UNITED STATES - PRELIMINARY DETERMINATIONS
WITH RESPECT TO CERTAIN SOFTWOOD LUMBER
FROM CANADA

Report of the Panel

(Continued) 

3. Claim 2: inconsistent determination of benefit

(a) arguments of the parties

(i) Canada

7.31 Canada argues that, even if one were to accept that the provision of stumpage constitutes a financial contribution in the form of the provision of a good under Article 1.1(a)(1)(iii) SCM Agreement, the USDOC erred in finding that the Canadian provincial stumpage programmes conferred a benefit. The USDOC compared the Canadian stumpage prices to those charged in the United States, which, in the USDOC's view, correspond with "commercially available world market prices". Canada is of the view that the USDOC's use of a benchmark for determining benefit based on conditions outside the country of the alleged provision of goods, was inconsistent with Article 14(d) SCM Agreement. Canada asserts that whether a "benefit" is conferred by the alleged provision of goods by the government (stumpage) depends on whether the Canadian producers were better off than other purchasers who buy the same good from other sellers in the country subject to the investigation.⁶⁴ Canada argues that a cross-border analysis using transactions in another country to determine the existence of a benefit, and to measure it, is inconsistent with Article 1 and Article 14 (d) SCM Agreement.⁶⁵

7.32 Canada further asserts that there was sufficient information on the record concerning private stumpage prices in Canada that could have been used by USDOC in determining the amount of the alleged benefit.⁶⁶ Canada submits that neither Article 1 nor Article 14 SCM Agreement presumes the existence of "distortion" of any kind simply because there is a financial contribution by the government, nor does either permit dismissing an in-country benchmark because of a presumption that that benchmark is or might be "tainted" or "distorted" by the financial contribution in question.⁶⁷

7.33 In addition, Canada argues, there are a number of factual differences between the rights and obligations involved in acquiring harvesting rights in the United States and in Canada which invalidate the use of the United States' system as a benchmark and which demonstrate that cross-border comparisons make no economic sense.⁶⁸ Canada asserts that a wide variety of complex factors affect stumpage rates, such as locational characteristics, timber characteristics, measurement systems, operating costs, differences in economic conditions and tenure holders' rights and obligations. In Canada's view, the USDOC failed to make proper adjustments and ignored other obvious differences between the two markets.

7.34 Canada, therefore, claims that the USDOC Preliminary Determination which measured the benefit under the "adequacy of remuneration standard" by reference to conditions in another country rather than the prevailing market conditions in Canada is inconsistent with Articles 1 and 14 SCM Agreement.

(ii) United States

7.35 The United States observes that, in accordance with Article 14 (d) SCM Agreement, the adequacy of the remuneration and the existence of a benefit must be determined "in relation to the prevailing market conditions in the country under investigation" and not necessarily "in the country under investigation".⁶⁹ The US argues that this "market" benchmark may refer to the entire market available to the subsidized producers.⁷⁰ According to the US, the concept of "prevailing market conditions in the country of provision" is sufficiently broad to permit consideration of prices for competitive goods commercially available on the world markets to purchasers in the country of provision.⁷¹ The US argues that commercially available goods, both imported and domestic, compete in the domestic market and together constitute the supply available to purchasers in the country of provision, that is, goods that the purchaser could obtain in the market absent the government's financial contribution.⁷²

7.36 According to the US, in this particular case, the provincial governments completely dominate the market in those provincial markets and are virtually the sole providers of timber. In the US' view, the evidence indicates that Canadian provincial governments so dominate the Canadian market for timber⁷³ that below�market government prices suppress prices in the small private market for timber in Canada. Therefore, the US submits, in the absence of market prices in Canada, the use of other prices commercially available to Canadian lumber producers on world markets is the only reasonable alternative. In light of the object and purpose of Article 14 SCM Agreement it would not make sense to use prices determined or influenced by the government to measure the adequacy of the remuneration for the goods. According to the US, the comparison in Article 14 (d) SCM Agreement is intended to identify the potentially trade-distorting artificial advantage resulting from the government's provision of a good, and the market conditions referred to in Article 14 (d) SCM Agreement thus relate to what the market price would have been absent the financial contribution. The US asserts that the USDOC used stumpage prices in various US states with comparable forests and adjusted such prices to reflect prevailing market conditions in Canada.⁷⁴ According to the US, the USDOC's use of US prices is supported by ample record evidence indicating that many Canadian companies do, in fact, import US logs and bid on US stumpage.⁷⁵ The US submits that US timber is therefore "commercially available" to Canadian lumber mills, and in light of the specific facts of this case, the USDOC's use of US prices for stumpage that is commercially available to lumber producers in Canada was appropriate and consistent with Article 14(d) SCM Agreement.

7.37 In any case, the US submits, only three provinces (Alberta, Quebec and Ontario) provided any information on private stumpage prices and such limited information was inadequate to serve as a benchmark for those provinces. Similarly, the US notes that it had no or insufficient information on stumpage prices in the Maritime provinces, the lumber originating from which the USDOC excluded from the investigation on the grounds that it was not subsidized.

7.38 For all these reasons, the US requests the Panel to reject Canada's claim concerning the inconsistency of the use of US stumpage prices as a benchmark to measure the amount of benefit conferred on Canadian lumber producers through the provincial stumpage programmes.

(b) Analysis

7.39 Article 1.1 SCM Agreement provides that a subsidy in the sense of the SCM Agreement shall be deemed to exist when a financial contribution confers a benefit. In light of our findings above concerning the USDOC Preliminary Determination that the provision of stumpage by the Canadian provincial governments constitutes a financial contribution by the government under Article 1.1(a)(1)(iii) SCM Agreement, we will therefore need to examine whether the USDOC determination that this financial contribution conferred a benefit was made in a manner consistent with the US obligations under the Agreement. The issue before us is whether the USDOC determined the existence and amount of benefit to the Canadian softwood lumber producers in a manner consistent with Articles 1.1 and 14 of the SCM Agreement. We recall that Canada raises two claims relating to the USDOC's benefit determination. First, Canada challenges the use of US stumpage prices as the benchmark to assess the adequacy of the remuneration and thus the existence and amount of the benefit. We will discuss this claim in this section. Second, Canada claims that the USDOC failed to properly examine and properly determine that the benefit was conferred on the producers of the subject merchandise, softwood lumber. This claim will be discussed in the following section.

7.40 In its Preliminary Determination, the USDOC compared the stumpage fees set by the Canadian provincial governments with stumpage sales prices in the United States in order to determine the amount of benefit to Canadian lumber producers. According to the USDOC, "The point of comparison for measuring the benefit from these types of subsidies is the market-place free of government interference".⁷⁶ The USDOC concluded that:

"there is no market determined price for stumpage within Canada that is independent of the distortion caused by the government's interference in the market. Therefore we preliminarily determine that we cannot use private transaction prices provided by the provincial governments.

Information on the record of this investigation indicates that there are prices from the world market for comparable goods which can be used to determine whether the provincial stumpage programmes provide a good or service to softwood lumber producers for less than adequate remuneration. For the reasons detailed below, we preliminarily determine that stumpage prices from the United States qualify as commercially available world market prices because it is reasonable to conclude that US stumpage would be available to softwood lumber producers in Canada at the same prices available to US lumber producers".⁷⁷

7.41 Canada argues that as a matter of principle a cross-border price analysis is not permitted under Article 14 (d) SCM Agreement. Even if it were in theory permitted, Canada submits that there was sufficient evidence on the record of private stumpage prices in Canada which, in accordance with Article 14 (d) SCM Agreement could have and should have been used by the USDOC in determining benefit. In addition, Canada argues that in any event, under the facts of this case, the specific US benchmark used, was not an appropriate benchmark because of the factual difficulty in comparing prices for standing timber in the US and Canada.

7.42 The US argues that the benchmark used by the USDOC is consistent with the guidelines of Article 14 (d) SCM Agreement since the US stumpage prices are commercially available to Canadian lumber producers and, therefore, constitute part of the prevailing market conditions for the sale of stumpage in Canada.⁷⁸ The US submits that the USDOC properly rejected private prices for domestic timber as a benchmark because the weight of the evidence at the time of the Preliminary Determination indicated that such prices were driven by the government provided timber, the financial contribution at issue.⁷⁹ Thus, in the US view, such prices are uninformative as to the adequate remuneration inquiry, i.e., a comparison of the government price to prices otherwise available in the market absent the financial contribution.

7.43 Article 14 (d) is the relevant provision in the SCM Agreement for measuring the amount of benefit to the recipient by determining whether the government has provided a good or service, within the meaning of Article 1.1(a)(1)(iii) SCM Agreement, for less than adequate remuneration. Article 14 (d) SCM Agreement provides as follows:

Article 14

Calculation of the Amount of a Subsidy in Terms
of the Benefit to the Recipient

For the purpose of Part V, any method used by the investigating authority to calculate the benefit to the recipient conferred pursuant to paragraph 1 of Article 1 shall be provided for in the national legislation or implementing regulations of the Member concerned and its application to each particular case shall be transparent and adequately explained. Furthermore, any such method shall be consistent with the following guidelines:

(a) government provision of equity capital shall not be considered as conferring a benefit, unless the investment decision can be regarded as inconsistent with the usual investment practice (including for the provision of risk capital) of private investors in the territory of that Member;

(b) a loan by a government shall not be considered as conferring a benefit, unless there is a difference between the amount that the firm receiving the loan pays on the government loan and the amount the firm would pay on a comparable commercial loan which the firm could actually obtain on the market. In this case the benefit shall be the difference between these two amounts;

(c) a loan guarantee by a government shall not be considered as conferring a benefit, unless there is a difference between the amount that the firm receiving the guarantee pays on a loan guaranteed by the government and the amount that the firm would pay on a comparable commercial loan absent the government guarantee. In this case the benefit shall be the difference between these two amounts adjusted for any differences in fees;

(d) the provision of goods or services or purchase of goods by a government shall not be considered as conferring a benefit unless the provision is made for less than adequate remuneration, or the purchase is made for more than adequate remuneration. The adequacy of remuneration shall be determined in relation to prevailing market conditions for the good or service in question in the country of provision or purchase (including price, quality, availability, marketability, transportation and other conditions of purchase or sale). (emphasis added)

7.44 Article 14 (d) SCM Agreement thus provides that the provision of goods by a government shall not be considered as conferring a benefit unless the provision is made for less than adequate remuneration. The adequacy of the remuneration charged by the government shall be determined "in relation to the prevailing market conditions for the good or service in question in the country of provision or purchase (including price, quality, availability, marketability, transportation and other conditions of purchase or sale)". We find that the text of Article 14 (d) SCM Agreement is very clear: the adequacy of remuneration is to be determined in relation to prevailing market conditions for the good or service in question in the country of provision or purchase. It further specifies market conditions that have to be taken into account: price, quality, availability, marketability, transportation and other conditions of purchase or sale. The text of the Agreement thus clearly requires that the prevailing market conditions to be used as a benchmark are those in the country of provision of the goods, in this case Canada. The words "in relation to" in our view, mean "on the basis of" or "in comparison with" and indicate in this context that the adequacy of the remuneration charged by the government is to be compared with the prevailing market conditions in the country of provision. We therefore disagree with the US that "the prevailing market conditions in the country of provision are a reference point, not necessarily an end-point for the market benchmark".⁸⁰ The prevailing market conditions in the country of provision are the benchmark and operate as the reference point or the point of comparison for the adequacy of the remuneration. We therefore find, on the basis of a plain reading of Article 14 (d) SCM Agreement, that the ordinary meaning of this provision excludes an analysis based on market conditions other than those in the country of provision of the goods, i.e. Canada.

7.45 We note that, in its arguments before us, the US "agree[d] with Canada that, as the text of Article 14 (d) states, the relevant 'prevailing market conditions' are those in the country of provision, not some other country".⁸¹ The US argues however that "[b]ecause US stumpage prices are commercially available to Canadian lumber producers, they fall within the universe of benchmarks that can be considered for purposes of measuring the benefit from provincial stumpage, consistent with Article 14(d) of the SCM Agreement."⁸² The issues in dispute are thus (i) whether the concept of "prevailing market conditions in the country of provision" includes world market prices, as the US is suggesting, and hence whether US stumpage prices are part of the prevailing market conditions in Canada, and, if so, whether a price benchmark for the purposes of Article 14 (d) SCM Agreement can be exclusively based on such prices; and (ii) whether the USDOC had valid reasons in this case to disregard Canadian private stumpage prices.

(i) Are US prices part of the prevailing market conditions in Canada?

7.46 In our view, there is no basis in the text of the SCM Agreement to conclude that the market conditions in the country of provision could mean anything else than the conditions prevailing in the market of that country, and not those prevailing in some other country. Article 14 (d) SCM Agreement does not just refer to "market conditions" in general, but explicitly to those prevailing "in the country of provision" of the good. The US argues, that because "conditions of purchase or sale" and "availability" are listed as market conditions in Article 14 (d) SCM Agreement, US stumpage which is available to Canadian producers and which may be purchased by Canadian producers is part of the market conditions in Canada. However, the fact that a good may also be bought on a market outside the country of provision, does not, in our view, imply that the prices for those goods in that other country become part of the market conditions "in the country of provision". To accept the US argument would imply that the phrase "prevailing market conditions in the country of provision" in fact refers to the world market conditions for the good in question, which is not what the text of the SCM Agreement says. In light of the clear language of Article 14 (d) SCM Agreement, the "availability" of the good, the "conditions of purchase or sale", the "price", are various aspects of the market conditions existing in the country of provision, and refer to the price for the good in that country, its availability in that country, the conditions of sale as they are prevailing in that country. In our view, the bracketed language in Article 14 (d) SCM Agreement specifies what the market conditions referred to in the preceeding sentence are, and, as is the case for the "market conditions", they also all relate to the country of provision, and not some other country.

7.47 We consider that a proper interpretation is one which gives meaning to all the terms in the treaty and does not reduce certain parts of the treaty to redundancy or inutility.⁸³ We consider that the US is reading Article 14 (d) SCM Agreement to mean that the benchmark to be used should be "market conditions" in general. In our view, to adopt the US approach that the market refers to the entire market available to the allegedly subsidized producers, would effectively read out of the SCM Agreement the explicit reference to the country of provision, as the country the prevailing market conditions of which have to be used as a benchmark.

7.48 In support of its argument that the "market", as generally referred to in the SCM Agreement, is not restricted to the exporting country, but rather "encompasses the entire market available to the subsidized producer or exporter", the US refers to item (d) of the Illustrative List of Export Subsidies of Annex I of the SCM Agreement. We note that item (d) of the Illustrative List establishes conditions under which certain instances of provision of goods constitute export subsidies. Thus, item (d) is concerned with a significantly different situation from that addressed by Article 14 (d). While it is true that item (d) of the Illustrative List contains explicit language requiring as a second part of the test that it establishes that the terms of provision of the good be more favourable "than those commercially available on world markets", we recall that item (d) of the Illustrative List is not before us in this dispute. The text of Article 14 (d) SCM Agreement, the provision that is before us, does not provide for such a world market test. The fact that in the different context of criteria for a similar measure to constitute a prohibited export subsidy there is an explicit requirement to look at commercially available world market prices, cannot mean that any reference to the "market" in the SCM Agreement necessarily refers to the world market, or some portion thereof, particularly when the language in the provision clearly states otherwise.⁸⁴ We note that the prices of imported goods in the market of provision can indeed form part of the prevailing market conditions in the sense of Article 14 (d) SCM Agreement.⁸⁵ But this is not the same as the price for those goods prevailing in the country of export. Nor does this imply that import prices necessarily can be the exclusive basis to determine prevailing market conditions. To us, the US chain of reasoning seems to be the following: Canadian lumber producers can purchase standing timber and logs in the United States, and some do, therefore, the prevailing US stumpage prices form "part of" the prevailing market conditions for stumpage in Canada, and therefore US stumpage prices represent and can be used as the exclusive benchmark for the prevailing market conditions for stumpage in Canada. For the reasons set forth above, we cannot accept such reasoning. In sum, we find that US stumpage prices cannot be considered to constitute prevailing market conditions in Canada, the country of provision of the good (standing timber).

(ii) Did there exist valid reasons not to use Canadian private stumpage prices?

7.49 The US justifies its reliance on US prices as the benchmark by arguing that, although the use of Canadian private stumpage prices would have been the preferred option to calculate the amount of benefit, in this particular case it was not possible to use such prices as the benchmark because they were distorted and suppressed by the very large number of government sales. According to the US, the trade-distorting potential of the government's provision of a good can be identified only by reference to an independent market price, i.e., a price that is unaffected by the very trade distortion the test is designed to identify.⁸⁶

7.50 In our view, however, the "prevailing market conditions" of Article 14 (d) SCM Agreement do not refer to a theoretical market free of government interference as the US seems to be suggesting. Article 14 (d) SCM Agreement provides that the "prevailing" market conditions in the country of provision of the goods are to form the basis for the comparison. The ordinary meaning of the term "prevailing" market conditions is the market conditions "as they exist" or "which are predominant".⁸⁷ Considering that the only qualifier used to the "market conditions" in question is that they be "prevailing", we are of the view that the text of Article 14 (d) SCM Agreement does not in any way require the "market" conditions to be those of a hypothetical undistorted or perfectly competitive market.

7.51 Moreover, the chapeau of Article 14 SCM Agreement clearly states that Article 14 SCM Agreement establishes guidelines for the calculation of "benefit" to the recipient.⁸⁸ We are of the view that in order to calculate the benefit to the recipient, an authority is to compare the price the recipient paid the government with the prices prevailing in other market transactions. We do not consider that the goal of the examination of the benefit enjoyed by the recipient is to determine what the market price would have been absent the government's financial contribution, as the US is suggesting⁸⁹, or to measure the trade distorting potential of the government's financial contribution.⁹⁰ The text of Article 14 SCM Agreement does not require a general "but for" test to the prevailing market conditions. We are thus of the view that Article 14 (d) SCM Agreement does not require that the authority construct a market price that could have existed but for the government's involvement, nor does it allow the authority to decline to use in-country prices because they may be affected by the government's financial contribution.

7.52 We consider that if the drafters of the SCM Agreement had wanted to exclude the use of market prices in case of price suppression due to the government's involvement, they would have explicitly provided so, but they have not. The opposite is the case. As we found above, when it comes to the market conditions, the only qualifier in the text of the Agreement is "prevailing". Thus, the market conditions are those that are actually existing in the country and are those faced by the recipient of the financial contribution. The reference prices are those that the producer would have had to pay if it had to buy the goods now provided by the government from a different and independent seller.

7.53 We wish to note that even if in certain exceptional circumstances it may prove difficult in practice to apply Article 14 (d) SCM Agreement, that would not justify reading words into the text of the Agreement that are not there or ignoring the plain meaning of the text. In our view, the text of Article 14 SCM Agreement leaves no choice to the investigating authority but to use as a benchmark the market, for the good (or service) in question, as it exists in the country of provision.

7.54 Canada also argues that the USDOC was not justified in rejecting private stumpage sales in Canada as providing a benchmark for market conditions in Canada. We note that the USDOC in its preliminary CVD determination found that

"the softwood harvest from Crown lands accounts for approximately 70 to 90 per cent of the stumpage sold in each province. Therefore, between 70 and 90 per cent of the good or service within each of the provinces is provided by the government."⁹¹

7.55 We conclude from this statement in the Preliminary Determination that between 10 and 30 per cent of the good within each province was not provided by the government. In the course of the proceedings before us, we requested further clarification from the parties as to the record evidence on the percentage of softwood lumber harvest in each province in Canada that took place on Crown land versus private land. The US stated that the evidence on the record at the time of the Preliminary Determination indicated that private stumpage sales represented 2 per cent in Alberta, 6 per cent in Manitoba, 8 per cent in Ontario, 10 per cent in British Columbia, 10 per cent in Saskatchewan and 17 per cent in Quebec.⁹²

7.56 Finally, the US argues that as a practical matter, the USDOC could not have calculated the benefit on the basis of the private sales prices even if it had wanted to, since the Canadian provinces failed to provide the necessary information. We note first of all that, while the US has argued before us that private sales data were not provided, it is clear from the Preliminary Determination that this was not an element in the USDOC's decision to use a cross-border price analysis. In fact the USDOC clearly acknowledged that the Canadian provinces reported private stumpage prices but that "an examination of the information on the record demonstrates that the private stumpage prices reported by the Provinces do not constitute market determined prices �".⁹³ The alleged lack of information is thus not mentioned by the USDOC as a consideration which led to its decision to use a cross-border price analysis. In any case, even the US acknowledges that at least two provinces (Quebec and Ontario) did provide private market standing timber prices.⁹⁴ Nevertheless, even for those two provinces no private stumpage sales data were used as a benchmark, because of the alleged price suppression.⁹⁵

7.57 Accordingly, based on our interpretation of the text of Article 14(d) SCM Agreement, which does not require the prevailing market conditions to be those of an "undistorted" market, we consider that the USDOC provided no rationale in the context of Article 14 (d) SCM Agreement for rejecting Canadian private stumpage prices as the basis for calculating the benefit.⁹⁶

7.58 We note, moreover, that the domestic markets of the member countries of the WTO are not identical - nor are they expected to be - and that there is nothing in the WTO or SCM Agreement indicating that, in order to qualify as such, markets must meet specific qualitative requirements, as the US would seem to suggest. A contrary conclusion would lead to a result in which the importing country would have a very broad scope to choose another market, including its own, in order to determine benefit. Such a result would clearly distort the letter and purpose of Article 14 (d) and vitiate its intended application. We further note that using the US methodology for determining a benefit from the provision of government-owned resources that are not in themselves tradable across borders and not sold at public auction would lead to the virtual automatic determination of the existence of subsidization in a resource-rich exporting country, even where the perceived price difference was simply a reflection of the exporting country's comparative advantage in the product

(c) Conclusion

7.59 We therefore find that by using prevailing US stumpage prices, which by definition do not constitute the prevailing market conditions in Canada, the USDOC acted inconsistently with Article 14 and 14 (d) SCM Agreement in determining the benefit to the recipient, and therefore also acted inconsistently with Article 1.1 SCM Agreement in determining the existence of a subsidy. As for the issue of the appropriateness of the actual adjustments used to calculate the benchmark stumpage prices in this case, our understanding is that Canada raised this point as factual support for its basic legal argument that the cross-border methodology used by the USDOC was invalid, and that Canada has raised no separate claim in this regard. Given this, and given our conclusion above, we do not need to consider this issue.

4. Claim 3: failure to examine and determine the existence of benefit to the producers of the subject product

(a) Arguments of the parties

(i) Canada

7.60 Canada asserts that the USDOC impermissibly assumed that the alleged financial contribution to timber harvesters through the stumpage rights conferred a benefit on the downstream producers of softwood lumber. Canada notes that standing timber is harvested and processed into logs which are then processed in sawmills and pulp mills to produce a wide variety of products including softwood lumber, and that the lumber may then be sold at arm's-length as an end-product or sold to re-manufacturing industries that make a vast array of products. Canada argues that it is not the case, as the US is suggesting, that all tenure or licence agreements require tenure/licence holders to own a sawmill. Nor are there in general any requirements either by law or by the terms of tenure that require tenure holders to sell to specific mills or to sell at specific prices or under specific terms and conditions. Canada argues that only in certain cases are tenure holders required to construct or maintain a facility with the ability to mill harvested timber.⁹⁷ Stumpage tenure holders are therefore in general free to sell their logs to unrelated sawmills.⁹⁸

7.61 Canada asserts that the evidence on the record demonstrates that a significant portion of harvesting is done by entities operating at arm's-length from lumber producers, and logs produced by such independent entities are sold to lumber producers in arm's-length transactions. In such cases, according to Canada, the USDOC should have conducted a pass-through analysis to determine whether the alleged benefit from the stumpage programmes was passed through to the lumber producers.

7.62 Canada argues that the record indicates that for British Columbia, 30.09 per cent of timber from Crown licenses was harvested by companies that do not have sawmills (33.0 per cent if private land logging is included); in Quebec, apart from a certain type of tenures (Timber Supply Forest Management Agreement (TSFMA) tenures in particular), 27 per cent of the provincial harvest was sold through arm's-length transactions from private lands and Forest Management Contracts (FMC); in Ontario, 30 per cent of harvested timber from Crown land was sold to third parties who processed it into forest products; in Alberta, 6 per cent of total softwood logs harvest were sold at arm's-length in the sense that the transactions were between totally unrelated parties. Canada argues in addition that at least 8 primary mills requested an exclusion as they purchased all their log inputs in arm's-length transactions. Canada further asserts that a large number of unaffiliated remanufacturers purchase lumber in arm's-length transactions from lumber producers. Canada argues that in those arm's-length transactions, any alleged benefit to the lumber producers would not be passed through to the remanufacturers and that, therefore, remanufactured products cannot be presumed to be subsidized.⁹⁹ In any case, Canada submits, even prices between related parties may and often are also arm's-length prices because they are freely negotiated without regard to any relationship of the parties.¹⁰⁰

7.63 According to Canada, when transactions take place at arm's-length, the recipient of a subsidy is presumed to have retained the benefit. Canada submits that since the USDOC failed to conduct an analysis of whether any benefits were passed along in such arm's-length transactions in the case at hand, its finding of a benefit conferred to the Canadian softwood lumber producers by the alleged provision of stumpage to the upstream log producers is inconsistent with the SCM Agreement.¹⁰¹

(ii) United States

7.64 The United States proffers two main arguments against the need for a pass-through analysis for determining benefit to the lumber producers. First, the US argues that no truly arm's-length transactions between independent harvesters and independent lumber mills existed. Second, the US asserts that a pass-through analysis was not required in this case as the investigation was conducted on an aggregate basis.

7.65 Concerning the first point, the US argues that no pass-through analysis was required because the subsidies were bestowed directly on the producers of the subject merchandise, since the entity receiving the financial contribution (standing timber) and the entity receiving the benefit (a below-market stumpage price) are "generally one and the same".¹⁰² The United States asserts that a very significant number of timber harvesters also own sawmills and are therefore at the same time producers of softwood lumber. In other words, according to the US, most lumber producers are also harvesters of standing timber, which they obtain through the conclusion of tenure agreements with the Canadian provincial governments. The United States asserts that in fact each Canadian province generally requires that tenure holders be sawmills or own a sawmill.¹⁰³ The small portion of Crown timber harvested by tenure holders that do not own sawmills is subject to restrictions that tie the timber to specific sawmills in Canada.¹⁰⁴ In such circumstances, the US argues, the loggers and the lumber producers (sawmills) are thus not operating at arm's-length with each other¹⁰⁵, and there is no need to examine the pass-through issue. Thus, according to the US, the evidence indicates that there are virtually no truly arm's length transactions between harvesters and lumber mills, and the independent logger is therefore "largely a myth".¹⁰⁶

7.66 The US argues that, in Canada's view, in two other situations involving arm's-length transactions a pass-through analysis was required:(i) in case logs are harvested by one sawmill and then sold at arm's-length to another sawmill, and (ii) when lumber is sold at arm's-length to companies that produce remanufactured lumber products. The US submits that in those two situations, no pass-through analysis is required in an aggregate case as all of the entities involved are producers of the subject merchandise.¹⁰⁷ The US is of the view that for remanufacturers who produce merchandise within the scope of the investigation, it is only in the context of determining a company-specific subsidy rate that a pass-through analysis may be necessary.¹⁰⁸ The US asserts that, as a result, the USDOC did not request any information on these types of transactions. According to the US, the precise amount of the benefit received by individual producers (for example, whether the benefit stayed with the original recipient or "travelled to" the purchaser) would only be determined in a company-specific review.

7.67 In sum, the US argues that, given this evidence, it was not necessary for the USDOC to conduct a pass-through analysis in order to determine the existence of benefit to Canadian softwood lumber producers by the government provision of stumpage.¹⁰⁹

(b) Analysis

7.68 Article 1.1 (b) SCM Agreement establishes that a subsidy shall be deemed to exist if there is a financial contribution by the government and a benefit is thereby conferred on the producers of this subject product, in this case softwood lumber. The USDOC determined that the provision of stumpage, standing timber, constituted a financial contribution by the government. However, the subject product of the USDOC's countervailing duty investigation is the downstream product, softwood lumber, which is processed from logs, which in turn have been processed from standing timber. The various forms of stumpage agreements in the form of tenures or licenses are all concluded between timber harvesting companies/loggers and the provincial governments.

7.69 We consider that the question before us is whether under the facts of this case, the USDOC was required to examine whether, in certain transactions covered by the investigation, some or all of the alleged benefit to the tenure holders from the stumpage programmes was passed through to the producers of the subject merchandise exported to the US. These transactions were not only the case where a sawmill buys logs in an arm's-length transaction but also where a re-manufacturer buys lumber for further processing in an arm's-length transaction. Thus, conceptually, for some remanufactured products that are subject merchandise the issue of quantifying pass-through of benefit could involve two separate arm's-length transactions.

7.70 We note that the US agrees that if a government makes a financial contribution to an entity which does not produce the subject merchandise, it would be necessary to analyze whether that financial contribution benefitted another entity that does produce the subject merchandise.¹¹⁰ The US also acknowledges that, in theory, an arm's length transaction between an independent tenure holder and an unrelated lumber mill may give rise to the issue of whether the financial contribution to the tenure holder conferred a benefit on the lumber producer.

7.71 We are of the view that an authority may not assume that a subsidy provided to producers of the "upstream" input product automatically benefits unrelated producers of downstream products, especially if there is evidence on the record of arm's-length transactions between the two.¹¹¹ Rather, we consider that in such circumstances the investigating authority should examine whether and to what extent the subsidies bestowed on the upstream producers benefited the downstream producers. In this respect, we recall that in the US �- Lead and Bismuth II case (WT/DS138/AB/R) the Appellate Body concluded in a different context involving subsidies provided to entities subsequently privatized, that an entity other than the original recipient of the subsidy should not be deemed to have received a benefit in the case where an arm's-length price was paid to acquire the entity that had received the subsidies:

"68. The question whether a "financial contribution" confers a "benefit" depends, therefore, on whether the recipient has received a "financial contribution" on terms more favourable than those available to the recipient in the market. In the present case, the Panel made factual findings that UES and BSplc/BSES paid fair market value for all the productive assets, goodwill, etc., they acquired from BSC and subsequently used in the production of leaded bars imported into the United States in 1994, 1995 and 1996. We, therefore, see no error in the Panel's conclusion that, in the specific circumstances of this case, the "financial contributions" bestowed on BSC between 1977 and 1986 could not be deemed to confer a "benefit" on UES and BSplc/BSES. (emphasis added)¹¹²

7.72 Canada and the US agree that the evidence on the record demonstrates that a large number of tenure holders own a sawmill and are thus at the same time softwood lumber producers. It is clear that in such circumstances of complete identity between the tenure holder/logger and the lumber producer, no pass-through analysis is required. Canada argues however that in a significant number of cases, lumber producers purchase logs at arm's-length from unrelated tenure holders. In addition, Canada argues that logs are traded at arm's-length between lumber producers and that lumber is sold at arm's-length by sawmills to remanufacturers who also produce the subject merchandise. According to Canada, in all these situations, USDOC failed to examine whether the subsidy passed through to the producers of the subject merchandise, softwood lumber. The US submits that the evidence does not support Canada's argument concerning independent loggers. As far as the two other categories is concerned, the US submits that the question of pass-through is moot in an aggregate investigation since all of the entities concerned are also producers of the subject merchandise.

7.73 We first examine the evidence on the record with regard to the relationship between the lumber producers and the tenure holders/loggers. The USDOC Preliminary Determination and additional clarifying statements of the US in the current proceedings reveal that not all logging companies are at the same time softwood lumber producers. For Ontario, the USDOC states in its Preliminary Determination that lumber producers may obtain the forest products they need in five ways. One such way is to pay the government stumpage fees and harvest the timber themselves, but another way is to purchase logs at arm's-length from a company that harvested them from Crown lands.¹¹³ As the US clarified before the Panel, the record shows that for British Columbia, between 9 and 17 per cent of Crown timber harvest is done by independent harvesters.¹¹⁴ For Alberta, the US states that 95 per cent of softwood timber was provided to tenure holders who own sawmills. This suggests that some 5 per cent of the timber provided would have been sold to sawmills by unrelated loggers. For Saskatchewan and Manitoba, the US gives a similar figure of 95 per cent of softwood timber which was provided to entities that also hold provincial licences to operate sawmills.¹¹⁵ The USDOC also acknowledged that it received a number of company specific requests for exclusion, thirteen of which from companies "purchasing logs at arm's-length, from private lands, or from US suppliers".¹¹⁶

7.74 We find that the US has thus conceded before us that in a certain number of cases, the lumber producers were independent from the tenure harvesters and may well have had to pay an arm's-length price to obtain the allegedly subsidized logs from the harvesters. Nothing in the record indicates, and the US does not argue, that in such cases where the harvesters and lumber producers are unrelated, the USDOC examined whether the prices charged between the harvester and its customer, the unrelated lumber producer, were at arm's-length, and hence whether any benefit passed through. We find that in such circumstances, where a downstream producer of subject merchandise is unrelated to the allegedly subsidized upstream producer of the input, an authority is not allowed to simply assume that a benefit has passed through. We further consider that conducting the investigation on an aggregate basis versus a company-by-company basis is irrelevant to this issue. The fact that in the large majority of cases the lumber producers and the harvesters are related, does not imply that it is no longer necessary in cases where there is no such relationship to examine and determine whether a benefit existed for the independent producers of the subject merchandise including independent re-manufacturers. By failing to examine whether the independent lumber producer paid arm's-length prices for the logs that they purchased, the USDOC determined the benefit to the producers of the subject merchandise inconsistently with the SCM Agreement.

7.75 The US argues, with regard to the situations of sawmills buying logs from other sawmills at arm's-length and of remanufacturers purchasing lumber from unrelated sawmills, that no pass-through analysis was required as all of the entities involved are producers of the subject merchandise. In sum, the US position, both in the investigation and before us, is that in an aggregate case, no pass-through analysis is necessary where all of the entitities involved are producers of the subject merchandise. We find no basis in the SCM Agreement for this argument, however. In our view, the purpose of the original investigation is precisely to determine whether and to what extent a subsidy has been granted to the producer of the subject merchandise. An authority cannot simply assume the existence of such a benefit in the original investigation. As the Appellate Body in the Lead and Bismuth Case stated:

"In an original investigation, the investigating authority must establish that all conditions set out in the SCM Agreement for the imposition of countervailing duties are fulfilled."¹¹⁷

7.76 Here we recall that any calculation of the rate of subsidization in a countervailing duty investigation involves three distinct steps. First is the calculation of the total amount of the subsidy provided during the period of investigation (the establishment of the numerator of the equation). Second is the calculation of the total amount of relevant sales to which the subsidy amount can be attributed (the establishment of the "sales denominator"). Third is the calculation of the rate of subsidization, by dividing the subsidy amount numerator by the denominator of relevant sales. The result is an ad valorem rate of subsidization of the subject merchandise.¹¹⁸

7.77 In this dispute, the US seems to be approaching the issue of pass-through of benefits as if it can be resolved solely by correctly identifying the value of relevant sales of the subject merchandise (softwood lumber) to be used as the denominator. This appears to be the reasoning behind the US position that a pass-through analysis for the purchases of inputs by unrelated producers of softwood lumber would be unnecessary where the calculation is performed on an aggregate basis. We disagree. In our view, the issue of pass-through has to do in the first instance with correctly establishing the amount of the subsidy benefiting the producers of the subject merchandise, i.e. the numerator. That is, alleged benefits from the stumpage programmes can only be included in the total subsidy amount to the extent that they benefit the producers of the subject merchandise, softwood lumber. Where a given producer of the subject merchandise itself harvests logs, and thus itself is the recipient of the alleged benefit, the entire amount of the alleged benefit to that recipient can be included in the subsidy amount (numerator) without further analysis. Where, however, a producer of softwood lumber does not itself harvest logs, but instead buys logs or lumber from unrelated suppliers, any alleged benefit from the stumpage programmes that may have benefitted the producer of the logs or lumber involved could only be included in the total subsidy amount to the extent that it has been established as a factual matter that the purchaser has received some or all of the benefit.

7.78 We find particularly important in this respect that, as Canada pointed out¹¹⁹, prior to the Preliminary Determination, the USDOC had applications for exclusion from 98 producers who purchased logs or lumber from unrelated suppliers.¹²⁰ Of these, 78 were remanufacturers unrelated to firms with harvesting rights and at least 8 were first mills that purchased all of their log inputs from unrelated suppliers. In light of the fact that, at the Preliminary Determination, the USDOC had knowledge of a sizeable number of lumber producers purchasing log and lumber inputs from unrelated suppliers allegedly at arm's-length, the USDOC should have examined, in calculating the total amount of the alleged subsidy, whether and to what extent any benefit passed through to those lumber producers.

(c) Conclusion

7.79 In sum, we find that in this case, there was evidence on the record of a certain amount of sales by producers of logs that did not own a lumber-processing facility, and were not related to the downstream lumber producers to which they sold their logs. In addition, there is evidence on the record that USDOC was aware of the fact that a number of lumber producers buy logs or lumber inputs from unrelated sawmills We therefore find that, given such record evidence, the absence of any examination of these transactions by the USDOC was not consistent with the SCM Agreement. The USDOC should have examined, in the original investigation, whether, in cases where the tenure holders were not at the same time processing the logs in their own sawmills, the lumber producers benefited from the financial contribution given to the tenure holders. We therefore conclude that the USDOC imposed a provisional countervailing duty without determining the existence and amount of the benefit conferred on the allegedly subsidized product, and, therefore, in a manner inconsistent with the SCM Agreement.¹²¹

64 Canada's First Written Submission, para. 43.

65 Canada also refers to the protocol of accession of China which explicitly, and in its view, exceptionally allows for the use of a benchmark outside of China. In Canada's view, there would be no need to include such a provision in a protocol of accession if, as the US is arguing, the language of Article 14 already allows for the use of cross-border benchmarks.

66 Canada provides a complete discussion of the data provided by the Canadian provinces in this respect in its Second Written Submission, paras. 53 -62.

67 Canada's Second Written Submission, para. 35. Canada notes that the USDOC did not have any evidence of alleged price suppression, but simply relied on a erroneous translation of a letter of the Quebec Minister of Natural Resources which formed part of the petition, which in fact merely stated that public land stumpage charges could have an indirect influence on the private market. Canada argues that a doctoral thesis further relied on in this respect by the United States before the Panel is outdated and the information on which it is based has earlier been rejected by the USDOC in the earlier investigation of imports of lumber from Canada (Lumber II) as evidence of alleged price suppression.

68 This, Canada argues, USDOC recognized in the first three investigations on imports of lumber from Canada (Lumber I -III) in which, in an identical factual setting, it rejected the use of cross-border benchmarks. Canada discusses the practical problems relating to the use of US benchmarks in this case in its Second Written Submission paras. 44 - 52.

69 According to the US, "in relation to" means "with reference to" and thus, under Article 14 (d) the prevailing conditions in the country of provision are a reference point, not necessarily an end point, for the market benchmark. US First Written Submission, para. 43

70 The US refers to the Appellate Body report in the case Canada - Measures Affecting the Importation of Milk and the Exportation of Dairy Products ("Canada -Dairy"), Recourse to Article 21.5 DSU (DS103AB/RW, para 184) and item (d) of the Illustrative List of Export Subsidies in Annex I SCM Agreement.

71 US Oral Statement at the Second Meeting of the Panel with the Parties ("US Second Oral Statement"), para. 7.

72 US Second Oral Statement, para. 8.

73 In para. 26 of its Answers to Questions from the Panel after the First Meeting, the US argues that the Canadian provincial governments' share of the market was 90 per cent or more.

74 The US refers to the Report of the Panel in Canada-Dairy in support of its argument in this respect. Panel Report, Canada - Dairy, WT/DS103/R, adopted as modified by WT/DS103/AB/R on 27 October 1999, para. 7.54.

75 US Second Written Submission, para. 27; US Second Oral Statement, para. 18.

76 USDOC Preliminary Determination, p. 43193.

77 USDOC Preliminary Determination, p. 43195.

78 US Answers to Questions from the Panel after the Second Meeting, para. 85

79 US Answers to Questions from the Panel after the Second Meeting, para. 92.

80 US First Written Submission, para. 43.

81 US Answers to Questions from the Panel after the Second Meeting, para. 82

82 US Second Written Submission.

83 As the Appellate Body stated in the United States - Standards for Reformulated and Conventional Gasoline case: "� One of the corollaries of the �general rule of interpretation� in the Vienna Convention is that interpretation must give meaning and effect to all the terms of the treaty. An interpreter is not free to adopt a reading that would result in reducing whole clauses or paragraphs of a treaty to redundancy or inutility." Appellate Body Report, United States - Standards for Reformulated and Conventional Gasoline case, WT/DS2/AB/R adopted on 20 May 1996, p.23.

84 The US has also referred to various statements of the Panel and the Appellate Body in the Canada - Dairy case. We are of the view that these reports are of little, if any relevance to the case in question as both dealt with certain provisions concerning export subsidies under the Agreement on Agriculture. We note that the Panel report was overturned on appeal, and that the Appellate Body itself rejected the use of world market prices as a benchmark. See Appellate Body Report, Canada -Dairy, WT/DS103/AB/RW, adopted on 18 December 2001, para. 84.

85 We note that in its Third Party submission, the European Communities submits that �a proper analysis of the �market conditions in the country of provision� may include all commercially available alternative sources for the recipient, including the price for imports into that market�. European Communities Third Party Submission, para. 26.

86 We note, however, that, as mentioned earlier (para. 7.36 of our Report), the United States has recognised that US timber is commercially available to Canadian lumber mills. This assertion would contradict the US rejection of the existence of a competitive timber and /or lumber market in Canada.

87 Concise Oxford Dictionary, Ninth Edition, p. 1084..

88 We note that the US agrees that "As stated in the chapeau to Article 14, and confirmed by the Appellate Body, the benefit for purposes of paragraph 1 of Article 1 is the benefit to the recipient." US Answers to Questions from the Panel after the First Meeting, para. 41. The Appellate Body in the Canada - Measures Affecting the Export of Civilian Aircraft case interpreted the term �benefit� in the SCM Agreement in the following manner:

"157. We also believe that the word "benefit", as used in Article 1.1(b), implies some kind of comparison. This must be so, for there can be no "benefit" to the recipient unless the "financial contribution" makes the recipient "better off" than it would otherwise have been, absent that contribution. In our view, the marketplace provides an appropriate basis for comparison in determining whether a "benefit" has been "conferred", because the trade-distorting potential of a "financial contribution" can be identified by determining whether the recipient has received a "financial contribution" on terms more favourable than those available to the recipient in the market." (emphasis added)

Appellate Body Report, Canada - Measures Affecting the Export of Civilian Aircraft, WT/DS70/AB/R, adopted 20 August 1999, para. 157.

121 n light of this finding, we see no need to address the issue raised by Canada in footnote 101.

To continue with 5. Claim 4: impermissible application of a provisional duty in excess of the subsidy rate

Return to Table of Contents