Like the Cloves Panel report, the Cloves AB report has crucial lessons for national treatment more broadly. Two stand out.
1. “Likeness” in all its variations is now about competition, not content or purpose: But who decides?
Already in Philippines – Spirits, the AB recently confirmed that also “likeness” under GATT Art. III:2, first sentence, is purely about competition (in that case, finding that even drinks with fundamentally different content such as different raw materials were “like” on the ground that they anyhow compete). In earlier cases, it had said the same for Art. III:2, second sentence (directly competitive or substitutable) and Art. III:4 (like).
The same competition test is now confirmed for TBT Art. 2.1. The AB rejected the Panel’s regulatory context approach which can be found also in investment arbitration cases (para. 112).
So, what matters is not that both clove and menthol cigarettes raise the same “regulatory concern” of attracting young smokers, but that both types of cigarettes are in a sufficient “competitive relation”. This focus on competition increases the importance of economic and survey evidence before panels.
However, whom should we ask to see whether products, here sub-groups of cigarettes, sufficiently compete? First, the AB says “all relevant consumers of the products at issue, not only … the main consumers of clove and menthol cigarettes [youth], particularly where it is clear that an important proportion of menthol cigarette smokers are adult consumers” (para. 137). But a few paras. later, the AB is nonetheless content with hypothetical evidence (the panel had refused to look at survey evidence) only in respect of a segment of the market, that is, young smokers: “young and potential young smokers perceive clove and menthol cigarettes as sufficiently substitutable. This, in turn, is sufficient to support the Panel's finding that those products are like” (para. 144); "actual competition does not need to take place in the whole market, but may be limited to a segment of the market" (para. 143).
I wonder what to make of this. Is there not a risk that “likeness” will be found too easily simply based on how the claimant identifies the specific market segment to be looked at (here youth smokers v. all smokers). So if you are a complainant, all you need to do is find a segment of consumers that considers, for example, high-carbon and low-carbon products to be like, or does not care about whether products were made with growth hormones or are genetically modified. If you can find such a segment, the products compete and you have “likeness” ?
2. "Less favourable treatment" test is now different in TBT as compared to GATT: Why?
The AB report said as much about the GATT national treatment standard as the TBT standard. Most importantly, the AB finds that under GATT, a national treatment violation occurs as soon as a measure modifies “conditions of competition in the marketplace to the detriment of the group of imported products vis‑à‑vis the group of domestic like products” (para. 179; that said, and very importantly, “all like products imported from the complaining Member” must be compared to “all like domestic products”; national treatment “does not require Members to accord no less favourable treatment to each and every imported product as compared to each and every domestic like product”, para. 193).
Unlike what many commentators and a number of panels had thought (e.g. EC-GMO), under GATT, there is no scope for an “inquiry as to whether such detrimental impact was related to the foreign origin of the products or explained by other factors or circumstances” (footnote 372).
Under TBT, in contrast, this type of additional requirement does apply: a measure will be excused if “the detrimental impact on imports stems exclusively from a legitimate regulatory distinction rather than reflecting discrimination against the group of imported products” (para. 182).
But why not apply the same (more deferential) test under GATT Art. III? If the vague preamble of TBT justifies this deference, then why not the explicit GATT Art. XX exception? One could argue that it is exactly because of GATT XX that we do not need the deferential approach under GATT Art. III itself. But this risks ending up with more deference in the agreement without exception (TBT, where e.g. the AB refers to any "legitimate regulatory distinction" and the burden remains on the complainant) than the deference offered in the agreement with an exception (GATT, where e.g. the list of justifications is limited, the burden is on the defendant and additional requirements in the chapeau apply).
Allowing a defendant to excuse detrimental effect with reference to legitimate purposes in both TBT Art. 2.1 and GATT Art. III may make GATT Art. XX less useful to justify de facto violations of national treatment. But Art. XX was most probably never written with these violations in mind and would remain useful to justify many other GATT violations (in particular de jure violations of Art. III, and Art. XI). Given the differential test now espoused by the AB, are we not likely to see inconsistencies and strategic filing under GATT instead of TBT or vice versa depending on what works best for the claimant?
As always, written in my personal, academic capacity, but interested in your views!
