(iv) Terms and conditions more favorable than those commercially available ... (Cont.)

4.420 New Zealand noted that the Appellate Body had indicated that whether something was a "like" or a "directly competitive" product was to be determined on a case-by-case basis. ²⁸⁰ Applying the relevant factors, such as the product's end use, consumer tastes and preferences, and the product's properties, nature and quality, to skim and whole milk powder, New Zealand was of the view that skim milk powder and whole milk powder were not "like or directly competitive" products with fluid milk in the context of the production of the products that are exported from Classes 5(d) and (e). In other words, processors would not import skim milk powder or whole milk powder at world prices in order to export the same products - skim milk powder and whole milk powder - at the same world prices. Skim milk powder and whole milk powder were two important products exported from Special Class 5(d) or (e) milk. For other products (principally butter and cheese) exported from Special Class 5(d) or (e) milk, while it might be technically possible to substitute skim milk powder or whole milk powder for milk in their production, it was highly unlikely that this would be an economically viable approach given the resultant products had to be exported at world prices. In general, such a reconstitution of milk only occurred for sales of liquid milk in domestic markets where domestic production was insufficient to meet demand. As Canada had stated "there have not been any imports of raw industrial milk in recent years under the Import for Re-Export Program". ²⁸¹ New Zealand considered that this was further evidence that the choice between domestic and imported milk was not unrestricted and did not depend only on commercial considerations in the meaning of Paragraph (d) of the Illustrative List of Export Subsidies.

4.421 New Zealand contended that the primary reference of the term "products" in brackets in the last part of Paragraph (d) was to the term "imported or domestic products" in the second line of the paragraph. New Zealand understood the paragraph to mean that, in order to constitute a subsidy, the imported or domestic products in question had to be made available on terms more favourable than the terms commercially available on world markets for "like or directly competitive" products. The bracketed reference to "products" made clear that there was no such limitation in the case of services.

4.422 The United States argued that the express reference to "like or directly competitive" products in Paragraph (d) of the Illustrative List was made in connection with the discussion of products provided "for use in the production of goods for domestic consumption". However, because Paragraph (d) contemplated a comparison between the "terms or conditions" applicable to products used in the production of exported goods and those "commercially available" on world markets to exporters, that comparison could only be meaningful if the products themselves were comparable. Thus, any products available on the world market used in such comparison had to also be "like or directly competitive" with the products used in the production of exported goods. The United States emphasized that it did not consider milk powders to be "like or directly competitive" with fluid industrial milk within the meaning of that phrase as used in Paragraph (d) of the Illustrative List of Export Subsidies.

4.423 The United States argued that the natural starting point for construing the meaning of the phrase "like or directly competitive products" was the definition of "like product" contained in the SCM Agreement. Footnote 46 of that Agreement stated that "[t]hroughout this Agreement the term 'like product' shall be interpreted to mean a product which is identical, i.e., like in all respects to the product under consideration, or in the absence of such a product, another product which, although not alike in all respects, has characteristics closely resembling those of the product under consideration." Although the definition of like product appeared in Part V of the SCM Agreement, use of the clause "throughout this Agreement" made clear that this definition was equally applicable to the provisions of the Illustrative List of Export Subsidies contained in Annex I to the Agreement. The United States noted that the definition of "like product" contained in the SCM Agreement had been applied in only one WTO proceeding to date: Indonesia - Certain Measures Affecting the Automobile Industry. ²⁸² There the Panel emphasized that the definition required not only that the characteristics of the products being compared resemble one another, but that the characteristics "closely" resembled one another. ²⁸³ The Panel had also found that an important element to be considered were the physical characteristics of the products involved, although its analysis need not be confined to physical characteristics alone. ²⁸⁴ In addition, the Panel had observed that tariff classification principles might be useful because they provided "guidance as to which physical characteristics between products were considered significant by Customs experts." ²⁸⁵

4.424 In light of the above, the United States first observed that for customs classification purposes fluid milk and milk powders were classified in different categories. Fluid milk was classified in item 0401, whereas milk powders were classified in 0402. Second, there were obvious differences in the physical characteristics of fluid and powdered milk, the foremost being that one was in liquid form while the other has been dried to powdered form. In addition, fluid industrial milk contained butterfat, whereas skim milk powder contained no butterfat at all. These physical differences resulted in constraints on the use of milk powders in particular end-uses, and/or require additional processing steps for their use. Skim milk powder, because it lacked butterfat, could not alone be used for any of the multitude of dairy end-uses where butterfat was required. That butterfat had been added in any formulation using skim milk powder where butterfat was required underscored the lack of competitiveness of skim milk powder in such end-uses. Fluid industrial milk because it contained butterfat was not subject to a similar constraint. This fact, in itself, suggested that skim milk powder and fluid industrial milk did not "closely resemble" each other in terms of physical characteristics and, therefore, were not like products. The United States noted that the same conclusion, i.e., that it was not a like product, was warranted with respect to whole milk powder. Again, whole milk powder and fluid milk were not classified in the same tariff category, and differed in terms of their physical form, one being in a liquid form, the other being a powder. Although both fluid industrial milk and whole milk powder contained butterfat, the fact that all liquid had been removed from whole milk powder meant that in almost all instances before it can be used it had to be rehydrated. This process was both time consuming and required additional costs, as well as additional equipment. The US industry estimated this cost to be approximately US$70 per hundredweight.

4.425 The United States further noted that whether the powder was produced using high intensity heat or lower temperatures also would affect the nature of the powder produced and the range of products in which it could be used without alteration of the final processed dairy product. For example, powdered milk produced with high temperatures required a lengthier period, often as much as 24 hours, for rehydration. The high temperatures used also frequently imparted a taste to the powder which altered the flavour of the finished product from that which would be obtained by using fluid milk. In many instances, the powder, once rehydrated, was used only in products where its taste would be obscured by means of flavouring, such as in ice cream.

4.426 The United States further noted that the differences in physical characteristics also evidenced itself in the manner in which the respective products were used. In broad terms, in Canada 40 per cent of fluid milk was used for beverage purposes, 40 per cent was used in the production of cheese, and 20 per cent was used in all other dairy products. In comparison, relatively small amounts of milk powder were used for beverage purposes in Canada or the United States. Only comparatively small amounts of powder milk were used in cheese, and primarily for fortification, i.e., to increase the protein level. For example, in two of the primary cheese products, mozzarella and cheddar, milk powder normally represented no more than 3 per cent of the product by weight. Powder could be used to fortify cheese because of its relatively high protein level compared to fluid milk. Another physical difference was the high perishability of fluid milk. Whereas fluid milk would spoil within a period of days even when refrigerated, milk powder could be stored for lengthy periods of time, measured in months, if not years.

4.427 The United States noted that two separate GATT Panel reports had concluded that the differences between fluid milk and various products derived from milk were sufficient that they generally did not constitute like products or compete directly within the meaning of Article XI:2 of the GATT 1947. In Japan - Restrictions on Imports of Certain Agricultural Products ²⁸⁶ , a Panel examined a broad variety of dairy products, including processed cheese, prepared whey, skim milk powder, and whole milk powder and determined that "a product in its original form and a product processed from it could not be considered to be 'like products'" for purposes of Article XI:2. ²⁸⁷ In Canada - Import Restrictions on Ice Cream and Yoghurt ²⁸⁸ , another Panel reached a similar conclusion, finding that neither ice cream nor yoghurt competed directly with raw milk.

4.428 The United States argued that the conclusions in the cited Panel reports were equally applicable in the present context. Although those Panels were interpreting the terms "like product" and "directly competitive" in the context of Article XI:2 of the GATT, the use of those terms in the SCM Agreement should be given an equally narrow meaning. In fact, the 1987 Panel Report on Japan - Customs Duties, Taxes and Labeling Practices on Imported Wines and Alcoholic Beverages stated that it was aware of the "very narrow definition" for the term "like product" for antidumping purposes. ²⁹⁰ Since the definition of like product in the Antidumping Agreement was identical to that currently found in the SCM Agreement, the Alcoholic Beverages Panel's observations regarding the narrow definition of "like product" were germane to this dispute. Given all of the foregoing, including the significant differences in physical characteristics described above, milk powders and fluid milk were not "like products".

4.429 The United States submitted that, for very similar reasons, milk powders and fluid industrial milk were not directly competitive products. ²⁹¹ Because the phrase "directly competitive products" was not defined in the SCM Agreement, the ordinary meaning of the term in light of its context, and the object and purpose of the provision were our starting point, consistent with Article 31 of the Vienna Convention. The purpose of the requirements in Paragraph (d) that a product be "like or directly competitive" and that the product be provided on "terms and conditions" no less favorable than the product supplied by the government of the country of exportation were to allow a determination whether a benefit was conferred to an exporter when the government provided for a product to be received on more favourable terms for export than for production for domestic consumption. This determination necessitated a comparison between products that closely resembled each other, otherwise the risk of an apples to oranges comparison might result. A determination whether products were provided on the same favorable "terms and conditions" would be impossible unless the products closely resembled each other. Products that were materially different would not be comparable for purposes of determining the equivalence of the applicable terms and conditions. Therefore, to construe the phrase "directly competitive" to encompass products that while substitutable for certain purposes were not interchangeable for most end-uses, either because of differences in physical characteristics or other reasons, would appear to undermine the ability to determine whether the applicable terms or conditions were similar.

4.430 The United States noted that, as indicated above, the Canada - Yoghurt Panel found that to be "directly competitive" a buyer would have to be indifferent in its choice between the products in issue. In the view of the United States, a buyer would not be indifferent to the choice between fluid industrial milk and milk powders because of the differences in physical characteristics, limited substitutability in end-uses, and the additional processes associated with reconstituting powders prior to their use in further manufacture of dairy products. Consequently, milk powders were not "like or directly competitive" with industrial milk provided through the special class system.

4.431 The United States submitted that cheese imported by Canada in the Import for Re-Export Program was not like or directly competitive with fluid industrial milk based on the significant and obvious physical differences between the products, as well as the inability to substitute cheese in most of the end-uses for milk. The United States also noted that beyond the question of whether milk powders were like or directly competitive products, there was the separate question of whether milk powder imports were available on the same favorable terms or conditions on which fluid milk was available to exporters under the special class system. The United States submitted that they were not. Import data supplied by Canada pertaining to the Import for Re-Export Program indicated that virtually all such imports consist of various manufactured dairy products, not raw milk, as was available under Special Classes 5(d) and (e). Moreover, those dairy ingredients were on a "milk equivalent" basis priced higher than industrial milk provided pursuant to Special Classes 5(d) and (e).

4.432 The United States used Canada's data ²⁹² on quantities and values of selected dairy products imported under the Import for Re-Export Program to calculate the equivalent price, or cost, of "milk" imported for calendar years 1995, 1996, 1997 and 1998 to date. ²⁹³ The products selected included dry milk, not containing sweetening matter, with greater than 1.5 per cent fat, or whole milk powder (WMP), cheese of all types, and fluid milk, not concentrated, containing no sweetening matter, with between 1 and 6 per cent fat. These three products accounted for approximately 75 per cent of the total value of 1997 dairy product imports for re-export. ²⁹⁴ Unit values of imports for re-export were computed in a straightforward manner (total value/total quantity), converted to US dollars per metric tonne, with international prices, as reported by the US Department of Agriculture, provided as a comparison. These unit values were then converted to a milk equivalent basis, using yield factors and manufacturing margins based on a combination of data from the CDC and factors used by the US Department of Agriculture, as shown in the immediately following table. Comparative data for Special Class prices 5(d) and (e) were computed for calendar year 1997 and for 5(e) for marketing year 1996/97. ²⁹⁵ The 1997 price data for Special Classes 5(d) and (e) were based on statistics from Agriculture and Agri-Food Canada ²⁹⁶ , and were weighted-averages based on 71 per cent and 81 per cent, respectively, of the 1997 volume of milk sold under Special Classes 5(d) and (e) in Canada (calculations also attached). The 1996/97 marketing year price data for Special Class 5(d) and (e) were computed based on Class 5(e) component values for Quebec, which represented 64 per cent of total Class 5(e) milk sales in 1997, using milk composition factors of 3.9 per cent for butterfat, 3.3 per cent for protein, and 5.21 per cent for other milk solids.

4.433 The United States contended that, based on the above-described calculations, contrary to Canada's assertion, Canadian processors did not have access to dairy products under the Import for Re-Export Program at prices comparable to prices for raw milk available under Classes 5(d) and (e). The "milk equivalent" prices for whole milk powder, cheese, and fluid milk imported through the Import for Re-Export Program were in each instance higher than the corresponding prices for special class 5(d) and (e) milk.

4.434 Canada noted that the reference to "products", within brackets, in the last part of Paragraph (d) of the Illustrative List, included "like or directly competitive products". Paragraph (d) referred initially to both "products or services" ²⁹⁷ and then subsequently again to both "products and services" ²⁹⁸ . The second reference to "products and services" was modified by the phrase "like or directly competitive". The use of the word "product" in the final condition of Paragraph (d) ²⁹⁹ followed the reference to "like or directly competitive products and services". Thus, the word "product" referred back to the antecedent reference to "like or directly competitive products and services." Accordingly, the purpose of the use of the word "product" in the bracketed portion in the final condition in Paragraph (d) was to indicate that this final condition applied only with respect to "products"; not to both "products and services". Since the antecedent reference was to "like or directly competitive products and services", then the qualifying term "like or directly competitive" similarly applied to "products" in the final condition.

4.435 Canada argued that in the dairy industry in Canada, and abroad, skim milk powder and whole milk powder were considered to be like or directly competitive with whole milk. In addition, raw milk could be a less economically efficient input for a manufacturer compared to dairy products that supplied the components needed for a particular manufacturing process. Milk powder was, of course, much cheaper to transport due to the fact that raw milk is over 80 per cent water. Milk powders could be reconstituted for use in the manufacture of some dairy products. They could also be used to standardize for protein and butterfat for a variety of products. Indeed, to the extent that there was a cost advantage to do so, dairy product manufacturers would use milk powders instead of raw milk. Thus, to a certain extent, milk powders competed in the same markets and fulfilled the same needs and uses as fluid industrial milk. These products could be used interchangeably in many applications. The differences between fluid industrial milk and milk powders primarily related to storage and transportation, not end-use.

4.436 Canada drew the Panel's attention to the US Dairy Export Council website where the US dairy processing industry claimed: "U.S. powdered milk can be easily reconstituted into wholesome drinking milk or used as an ingredient in baked goods, dairy products, confections and other prepared foods". ³⁰⁰ Similarly, a study prepared by the United States Department of Agriculture (USDA) had stated, in respect of SMP that it "can be substituted for fresh skim milk or condensed skim milk in many manufactured products" and "converted easily back to milk, and with little formulaic adjustment...[and] can be substituted for condensed skim in many processing operations". ³⁰¹ Canada further noted that the United States claimed that SMP normally represented no more than 3 per cent of (mozzarella and cheddar cheese products) by weight (paragraph 4.426). However, the same USDA study just referred to stated that, on a skim equivalent volume basis, skim milk powder usage had grown to "25.4% of the dairy ingredients in the manufacture of mozzarella and other Italian-cheeses in one region of the United States and 16.1% nationally". ³⁰² Perhaps it was for this reason that the US Dairy Export council proudly proclaimed on its website that "the US milk powder industry is capable of unrestrained growth." ³⁰³

4.437 Canada argued that the GATT Article XI Panel Reports relied on by the United States (paragraph 4.427 and following) had little to do with whether, for the purposes of Paragraph (d) of Annex I of the SCM Agreement, milk powders and fluid industrial milk were "like or directly competitive products." GATT Article XI:2(c)(i) was concerned with restrictions on the importation of primary products "in any form". This expression was subject to the definition in Ad Article XI:2(c). Furthermore, GATT Article 2(c)(i) created an exception to the general prohibition created in Article XI:1, and thus had been construed narrowly. ³⁰⁴ The "like or directly competitive products" concept in Paragraph (d) of the Illustrative List of Annex I to the SCM Agreement was a defining part of that list and thus was to be given its full and ordinary meaning in accordance with the principles of treaty interpretation.

4.438 Canada argued that the United States had given scant attention to more authoritative and more relevant Appellate Body jurisprudence on concepts of "directly competitive or substitutable products"; jurisprudence that plainly supported the conclusion that milk powders directly competed with fluid industrial milk. ³⁰⁵ Apart from being of minimal relevance, the Panel Report in Canada - Yoghurt relied on by the United States was readily distinguishable on its facts. That Panel had considered whether, from the standpoint of the end-use consumer, ice cream and yoghurt competed directly with fresh milk. This was an entirely different issue from assessing, from the standpoint of processors, whether milk powders and fluid industrial milk were "like or directly competitive" products. In respect of the other GATT Article XI case cited by the United States (Japan - Restrictions on Imports of Certain Agricultural Products) seemed to lend support to the common-sense proposition that milk powders competed directly with fluid milk. What the United States failed to mention was that the Panel in that case appeared to accept that milk powders and certain other dried and condensed milk products "could compete directly with fresh milk for manufacturing use." ³⁰⁶

4.439 Canada refuted the validity of the "milk equivalent" prices calculated by the United States as provided in US Exhibit 56 (paragraph 4.432). Canada argued that dividing the total reported trade value by the total reported trade volume in trade statistics was an unreliable guide to actual prices. Thus, the data submitted by the United States (the Dairy Market Review attached to Exhibit 36) showing the highest Canadian domestic milk price for Class 1 use was C$65.35/hl (US$48/hl), together with an estimate of import costs of US$116/hl. Canada argued that it was obvious that either the estimated unit import value was wrong, or the data was not comparable. The alternative was to assume that Canadian importers freely chose supplies that were twice as expensive as the most expensive domestic alternative. Canada noted that in the case at issue, much of the fluid milk imported under the Import for Re-export Program was imported in retail packages for use on cruise ships. It was hence, clearly inappropriate to make price comparisons between retail packages on one hand and milk supplies from farmers to processors on the other.

4.440 Similarly, Canada argued that the United States estimates of unit import value for "cheese of all types" was simply an average of high and low value cheese of vastly different types. These were average import values, together with unsubstantiated conversion factors were then used to generate a "milk equivalent cost of cheese imported for re-export". This approach was invalid because the yield of cheese from a hectolitre of milk varied widely between varieties of cheese. As the mix of varieties imported varied from year to year, a conversion using some "standard" yield of "generic" cheese from a hectolitre of milk would bear no consistent relationship to the actual value of milk used in the production of imported cheeses. In the case of whole milk powder, which was a much more uniform commodity product for which it may be valid to use a simple conversion formula, the United States data showed that, the US estimated average Class 5(e) price was very similar to the estimated "milk equivalent cost of WMP imported for re-export" in 1997 and 1998. While the problems noted above with respect to the use of unit import values remained, this gave an indication that, in the one situation in Exhibit 56 where the construction of milk equivalent prices on the basis of formulae was most defensible, the resulting data supported Canada's argument.

4.441 Canada further noted that on page four of Exhibit 56 the United States estimated milk prices in Classes 5(d) and (e) on the basis of prices estimated for individual provinces (except Ontario). The resulting estimated prices were incorrect. The actual Canadian average return for Class 5(d) sales in the 1997 calendar year was $26.22/hl. For Class 5(e), the return was $22.98/hl. ³⁰⁷ Canada argued that the discrepancy between these real figures and the values constructed by the United States was partly explained by the fact that the United States included a Class 5(d) price of $11.98/hl for Manitoba, although it was self-evident that this was a data error in the 1997 Dairy Market Review. (For the record, the correct datum should have read $22.01). This should have been obvious from the fact that the Class 5(d) prices reported in the Review for all other provinces were between $31.31/hl and $21.66/hl. It was questionable statistical practice to simply use what was clearly an anomalous figure without verifying from other sources. In addition, the data set used by the United States did not include prices from Ontario. The flaws in the data presented by the United States meant that no weight can be attached to them.

4.442 The United States noted that Canada carefully qualified its statement in respect of the "like or directly competitive" nature of milk powders with fluid milk (paragraph 4.435). Canada stated that milk powders could be used in the manufacture of some dairy products. Similarly, Canada stated that to a certain extent milk powders competed in the same markets. The United States argued that direct competition contemplated more than a certain degree of competition or the ability to substitute for fluid milk in some uses.

(d) Article 3

4.443 The United States argued that once a violation of Part V of the Agreement on Agriculture had been found, Canada had to bring the operation of its export subsidies into conformity with its obligations, particularly those establishing its export reduction commitments respecting the quantity of subsidized exports. This was true, regardless of whether a separate violation of the SCM Agreement was also found.

4.444 The United States noted that Article 3 of the SCM Agreement was addressed by Article 13 of the Agreement on Agriculture. That provision directed that only export subsidies that conformed with Part V of the Agreement on Agriculture were exempt from challenges pursuant to Article 3 of the SCM Agreement. Furthermore, the United States claimed that Article 3 of the SCM Agreement was within the Panel's terms of reference. Consequently, the United States requested that the Panel make separate findings respecting the consistency of the Special Class system with the SCM Agreement, including a finding whether the Special Class system provided a prohibited export subsidy within the meaning of Article 3.

4.445 The United States claimed that as Canada's Special Milk Classes Scheme was an export subsidy of dairy products in excess of the limits for Canada under the Agreement on Agriculture, whether they fell under Article 9 or Article 10 of that Agreement, as a result, those subsidies did not benefit from the exemption in Article 13(c)(ii) of that Agreement on Agriculture. Consequently, the United States claimed that these export subsidies were also inconsistent with Canada's obligations under Article 3 of the SCM Agreement.

To continue with Protected Markets and Export Subsidies

²⁸⁰ Appellate Body Report on Japan - Liquor Tax, op. cit., pp.20-21; Appellate Body Report on Canada - Periodicals, op. cit., p.21.

²⁸¹ Canada's Oral Submission of 17 November 1998, para. 74.

²⁸² Panel Report on Indonesia - Automobile Industry, op. cit. The United States further noted that the term "like product" appeared in various provisions of the GATT 1994, as well as the WTO Agreements, and its meaning had been determined to depend on its precise usage in a particular agreement. Thus, the Appellate Body in Japan - Liquor Tax, op. cit., p. 21, stated: "The concept of 'likeness" is a relative one that evokes the image of an accordion. The accordion of "likeness" stretches and squeezes in different places as different provisions of the WTO Agreement are applied. The width of the accordion in any one of those places must be determined by the particular provision in which the term "like" is encountered as well as by the context and the circumstances that prevail in any given case to which that provision may apply."

²⁸³ Panel Report on Indonesia - Automobile Industry, op. cit., para. 14.172.

²⁸⁴ Ibid, 14.172. 14.173.

²⁸⁵ Ibid. The Appellate Body in Japan - Liquor Tax, op. cit., also found the tariff classification of a good to be relevant to the determination of "likeness", p. 21.

²⁸⁶ Report on Japan - Certain Agricultural Products, op. cit.

²⁸⁷ Ibid, p. 231, para. 5.3.1.4.

²⁸⁸ Panel Report on Canada - Yoghurt, op. cit.

²⁸⁹ Ibid, p. 89, para. 73.

²⁹⁰ Panel Report on Japan - Alcoholic Beverages, (hereafter "Japan - Alcoholic Beverages") L/6216, BISD 34S/83, adopted 10 November 1987, pp. 115-116, para. 5.6.

²⁹¹ The United States noted that in the different context of Article III:2 of the GATT 1994, the Appellate Body in Japan - Liquor Tax, op. cit., indicated that the category of "directly competitive or substitutable products" may be broader than the class of "like products", but stated that this determination was "a matter for the panel to determine based on all relevant facts in that case" and had to be determined on a case-by-case basis (p.25). The Appellate Body further observed that consideration of such factors as physical characteristics, common end-uses, and tariff classifications were appropriate for this purpose. Ibid.

²⁹² The United States referred to the data attached to Canada's comments on the US Oral Statement at the Second Substantive Meeting of the Panel ("Comments by Canada on Oral Statement of United States, Second Panel Hearing, November 17-18, 1998").

²⁹³ Contained in US Exhibit 65.

²⁹⁴ The United States claimed that imports of skim milk powder under the Import for Re-Export Program were at such extremely low levels during the last two calendar years that their presence could not be considered to provide an alternative to milk provided through the Special Classes Scheme.

²⁹⁵ The United States noted that the most complete and reliable data available was for 1997 and marketing year 1996/97.

²⁹⁶ Attached to United States, Exhibit 56.

²⁹⁷ " ... of imported or domestic products or services for use in the production of exported goods ... "

²⁹⁸ " ... for provision of like or directly competitive products or services ... "

²⁹⁹ " ... if (in the case of products) such terms or conditions are more favorable than those commercially available on world markets to their exporters."

³⁰⁰ http://www.usdec.org/cgi_win/usdec.exe/section24  

³⁰¹ A Review of Class III - A Pricing under Federal Milk Marketing Orders, Dairy Division, Agricultural Marketing Service, United States Department of Agriculture, March 1995, pp.1-2.

³⁰² Ibid., at pp. 17 and 22, referring to the years 1992 to 1994.

³⁰³ A Review of Class III - A Pricing under Federal Milk Marketing Orders, Dairy Division, Agricultural Marketing Service, United States Department of Agriculture, March 1995, pp.1-2.

³⁰⁴ Panel Report on Japan -Certain Agricultural Products, op. cit., para. 5.1.3.7; Panel Report on Canada -Yoghurt, op. cit., para. 59.

³⁰⁵ Canada referred to the Appellate Body Report on Japan - Liquor Tax, op. cit. and the Appellate Body Report of Canada -Periodicals, op. cit.

³⁰⁶ Appellate Body Report on Japan - Liquor Tax op. cit., para. 5.3.1.4. Canada noted that the Panel went on to find that milk powders did not meet the perishability requirements of Ad Article XI:2(c), which had no counterpart in Paragraph (d) of Annex I to the SCM Agreement.

³⁰⁷ Canada noted that as dairy farmers in Canada were actually paid based on the butterfat, protein and other solids components of milk, it was possible to generate differing estimates of the price of a hectolitre of milk depending whether one used the average composition of milk delivered or a standard milk composition. At the individual farm level, payment was based on the components delivered by the farmer, so, while the problem of translating to hectolitres of milk complicated life for analysts, it had no impact on the farm-level economics of the system.