Since the AB's ruling in US-Cloves on TBT 2.1, where it introduced the notion that a measure does not violate National Treatment or MFN if the detrimental effects on competition are due exclusively to a legitimate regulatory distinction, there has been much discussion on this blog and elsewhere of whether the AB was intending a symmetrical approach to the interpretation of GATT Article III:4. One view was that the AB had introduced this step into the application of Article 2.1 because of the absence of an Article XX-like exceptions provision in TBT; it was thus inapplicable to GATT III:4. The alternative view was that, because of its concern with preserving an appropriate balance between trade liberalization and the right to regulate under both the GATT and the TBT Agreement, the AB also saw the interpretation of IIII:4 as involving the challenge of ensuring that neutral, objective measures with incidental trade effects, but without any un-evenhanded design features, are not impugned as violations of the GATT (even if the regulating Member might be able to justify them under Article XX in the last resort). On this latter view, which I hoped the AB was moving towards, it seems questionable, given the purpose stated in GATT Article III:1, namely anti-protection, to characterize the enactment of normal government regulation with some incidental negative effects on the competitive opportunities of foreigners as a presumptively internationally wrongful act requiring strict justification if international wrongfulness is to be procluded.
The AB has now (largely) ended the speculation, indicating that it takes the first view, not the second. Policy space is, quite simply, it seems at first glance from Seals a matter for the exceptions under GATT. There is a certain clarity to this approach, which is one reason why it has been favoured by many scholars, lawyers and particularly economists. It has often been noted that the list of legitimate policy objectives in GATT Article XX is a closed one, and the EU raised this as an issue in determining whether policy space might have to be addressed in some way within Article III:4. The AB retorted that the EU was unable to name a legitimate policy objective that would not be covered by Article XX but would be legitimate withinthe meaning of TBT 2.1. Now consider whether this is a satisfactory way of dismissing the closed list problem. If the EU legal counsel names such a policy objective they are basically compromising the EU's ability to claim in a future case when defending their regulation, that the objective in question somehow fits within one paragraph or another in Article XX. The AB should not be letting the parties (with their particular litigation strategies) do all its thinking for it on such an important systemic question, especially where there is extensive scholarly literature.
Perhaps the AB is saying:trust us, if there is a legitimate regulation that nevertheless fails III:4 we will find a way of allowing it be be saved under XX. Could public morals now be viewed by the AB as a kind of residual clause of Article XX, covering public policies that don't specifically go to the objectives in the other paragraphs, as long as standards of right and wrong in the community are at issue (take consumer fraud, for example)? XX (d) may have potential along similar lines.
There is however one safety valve to preserve legitimacy that the AB has left open within III:4 itself, and this emerges from the discussion of DR-Cigarettes at paragraph 5.105. The detrimental impact on the competitive opportunities must be attributable to, i.e. have a "genuine relationship" to, the measure. This, surely, entails examining and eliminating other factors to which the detrimental impact might be atributable: climatic or geographical or demographic factors, changing consumer tastes or ethical values, policies of the governments of the exporting countries that could effect how the market is structured, etc. To establish attribution, or a genuine relationship to the measure, a panel must arguably be sure that the detrimental competitive impact in question is not due to the other factors, largely or entirely. In footnote 1071, the AB was careful to note that the EU did not appeal the panel conclusion that the detrimental impact was attributable to the measure. Instead, the EU focused on its argument that there is a third step, the inquiry into "legitimate regulatory distinction." In future cases, the regulating state will surely want the panel to consider whether detrimental impact is attributable to factors other than the measure itself.
The second safety valve for policy space within III:4 that remains is the flexible or imaginative use of the criteria for likeness, an issue that the AB essentially ignores in Seals. In this respect we should all re-read carefully the likeness discussion in EC-Asbestos.