(b) The GM Opel Blazer is comparable to the Kia Sportage

8.149 As noted above, Kia Timor�s plans call for the assembly and sale in Indonesia of the Kia Sportage, to be known as the Timor J520i. When introduced, the Timor Kia Sportage will compete directly with GM�s Opel Blazer, which GM continues to import and assemble in Indonesia.

8.150 A comparison of specifications demonstrates that the Blazer and the Timor Kia Sportage are comparable products: ⁴⁷⁵

(c) Indonesia�s arguments concerning "like product" are incorrect as a matter of law

8.151 Finally, let us consider the heart of Indonesia�s defense; namely, that none of the passenger cars that would have been sold by US manufacturers can be considered a "like product" to the Timor Kia Sephia sedan for purposes of footnote 46 of the SCM Agreement. Putting aside the fact that Indonesia did not make this argument in connection with the United States claim of serious prejudice, Indonesia nevertheless is, once again, wrong on the facts and the law.

8.152 Indonesia begins its like product legal analysis with an accurate quotation of footnote 46, noting that in the absence of an identical product, a like product can consist of a product which "has characteristics closely resembling those of the product under consideration." (Emphasis added). Because this language obviously undermines Indonesia�s entire case, Indonesia dismisses the quoted phrase as "amorphous and skeletal" and never refers to it again.

8.153 However, as the Appellate Body has so often noted, a treaty interpreter must start with the language of the treaty. ⁴⁷⁶ The SCM Agreement does not require identity of products, but only a close resemblance. All of Indonesia�s subsequent legal argumentation is designed to do nothing more than divert the Panel�s attention from what the SCM Agreement actually says.

8.154 Next, Indonesia makes the extraordinary argument that the ability to bring a serious prejudice case against a developing country Member constitutes a so-called "exception" to the "right" conferred on Indonesia by the SCM Agreement to subsidize. As such, Indonesia argues, this exception, including the definition of "like product," must be narrowly construed. This is utter nonsense.

8.155 As a general matter, the United States does not agree with the principle that simply because a particular provision or agreement can be labelled as an "exception" to something else, that provision or agreement must be mechanically and narrowly construed. In the Wool Shirts case, India made a similar argument, and the Appellate Body rejected it. ⁴⁷⁷ Moreover, it is ludicrous to suggest that like product analyses differ depending upon whether the complainant is challenging the subsidies of a developed or a developing Member.

8.156 However, even assuming that this mechanical rule of treaty interpretation applies, Indonesia�s application of it is 180 degrees backward. If there is an "exception" in Article 27 of the SCM Agreement, it is that developing country Members are excepted from the general prohibition against the use of local content subsidies and the remedies available to combat such subsidies that require no showing of adverse trade effects. Thus, if the principle suggested by Indonesia is to be correctly applied, this Panel must interpret all relevant provisions in such a way as to limit the exceptional treatment afforded by Article 27.3. In the context of this case, this means, among other things, interpreting the term "like product" broadly, not narrowly. ⁴⁷⁸

(d) Indonesia�s arguments concerning "like product" are incorrect as a matter of fact

8.157 Having disposed of Indonesia�s legal arguments concerning like product, let us turn to Indonesia�s factual arguments. As discussed above, Table 23 compares the specifications for the Timor Kia Sephia, the Ford Escort, and two types of Opels, the Optima and the Vectra. This comparison establishes the close physical resemblance between these products. However, notwithstanding the fact that physical characteristics are a standard criterion in identifying like products, Indonesia simply dismisses this comparison as "unduly restrictive".

8.158 With respect to end-use, another standard like product criterion, the United States noted that these cars all have the same end-use; namely, the transport of persons. Here, too, Indonesia dismisses this criterion as irrelevant.

8.159 Instead, Indonesia recycles the same argument it made in connection with Article I of GATT 1994, which is that consumers consider numerous physical and non-physical characteristics in making their purchasing decisions. While this is true in a general sense, it proves nothing, and Indonesia offers no evidence as to how this factor justifies the conclusion that Escorts, Neons, and Opel Optimas and Vectras are not "like" the Sephia. Of course, consumer products, including passenger cars, are different from each other, because product differentiation is one of the means by which products compete. However, the fact that consumer products inevitably will not be identical to one another does not mean that they cannot be considered as "closely resembling" one another for purposes of identifying a "like product" under the SCM Agreement. If products had to be "identical" (or virtually identical) in order to be considered "like," it would be impossible to bring a serious prejudice case (or a countervailing duty case, for that matter) against a subsidized consumer product.

8.160 There are a few "facts" that Indonesia throws out in order to confuse the issue. First, it notes that the DRI/McGraw-Hill market segmentation categories were developed in the context of the European market. Even if true, Indonesia does not explain how this renders these categories inappropriate for purposes of this case or the Indonesian auto market. Indeed, DRI/McGraw-Hill uses these categories for purposes of its authoritative Asia Automotive Industry Forecast Report. ⁴⁷⁹

8.161 Second, Indonesia notes that cars with the same nameplate often differ significantly from market to market. Even if true, Indonesia offers no evidence that the passenger cars in question do differ significantly.

8.162 Third, Indonesia cites the fact that for HTS purposes, passenger cars are distinguished on the basis of total cylinder capacity, with breakpoints at 1000cc, 1500cc, and 3000cc. However, these breakpoints are arbitrary, and provide no basis for concluding that the US passenger cars are not like products to the Sephia simply because the Sephia, at 1,498cc, falls two "cc�s" below the breakpoint. Moreover, Indonesia does not explain why, if 1500cc is such a significant breakpoint, it chose 1600cc as the breakpoint for conferring tax incentives under Government Regulation No. 36/1996.

8.163 Finally, notwithstanding all of the money and resources that Indonesia reportedly has devoted to this case, Indonesia apparently has been able to find only one document, (Indonesia Attachment 12), that groups the Sephia in a different category from the Escort and Neon. However, neither the document itself nor Indonesia explains how these categories were drawn.

8.164 The United States has provided more than sufficient evidence to establish that the US passenger cars in question "closely resemble" the Timor Kia Sephia and, thus, are "like" the Sephia. In response, Indonesia has presented only flawed legal arguments and virtually no evidence.

(e) Market segment

8.165 It is important to put the issue of "market segments" in a proper and legally relevant context. Article 6.4 of the SCM Agreement indicates that displacement or impedance includes a situation where there has been a change in relative shares of the market to the disadvantage of the non-subsidized like product. In analysing changes in relative market shares, the United States generally agrees with the statement by Indonesia that, to be meaningful, market share data must be calculated for the market segment into which the like products fall.

8.166 Having said that, however, it also is important to note that, based on Indonesia�s own data, by the end of May, 1997, the Timor Kia Sephia had achieved a market share of 26.53 per cent. See AV/14, Attachment U-21/6. Although it is not readily apparent from Attachment U-21/6 itself, when one examines Attachments U-21/4-B and U-21/5-B to AV/14, it is clear that the 26.53 percent figure represents the Timor Kia Sephia�s share of the market that includes all passenger cars sold in Indonesia. ⁴⁸⁰

8.167 In the view of the United States, the attainment of a 26.53 per cent market share in less than one year more than satisfies the requirements of Article 6.4. Eliminating certain passenger cars from the market segment used by Indonesia in its calculations (i.e., all passenger cars) simply increases the market share attained by the Timor Kia Sephia, thereby making the serious prejudice case against Indonesia all the more damning. Thus, unless the Panel were of the view that the attainment of a 26.53 per cent market share did not demonstrate serious prejudice, attempting to place all of the passenger cars sold in Indonesia into their appropriate market segments would appear to be an exercise that is somewhat tangential to the issues raised in this case.

8.168 Nevertheless, in order to be responsive to the Panel�s questions, the United States suggests that instead of "reinventing the wheel", the Panel might want to rely on the market segmentation categories actually used by the automobile industry. Such an approach seems particularly appropriate in light of the fact that "[t]he GATT 1994 is a commercial agreement, and the WTO is concerned, after all, with markets." Japan - Taxes on Alcoholic Beverages, supra, page 26.

8.169 Applying this approach, the Timor Kia Sephia falls into the "C" segment, which includes "lower medium class" cars. (See above for a discussion of automobile market segment categories.) Other C segment passenger cars that are sold (or that, but for the National Car Programme, would have been sold) in Indonesia include: Opel Optima, Ford Escort, Chrysler Neon, Peugeot 306, Toyota Corolla, Honda City, Honda Civic, Nissan Sunny, Mitsubishi Lancer, Daewoo Nexia, Hyundai Accent, and the Mazda 323. It is possible that the Bimantara Cakra or Nenggala models also may fall into the C segment, but the United States lacks sufficient information at this time to make any definitive statements regarding these two models.

8.170 Alternatively, it would be reasonable for the Panel to find that the relevant market segment also includes the "D" segment of "upper medium class" passenger cars. If the market segment were defined as comprising all medium class cars in both the C and D segments, additional cars that are sold (or that, but for the National Car Programme, would have been sold) in Indonesia include: Opel Vectra, Honda Accord, Peugeot 405, 406, and 606, Toyota Corona and (possibly) the Crown, Nissan Sentra, Mitsubishi Galant, and Mazda 626.

8.171 Indonesia makes much of the fact that the passenger car market is segmented and that car manufacturers devote substantial resources to establishing and targeting different market segments. (See Section VIII.B.3.) However, this fact is not new, and it is consistent with the United States position that the Timor Kia Sephia falls within the "C" segment of the passenger car market. Likewise, it is not a new fact that an individual auto manufacturer will try to produce and sell one or more cars within each segment of the passenger car market.

8.172 Consider, for example, Kia Motors, TPN's joint venture partner in the National Car Programme. US Exhibit 30 consists of page 211 of DRI's Asian Automotive Industry Forecast Report, November 1997. Exhibit 30 demonstrates that Kia offers products in passenger car segments A through E, and usually offers more than one car in each segment. Strikingly, however, the Sephia is Kia's sole entry in the "C" segment.

8.173 Indonesia also makes much of the fact that there are numerous differences between passenger cars. However, this, too, is not a new fact, and it is exactly the point the United States has been making throughout this case; namely, that consumer products, including passenger cars, compete on the basis of numerous differences between products. Product characteristics within similar categories will vary among competing models within general bands or market segments. However, this is not enough to render cars within the same market class as "unlike" each other.

8.174 As an example, consider US Exhibit 31. Exhibit 31 is a summary of the various features that Ford has touted as making the Escort preferable to, and distinguishable from, its competition. The United States submits that while these features may distinguish the Escort from its competition, they are not enough to render the Escort "unlike" its competition. By the same token, the fact that the Timor Kia Sephia may have features that distinguish it from its competition is insufficient to render the Timor "unlike" its competition.

8.175 Indonesia argues that so-called "non-physical characteristics" render the Timor "unlike" and non-competitive with the passenger cars of the US manufacturers that are at issue in this case. The United States questions whether, on balance, any such differences exist, and Indonesia has not presented any evidence of such characteristics.

8.176 In this regard, TPN, Ford, and Chrysler each were (or would have been) new entrants to the Indonesian passenger car market (and General Motors was only a relatively new entrant) with little, if any, developed brand image or loyalty. While Indonesian consumers may well have formulated a perception that the cars of US manufacturers are of superior quality as compared to the Timor, such perceptions likely would be balanced by the undoubted patriotic feelings that Indonesian consumers would have with respect to the Timor as the first Indonesian National Car.

8.177 On balance, Indonesia appears to be arguing that if consumer choices were determined solely on the basis of the market, the Timor would have lost out to the Opel Optima, the Ford Escort, and the Chrysler Neon. With this, the United States wholeheartedly agrees, and that is the very crux of this dispute. What this dispute is about is that the choices of Indonesian consumers were not left to the market, but instead were distorted by the massive subsidies provided by the Government of Indonesia. These subsidies enabled the Timor to undercut the prices of competitive products so significantly, that a significant number of consumers appear to have been willing to make a trade-off between quality and price. However, the fact that consumers made this choice is not enough to render the Timor "unlike" the other cars in its class for the purposes of the SCM Agreement.

8.178 Now let us turn to the specifics of Indonesia's model comparisons. At the outset the United States takes issue with Indonesia's assertion that there are four basic physical attributes or specification groupings that differentiate passenger cars and that contribute to segmenting the market. (See Section VIII.B.3, Tables 26 and 27.) Indonesia claims that these attributes are: power plant; steering and suspension; safety features; and passenger compartment. In the view of the United States, there are three factors that segment the market: power plant, vehicle size, and overall features.

8.179 Having said that, however, let us assume that Indonesia's approach is correct and consider Indonesia's product comparisons. A close look at the comparisons reveals that (1) they fail to rebut the evidence presented by the United States that the GM Opels, the Ford Escort, and the Chrysler Neon are like the Timor Kia Sephia; and (2) Indonesia's comparisons are not even accurate.

8.180 First, however, the United States should note that Indonesia's product comparisons in Table 27 are inaccurate, because Indonesia compares the Timor S515, the base model for the Timor Kia Sephia. Indonesia fails to mention the Timor S515i, the fuel injected model which was imported under the one-year authorization and which, according to Indonesia's own Annex V response is "[t]he only Timor car produced in Indonesia." See AV/14, Attachment 28.

8.181 With respect to the category of"power plant," Indonesia finds the US cars in question to be non-comparable to the Timor. (See Table 26.) The United States does not agree that these differences render the cars in question non-comparable to the S515. However, let us insert the Timor S515i into the comparison. The Timor S515i has a double -overhead cam (DOHC), 16 valve engine. By comparison, the Chrysler Neon has only a single -overhead cam (SOHC) engine, although it also has 16 valves. The Opel Optima has only an SOHC, 8 valve engine. Based on industry estimates, these differences between the Timor S515i and the Opel Optima are worth approximately Rp. 2.5 million. ⁴⁸¹

8.182 Thus, while the Opel Optima and Chrysler Neon have a larger engine displacement than the Timor S515i, the S515i has a DOHC, 16 valve engine, whereas the Neon has only an SOHC, 16 valve engine and the Optima only an SOHC, 8 valve engine. As for the Escort, its engine displacement is only 99 cc's more than that of the S515i. (In this regard, the United States notes that Indonesia incorrectly lists the Neon's engine displacement as 2,000 cc. As clearly set forth in evidence submitted by the United States (Table 23), the Neon's engine displacement is 1,996 cc.)

8.183 In the view of the United States, these differences are insufficient to warrant the conclusion that, in terms of power plant, the US cars do not "closely resemble" the Timor S515i or the S515 within the meaning of the "like product" definition in the Subsidies Agreement.

8.184 Turning to what Indonesia refers to as the category of"steering and suspension", Indonesia finds the US cars to be non-comparable based on the presence of power steering and tire size. (See Table 26 and accompanying notes.) However, according to Timor sales brochures, both the S515 and the S515i have power steering as a standard feature. (See European Communities Exhibit D-4.) Moreover, according to Ford officials, power steering on the Escort is an option, not a standard feature, worth approximately US$200.

8.185 As for larger tire size, this is a designed option that depends upon anticipated road conditions and generally is defined as a standard feature of minimal value to consumers.

8.186 In short, the US cars cannot be considered "unlike" the S515 or the S515i on the basis of steering and suspension.

8.187 Turning to what Indonesia calls "safety features", with respect to brakes, Indonesia cites the fact that the Neon has antilock brakes. However, this is an optional feature on the Neon, as clearly set forth in Table 27. With respect to the Escort, Indonesia claims that the Escort has rear disc brakes, but this is incorrect. As set forth in US Exhibit 33, page 3, the Escort has rear drum brakes, just like the Timor. Finally, there are no differences between the Opel Optima and the Timor with respect to brakes.

8.188 With respect to airbags, according to Ford officials, Ford did not include airbags in its Indonesia launch plans for the Escort. The Neon, however, would have had airbags.

8.189 With respect to fuel economy, the United States has some difficulty understanding how this constitutes a "safety feature." Nevertheless, according to the US Department of Energy, the fuel economy of the Sephia is comparable to, and in some cases better than, that of the Escort and the Neon. US Exhibit 34 contains excerpts from the US Department of Energy Fuel Economy Guide: Model Year 1997. US Exhibit 35 summarizes the data from this guide with respect to the Sephia, the Escort, and the Neon. As can be seen, the Sephia actually has superior mileage to the versions of the Escort and Neon that have automatic transmissions. Note also that Indonesia does not provide any source for the fuel economy figures contained in Table 27.

8.190 Turning from fuel economy to vehicle weight, in a note to Table 27, Indonesia cites only the fact that the Opel Vectra is heavier than the Timor, thus conceding that Timors, Neons, Escorts, and Optimas have comparable weights.

8.191 Finally, in the same note to Table 27, Indonesia cites the fact that the Vectra has a larger fuel tank than does the Timor, thereby conceding that Timors, Neons, Escorts, and Optimas have comparably sized fuel tanks. In this regard, the United States would note that it does not believe that fuel tank size plays much of a role, if any role, in consumer purchasing decisions.

8.192 In summary, insofar as "safety features" are concerned, Indonesia's analysis is inaccurate and fails to rebut the US evidence that the cars of US manufacturers are "like" the Timor Kia Sephia.

8.193 Finally, with respect to the category of "passenger compartment," in Table 26, Indonesia concedes that Timors, Neons, Escorts, and Optimas are comparable.

8.194 Thus, Indonesia's own comparisons, when corrected for inaccuracies, demonstrate that the US cars in question are "like" the Timor. However, the United States should note that Indonesia ignores or downplays certain facts that, according to Indonesia, should form a part of any "like product" analysis. We already have mentioned the fact that the engine of the Timor S515i has certain attributes that make it more desirable than the engines in the Optima and the Neon. In addition, Indonesia fails to cite the fact that the Timor S515i has tilt steering, mud flaps, colour key mirrors and bumpers, and a foldable mirror, all of which are attractive features to consumers.

8.195 Moreover, Indonesia ignores the fact that the US manufacturers, as new (or relatively new) entrants into the Indonesian passenger car market, have little, if any, brand loyalty base or a well-developed brand image. Ford, for example, has only one after-sales service outlet in Indonesia.

8.196 Finally, the United States would like to call the Panel's attention to US Exhibit 36, which contains excerpts from the Catalogue de la Revue Automobile '96, described by industry officials as an authoritative source. In the entry for the Sephia, the Sephia is described as "techniquement parenté aux Mazda 323/Ford Escort USA/Mercury Tracer." Translated, this means that the Sephia is, among other things, technologically related to the Escort.

8.197 As the United States has explained, the Escort is Ford's entry in the "C" segment of the passenger car market. Given the fact that the Sephia is technologically related to the Escort, it seems fair to say that the Timor Kia Sephia falls within that same segment and is a "like product" to the other cars in that segment.

(f) Indonesia's arguments concerning the burden of proof are incorrect

8.198 Indonesia claims that the United States has failed to satisfy its burden of proof. In the view of the United States, it has more than satisfied its burden as articulated by the Appellate Body in Wool Shirts, of presenting a prima facie case of serious prejudice. The first and second submissions of the United States establish all of the elements for a serious prejudice case. Instead, it is Indonesia that has failed to rebut the US case. Of course, Indonesia is hard put to rebut the evidence cited by the United States, because much of it is Indonesia's own evidence, submitted in the course of the Annex V process.

8.199 Indonesia presents a new argument that the burden of proof on complainants is higher in this case because Indonesia is a developing country and the ability to bring a serious prejudice case against it allegedly is a derogation of its "right" to provide local content subsidies. (See section VIII.B.3(a)(3).) This is an expansion on Indonesia's earlier argument that "like product" must be interpreted differently in this case because of Indonesia's developing country status.

8.200 The United States addressed Indonesia's earlier argument in section VIII.B.3(a)(2). Indonesia's expansion of its argument brings the argument even more clearly within the scope of the Wool Shirts case, where the Appellate Body rejected India's argument that the burden of proof changes simply because a particular provision may be labeled as an exception to something else. Significantly, in its submission, Indonesia does not mention the Wool Shirts report when it advances its "derogation" argument.

To continue with Indonesia's arguments

⁴⁷⁵ Source: US companies. The United States also believes that Jeep Cherokees and Wranglers are comparable to the Sportage, but lacks detailed specifications.

⁴⁷⁶ India - Patent Protection for Pharmaceutical and Agricultural Chemical Products, WT/DS50/AB/R, Report of the Appellate Body issued 19 December 1997, para. 45.

⁴⁷⁷ United States - Measure Affecting Imports of Woven Wool Shirts and Blouses, WT/DS33/AB/R, Report of the Appellate Body adopted 23 May 1997, page 19.

⁴⁷⁸ Carried to its logical conclusion, a correct and consistent application of Indonesia�s argument would result in Indonesia bearing the burden of proving that serious prejudice has not occurred. Of course, this is the type of burden shifting that India proposed in Wool Shirts, and that the Appellate Body rejected.

⁴⁷⁹ US Exhibit 29 presents two pages of DRI promotional materials that attest to the authoritative nature of the DRI Global Automotive Group. Note in particular that the selected list of clients includes familiar names in this case, such as Chrysler, Ford, GM, Honda, Hyundai, Kia, Mitsubishi, Nissan, and Toyota. The United States does not take the position that the DRI market segmentation is dispositive with respect to the like product issues in this case. Rather, DRI's market segmentation confirms what a traditional like product analysis demonstrates; namely, that the Timor Kia Sephia is "like" the products that US manufacturers would have sold in Indonesia but for the National Car programme.

⁴⁸⁰ AV/14, Attachment U-21/4-B provides sales figures for passenger cars that benefited from subsidies during the period 1995-1997. The "Total" column for 1997 indicates that as of May, 1997, the total sales volume for these passenger cars was 12,413, of which the Timor Kia Sephia accounted for 7,058. AV/14, Attachment U/21-5-B provides sales figures for passenger cars that did not benefit from subsidies during the period 1995-1997. The "Total" column for 1997 indicates that as of May 1997, 14,038 of these cars were sold. Thus, the total number of passenger cars (both subsidized and non-subsidized) sold in Indonesia during the period January - May, 1997 was 26,451. The 7,058 Timor Kia Sephias sold during this period accounted for 26.68 percent of this total. Presumably, an arithmetic error on the part of the individual preparing AV/14 accounts for the difference between the 26.53 percent figure in Attachment U/21-6 and the 26.68 percent figure derived from Attachments U-21/4-B and U-21/5-B.

⁴⁸¹ See US Exhibit 32.