the principles and procedures of Article 22.3(b), in making this determination.

E. REVIEW OF ANTIGUA'S DETERMINATION IN RESPECT OF "OTHER SECTORS" UNDER THE GATS

4.66.
As noted above, Antigua sought an authorization to retaliate both under the GATS and the TRIPS Agreement, but then indicated in the course of the proceedings that it considered that it had no effective remedy under the GATS and was now seeking to suspend obligations under the TRIPS Agreement only.
4.67.
The United States considers that Antigua has not provided an adequate explanation of why it cannot seek to suspend concessions under the GATS, as opposed to under the GATS and the TRIPS Agreement, and that Antigua's trade figures indicate that Antigua would have the option of suspending concessions or other obligations in the services area.
4.68.
The relevant provision, Article 22.3 (c), provides that:

"[I]f [the complaining] party considers that it is not practicable or effective to suspend concessions or other obligations with respect to other sectors under the same agreement, and that the circumstances are serious enough, it may seek to suspend concessions or other obligations under another agreement."

4.69.
Two cumulative conditions therefore have to be met for a complaining party to be able to seek suspension under another agreement:

(a) that complaining party considers that it is not practicable or effective for it to suspend concessions or other obligations with respect to other sectors within the same agreement; and

(b) that party considers that the circumstances are "serious enough".

4.70.
In addition, subparagraph (d) requires, in the application of the principles of subparagraph (c), that the two factors considered above in the context of paragraph (b) suspension also be taken into account.
4.71.
We consider these elements in turn.

1. Whether suspension of concessions or other obligations is practicable or effective in "other sectors" under the GATS

4.72.
The first condition under Article 22.3(c) of the DSU for the suspension of concessions to be sought in another agreement than that in which a violation was found, is that the complaining party has determined that it is "not practicable or effective" to suspend concessions or other obligations in other sectors within that agreement.
4.73.
In determining whether suspension is practicable or effective, the complaining party is also required to take into account the two factors identified in subparagraph (d), namely, in the case of an application of subparagraph (c):

(a) the trade under the agreement under which a violation has been found and the importance of such trade to the complaining party (i.e., in this case, trade under the GATS and the importance of such trade to Antigua); and

(b) the broader economic elements related to the nullification or impairment and the broader economic consequences of the suspension of concessions or other obligations (which as noted above, could include a consideration of the impact of the retaliation both on the party seeking to retaliate and on the party against whom retaliation is being sought).

4.74.
In approaching this part of our determination, we are guided, mutatis mutandis, by the interpretations that we have developed in Section IV.C.1 above in relation to the same principles and procedures, as they relate to "same-sector" suspension in the context of Article 22.3(b).
4.75.
We must therefore determine whether Antigua has considered the relevant facts objectively and whether, on the basis of these facts, it could plausibly arrive at the conclusion that it was not practicable or effective to seek suspension of concessions or other obligations with respect to other sectors under the GATS.
4.76.
In this part of our assessment, the relevant sectors to be considered are all principal sectors, within the meaning of MTN.GNS/W/120, other than Sector 10, under the GATS. In respect of such sectors, the relevant "concessions or other obligations" under the GATS include, as determined above in relation to Article 22.3(b), not only any specific commitments made by Antigua with respect to those sectors, but also other obligations under the GATS that apply to this sector, such as the MFN obligation contained in Article II of the GATS.312
4.77.
We first address Antigua's general argument that its volume of imports is insufficient to support suspension of obligations under the GATS alone at the level it is entitled to, before turning to the detail of Antigua's consideration of the practicability and feasibility of such suspension.

(a) Level of suspension and availability of suspension under the GATS

4.78.
As noted above, Antigua has explained that the "first and most fundamental reason" for which it is not practicable or effective to suspend concessions or other obligations under the GATS is that the volume of its imports from the United States in services is nowhere near sufficient to absorb the level of suspension of concessions that it is entitled to.313 The United States, for its part, considers that Antigua's argument is flawed and that once the level is recalculated, it is in line with Antigua's services imports and suspension of services concessions is possible (and, in the US view, required).314
4.79.
As we noted in our preliminary observations, we will take into account, in assessing Antigua's arguments, our own earlier determination that the annual level of nullification or impairment of benefits accruing to Antigua is in the amount of US$21 million, significantly lower than Antigua's estimation. We will compare this amount to the annual level of Antigua's services imports from the United States in order to determine the extent to which suspension of concessions or other obligations entirely under the GATS is at least conceivable.
4.80.
Antigua indicates that its total imports of services amount to US$208.11 million for 2005.315 In response to a question by the Arbitrator, the United States observes that the figures presented by Antigua on its services imports appear to be consistent with the IMF Balance of Payment statistics that the United States has itself presented to the Arbitrator as evidence.316 These statistics indicate a total value of US$204.85 for imports of services by Antigua.317
4.81.
Antigua indicates that it has not been able to allocate these services imports among its trading partners, and cites the US CIA's World Factbook in support of the proposition that the United States accounts for 21.1 per cent of Antigua's services imports. The notes and definitions of this Factbook suggest, however, that these figures relate to merchandise imports, rather than services imports.318 In its request for authorization to suspend obligations, Antigua indicated that "[o]n an annual basis, approximately 48.9 per cent of these imported goods and services come from single source providers located in the United States".319 The United States does not dispute this figure and refers to it in order to estimate the total value of Antigua's imports of services for the United States.320 In the absence of any reliable alternative estimation provided by either party, we rely on Antigua's initial statement that the United States share in Antigua's imports of services is "approximately 48.9 percent".
4.82.
Combining the amount of US$204.85 million with this percentage, we arrive at a total annual value of services imports from the United States to Antigua of US$101.77 million. This thus represents the maximum value of services trade in respect of which Antigua could potentially seek to suspend concessions or other obligations. In light of our earlier determination that the annual level of nullification or impairment of benefits accruing to Antigua in this case is in the amount of US$21 million, this means that it would in principle be possible for Antigua to seek suspension of concessions or other obligations entirely within the GATS.
4.83.
With this initial determination in mind, we turn to a consideration of whether such suspension is, nonetheless, not practicable or effective within the meaning of Article 22.3(c).

(b) Practicability and effectiveness of suspension in other sectors

4.84.
In considering this question, we first recall our determinations in paragraph 4.29above on the nature of the assessment to be conducted in applying the principles of Article 22.3. Specifically, we recall our determinations that an examination of the "practicability" of an alternative suspension concerns the question whether such an alternative is available for application in practice as well as suited for being used in a particular case. We also recall our determination that in contrast, the thrust of the "effectiveness" criterion empowers the party seeking suspension to ensure that the impact of that suspension is strong and has the desired result, namely to induce compliance by the Member which fails to bring WTO-inconsistent measures into compliance with DSB recommendations and rulings within a reasonable period of time.
4.85.
We also bear in mind our related determination that a consideration of the practicability and the effectiveness of an alternative suspension within the same sector or under the same agreement does not need to lead to the conclusion that such an alternative suspension is both not practicable and not effective in order to meet the requirements of Article 22.3. This implies that it would be sufficient for Antigua to have determined that suspension under the GATS in respect of other sectors is not effective, even if it is practicable.
4.86.
Antigua has provided some general arguments relating to the size of its economy and the relative sizes of the economies of the United States and Antigua, as well as some more specific indications why it would not be practicable or effective to suspend concessions in specific services sectors. We consider these in turn.
4.87.
Antigua has first stressed the fact that, as a small economy with limited natural resources, it is heavily dependant on imports of goods and services, including from the United States, so that the imposition of additional duties on products imported from the United States or on the provision of services from the United States would have "a disproportionate adverse impact on Antigua and Barbuda by making these products and services materially more expensive to the citizens of the country".321 Antigua also notes that with respect to most services other than "Sector 10.D", covered by Antigua's GATS schedule, suspension of concessions in the form of higher duties, tariff, fees or other restrictions would have a disproportionate impact on the economy of Antigua and virtually no impact of the United States.
4.88.
As a preliminary matter, we note that Antigua's arguments include references to the potential impact of a suspension in respect of trade in goods. Our assessment is however limited to a consideration of suspension of obligations under the GATS. We therefore make no determination in relation to the potential impact of a possible suspension of obligations in relation to trade in goods.
4.89.
To the extent that it relates to the suspension of obligations under the GATS, Antigua's concern that, as a small import-dependent economy, it could suffer an adverse impact from such suspension, is in our view pertinent to an assessment of whether suspension is practicable or effective. Specifically, we consider that these circumstances have the potential to affect significantly the effectiveness of the proposed suspension.
4.90.
We note in this respect the determination of the arbitrators in EC – Bananas III (Ecuador), that, in a situation "where a great imbalance in terms of trade volume and economic power exists between the complaining party seeking suspension and the other party which has failed to bring WTO-inconsistent measures into compliance with WTO law" "and in situations where the complaining party is highly dependent on imports from the other party, it may happen that the suspension of certain concessions or certain other obligations entails more harmful effects for the party seeking suspension than for the other party".322 In the view of that arbitrator, "in these circumstances, a consideration by the complaining party in which sector or under which agreement suspension may be expected to be least harmful to itself would seem sufficient for us to find a consideration by the complaining party of the effectiveness criterion to be consistent with the requirement to follow the principles and procedures set forth in Article 22.3".323
4.91.
At the same time, we agree with the United States that general statements relating to the size of the complaining party's economy or the relative size of the economies of both parties do not justify a departure from the requirements of Article 22.3 of the DSU. Rather, the complaining party is required, in all cases, to follow the specific sequence of steps provided in this provision and to make a determination in respect of each of its elements as relevant. In this context, an explanation of how such circumstances affect the practicability or effectiveness of the proposed suspension would in our view be required, in order for these considerations to validly support a determination that it is not practicable or effective to suspend concessions or other obligations in other sectors under the same agreement within the meaning of Article 22.3.
4.92.
In this case, Antigua has not limited itself to general assertions as to the size of its economy, in itself or relative to the US economy. It has also provided arguments to explain how the potential adverse impact of suspension of obligations would manifest itself in various specific sectors under the GATS in respect of which suspension might be considered.
4.93.
Antigua thus explained that it initially envisaged the suspension of concessions or other obligations in telecommunication services, but that, upon more detailed review of the import and use of telecommunications services in Antigua, it had determined not only that the volume of the trade was low (US$5.03 million) but that disruptions in changing services and suppliers and increased cost to Antiguan consumers would result in a heavier burden on Antiguan citizens as the result of suspending concessions in this area while having no perceptible impact on the United States.324
4.94.
Antigua has also explained that the arguments it made in respect of Sector 10 ("Entertainment services") also hold true for other sectors of the GATS in which Antigua has made specific commitments in its Schedule. Antigua explains that, of the total value of services imports to Antigua in 2005, the main services imported were transportation (US$ 70.7 million), travel (US$ 40.1 million) and insurance services (US$35.5 million). Antigua argues that if it were to suspend concessions with respect to these services, given the low level of these imports, it is clear that the suspension would have virtually no impact upon the United States, while having to replace these services from other service providers, if reasonably practicable at all, would most likely prove to be more expensive to Antiguan consumers.325
4.95.
The United States considers that WTO statistics indicate that Antigua's "other services imports" (services other than travel and transport) are in fact larger than its "other services exports", and thus are "an obvious candidate" for a request for suspension of concessions.326 It is not clear to us why the United States considers that the fact that Antigua's exports of services other than travel and transport are larger than its imports of services in the same category makes such imports "an obvious candidate" for suspension of concessions. To the extent that the United States assumes that, in these circumstances, the provision of the imported services could be replaced with the provision of the same services by local suppliers, this assumption would seem to be unverifiable, in the absence of any specific indication on the composition and distribution of such imports and exports.
4.96.
The United States further notes that Antigua has even higher levels of services imports for travel and transport services. The United States argues that although Antigua's economy does depend to a substantial extent on tourism, it does not necessarily follow that suspension of concessions in the areas of travel and transport would adversely affect Antigua's tourism industry.327 The United States also notes, however, that, in light of the reliance of the Antiguan economy on tourism, it has not contested Antigua's assertion that it would not be practicable or effective for Antigua to suspend concessions in a manner that might discourage US tourists.328 Thus, the United States notes, it has not argued, for example, that Antigua should suspend concessions with regard to US travel services.329
4.97.
The United States therefore seems to accept that the suspension of concessions or other obligations in respect of tourism services, or of other services that may have an impact on the tourism industry, may be such as to adversely affect Antigua, and that this may render such suspension not practicable or effective within the meaning of Article 22.3(c). We agree that suspension of obligations in respect of such services would have the potential to adversely affect Antigua's economy, while not necessarily having any perceptible impact on inducing compliance by the United States.
4.98.
We also note that, by their very nature, services transactions are closely woven into a country's domestic economic fabric, with many cross-sectoral linkages. This is typically the case for so-called infrastructural services, such as large segments of the transport, telecommunications and financial services sectors. In our view, it is plausible that suspension of obligations against US firms established in Antigua in such sectors could entail a negative impact for the Antiguan local economy. The risk of economic disruption in case of forced divestiture, or similar measures, is not negligible, in particular for small economies. For instance, it is not clear to what extent Antiguan services suppliers would be able or willing to step in for US services suppliers obliged to suspend their operation or even leave the country. Moreover, legislative and other measures protecting foreign investors may make it difficult in practice to take and enforce action against them.
4.99.
Overall, in light of all these elements, we consider that Antigua has provided sufficient elements in support of its determination, to satisfy us that it has considered objectively the relevant elements and that it could plausibly, on the basis of these elements, reach the conclusion that it was not practicable or effective to suspend concessions or other obligations with respect to "other sectors" under the GATS. By contrast, the United States has not demonstrated to us that Antigua has not met the requirements of Article 22.3(c) of the DSU. In fact, even if it considers Antigua's explanations to be overall insufficient, the United States apparently recognizes the legitimacy of Antigua's concerns, and their relevance to this determination, with respect to at least some sectors. The United States has not, on the other hand, persuaded us that there are services sectors in which Antigua could suspend concessions or other obligations practicably or effectively.
4.100.
In making this determination, we recognize that the elements provided by Antigua in explaining how it arrived at its determination are somewhat limited. Specifically, we have been provided with only limited information on the exact composition of Antigua's services imports, including the share of US imports in various sectors. However, we are not persuaded that it would be reasonable to expect Antigua to be in a position to produce, in the context of these proceedings, such detailed sector-specific statistics and data relating to its bilateral services trade. In this respect, we find it appropriate to take into account, in reviewing Antigua's determination, that it may not maintain such data on a regular basis, in the same way as some developed countries or larger developing countries may do. Indeed, the United States itself has also not provided us with any more detailed information on the volume and distribution of its services trade with Antigua.

(c) Antigua's consideration of the elements in subparagraph (d)

4.101.
We must now consider whether Antigua has taken due account, in making its determination, of the factors identified in subparagraph (d) of Article 22.3.