I found this to be an excellent discussion of "likeness", and the non-discrimination principle more generally, under the GATS: http://www.wto.org/english/res_e/reser_e/ersd200608_e.pdf . It's by Mireille Cossy of the WTO's Trade in Services division. I felt like she really addressed some of the hard issues involved in an intelligent way. There are no easy answers, of course, but she asks the right questions. Unfortunately, there's too much packed in there for me to give a quick summary, but I recommend taking a look if the subject interests you. Here's the summary provided on the WTO web site:
The concept of “like services and service suppliers” used in the General Agreement on Trade in Services (GATS) is still very much uncharted territory. The few dispute cases involving national treatment and most-favoured-nation treatment claims under the GATS are vague concerning the criteria which should be used to establish “likeness”. Discussions among WTO Members on this subject have remained limited and inconclusive. Perhaps the only point on which everybody agrees is that a determination of “likeness” under the GATS gives rise to a wider range of questions — and uncertainties — than under the GATT. The intangibility of services, the difficulty to draw a line between product and production, the existence of four modes of supply, the combined reference to like services and like service suppliers, and the lack of a detailed nomenclature are some of the factors which complicate the task of establishing “likeness” in services trade.
This contribution focuses on the concept of “likeness” in the context of the national treatment obligation (Article XVII of the GATS). It discusses the possible implications of the combined reference to “like services and service suppliers”, as well as the relevance and role of the modes of supply in determining “likeness”. It also examines whether the criteria developed by GATT case-law (physical properties, classification, end-use and consumer tastes) can be mechanically transposed to services trade and how far they may contribute to establishing “likeness” under the GATS. It then discusses whether other parameters, such as the regulatory context or an “aim and effect” type approach could be relevant.
