A few months ago, I did a post praising a paper by the WTO's Mireille Cossy, which was on "likeness" and non-discrimination under the GATS. I stand by my statement that she "addressed some of the hard issues involved in an intelligent way." However, I also agree with some criticisms of her paper offered in a new paper by Joost Pauwelyn. Joost writes:
although it follows most GATT and WTO jurisprudence, as well as most commentators, Cossy's focus on "likeness" as the core element of a national treatment test, is misguided. It begs more questions than it resolves. Instead, the focus of a national treatment analysis ought to be on how to define "less favorable treatment of foreign services or service suppliers", in particular: What qualifies as de facto discrimination?
He further explains:
The obsession with “likeness” in the GATT/WTO is unheard of in any other legal arena that needs to decide cases of alleged discrimination, be it in EU law, US constitutional law or before the European Court of Human Rights. In each of these other fora the big question is rather: How to define de facto discrimination? If a law or regulation explicitly differentiates based on sex, race or nationality, de jure discrimination is easily found (even though such discrimination might still be justified under exceptions). The sticking point, of course, is to decide whether laws or regulations that are sex, race or origin-neutral on their face amount to de facto discrimination of, for example, women, blacks or foreigners.
Instead of tackling this question head-on – i.e., although the law or regulation is origin-neutral on the books, is it nonetheless more burdensome on imports? –, especially WTO panels and the Appellate Body have spent most of their energy on deciding whether the domestic products and imports compared are “comparable” or “like” in the first place. Once they find “likeness”, discrimination almost follows naturally, as an afterthought. Yet, this bottom-up approach of deciding, in the abstract, whether two products referred to by the complainant are “like” -- be it shochu and vodka; asbestos and other fibres; GMOs and conventional crops – is tackling the question from the wrong end. Instead of examining whether products are “like” with reference to some purportedly neutral criteria of physical characteristics, end use, consumer tastes and tariff classifications -- all of which are portrayed as permissible features on which to base regulatory distinctions -- panels ought to focus on the specific law or regulation that is at issue, and ask the following question: Does the law or regulation differentiate, de jure or de facto, based on the one and only impermissible criterion, namely: national origin. Instead of approaching the question from what is permissible (the universe of features that validly make products unlike), panels ought to approach the question from what is not permissible (the specifics of whether the regulation distinguishes based on origin).
Now, as most readers of this blog are probably aware, there is a long history of debate over the proper interpretation of the WTO's non-discrimination provisions. People hold very different views on fundamental questions such as (1) which factors should be taken into account and (2) under which parts of the relevant provisions should these factors be considered. I agree with what I see as Joost's view: The core issue in the de facto discrimination cases (which are the hard cases) should not be likeness, but rather "less favourable treatment." (For those provisions that use this term, anyway. For those that do not, the situation is more complicated -- too complicated to get into in this blog post!).
Rather than delve any deeper into this, I'm going to stop for now and see if anyone has any reactions to his paper. For those wondering why Joost didn't do a post himself, I asked him whether he'd be willing, but he was too busy with other projects. But, time permitting, he will probably respond if people offer comments.
