This is from the second EU written submission in seal products:
199. Canada considers that the EU Seal Regime, through the IC exception, modifies the conditions of competition to the detriment of its imports when compared with imports of Greenland. Specifically, Canada suggests that, in the case of an origin-neutral measure, like the EU Seal Regime, a quantitative analysis of the treatment provided to the group of products in question as a whole shows de facto less favourable treatment.221 Canada thus considers that the detrimental impact is shown because most of its seal products cannot be placed on the EU market whereas most of Greenlandic seal products can be placed on the EU market through the operation of the IC exception.
200. The European Union considers that Canada's reliance on its "quantitative approach" to show de facto discrimination is manifestly incorrect.222 To illustrate Canada's arguments with the following drawing, Canada submits that where a measure is designed, is expected to result or indeed results in a specific data pattern (as the one shown in the drawing), such fact alone is sufficient to show detrimental impact.
[The EU has a nice graphic here, but it was not easy to reproduce in this post]
204. In order to ensure that the comparison required at this stage of the analysis is not skewed towards finding less favourable treatment (i.e., by taking individual products in the group of imported products and comparing them with other individual products in the other group) and yet preserves the analysis of the group as a whole, it is necessary to compare the aggregate competitive opportunities afforded. This requires an analysis of the various categories within each of the groups of products at issue. Specifically, the Panel would have to compare the treatment accorded by the EU Seal Regime to Canadian seal products derived from hunts for commercial purposes with the treatment provided to "like" products derived from hunts for commercial purposes (e.g., Greenland/Norway). Such comparison leads to the result that all those products are subject to the same treatment, i.e., the General Ban. Further, the Panel would have to compare the treatment accorded by the EU Seal Regime to Canadian seal products derived from hunts for subsistence purposes with "like" products of other origins derived from hunts for subsistence purposes (e.g., Greenland/Norway). Such a comparison leads to the conclusion that all those products are subject to the same treatment, i.e., they are allowed to be placed on the EU market under the IC exception. The categories of "like" products used to make the above comparison are not "discretionary" or "arbitrary". They are based on the "regulatory distinction" embedded in the EU Seal Regime, specifically, in the IC exception, and thus are comparable.
205. Thus, when making a category-to-category comparison within the groups of products in each side of the comparison (i.e., Canada versus other origins, such as Greenland/Norway), it should be concluded that there is no alteration of the aggregate competitive opportunities in favour of those other origins' groups of products. Put in simple terms, this approach shows that there is no discrimination since each category in the same situation (by reference to the "purpose" of the hunt) is treated equally and has identical access to (or prohibition to access) the EU market.
If I understand all this correctly, the EU is saying that in order to determine whether detrimental impact exists, you cannot compare the distribution of imported and domestic products across the different regulatory categories that exist. Instead, you have to compare only within an individual category.
I'm not sure how well that approach works, though. If the argument is that imports disproportionately fall into the disfavored categories, how would you evaluate the issue without a comparison that looks across categories?
Also, in Chile - Alcohol, the AB said this:
51. Chile has argued that there is "similar taxation" of domestic and imported production under the New Chilean System because all beverages with a specific alcohol content are subject to identical ad valorem tax rates, irrespective of their origin. In making this argument, Chile invites us to focus exclusively on a comparison of the relative tax burden on domestic and imported products within each fiscal category and to disregard the differences of tax burden on distilled alcoholic beverages which have different alcohol contents and which are, therefore, in different fiscal categories. In other words, Chile asks us to disregard the comparison which the Panel undertook of the relative tax burden on domestic and imported products located in different fiscal categories.
52. It is certainly true that, as Chile claims, if we were to focus the inquiry under the second issue solely on a comparison of the taxation of beverages of a specific alcohol content, we would have to conclude that all distilled alcoholic beverages of a specific alcohol content are taxed similarly. However, as we stated at the outset, in our analysis we must assume that the group of directly competitive or substitutable products in this case is broader than simply the products within each fiscal category. Chile's argument fails to recognize that the Panel has found, and Chile has not appealed, that imported beverages of a specific alcohol content are directly competitive or substitutable with other domestic distilled alcoholic beverages of a different alcohol content. To accept Chile's argument on appeal would, we believe, disregard the objective of Article III, which is to "provide equality of competitive conditions" for all directly competitive or substitutable imported products in relation to domestic products, and not simply for some of these imported products. The examination under the second issue must, therefore, take into account the fact that the group of directly competitive or substitutable domestic and imported products at issue in this case is not limited solely to beverages of a specific alcohol content, falling within a particular fiscal category, but covers all the distilled alcoholic beverages in each and every fiscal category under the New Chilean System.
It seems to me the AB has already rejected a view that comparisons within a category would be appropriate in this part of the analysis.
Of course, this is all just about the initial step of the analysis. Later, you get to the "discrimination" part, and there you can look at the legitimacy of the different categories.