3. Product comparisons

(a) General

30. The next step is to consider the various attributes of the products at issue to determine whether these attributes support a conclusion that there is a directly competitive or substitutable relationship between the imported and domestic products. In this regard, we will examine the end-uses of the products, their physical characteristics, the channels of distribution, price relationships (including cross-price elasticities), and other relevant characteristics. ³⁶⁷

(b) End-uses

31. Overlap in end-use determines to a great extent direct competitiveness or substitutability. The Appellate Body in Japan � Taxes on Alcoholic Beverages II specifically agreed with the panel finding to the effect that:

[T]he decisive criterion in order to determine whether two products are directly competitive or substitutable is whether they have common end-uses, inter-alia, as shown by elasticity of substitution. ³⁶⁸

32. In other words, the overall inquiry focuses on whether there are common end-uses by examining a number of factors which can include elasticity of substitution. It is worth examining the extent of the current overlap of the end-uses as well as the appropriate definition of what a common end-use is for purposes of this inquiry. The current overlap of end-uses can be limited due to, inter alia, the very measures at issue, protective tariffs, resulting low volumes and high sales costs or other factors. It is also possible that the inquiry in some cases can include an examination of other relevant markets than the one in question to see if there is evidence of common end-uses of the products and take that into consideration.
33. In this regard, it is worth noting that the panel in Korea � Taxes on Alcoholic Beverages observed that:

End-uses constitute one factor which is particularly relevant to the issue of potential competition or substitutability. If there are common end-uses, then two products may very well be competitive, either immediately or in the near and reasonably predictable future. ³⁶⁹

34. The European Communities asserts that pisco and the imported distilled spirits are already used by Chilean consumers for similar end-uses. The European Communities refer to a market research done on the drinking habits of a representative sample of consumers. ³⁷⁰ The European Communities argues that on the basis of the 1997 SM Survey, pisco and the imported distilled spirits are consumed in the same way (straight, diluted with water, ice, soft drinks or fruit juice, and in cocktails). The European Communities thus argues that there is a substantial overlap of end-use between whisky and pisco, the two spirits which Chile emphasises as being the most different.
35. Chile raises questions on the probative value of the consumer surveys being relied on by the European Communities. With respect to the discussion of end-uses, Chile disagrees with the results of the 1997 SM survey. In Chile's view, questions that ask consumers what they would choose if their preferred spirit were unavailable can only lead to abnormal and unpredictable answers. Chile also points out that the data in the 1998 SM Survey ³⁷¹ is a purported quantitative analysis, in which consumers were faced with a hypothetical change in the prices of whisky, pisco and other spirits, the results of which can only be misleading.
36. Chile argues that it is oversimplistic to conclude that all spirits whose basic constitution is water and alcohol, and which are drunk mixed in not too different a manner, have necessarily the same end-uses. In Chile's view, this argument is tantamount to saying that consumers' only consideration is simply to have alcohol, irrespective of the form in which it is contained. Chile argues that even wine and beer share these characteristics and end-uses despite being different forms of alcoholic beverages.
37. Chile further argues that assertions by the European Communities that the 1997 SM Survey shows that both pisco and imported distilled spirits are consumed by Chileans in roughly similar percentages at various occasions and in various places, e.g., discos, bars, at home after work, at friends' homes etc., has no probative value. According to Chile, the categories of end-uses offered by the European Communities are simply too broad. Chile argues that pisco is more of a popular spirit in Chile than the imported spirits, such as whisky, which tend to be more expensive and, consequently, are consumed by the wealthier segment of the population.
38. The parties disagreed over the appropriate breadth of categories of end-uses. The European Communities' position is that if two spirits have similar end-uses, it is a factor tending to show that the two products are directly competitive or substitutable. On the other hand, Chile asserts that similar end-uses are common features applicable to every alcoholic beverage without being in any way determinative of the question of direct competitiveness or substitutability.
39. In our view, the 1997 SM Survey provides some useful evidence about overlapping end-uses. It tends to confirm the observation that distilled alcoholic beverages are used for relaxation and socialisation in appropriate social settings. Chile's rebuttal largely turns on the observation that the evidence submitted by the European Communities proves too much due to the broad nature of the proposed categories of end-uses. Indeed, it may very well be that beer and wine can be used for some of the same purposes and may very well share some of the same end-uses as distilled alcoholic beverages. Beer and wine might be directly competitive or substitutable with some or all distilled alcoholic beverages in the Chilean market, but that is not the subject of our inquiry. Moreover, that does not detract from the probativeness of the evidence in regard to the substitutability of distilled alcoholic beverages.
40. We note that, while we find the 1997 SM Survey useful, we do not rely on this single piece of evidence for our analysis. Rather, in the process of weighing all the evidence presented, we take the 1997 SM Survey into consideration in determining whether the imported and domestic distilled alcoholic beverages are directly competitive or substitutable. That some portions of the survey are broad and if applied to other non-distilled alcoholic beverages might also show overlap of end-uses, does not mean that the survey is irrelevant to our inquiry. To put it another way, it may be that distilled alcoholic beverages are a subset of a broader category of directly competitive or substitutable products (and we make no findings in that regard), but that does not lead to the conclusion that the subset at issue here is not made up of directly competitive or substitutable products.
41. We also note that there was a survey produced by the Adimark company (the "Adimark Survey") for the Chilean industry and provided to the Chilean legislature during its deliberations on changing the tax system for distilled alcoholic beverages. This survey was based on a very limited number of persons, but persons who were chosen presumably quite carefully to be representative of specific categories of Chilean customers. According to the survey, certain categories of consumers found whisky and pisco quite substitutable and would shift consumption to whisky in favorable price conditions which could result from tax equalisation. This was particularly pronounced for young consumers. Another category showed a willingness to increase whisky consumption initially in response to these price changes, but could return more to pisco over the long run because of its identification as a traditional Chilean drink.
42. While we note the limited nature of the samples in this study, it confirms and highlights the data and conclusions of other evidence such as the 1997 and 1998 SM Surveys. There appears to be no question that pisco is identified as the traditional drink of Chile. It is probable that pisco will retain such identity regardless of the tax structure. However, the argument becomes tautological if it should be claimed that pisco is the traditional drink and is, therefore, perceived somewhat differently so that it may receive favorable tax treatment based on its character as a traditional drink. This would amount to a difference in perception that is reinforced by the tax system and then used as justification to maintain the favourable tax treatment itself.
43. Products do not have to be substitutable for all purpose at all times to be considered competitive. It is sufficient that there is a pattern that they may be substituted for some purposes at some times by some consumers. The Adimark Survey shows this. It is not extensive enough to show the extent of such willingness, but we note that it was based on representative groups. The Adimark Survey shows the price sensitivity of the demand for the product. There is an increased willingness to try imports as the price changes in a manner reflective of tax equalization. The Adimark Survey also shows the nature of alcoholic beverages as an experience good. The sections of Chilean society most reluctant to switch from pisco to imports is the group with least exposure to such products. They show a willingness to try imports if available at lower prices, but think they would go back to pisco in the long run. Those with the greatest experience already with imports showed the willingness to switch more readily. Furthermore, the Adimark survey shows the nature of the consumption decisions at issue here. Distilled alcoholic beverages are products with low prices relative to income, and therefore it is relatively easy to switch consumption to another product for a portion of needs and still maintain loyalty to familiar brands on a broader basis. There can be some level of substitution without fundamental changes in consumption patterns such as might be required with respect to high priced consumer durables.
44. We do not wish to over-emphasise the Adimark Survey. It is useful evidence, particularly given that it was produced for use by the legislature and not commissioned strictly for the purposes of this dispute, and that it was based on representative samples of Chilean society. It is consistent with other evidence.
45. As the Appellate Body has noted, Article III cases deal with markets. ³⁷² The panel in Korea � Taxes on Alcoholic Beverages noted the usefulness of examining marketing strategies in determining whether products are substitutable. ³⁷³ Marketing strategies that highlight fundamental product distinctions or, alternatively, underlying similarities may be useful tools for analysis.
46. The evidence of trends towards increasing overlap in end-uses is supported by the marketing strategies of the domestic Chilean companies. These companies met the threat of imports of distilled alcoholic beverages by, among other things, creating and selling premium pisco, which is more expensive than ordinary pisco and is usually colourfully presented as an up-market distilled spirit generally having an alcohol content comparable to whisky, cognac or up-market brandy. The complainants also produced evidence that these products were being advertised as competitive with up-market imported distilled spirits. ³⁷⁴
47. There is evidence that imported spirits and pisco are used similarly in various social settings � homes, bars, discos etc. The advertising of pisco indicates to consumers that it is suitable as an up-market distilled spirit which shows that the intention by the producers is to put it in the same competitive category with such up-market imports as whisky, cognac, brandy, etc. The various surveys reviewed also show that consumers have an increased willingness to shift between domestic and imported spirits for at least some purchases and some occasions. The current actual overlap in end-uses plus the evidence of potential overlap, is supportive of a conclusion that pisco and the imported distilled spirits are directly competitive or substitutable. ³⁷⁵

(c) Physical characteristics

48. It is necessary to examine the physical characteristics of the products at issue. In our view, the closer the physical similarity the greater the likelihood of a directly competitive or substitutable relationship.
49. The European Communities argues that pisco and the imported distilled spirits share the same basic physical characteristics in that all have the essential feature of being beverages containing alcohol obtained using naturally fermented ingredients by similar distillation processes. The differences between pisco and whisky, according to the European Communities, are no greater than, for example, differences between brandy and whisky. This implies that the different substances from which brandy and whisky are distilled, that is grape wine and malted barley, are not fundamental physical characteristics in determining substitutability. Other differences arise from post-distillation processes such as ageing, colouring or flavouring that confer on each type of distilled spirit its own identity. In the EC's view, however, the differences are not so important as to render the various types of distilled spirits incapable of being directly substituted with each other by consumers.
50. Chile's response is that pisco and the imported distilled spirits share virtually no common physical characteristic other than containing alcohol and water. According to Chile, the ingredients of pisco and say, whisky, are markedly different. Chile points out these two spirits are made from different ingredients, pisco from grapes and whisky from grain. Chile further argues that the basic similarities the European Communities refers to are mere characteristics of all distilled spirits and that the ultimate physical characteristics, which consumers use to distinguish between different types of spirits, are determined by the other processes involved in the production of spirits.
51. We are of the view that an examination of the physical characteristics of products is more critical in determining whether two products are "like" than in the determination of whether two products are directly competitive or substitutable. This does not mean, however, that products' physical similarities should not be examined when determining whether products are directly competitive or substitutable. The Appellate Body has noted that:

"Like" products are a subset of directly competitive or substitutable products: all like products are, by definition, directly competitive or substitutable products, whereas not all directly competitive or substitutable" products are "like".³⁷⁶

52. Consequently, if two products are nearly physically identical, they are "like". ³⁷⁷ Because it necessarily follows that they are also then directly competitive or substitutable, physical similarity is a useful category of examination for our analysis in this case. This is relevant where such activities as marketing campaigns or government tax regimes, have created a distinction in consumer perceptions between very similar products. Such distinctions that result from consumer perception are relevant but not determinative of the nature of an actual or potential competitive relationship.
53. These physical similarities are relevant to the inquiry, particularly with respect to potential competition. We regard the aspect of a product being a potable distilled spirit with a high alcohol content as an important defining characteristic. ³⁷⁸ We note that all the products presented to the Panel have this significant common feature.
54. In our view, the post-distillation differences due to the filtration, colouring or aging processes of the beverages are not so important as to render the products non-substitutable. ³⁷⁹ We find these differences relatively minor. There are some differences imparted from such things as aging in wooden barrels. Some spirits have added flavourings such as juniper berries in gin. But we also note that pisco shares many identical physical characteristics with other spirits made from grapes such as grappa, cognac, brandy or "Peruvian pisco". ³⁸⁰ Overall, weighing the evidence presented, we find that the common physical features of the imported and domestic products are supportive of a finding that the imported and domestic products in question are directly competitive or substitutable. ³⁸¹

(d) Channels of distribution and points of sale

55. The European Communities argues that the 1997 SM Survey shows that all types of premises market both pisco and the imported distilled spirits together. For both categories, the preferred outlets are the same, supermarkets and liquor stores. The European Communities also argue that their presentation in the retail outlets in the same shelf space is evidence of their substitutability.
56. Chile does not dispute the factual assertion that pisco and imported distilled spirits are sold in the same sales channels and can even share shelf space, but does not agree that this is evidence of substitutability. According to Chile, such an argument is as unreasonable as an assertion that toothpaste and soap are substitutable because they are sold in the same channels and share shelf space. Chile also presented evidence showing that imports, and not pisco, are more likely to be distributed in supermarkets than pisco, while pisco is more commonly available in traditional stores.
57. The Panel in Korea � Taxes on Alcoholic Beverages noted that:

[C]onsiderable evidence of overlap in channels of distribution and points of sale�..is supportive of a finding that the identified imported and domestic products are directly competitive or substitutable. ³⁸²

We agree with that panel's finding on this point. In the case before us, there is no dispute that the two categories of distilled spirits are sold in similar sales channels, albeit arguably in somewhat different proportions, and indeed actually share shelf space. Chile has pointed out that many products share shelf space, but cannot be considered substitutable because of that. That is of course true. However, the consistent practice of putting these products on adjoining shelf space in similar outlets is one piece of evidence supporting a finding of substitutability. If it is a coincidence that products happen to be next to each other on shelves, one would not expect it to be repeated consistently.

58. It is also the case that complementary products are often grouped together to help in their marketing. However, as revealed in our discussion of overlapping end-uses above, we find no evidence that pisco and the imports are considered as complements by consumers in the way, in Chile's example, soap and toothpaste might be.
59. In our view, if products have quite distinct channels of distribution that could be a negative indicator with respect to substitutability. For example, if the products were regularly presented separately, it would be one piece of evidence that perhaps consumers did not group them together in their perceptions. In our view, the facts before us indicate an overall pattern of use of channels of distribution, including the presentation of the products within those channels, that is supportive of a finding that the domestic and imported products are directly competitive or substitutable. ³⁸³

To continue with Prices

³⁶⁷ These are the criteria we have examined in this case. There may be other criteria more or less relevant in other situations depending on the facts available.

³⁶⁸ Appellate Body Report on Japan �Taxes on Alcoholic Beverages II, supra., p. 25, quoting the Panel Report, para. 6.22.

³⁶⁹ Panel Report on Korea � Taxes on Alcoholic Beverages, supra., para. 10.78.

³⁷⁰ Comprehensive survey done by Search Marketing S.A., Santiago, December 1997 ("the 1997 SM Survey"), EC Exhibit 21. See also the 1998 SM Survey (EC Exhibit 22) discussed more fully below.

³⁷¹ See Table 23 in Descriptive Part of this Report.

³⁷² Appellate Body Report on Japan � Taxes on Alcoholic Beverages II, supra., p. 25.

³⁷³ Panel Report on Korea � Taxes on Alcoholic Beverages, supra., para. 10.95.

³⁷⁴ See EC Exhibit 51, Control's Recipe Brochure, EC Exhibit 52, Capel's Internet home page & Exhibit 54, Capel's advertising brochure.

³⁷⁵ We note that these conclusions with respect to end-uses support our conclusion that the identified imported products should be considered as a single category.

³⁷⁶ Appellate Body Report on Korea � Taxes on Alcoholic Beverages, supra., para. 118.

³⁷⁷ Panel Report on Japan � Taxes on Alcoholic Beverages II, supra., para. 6.22.

³⁷⁸ We note that alcoholic beverages containing only fermented ingredients cannot achieve as high a concentration of alcohol as is possible through the distillation process.

³⁷⁹ See Tables 7 and 8 of the Descriptive Part of this Report. Indeed the panel in Korea � Taxes on Alcoholic Beverages found that post-distillation differences between soju and imported spirits at issue in that case were relatively minor compared to the common feature of being potable distilled spirits. Panel Report on Korea � Taxes on Alcoholic Beverages, supra., para. 10.67.

³⁸⁰ We note that the product labelled pisco by Peru is not allowed to use that appellation in Chile. Peru claims that it has a disagreement with Chile over who has rights to the appellation. That question is outside our terms of reference and we take no position on it and no implication should be drawn from our use of the term "Peruvian pisco."

³⁸¹ We note that these conclusions with respect to physical characteristics support our conclusion that the identified imported products should be considered as a single category.

³⁸² Panel Report on Korea � Taxes on Alcoholic Beverages, para. 10.86.

³⁸³ These conclusions with respect to channels of distribution and points of sale support our conclusions that the identified imported products should be considered as a single category.