Article contents
The meaning of ‘necessary’ in GATT Article XX and GATS Article XIV: the myth of cost–benefit balancing
Published online by Cambridge University Press: 31 October 2007
Abstract
Conventional wisdom tells us that in Korea–Beef, the Appellate Body interpreted the word ‘necessary’ in GATT Article XX to require a cost–benefit balancing test. The Appellate Body is supposed to have applied this test also in EC–Asbestos, US–Gambling (involving GATS Article XIV), and Dominican Republic–Cigarettes. In this article I demonstrate, by detailed analysis of the opinions, that the Appellate Body has never engaged in such balancing. They have stated the balancing test, but in every case they have also stated the principle that Members get to choose their own level of protection, which is logically inconsistent with judicial review by cost–benefit balancing. And they have decided every case by reference to the ‘own level of protection’ principle. The Appellate Body is right not to balance. Balancing is not authorized by the treaty texts, and it is not needed to prevent inefficient harm to foreign interests.
- Type
- Research Article
- Information
- Copyright
- Copyright © Donald H. Regan 2007
References
1 Korea – Measures Affecting Imports of Fresh, Chilled and Frozen Beef, WT/DS161 and 169/AB/R (adopted 10 January 2001).
2 E.g., Marceau, and Trachtman, , ‘The Technical Barriers to Trade Agreement, the Sanitary and Phytosanitary Measures Agreement, and the General Agreement on Tariffs and Trade: A Map of the World Trade Organization Law of Domestic Regulation of Goods’, 36 Journal of World Trade (5, 2002) 811, at 826–828Google Scholar, 851–853; WTO, World Trade Report 2005: Exploring the Links between Trade, Standards and the WTO, at 135, 137–139.
3 Among those who were not deceived are Howse and Tuerk, ‘The WTO Impact on Internal Regulations – A Case Study of the Canada–EC Asbestos Dispute’, in G. de Búrca and J. Scott (eds.), The EU and the WTO: Legal and Constitutional Issues (2001), 283, at 324–325; F. Ortino, ‘From Non-Discrimination to Reasonableness: A Paradigm Shift in International Economic Law?’, Jean Monnet Working Papers 01/2005.
4 ¶¶162–164, discussed in detail in Section 3.
5 ¶¶176, 178, discussed in detail in Section 3.
6 ¶180, discussed in detail in Section 3.
7 The reader might object that Article XX itself implicitly ranks possible goals, by mentioning some and not others. There is some truth in this, although the significance of the positive list in Article XX is less than many people think, if we interpret Article III correctly – that is, as violated only by origin-specific regulations and by origin-neutral regulations with a protectionist purpose – see Regan, , ‘Regulatory Purpose and “Like Products”in Article III:4 of the GATT (With Additional Remarks on Article III:2)’, 36 Journal of World Trade (3, 2002) 443CrossRefGoogle Scholar, and ‘Further Thoughts on the Role of Regulatory Purpose Under Article III of the General Agreement on Tariffs and Trade – A Tribute to Bob Hudec’, 37 Journal of World Trade (4, 2003) 737. But even if Article XX does imply some ranking, it still does not authorize the Appellate Body to do further ranking on its own, which is what the Appellate Body here proposes to do.
8 European Communities – Measures Affecting Asbestos and Asbestos-Containing Products, WT/DS135/AB/R (adopted 5 April 2001).
9 United States – Measures Affecting the Cross-Border Supply of Gambling and Betting Services, WT/DS285/AB/R (adopted 20 April 2005).
10 Dominican Republic – Measures Affecting the Importation and Internal Sale of Cigarettes, WT/DS302/AB/R (adopted 19 May 2005).
11 There is passing mention of the ‘necessity’ issue under Article XX, with a cite to Korea–Beef and a mention of balancing, in Mexico – Tax Measures on Soft Drinks and Other Beverages, WT/DS308/AB/R (adopted 24 March 2006), ¶74. But there is no discussion of this issue, and it is perfectly clear that the action in Mexico – Soft Drinks is elsewhere, so I shall ignore this case.
12 See Section 5.
13 The claim in the text that passing the balancing test is itself sufficient for legality presupposes that the balancing test is applicable not only to the overall benefits and costs of the measure as compared to inaction, but also to the (marginal) benefits and costs of the measure as compared to any alternative. For a fuller treatment of the logic of this family of tests, see Appendix 2 to Regan, ‘Judicial Review of Member-State Regulation of Trade Within a Federal or Quasi-Federal System: Protectionism and Balancing, Da Capo’, 99 Michigan Law Review (8, 2001) 1853, at 1899–1902.
14 With regard to whether Article XX itself imposes some ranking, see n. 7 above. Incidentally, there is a threshold question, which arises before we even get to the ‘margin of appreciation’ issue, whether the claimed purpose of a measure is the Member's true purpose. General views about the value of various purposes, and thus the likelihood of Members' pursuing them, are relevant here too, but with the same caveat that they are relevant only as heuristics.
15 United States – Section 337 of the Tariff Act of 1930, BISD 36th Supp. 345 (adopted 7 November 1989), ¶5.26.
16 Korea–Beef, ¶176, quoting US – Section 337, ¶5.26.
17 The Appellate Body wrestles unsatisfactorily with a similar sort of question in Australia – Measures Affecting Importation of Salmon, WT/DS18/AB/R (adopted 6 November 1998), part V.D. I am not sure why they do not see what seems clear after the briefest reflection: that when we are applying a less-restrictive alternative test, the ‘level of protection’ we should require from proposed alternatives is (a) the level achieved by the actual measure, or (b) the level the Member asserts as its desired level, whichever is lower (that is, whichever is less protective).
18 We shall see below that after deciding four cases on ‘necessity’, the Appellate Body has yet to say of any legitimate purpose that it is not important; and in US–Gambling they seem to imply that fraud-prevention in particular is important. Actually, the Appellate Body does seem to have a somewhat jaundiced view of Korea's purpose in the present case; but the best explanation is not that they see fraud-prevention as unimportant. Rather, they suspect that Korea's actual purpose is not origin-neutral fraud-prevention. Remember the observation from the Section 337 Panel that the regulating Member can choose its own level of protection, but it must seek the same level of protection against harm from foreign products and like domestic products. The Korea–Beef Appellate Body was plainly troubled by the fact that Korea seemed to care less about domestic dairy beef being passed off as Hanwoo beef than about foreign beef being passed off (¶168). This looks like having a higher level of protection when the harm is from a foreign product. (Related to the idea that Korea seems to want a higher level of protection against harm from foreign beef is the Appellate Body's complaint that the dual retail system puts ‘all, or the great bulk’ of the costs of fraud-prevention on the foreign beef, as opposed to spreading the costs between domestic and foreign beef (¶181).)
19 US–Gambling, ¶308 n. 379.
20 DR–Cigarettes, ¶71. The Panel did suggest that this sort of purpose might be specially important for developing countries, but we can hardly make such a distinction part of the legal test of ‘necessity’.
21 US–Gambling, ¶323.
22 See Section 3 above.
23 Notice incidentally that the Appellate Body's approach, as I have elaborated it, does not involve shifting the final burden of proof on the ‘necessity’ exception to the complaining Member. Once the respondent Member has made a prima facie case for the necessity of its measure, as described in the text, it is up to the complaining Member to propose some alternative; but once the complaining Member proposes a minimally plausible alternative measure, the burden is still on the respondent Member to show that the alternative does not achieve the desired level of protection, or is not less trade-restrictive, or involves excessive administrative/enforcement costs.
24 See the discussion of the Section 337 case in Section 3 above.
25 E.g., SPS 3.3, 4.1, 5.5, 5.6, TBT Preamble.
26 Notice that even subsidies to export industries, sometimes distinguished as ‘export promotion’, can be protectionism in the broad sense of regulation designed to improve the competitive position of some group of domestic economic actors vis-à-vis their foreign competitors.
27 Of course there are numerous more specific disciplines in the SPS and TBT Agreements, but even with regard to these, the ‘least restrictive alternative’ principle is what they seem primarily designed to implement, insofar as they are applied to non-discriminatory measures. And whatever the purpose of these more specific disciplines, an inquiry into whether a measure is based on a risk assessment, or whether it is based on an international standard, plainly does not require balancing.
28 Regan, , ‘What Are Trade Agreements For? – Two Conflicting Stories Told by Economists, With a Lesson for Lawyers’, Journal of International Economic Law, 9(4), 951–988 (2006CrossRefGoogle Scholar).
29 Ibid.
30 Other assumptions, aside from purely mathematical ones about convexity and the like, are (a) competitive domestic markets, (b) constant returns to scale, and (c) domestically rational behavior by all other countries. Again, I have explained elsewhere why these are appropriate assumptions for the dispute settlement system to rely on, even though they are obviously idealizations. Ibid. at 982–983.
31 This is just a miscalculation by Home; by hypothesis, Home is not setting the tax above the efficient level as a means of purposeful terms-of-trade manipulation.
32 This assumes there is efficient regulation in Foreign, so that Home's too-high tax does not merely compensate for, say, an export subsidy in Foreign.
33 This assumes Home has market power as a buyer in the world market for widgets; if they do not, their policy cannot hurt Foreign in any event.
- 23
- Cited by