1 Empirical estimates for selected countries and sectors suggest that current GATS commitments are 2.3 times more restrictive than actual trading conditions. The offers submitted in the Doha Round between 2003 and 2006 would result in a modest reduction only; the commitments would still be 1.9 times more restrictive than the applied regimes. Gootiiz, Batshur and Mattoo, Aaditya (2009), ‘Services in Doha: What's on the Table?’, Journal of World Trade, 43(5): 1013‒1030Google Scholar.

2 WTO, Eighth Ministerial Conference, Chairman's Concluding Statement, WTO document WT/MIN(11)/11 of 17 December 2011, p. 3.

3 Australia, Canada, Chile, Taipei, Columbia, Costa Rica, European Union (and member states), Hong Kong, Iceland, Israel, Japan, Korea, Liechtenstein, Mexico, New Zealand, Norway, Pakistan, Panama, Paraguay, Peru, Switzerland, Turkey, United States, and Uruguay (participants as of May 2015).

4 Some WTO Members, including in particular Venezuela, even criticized that such de-briefings on a ‘plurilateral initiative’ took place in the Council, which, in their view, did not constitute an appropriate forum. See, for example, WTO document S/C/M/115 of 27 November 2013, p. 9.

5 See, for example, Gantz, David A. (2008), Liberalising International Trade after Doha – Multilateral, Plurilateral, Regional, and Unilateral Initiatives, Cambridge University PressGoogle Scholar; and Hufbauer, Clyde Gary, Jensen, J. Bradford, and Stephenson, Sherry (2012), Framework for the International Services Agreement, Peterson InstituteGoogle Scholar, pp. 23f.

6 WTO document WT/L/847 of 19 December 2011.

7 See also the quote in below n. 63.

8 Article XVII of the GATS lays down the Agreement's national-treatment obligation (see Appendix). For a detailed discussion of Article V, see Cottier, Thomas and Molinuevo, Martin (2008), ‘Article V GATS: Economic Integration’, in Wolfrum, Rüdiger, Stoll, Peter-Tobias, and Feinäugle, Clemens (eds.), WTO–Trade in Services, Max Planck Commentaries on World Trade Law, pp. 125–151Google Scholar, at p. 136.

9 Appellate Body Report, Turkey – Restrictions on Imports of Textile and Clothing Products (Turkey–Textiles), WT/DS34/AB/R, 22 October 1999, para. 48.

10 Only one of the 120 odd RTAs notified to the WTO under the relevant provisions, the Agreement on the European Economic Area (EEA), signed by the EU, Iceland, Norway, and Liechtenstein, is a services-only agreement.

11 In December 2006, a modified schedule for the EU, covering 25 member States was adopted by the Council for Trade in Services (WTO documents S/C/W/273, 9 October 2006 and S/L/286, 18 December 2006). The modification process had been launched because the initial schedules of the new EU members contained some more liberal entries than those of the initial EU members which were retained for the enlarged Union. However, the results have not entered into force since they have not yet been ratified by all EU members. In three additional cases, negotiations pursuant to Article XXI were launched, but not (yet) concluded. They concerned the EU's enlargement to 27 members, the proposed exclusion of gambling services from the United States’ schedule of commitments (in the wake of a dispute with Antigua and Barbuda), and Bolivia's intention to withdraw its commitments on hospital services.

12 Appellate Body Report, Turkey–Textiles, supra n. 9, para. 58. See also the discussion in Trebilcock, Michael, Howse, Robert, and Eliason, Antonia (2013), The Regulation of International Trade, 4th edn, RoutledgeGoogle Scholar, pp. 113ff.

13 As has been noted elsewhere, the dispute concerned a customs union, and there has not yet been a comparable case dealing with a free trade area. (de Mestral, Armand C. M. (2013), ‘Dispute Settlement Under the WTO and RTAs: An Uneasy Relationship’, Journal of International Economic Law, 16(4): 777–825Google Scholar, at 791.) However, it is difficult to see whether and how this distinction would matter in the current context.

14 Appellate Body Report, China – Measures Affecting Trading Rights and Distribution Services for Certain Publications and Audiovisual Entertainment Products (China–Publications and Audiovisual Products), WT/DS363/AB/R, 19 January 2010, para. 394.

15 Panel Report, Canada – Certain Measures Affecting the Automotive Industry (Canada–Autos), WT/DS139/R und WT/DS142/R, 11 February 2000, para. 10.270. The panel further opined that ‘the purpose of Article V is to allow for ambitious liberalization to take place at a regional level, while at the same time guarding against undermining the MFN obligation by engaging in minor preferential arrangements'. The Appellate Body has not pronounced on these considerations in its subsequent report (Appellate Body Report, Canada–Autos, WT/DS139/AB/R und WT/DS142/AB/R, 19 June 2000).

16 For example, the MFN clauses in the agreements between EFTA and, respectively, Chile, Korea, and Mexico do not cover other RTAs concluded by the parties, while the clauses in the EU–Korea FTA and the EU–CARIFORUM EPA only apply to future RTAs the parties might enter into. Other agreements, including an FTA between the EU and Chile, do not contain any MFN obligations. For a more detailed presentation of various MFN clauses in trade agreements, see Kim, Jong Bum (2014), ‘Entrenchment of Regionalism: WTO Legality of MFN Clauses in Preferential Trade Agreements for Goods and Services’, World Trade Review, 13(3): 443–470Google Scholar.

17 For example, in a recent collection of such studies no more than two out of a total of 11 contributions explicitly refer to the existence of GATS minus commitments (Rupa Chanda (2014), ‘Mapping the Universe of Services Disciplines in Asian PTAs', and Bosworth, Malcolm and Trewin, Ray (2014), ‘Much Ado About Nothing Much? Australian PTAs in Services’, in Sauvé, Pierre and Shingal, Anirudh (eds.), The Preferential Liberalization of Trade in Services – Comparative Regionalism, Edward Elgar, pp. 227ffGoogle Scholar, at pp. 357 and 373).

18 Sébastien Miroudot, Jehan Sauvage, and Marie Sudreau (2010), Multilateralising Regionalism: How Preferential are Services Commitments in Regional Trade Agreements?, OECD Trade Policy Working Paper No. 106, TAD/TC/WP(2010)18/FINAL, pp. 36ff.

19 Summaries are provided in Adlung, Rudolf and Miroudot, Sébastien (2012), ‘Poison in the Wine? Tracing GATS-minus Commitments in Regional Trade Agreements’, Journal of International Economic Law, 46(5): 1045–1082Google Scholar, at 1056ff.

20 The table contains the agreements covered by the underlying study, but does not give a complete picture of all services RTAs between TISA participants (see also Adlung, Rudolf and Mamdouh, Hamid (2014), ‘How to Design Trade Agreements in Services: Top Down or Bottom-Up?’, Journal of World Trade, 48(2): 191–218Google Scholar, at 212ff).

21 See Appendix for a brief explanation of relevant GATS provisions and scheduling practices.

22 Adlung and Miroudot, supra n. 19, pp. 1062 and 1064. For the purposes of this study, the initial OECD dataset of 56 RTAs was extended by ten more agreements to cover all agreements that had been notified to the WTO by August 2011 and involved either an OECD country, India, or China.

23 The Agreement with Chile contains at least a review provision related to the mandated negotiations on subsidy disciplines under Article XV:1 of the GATS. These started immediately after the Agreement's entry into force and are still ongoing without any agreed outcome in sight.

24 However, the sectoral part of Chile's schedule exempts virtually all listed services from national-treatment disciplines under modes 1 and 2, regardless of the measures concerned. In contrast, there are far fewer such exemptions under mode 3. Concerning Mexico's GATS schedule, it contains a very small number of national-treatment limitations under any of these three modes.

25 See also below n. 57. There are considerable variations around these averages, however. For example, according to the OECD Secretariat's dataset, the shares of sectoral GATS minus commitments on national treatment amount to 12% for the US under both NAFTA and the Free Trade Agreement with Singapore, but are zero in the agreements with Chile, Bahrain, and Peru (Miroudot et al., supra n. 18, p. 38).

26 According to Article XXI:3, a Member that intends to modify its GATS commitments must negotiate ‘any necessary compensatory adjustment’ at the request of ‘any Member the benefits of which under the Agreement may be affected’. No further definition of such Members is provided, possibly reflecting the difficulties of specifying pertinent criteria per mode of supply. However, the prospect of having to deal with significant numbers of affected Members, self-selected, certainly results in these provisions being invoked very reluctantly and only if there is an obvious need. In contrast, the obligation to re-negotiate tariff schedules under the GATT, pursuant to Article XXVIII of the Agreement, relates directly to those trading partners with which a tariff concession was initially negotiated and those which have been determined by the CONTRACTING PARTIES (now: WTO Membership) to have ‘a principal supplying interest’. An interpretative note further defines the criteria to identify such principal suppliers, while providing that not more than one applicant should normally be determined to qualify. According to the most recent overview prepared by the WTO Secretariat, the Article XXVIII provisions were used in 18 cases between January 1995 and May 2011 (ignoring invocations related to successive enlargements of the EU). This compares with two renegotiations under GATS Article XXI in cases not related to EU enlargements, which continue to linger on (see supra n. 11).

27 See Adlung and Miroudot, supra n. 19, pp. 1062f.

28 Ibid., Table 1 (p. 1062) and Table 2 (p. 1064). A concrete example of the latter limitations is the introduction, in the EU–Korea FTA and other RTAs recently promoted by the EU, of a requirement that renders the ‘[p]articipation of private operators in the education network … subject to concession’. No such requirement, which provides virtually unlimited scope for policy discretion, is contained in the initial EC 12 services schedule.

29 These and the following observations are based on the dataset examined by Adlung and Miroudot, supra n. 19. It covers 28 RTAs enacted before 2005 and 38 later agreements.

30 For example, while some 70% of the North–North RTAs (involving 42 partners) contain one or more minus commitments at horizontal level under modes 1 to 3, the shares for both North–South and South–South RTAs (involving 82 and ten partners, respectively) are in the order of 80%. Because of the limited number of observations, the data provided for the South–South RTAs are possibly less significant.

31 Adlung and Miroudot, supra n. 19, p. 1066.

32 Ibid.

33 Financial services and business services have been committed, respectively, by some 80% of Members, communication services by some 70%. See Adlung, Rudolf and Roy, Martin (2005), ‘Turning Hills into Mountains? Current Commitments Under the General Agreement on Trade in Services and Prospects for Change’, Journal of World Trade, 39(6): 1161–1194Google Scholar, at 1168.

34 There are also counter-examples, such as NAFTA Article 103: ‘In the event of any inconsistency between this Agreement and … other agreements, this Agreement shall prevail to the extent of the inconsistency, except as otherwise provided in this Agreement.’ It needs to be acknowledged, however, that the NAFTA negotiations were concluded well before the conclusion of the Uruguay Round; NAFTA was signed on 17 December 1992.

35 For example, in the agreements between the US and, respectively, Chile and Peru, the parties ‘affirm their existing rights and obligations with respect to each other under the WTO Agreement and other agreements'. Similar provisions can be found in quite a number of other RTAs, including the EU's CARIFORUM EPA.

36 See Adlung and Miroudot, supra n. 19, p. 1072. Again, this assessment is based on an examination of the 66 RTAs involving either OECD countries, India, or China that had been notified to the WTO by August 2011.

37 There are at least two publications, nevertheless, that refer to some cases of GATS-minus framework provisions. See Mario Marconini (2009), Revisiting Regional Trade Agreements and Their Impact on Services Trade, ICTSD Issue Paper No. 4, pp. 10ff, and Pierre Latrille and Juneyoung Lee (2012), Services Rules in Regional Trade Agreements: How Diverse and How Creative as Compared to the GATS Multilateral Rules?, WTO Staff Working Paper ERSD-2012-19, pp. 15 and 24ff.

38 All following references to this agreement are based on the consolidated text published by the European Commission on 26 September 2014. As noted by the Commission, this text is not binding under international law and will only become so on completion of the ratification process (available at http://trade.ec.europa.eu/doclib/docs/2014/september/tradoc_152806.pdf).

39 Of course, it would also be possible to exclude certain government levels from specific commitments under the GATS. However, such exclusions would relate only to market access and national treatment, but would not affect the scope of the MFN obligation which is horizontally applicable regardless of the existence of commitments (see Appendix).

40 EU–Central America Association Agreement, Chapter on Establishment, Article 162(b). Surprisingly, the modification is contained only in this chapter, overlapping with mode 3 of the GATS (commercial presence), but not in the chapter dealing with cross-border supply of services, which covers modes 1 and 2.

41 See, for example, Leroux, Eric (2006), ‘What is a “Service Supplied in the Exercise of Governmental Authority” Under Article I:3(b) and (c) of the General Agreement on Trade in Services?’, Journal of World Trade, 40(3): 345–385Google Scholar.

42 ‘Where, in exceptional circumstances, capital movements and payments, including transfers, cause or threaten to cause serious difficulties for the operation of the economic and monetary union of the European Union, safeguard measures that are strictly necessary and do not constitute a means of arbitrary or unjustified discrimination between a Party and a non-Party may be taken by the European Union with regard to capital movements and payments, including transfers, for a period not exceeding six months. The European Union shall inform Canada forthwith and present, as soon as possible, a time schedule for the removal of such measures.’

43 Para. 2(a) of the Annex on Financial Services reads: ‘Notwithstanding any other provision of the Agreement, a Member shall not be prevented from taking measures for prudential reasons, including … or to ensure the integrity and stability of the financial system. Where such measures do not conform with the provisions of the Agreement, they shall not be used as a means of avoiding the Member's commitments or obligations under the Agreement.’ For a more detailed analysis, see Marchetti, Juan A. (2011), ‘The GATS Prudential Carve-out’, in Delimatsis, Panagiotis and Herger, Nils (eds.), Financial Regulation at the Crossroads: Implications for Supervision, Institutional Design and Trade, Wolters Kluwer, pp. 279ffGoogle Scholar.

44 See, for example, the discussion in Delimatsis, Panagiotis (2014), ‘Who's Afraid of Necessity? And Why It Matters?’, in Lim, Aik Hoe and de Meester, Bart (eds.), WTO Domestic Regulation and Services Trade, Cambridge University Press, pp. 95–109Google Scholar.

45 Thus, 70% of the 80 RTAs examined in a recent study either completely dispensed with the requirement that new regulations avoid nullification and impairment of existing commitments or transformed it into a best-endeavours clause. This was the case in all ‘EU-type’ agreements and, barring one exception, all ‘NAFTA-type agreements' reviewed by the authors. (Latrille and Lee, supra n. 37, p. 27.)

46 According to an unnamed official, the US proposal ‘in particular excludes reference to technical standards and modifies the language of Article VI:4 by removing prescriptive language in favor of nonbinding “best endeavour” commitments … adding that it represents a “downgrade of GATS”’ (WTO Reporter, 29 April 2014).

47 According to relevant WTO jurisprudence, it would be for a complainant to argue and provide evidence that a particular measure is more trade restrictive than necessary to fulfil ‘a legitimate objective’ (TBT Agreement, Article 2.2). For further information, see Delimatsis, supra n. 44.

48 For a discussion of the definitional subtleties associated with references to ‘like circumstances', see, for example, Ortino, Federico (2009), ‘Services’, in Lester, Simon and Mercurio, Bryan (eds.), Bilateral and Regional Trade Agreements: Commentary and Analysis, Cambridge University Press, pp. 184–214Google Scholar, at pp. 184ff, as well as Diebold, Nicolas F. (2010), Non-discrimination in International Trade in Services – ‘Likeness’ in WTO/GATS, Cambridge University PressGoogle Scholar, pp. 143ff.

49 For instance, as noted by Ortino (supra n. 48, p. 202), the concept of ‘like circumstances' has been employed by NAFTA tribunals not only to determine the competitive relationship between foreign and domestic suppliers, in tune with the interpretation given by WTO panels, but also ‘as a de facto policy justification mechanism’. If so, however, NAFTA's national-treatment standard is weaker than that of the GATS.

50 European Commission, Memo 12/107, available at http://trade.ec.europa.eu/doclib/docs/2013/february/tradoc_150552.pdf.

51 Available at http://trade.ec.europa.eu/doclib/press/index.cfm?id=1155 (emphasis added).

52 The basis for comparison is still the GATS schedule for EC 12, which entered into force on January 1995 and essentially continues to govern the relations of the then member States across all sectors, except basic telecommunications and financial services. Commitments in the two latter sectors were agreed for EC 15 in negotiations extended beyond the timeframe of the Uruguay Round and concluded in 1997.

53 Concerning the CARIFORUM EPA, see Rudolf Adlung and Peter Morrison (2010), ‘Less than the GATS: ‘Negative Preferences' in Regional Services Agreements', Journal of International Economic Law, 13(4): 1103–1143, at 1128ff.

54 There is no further clarification of structure and functions of an ‘education network’ (or: ‘health and social network’) and of the terms and conditions governing the award of the respective concessions and authorizations.

55 See, for example, Hamid Mamdouh (2014), ‘Services liberalization, negotiations and regulation: some lessons from the GATS experience’, in Lim and de Meester, supra n. 44, 325–331, at 328f.

56 Reciprocity provisions have been included for Bulgaria in legal services as well as accounting services, France in veterinary services, and Italy for advisory and consulting services incidental to agriculture. It is this latter entry that explicitly subjects the enrolment of third-country nationals in the professional register to reciprocity. Are these supposed to be nationals of non-TISA participants? If so, the measure concerned should have been listed, at the Agreement's entry into force, as an exemption from MFN treatment under GATS Article II:2. An RTA is certainly not the appropriate medium to seek cover for any such departures from the MFN obligation. The three other reciprocity provisions might be destined for RTA-internal use only. Even then, however, they would be inconsistent with the panel's interpretation, in Canada–Autos, of the requirement in Article V:1(b) to provide for ‘the absence or elimination of substantially all discrimination’ (see Section 2.2, indent (iv)).

57 The consolidated national-treatment gains for two samples of RTAs, covering each some 20 agreements, were 62% of commitments under negative-list agreements as compared to only 22% for positive-list agreements. See Adlung and Mamdouh, supra n. 20, p. 212ff.

58 The respective shares of GATS minus commitments are 4.7% for negative-list agreements and 3.5% for positive-list agreement – not counting minus commitments scheduled on a cross-sectoral (horizontal) basis. See Adlung and Mamdouh, supra n. 20.

59 In the absence of reservations, the national-treatment commitment under TISA is envisaged to trigger a standstill (binding current levels of liberalization) and ratchet obligation (binding any future improvements).

60 Relatively frequent examples are regulations which limit the share of foreigners in a company's governing board. A number of Members have inscribed such measures as limitations on national treatment under mode 3 (commercial presence), while others considered them to constitute a market access limitation under mode 4 (presence of natural persons). For a more detailed analysis of interpretational divergences and their reflection in current GATS schedules, see Adlung, Rudolf, Morrison, Peter, Roy, Martin, and Zhang, Weiwei (2013), ‘FOG in GATS Commitments: Why WTO Members Should Care’, World Trade Review, 12(1): 1–27Google Scholar, at 13ff.

61 The EU‘s list is contained in WTO document GATS/EL/31 of 15 April 1994 and a consolidated list for EU 28 has been attached to the TISA offer. The possibility to list such exemptions from MFN treatment existed only at the Agreement's entry into force, or, in the case of accessions, the time of acceptance. Over 100 WTO Members used that opportunity and listed over 500 exemptions in total.

62 As noted in Section 2.2 (last paras.), however, there is still an element of uncertainty surrounding the question of whether and to what extent a Member is free to vary the levels of liberalization across its RTAs.

63 As has been noted elsewhere: ‘The pessimistic view suggests that steady erosion of the WTO's centricity will sooner or later bring the world to a tipping point – a point beyond which expectations become unmoored and nations feel justified in ignoring WTO norms since everyone else does.’ Richard Baldwin and Teresa Carpenter (2009), Why not in the WTO? Erosion of WTO Centricity in Trade Liberalisation, Background Paper at the Conference of Thinking Ahead on International Trade, www.wto.org/english/res_e/statis_e/tait_sept09_e/baldwin_e.doc.

64 See the estimates in Gootiiz and Mattoo, supra n. 1.

65 Adlung, Rudolf (2012), ‘Trade in Services in the WTO: From Marrakesh (1994), to Doha (2001), to …?’, in Narlikar, Amrita, Daunton, Martin, and Stern, Robert M. (eds.), The Oxford Handbook on the World Trade Organization, Oxford University Press, pp. 370–393Google Scholar, at p. 386.