(d) Channels of Distribution

130. The European Communities notes that Table 13 below sets out the findings of the 1997 SM survey ⁸⁹ with respect to the availability of pisco and other distilled spirits in different sales channels. It shows that all different types of premises market both pisco and all the other spirits and that, for all of them, the preferred sales outlets are the same (supermarkets and liquor stores).

Table 13 ⁹⁰

Retail purchase outlets

	Pisco	Whisky	Liqueur	Brandy	Tequila	Gin	Rum	Vodka
Super market	61	65	61	47	50	58	46	47
Liquor store	41	26	32	39	40	43	34	43
Gift shop	3	8	6	9	5	7	6	-
Duty free	1	7	3	-	2	-	9	2
Grocery	4	1	-	5	4	-	-	3
Air lines	-	1	-	-	1	3	2	-
Other	3	6	9	-	8	-	8	7

Basis: % of consumers of each spirit

131. According to the European Communities, a further indication of substitutability between pisco and other distilled spirits is the similarity of their presentation in retail outlets. A selection of photographs taken from 6 retail outlets in Santiago in mid 1997 ⁹¹ evidence that pisco and other spirits are shown to consumers in the same shelf space areas. This shelving is responsive to the consumers' need to make choices among substitutable products.
132. Chile points out that as to distribution channels, the European Communities puts great weight on the tendency of supermarkets to show pisco on shelves near whisky and other distilled spirits. One might on that basis argue that toothpaste and soap are substitutable, because they share shelf space.
133. The European Communities states that it does not put "great weight" on the fact that Chilean supermarkets tend to show pisco on the same shelves as whisky and other distilled spirits. The European Communities referred to that tendency as an additional indication of competition, among many other indications mentioned in its submission. For instance, the European Communities has produced detailed survey evidence showing commonality of distribution channels. The European Communities doubts that Chile would consider it irrelevant if whisky and pisco were generally sold at opposite corners of supermarkets. Finally, although toothpaste and soap may share shelf-space in some cramped night shop patronised by busy lawyers in Washington DC, they tend to be displayed separately in larger supermarkets.
134. Chile contests the EC statement that "evidence that two products are sold in the same channels .... can be taken as evidence that they are 'directly competitive and substitutable'." Common distributions channels are not strong evidence, and multipurpose channels are weaker still than dedicated channels, such as gas stations or pharmacies. Supermarkets are, by no means, a "dedicated channel", and finding that two products are sold mainly at supermarkets, doesn't prove anything about their competitiveness.
135. Chile further states that even if it considers dedicated distribution channels, like gas stations or pharmacies, it is hard to conclude that two products sold through those channels are substitutable. For instance, medicines sold only through pharmacies are usually non-substitutable, as are gas and lubricants sold primarily at gas stations.
136. In Chile's view, in the case of liquor stores, the only dedicated channel that could be considered, besides spirits, they sell wine, beer and soft drinks (snacks and some confectionery could also be added to this list).
137. Chile explains that it has requested AC Nielsen, a private marketing research firm, to provide information on distribution channels employed in Chile by different type of food industries, pisco and whisky. The results are shown in Table 14 below.

Table 14 ⁹²

Distribution Channels

Product	Supermarket	Traditional
tea	76.6%	23.4%
tomato sauce	76.0%	24.9%
rice	83.1%	16.9%
pasta	76.7%	23.3%
hot pepper sauce	77.3%	22.7%
pisco	46.2%	53.8%
whisky	66.0%	33.3%

Source: AC Nielsen

138. Chile considers from these data that it can observe, that roughly 80% of tea, tomato sauce, rice, pasta and red pepper is sold in supermarkets, while the remainder is sold through what they call "traditional" channels. It is more than obvious that these products are not competitors or substitutable among themselves, yet they share basically the same distribution channel.
139. Chile further states that other important information provided by AC Nielsen, is that 46.2% of pisco is sold in supermarkets, while 53.8% is sold by the "traditional" channels. In the case of whisky, 66% is sold through supermarkets and only 33% through "traditional" channels. This is not surprising, because whisky is primarily consumed by the wealthier segment of population, while pisco is a more popular spirit.
140. According to Chile, comparing AC Nielsen's and Search Marketing's outcome results, one could question the accuracy of Search Marketing's findings, that tell us that 61% of pisco is sold through the supermarkets.
141. Chile indicates that it disagrees with the EC's point regarding supermarket shelving. Despite the EC's argument, the fact is that in Chilean Supermarkets lot of products that are, by no means, substitutable (as soap and toothpaste) share the same shelves.

(e) Price Differentials

142. In rebuttal to the EC argument, Chile stresses the price differentials between pisco and other distilled spirits using Table 15 below.

Table 15 ⁹³

Product	Duty Free Price US$	Alcohol Content (�)	Chilean Custom Duties	Price including Chilean Duties US$	New Chilean System	Price including new tax system US$
1	2	3	4	5	6	7
Whisky J W Red	10.96	43	11%	12.17	47%	17.88
Whisky J W Black	24.11	43	11%	26.76	47%	39.34
Whisky J W Gold	52.00	43	11%	57.72	47%	84.85
Whisky J W Blue	144.00	43	11%	159.84	47%	234.96
Canadian Club	11.41	40	0%	11.41	47%	16.77
Jack Daniels	14.07	45	11%	15.62	47%	22.96
Tequila (Cuervo)	11.97	38	0%	11.97	39%	16.64
Grappa	10.80	40	11%	11.99	47%	17.62
Pisco Especial	2.86	35	11%	2.86	27%	3.63
Pisco Reservado	3.89	40	11%	3.89	47%	5.72

Sources: Pisco prices from Chilean Industries.

Prices of other products, from Peter Justessen Catalogue, 1998 edition.

Note: All prices are referred to 1 litre bottle.

143. Chile emphasises that pisco, in any of its different varieties, has a pre-tax price (i.e., duty free) that is substantially lower than that of whisky. As column five of Table 15 shows, whisky (at a price including Chilean custom duties of US$12.17) is 3.1 times more expensive than a pisco of same alcohol strength (price of US$ 3.89). The same column shows that when comparing whisky with Pisco Especial (35� strength) the price of the former is more than 4.2 times more expensive.
144. Chile further points out that the price difference between whisky and pisco of the same strength is not altered when comparing prices including the New Chilean System. Indeed, column seven of the chart, shows that the price of whisky (US$ 17.88) remains 3.1 times more expensive than pisco of same strength (price US$ 5.72). The price differential is so big that it is no wonder that consumers, particularly in a country of relatively low incomes, prefer low cost spirits.
145. The European Communities responds that Table 15 is incomplete and misleading. It compares the prices of a relatively expensive brand of whisky (Johnnie Walker) to what appears to be the price of a relatively inexpensive brand of pisco. Furthermore, the prices are at different levels of trade and, therefore, not comparable. The prices for whisky are retail prices (presumably in Chile), as shown in the catalogue of a supplier of duty free goods to the diplomatic trade, and include not only freight and insurance but also what appears to be a rather substantial margin for the distributor. In contrast, the prices for pisco are producers' prices at ex factory level.
146. The European Communities argues that both pisco and the other spirits are sold within a wide range of prices. As shown in Table 16 below, the price differences within the category of pisco may be as large, both in absolute and in relative terms, as the price differences between pisco and whisky shown in Chile's table. The European Communities claims that Chile does not appear to contest that all types and brands of pisco constitute a single product and compete with one another.

Table 16 ⁹⁴

Retail prices of pisco, February 1997

Brand	Price (pesos)	Price difference index
Valle del Limarí 30 %	869	1
Capel 30 %	999	1.15
Tres Erres 32 %	1,459	1.69
Bauza 30 %	2,468	2.84
Control 35 %	1,295	1.49
Capel 35%	1,458	1.68
Control de Guarda 35 %	2,389	2.75
Alto del Carmen 35 %	2,458	2.83
Bauza 35 %	2,480	2.85
Control 40 %	1,519	1.75
Capel 40 %	1,588	1.83
Control de Guarda 40 %	2,699	3.11
Alto del Carmen 40 %	2,998	3.45
Bauza 43 %	3,628	4.17
Alto del Carmen 46 %	3,798	4.37
Chenaral 46 %	4,790	5.51

Source: ISWR Report, EC Exhibit 19, pp. 80-81

147. The European Communities further argues that pisco is not less expensive than other spirits. Table 17 below evidences that, despite the price distortions caused by the differences in taxation, the prices (tax included) of pisco overlap with the prices (tax included) of other types of spirits (including whisky) which are more heavily taxed.

Table 17 ⁹⁵

Range of retail prices (including tax), February 1997

Spirit	Lowest price (pesos)	Highest price (pesos)
Pisco	869	4,790
Gin	2,175	5,580
Vodka	2,545	8,050
Rum	1,490	6,838
Brandy	1,375	114,900
Tequila	3,180	5,490l
Whisky	3,550	34,690

Source: ISWR Report, EC Exhibit 19, pp. 53-54, 58, 62, 65, 70, 75-76, 80-81

148. The European Communities also argues that in any event, a comparison of absolute price differences is of limited value in order to establish whether two products are actually competing on a given market. Basic economic theory tell us that it is more relevant to look at the response of consumers to changes in the relative prices of the products, i.e., to their rate of cross-price elasticity. ⁹⁶ The European Communities claims that it has provided to the Panel ample evidence (including two studies commissioned by the pisco industry itself) showing that there is a significant degree of cross-price elasticity between pisco and the other spirits and, therefore, that they are directly competitive or substitutable.
149. The European Communities adds that in the present case, absolute price differences are even less relevant in view of the nature of the products concerned. Distilled spirits are consumer goods which have a small value relative to income and are purchased many times over a short period of time. This means that, even if one spirit was much more expensive than the other, a relatively small decrease in the price of the more expensive one could be sufficient for consumers to increase the number of occasions in which they drink that spirit instead of the less expensive one. ⁹⁷
150. The European Communities further alleges that absolute price differences can be the consequence of the measures in dispute. A comparison of pre-tax prices is not sufficient to remove all the distortions which a protective system of taxation may have caused over a long period of time. For example, one of the effects of a protective system of taxation may be to favour the sale of premium brands of imported spirits over less expensive brands. Also, protective taxes limit the sales growth of the imported spirits and keep their selling and distribution expenses at an artificially high level as compared to domestic products sold in larger volumes. ⁹⁸
151. The European Communities finally recalled that absolute price differences can be the consequence of conjunctural factors such as movements in the exchange rates. ⁹⁹
152. In response, Chile explains that its intention in providing Table 15 above was twofold: (1) to illustrate for the Panel the significant price difference between pisco and whisky; and (2) to illustrate for the Panel the disproportionately greater burden imposed on a low-priced product by the use of a specific tax (in this case the tax levied by some EC countries on alcohol), measured in ad valorem terms.
153. Chile considers that the most accurate basis for comparing the prices in Chile of pisco and whisky should be between ex-factory price of pisco, and the after customs duty price of whisky (CIF + import duties) in Chile. That comparison excludes any mark-up in the distribution channel.
154. Chile adds that the ex-factory prices of pisco are, on average, US$ 2.60 for pisco of 30� to 35�, and US$ 3.60 for pisco of 40� to 46�. For imported whisky the average CIF price, in 1997, according to Central Bank of Chile statistics, was US$ 5.55. Adding an 11% ad valorem import duty, the average after customs price of whisky was US$ 6.16.
155. Chile goes on to argue that these price differences between whisky and pisco are thus very significant, almost double between whisky and high alcohol pisco, and more than double between whisky and low alcohol pisco.
156. Further, Chile points out that it is unlikely that these prices are distorted by levels of trade in any material way. While much more pisco than whisky is sold in Chile, whisky is still imported in substantial quantities. Further, far more whisky than pisco is produced in the world, and that ample production would presumably allow superior economies of scale for whisky.
157. Also, in Chile's view, the European Communities also affirms that "Distilled spirits are consumer goods which have a small value relative to income ..." and interprets this to mean that "even if one spirit was much more expensive that the other, a relatively small decrease in price of the more expensive one could be sufficient for consumers to increase the number of occasions in which they drink this spirit instead of the less expensive one". This is highly speculative on the part of the European Communities and thus far remains unproven.
158. Further, Chile contests the EC statement with respect to "... the distortions which a protective system of taxation may have caused over a long period of time. For example, one of the effects of a protective system of taxation maybe to favour the sale of premium brands of imported spirits over less expensive brands". First, the effects of a repealed system are not relevant. Second, because even the previous Chilean tax system operated on an ad valorem bases, there was no distortion in favor of more expensive brands.
159. Finally, Chile disagrees with the EC's table of retail prices of pisco, which purports to show an overlap between pisco and whisky. Chile argues that this is misleading, and points out that the table does not reflect the fact that most pisco sold in Chile has a price of less that 3,000 pesos. In fact, there is very little overlap except for premium piscos, which will also generally be taxed at the same rate as whisky when the New Chilean System takes effect in December of 2000.

To continue with Cross-Price Elasticity

⁸⁹ 1997 SM survey, p. iii and section 4.3 (EC Exhibit 21).

⁹⁰ EC First Submission, Table 16.

⁹¹ EC Exhibit 56.

⁹² Chile Oral Statement at the Second Substantive Meeting, Table IV.

⁹³ Chile Oral Statement at the First Substantive Meeting, Annex I.

⁹⁴ EC Response to Questions asked at the First Substantive Meeting, Table 1.

⁹⁵ EC Response to Questions asked at the First Substantive Meeting, Table 2.

⁹⁶ Panel Report on Korea - Taxes on Alcoholic Beverages, supra., para. 10.94.

⁹⁷ See ibid., paras. 10.74 and 10.91.

⁹⁸ Ibid., para. 10.93 and fn. 410.

⁹⁹ Ibid., fn. 410.