The Cloves panel analysis of TBT Art. 2 is fascinating.
Given the structure of TBT Art. 2.1-2.2, and the absence of an explicit GATT Art. XX exception in TBT, I have long wondered how to read TBT 2.1 (national treatment) with TBT 2.2 (necessity) (see my case book with A. Guzman, p. 533).
I saw three options:
(i) interpret Art. 2.1 narrowly (so as to avoid finding a 2.1 violation too easily, given the absence of an explicit Art. XX exception);
(ii) interpret Art. 2.1 “normally” (broad competition test) but then somehow construe 2.2 as a provision that may justify certain 2.1 violations;
(iii) interpret Art. 2.1 “normally” and allow Art. XX to justify 2.1 violations.
The Cloves panel has chosen option 1 (narrow interpretation of 2.1), and also left the possibility open that even where 2.1 is violated, XX could come to the rescue (the US did not invoke XX under TBT, so the panel left this question open).
According to Cloves:
- TBT 2.2 (necessity) is about what the US does ban (cloves cigarettes), a ban that was, however, found to be “necessary” to combat youth smoking.
- TBT 2.1, in contrast, is about what the US is not banning: like products (menthol cigarettes), an exclusion that was found to be “discriminatory”.
So while the measure/ban is “necessary”, it is still “discriminatory”. This at first sight weird result (can a “discrimination” really be “necessary”?) is, of course, only arrived at because under each standard different elements of the measure were examined: Under 2.2, it is the ban on cloves that is “necessary” (not the discrimination of menthols). Under 2.1, it is the exclusion of menthols that is “discriminatory” (not the ban on cloves).
Put differently, TBT 2.2 has been construed like the paragraph analysis under GATT XX (focus on the measure as a whole); 2.1 like the chapeau analysis under XX (focus on the discrimination).
TBT 2.1 has been “narrowly” defined by Cloves for two main reasons.
First, under “likeness”, the panel shifted from what I called in a 2008 AJIL article (with Dimascio) a “competition test” towards a “regulatory context” test: Even where two products compete (say, plastic cups and steel cups) they may not be “like” because of factual differences in terms of the regulatory justification invoked (e.g. if the measure relates to what can be put in a microwave, steel cups are not “like” plastic cups; see para. 7.245). Here, the panel found that menthols are “like” cloves not so much because they compete but because they both raise the same “regulatory concern” invoked by the US itself, namely: they are both flavoured cigarettes that are particularly attractive to youth (see paras. 7.231 and 7.247).
Second, under “less favourable treatment”, it is not enough that the measure worsens conditions of competition for one imported product (one cloves from Indonesia) as compared to one “like” domestic product (one US menthols). Two additional requirements apply. First, the group of imports must be compared to the group of like domestic products. Second, if detrimental effects on imports “can be explained by factors or circumstances unrelated to the foreign origin of the product” no violation will be found (para. 7.269).
Here, the panel did find a violation because under (i) “the vast majority of cigarettes that were imported from Indonesia into the United States were clove cigarettes” (7.275); before the ban only one US company produced cloves, whereas many US producers sell (permitted) menthols. And under (ii) US justifications for not banning menthols (impact on US health system and risk of smuggling) did not convince the panel (para. 7.289; this may be one of the weak links in the report: the panel equated US justifications with a US concern not to impose “costs” on US menthol producers; however, this is arguably not what the US was saying; it is not that they were unwilling to impose costs on (mainly US) menthol producers but rather that they saw concerns with banning a cigarette (menthols) that 25% of smokers now smoke, concerns which according to the US are not present for cloves, which only 0.1% smoke, and this irrespective of whether menthols are domestically produced or imported).
Two things came to mind when reading this crucial and novel interpretation of TBT 2.1.
First, should the same approach be followed for national treatment under GATT Art. III? Although the panel spends a lot of time trying to distinguish TBT from GATT in this respect, are its arguments really convincing? The panel may even secretly hope that its analysis will, one day, find its way back into the AB’s standard Art. III analysis (at this point the panel was probably more concerned with not being overruled by the AB).
Second, the Cloves panel’s “regulatory context” test for likeness, and reference to possible policy/factual justifications that may allow for differential treatment between like products, is very similar to how investor-state tribunals have interpreted national treatment in investment agreements (in e.g. Pope & Talbot, Feldman and UPS). From our 2008 AJIL article:
- "... it is not the investment or the investor that must be "like"; the provision calls, instead, for a comparison of the circumstances in which a government regulates foreign and domestic investments. What matters is not the positioning of those investments in relation to each other within the market ("competition test"), but rather the factual support for the government's distinction between the two when taking regulatory action ("regulatory context test")". (p. 81)
- "... the extent of consideration given to legitimate domestic policy concerns in trade and investment law is converging: in investment law, the test is already one of "regulatory context"; in trade law, the "competition test" is gradually being - and, in our view, should be - supplemented by an examination of the policy justifications for the regulation in question. Such an examination may demonstrate that the differential treatment between imported and like domestic products is based on a legitimate objective, and not nationality". (p. 83)
One element that may explain this convergence around reading regulatory context/goals into the test of national treatment itself, is the absence of general exceptions in both TBT (discounting for a possible application of XX to TBT) and (most) investment agreements. If one reads policy justifications into the test of discrimination itself, there is less/no need for general exceptions. Put differently, reading them into TBT 2.1 then makes up for the absence of XX in TBT.
Note, however, that a “regulatory context” test may both narrow and broaden the scope of national treatment violation: It limits violations by looking at possible policy justifications. It may broaden violations (especially in the investment context) by comparing products/investors that do not really compete (in Occidental, exporters of oil v. exporters of flowers) but present the same regulatory concerns. From our 2008 AJIL article:
- "There is nothing to preclude a complainant from presenting a fact-based argument that, in particular regulatory circumstances, its investment should be treated the same as an investment in a completely separate sector. If one investor is a toy manufacturer, and another a maker of dies for fabric, but each uses the same potentially harmful chemical in their production processes, one would not necessarily expect them to be treated differently by a chemical regulation simply because they are in different sectors of the economy". (p. 85)
Whether the Cloves approach to national treatment in TBT will eventually cross-over into GATT and, in turn, converge with investment jurisprudence on national treatment remains to be seen. Let us first await the appeal …
