The AB report in Tyres reversed the Panel’s interpretation of “arbitrary or unjustifiable” discrimination under the chapeau of Article XX with respect to the Mercosur exception and the Panel’s interpretation of “arbitrary” discrimination with respect to the used tyres exemption. Upon a preliminary reading of the report, I have several observations and comments. [More thoughts are added at the end of the post after considering good comments from others.]
1. The AB has clarified for the first time that the policy objective of the measure at issue should be considered in the chapeau analysis:
The assessment of whether discrimination is arbitrary or unjustifiable should be made in the light of the objective of the measure. (Para. 227)
This is an important development. In my view, the chapeau jurisprudence prior to Tyres was largely devoiced from the policy objective of the measure challenged. Just compare the above statement to the following statement of the AB in Shrimp:
The policy goal of a measure at issue cannot provide its rational or justification under the standards of the chapeau of Article XX. The legitimacy of the declared policy objective of the measures, and the relationship of that objective with the measure itself and its general design and structure, are examined under Article XX(g). (Shrimp, para. 149)
In fact, I believe that it is the Tyres Panel that first grounded the chapeau analysis on the policy objective of the measure. See my earlier post. The Panel’s reasoning regarding “unjustifiable” discrimination was based entirely on whether the import ban was applied in a manner that “significantly undermines the objective” of the ban. (It is true that the Panel did not link “arbitrary” discrimination to the objective of the ban, which the AB reversed. More on this later.)
2. The AB has held that it is the rationale, rather than the effect, of the discrimination that should determine the chapeau analysis. It criticized the Panel’s approach (paras. 229-230):
[T]he Panel's interpretation of the term "unjustifiable" does not depend on the cause or rationale of the discrimination but, rather, is focused exclusively on the assessment of the effects of the discrimination. The Panel's approach has no support in the text of Article XX and appears to us inconsistent with the manner the Appellate Body has interpreted and applied the concept of "arbitrary or unjustifiable discrimination" in previous cases. (In a footnote to this sentence, the AB observed that the Panel's approach was based on a logic that is different in nature from that followed by the AB under GATT Article III, i.e., Article III protects competitive conditions rather than trade effects.)Having said that, we recognize that in certain cases the effects of the discrimination may be a relevant factor, among others, for determining whether the cause or rationale of the discrimination is acceptable or defensible and, ultimately, whether the discrimination is justifiable. The effects of discrimination might be relevant, depending on the circumstances of the case, ...
While it seems reasonable to focus on the rationale, rather than the effect, of discrimination to determine whether the discrimination is justifiable, the AB’s approach risks turning the chapeau analysis into a separate inquiry on whether a measure in violation of Article XI also violates MFN or NT and whether the resulting discrimination is otherwise justifiable on the policy grounds of XX. (See Point 3 below) The footnote reference to the Article III logic also indicates this potential confusion between the chapeau and NT analysis.
Moreover, the AB’s criticism of the Panel struck me as neither fair nor convincing. As pointed out above, the Panel’s entire analysis of “unjustifiable” was based on whether the ban was implemented in a manner that has significantly undermined the “policy objective” of the ban. It may well be that the Panel’s approach has no support in the text of Article XX, but where is the textual support for the AB’s approach?
ADD: It is one thing to say that the Panel's approach has no support in the text, and it is another to judge whether that approach is permissible under the text. See below ADD for further explanation.
3. The AB’s interpretation leaves “arbitrary” or “unjustifiable” discrimination undefined, confusing the chapeau standards with the primary discrimination standards under Article I or III. In this case, Article XX was invoked to justify an Article XI violation, and the Panel skipped the examination of Articles I and III claims on the ground of judicial economy. Under the AB’s approach, the chapeau examination of “arbitrary” and “unjustifiable” discrimination effectively provides a second chance to examine I and III claims.
Think what would have happened if the Panel had not exercised judicial economy on EC’s Mercosur claim under Article I (the AB has criticized the Panel for not having done so). The Panel would likely have found that Brazil’s ban violated MFN and the violation was not justified by XXIV. Then the Panel would need to examine whether the MFN violation was nonetheless justified under XX. Assuming the measure was found “necessary” under (d) or (b), the Panel would still have had to examine whether the measure also met the chapeau requirement of not constituting “arbitrary or unjustifiable” discrimination. Under the AB’s approach, what could the Panel have said about the chapeau standards that would not be merely repeating the Article I claim?
In other words, the AB’s approach may effectively turn the chapeau interpretation into an examination of whether a measure violating MFN or NT can be justified by the policy objectives recognized by XX. This, however, is inconsistent with what the AB held in Gasoline that the chapeau standard of arbitrary or unjustifiable discrimination should be different from that of NT under Article III.
By contrast, the Panel seemed to be aware of this pitfall and hence tried hard to make some sense out of the “arbitrary or unjustifiable” standards. Thus, it interpreted “arbitrary” as capricious, and “unjustifiable” as the existence of other measures that significantly undermine the policy objective of the measure at issue (i.e., inconsistent with the policy objective of the ban). Unfortunately, the AB rejected the Panel’s interpretation without providing a better alternative.
In sum, while the AB has clarified that the policy objective of the measure at issue should be considered in the chapeau analysis, it has left the standards of “arbitrary or unjustifiable” discrimination as vague and confusing as ever. One way to get out of this conceptual confusion, in my view, is to focus on the concept of the “same conditions prevailing”, because this concept defines “arbitrary or unjustifiable discrimination” and can be readily distinguished from the concept of “like product” that defines MFN and NT discrimination under Articles I and III. For further discussion, see my article.
ADD: Upon further consideration, I am now more convinced that the Panel's approach of using an effect test to interpret "unjustifiable" should be permissible under the chapeau. The text of the chapeau does not prohibit all discrimination between similarly-situated countries resulting from the measure at issue. Rather, it prohibits only the kind of discrimination that has reached a certain level or degree of seriousness, characterized as "arbitrary or unjustifiable". Conceptually, any discrimination, regardless of the kind or degree, is to be identified by reference to the purpose of the comparison between treatment of two or more objects. It is for this reason that the policy objective must be considered in determining whether the measure provisionally justifiable under XX results in (any kind of) discrimination between similarly-situated countries. Once such discrimination is found, however, it is necessary to further determine whether it has reached the level or degree of "arbitrary or unjustifiable". One way to gauge the seriousness of the discrimination is to apply an effect test.
In other words, the text of the chapeau warrants a two-part inquiry. First, whether the application of the measure that is otherwise justifiable by the policy objective under XX has resulted in any discrimination between similarly situated countries. (In Tyres, the answer is probably "yes" because the Mercosur exemption and the used tyres exemption both discriminate between Members by design, unless it can be demonstrated that different conditions prevail in them.) Second, whether such discrimination has reached a certain requisite level of seriousness that can be called "arbitrary or unjustifiable". Here, an effect test is a natural and logical way to gauge the serious level of the discrimination. (In Tyres, the Panel found the discrimination resulting from the Mercosur exemption had not reached such degree of seriousness, whereas the discrimination resulting from the used tyre exemption had.)
In short, the word "unjustifiable" in the chapeau context should be understood as referring to the serious nature of the discrimination, and not merely as a general inquiry of whether any discrimination resulting from the application of the measure is justifiable by the policy objective. The AB in Tyres interpreted "unjustifiable" as such a general inquiry, thus confusing the chapeau standards with the MFN and NT standards.
It is interesting to note that the AB in Gasoline and Shrimp did distinguish any discrimination from arbitrary or unjustifiable discrimination, but failed to grasp the role of policy objective in determining the existence of any discrimination. The AB in Tyres corrected that failure, but at the same time threw away the two-part analysis.
