5.106 According to Brazil, neither these statements by senior Canadian Government officials nor the statements about the industry in TPC's promotional materials were crafted in a vacuum. Brazil maintains that at the time the Canadian Government was making these statements, it was eminently aware of the fact that every single sale of Dash 8 series aircraft made since 1992, and every single sale of the CRJ since its development and commercialization, was for export. For Brazil, the industry is devoted to exports, and the Canadian Government has therefore made it very clear that it maintains massive amounts of support to the Canadian regional aircraft industry precisely because it is an export industry and precisely because it anticipates that the Canadian regional aircraft industry will continue to be an export industry.

5.107 Regarding Canada's arguments (para. 5.86) concerning anticipated exportation or export earnings, Brazil indicates first that it agrees with Canada's statement that "'[t]he prohibition in Article 3 is both forward looking and backward looking'", noting that it is for this reason that Brazil cites the overwhelming export orientation of the Canadian regional aircraft industry as one reason supporting the conclusion that assistance granted under or through TPC, the Subsidiary Agreements, SDI/IQ and the sale of de Havilland constitute subsidies contingent upon export. Brazil argues that in disbursing funds to the Canadian regional aircraft industry under the auspices of these programmes, the federal and provincial governments, based on the past export orientation of the industry, anticipate or expect that the this orientation will continue in the future.

5.108 Second, Brazil also agrees with Canada's statement that "'tied to . . . anticipated exportation'" means that "'one of the conditions for the grant of the subsidy is the expectation that exports will flow thereby'" (para. 5.90). Brazil contends that the evidence before the Panel demonstrates that assistance under or through TPC, the Subsidiary Agreements, SDI/IQ and the sale of de Havilland would not have been granted were it not for the virtually total export orientation of the Canadian regional aircraft industry. In this sense, exportation is a condition of the Canadian subsidies.

5.109 According to Brazil, there is no requirement in Article 3.1 that for a subsidy to be contingent "in fact" upon export, the condition be express, and were that the case, there would be no distinction between a subsidy contingent "in law" and a subsidy contingent "in fact" upon export. Members thus would be able to avoid the prohibition of Article 3.1 by simply leaving the word "export" out of their laws, regulations and dealings.

5.110 Third, Brazil emphasizes that it is challenging the application of TPC, the Subsidiary Agreements and SDI/IQ to the Canadian regional aircraft industry, which is overwhelmingly export oriented. As a result, Brazil submits that it is not relevant to the Panel's decision in this case that TPC, the Subsidiary Agreements and SDI/IQ may also grant assistance to other industries with sales in export markets, or to other industries with sales in domestic markets. Brazil argues that when assistance under these programmes is granted to the regional aircraft industry, it is granted because the industry is totally export oriented, and because the Canadian federal and provincial governments anticipate or expect that it will remain so. Brazil states that it has identified many factors apart from export performance or orientation as contributing to its conclusion that that assistance granted to the regional aircraft industry under these programmes is therefore contingent "'in fact'" upon export (e.g., paras. 6.227, 6.231). Thus, Brazil disagrees with Canada's argument that the export performance of the regional aircraft industry is the only factor called upon by Brazil in support of its case.

5.111 Fourth, Brazil disagrees with Canada's statement that the terms "'actual'" and "'anticipated'" from footnote 4 to the SCM Agreement are nothing more than a "'reformulation'" of the term "'export performance'" in the body of Article 3.1(a). According to Brazil, the implication of Canada's statement is that the terms "'actual'" and "'anticipated'" should not be accorded their ordinary meaning. Brazil submits that footnote 4 discusses the meaning of the term contingent "'in fact,'" and therefore the ordinary meaning of the terms included in that footnote are crucial to an understanding of contingent "'in fact.'" Brazil notes that footnote 4 provides that subsidies are considered contingent in fact on export if they are "tied to actual or anticipated exportation or export earnings". Brazil recalls its argument regarding the "'ordinary meaning'" of the word "'anticipate'" and Brazil contends that funding to the Canadian regional aircraft industry under or through TPC, DIPP, the Subsidiary Agreements, SDI/IQ, and the de Havilland sale was granted because the Canadian federal and provincial governments anticipated, or expected, based on ample historical evidence, that the industry would remain virtually totally export oriented.

5.112 Regarding Canada's argument that Brazil's view of Article 3.1(a) as applying to situations of "'anticipated'" exportation or export earnings would lead to "'a manifestly absurd or unreasonable result'" (para. 5.4), namely that "'that there would be one law for larger economies that are not dependent on international trade and another for smaller economies that are'", Brazil states that Canada is effectively arguing that despite the terms of Article 3.1(a), Canada's alleged position as a "'small economy'" warrants that it be accorded special treatment. Brazil doubts Canada's claim of "'small economy'" status - as a founding Member of the WTO, a member, with the United States, Japan and the European Communities, of the WTO "Quad," and a member of the OECD. For Brazil, moreover, Canada's argument must fail as a matter of law, because there is only one law, applying equally to large and small economies alike, and it is found in the ordinary meaning of Article 3.1(a).

5.113 Brazil argues that this is not a case of a small economy of necessity having a small share of a global market, as Canada claims, or of support being given to an industry by a government indifferent to whether that industry sells in domestic or export markets, but is a case in which, by the deliberate action of the Canadian Government, whatever demand existed in the domestic market was satisfied by what were legally export sales, not domestic sales.

5.114 In this regard, Brazil recalls the EDC President's comments to Parliament about EDC's support, with export funds, of a sale to a foreign company of planes developed with TPC funds and ultimately leased to a domestic carrier, in order, according to the EDC President, to "'launch an aircraft that has a world market.'"

5.115 For Brazil, these facts belie Canada's "small market" claim, and demonstrate that the Panel is not faced with a situation in which domestic sales are a small percentage of total sales because the domestic market is small, but is rather faced with a situation in which domestic demand, however large or small, was satisfied by the Export Development Corporation with export development funds for the purpose of launching an export product which had benefited from TPC development funds. To Brazil, these facts establish that Canada's use of TPC support for the aircraft industry was, de facto if not de jure, an export subsidy.

3. Response of Canada

5.116 In response to Brazil's argument that Canada's proposed tests for de facto export contingency do not find textual support in the SCM Agreement (paras. 5.101-5.102), Canada notes that it proposed three tests for determining whether a subsidy is in fact tied to exports. Canada submits that the tests Canada proposed find their support in the text of the SCM Agreement, in the words "contingent upon" and "tied to", and that though anchored in legal provisions, tests, by definition, attempt to elaborate and clarify those provisions. Thus, Canada argues, tests are used as an aid to the application of laws. Canada notes that Brazil has not shown what other criteria would be used, or useful, in determining whether a subsidy is in fact contingent on export performance. Canada submits that it is because of that, and to help the Panel in its determination, that Canada has put forward these suggested tests. Canada acknowledged that regardless of the test used, the legal requirement is still the same: one of the conditions for granting a subsidy must be that exports take place. Regarding the United States's (paras. 7.35-7.37) and Brazil's (paras. 5.101-5.102) objection that Canada's argument on Article 3.1(a) applies only to actual exports, rather than anticipated or expected exports, Canada asserts that the objection is based on a reading that the test proposed in footnote 4 is one of intent rather than conditionality. Canada submits that this reading of Canada's arguments is not accurate, in that Canada's view is that footnote 4 prohibits subsidies that are conditional on, or tied to, exportation, whether anticipated or actual.

5.117 For Canada, Article 3.1 refers to both actual and anticipated exports, but whether exports are taking place or are expected to take place, the key verbs in Article 3.1, are contingent upon in the body and tied to in footnote 4. Canada submits that each of these denotes the existence of a condition; each connotes a "but for" test. Canada notes that exports that were expected might not happen, but that this does not mean that a subsidy that had been contingent upon such expectations is saved. Canada observes, however, that if a government wants to grant export subsidies, and exports do not result, the government would change the programme to ensure exports do follow. For Canada, that is how a de facto export subsidy can be detected, even if initially no exports take place.

5.118 Regarding Brazil's arguments that Canada's interpretation "lacks support" in the relevant rules of treaty interpretation, Canada recalls that the principles of treaty interpretation require that the ordinary meaning of the words of a treaty be viewed in the light of its context and object and purpose. For Canada, the phrase "anticipated exportation or export earnings" should therefore be read in context - that is to say, in the context of the condition explicitly provided for in the verb tied to. In Canada's view, nothing in footnote 4 or in Article 3 changes the nature of the condition: whether it is the only one or one of many, and therefore whether anticipated or actual, exportation must be a requirement of a subsidy for it to fit within Article 3.

5.119 In this sense, Canada asserts, "actual or anticipated exportation" is simply a reformulation - an elaboration - of "export performance", just as "tied to" is a reformulation of "contingent upon." That is, the first sentence of footnote 4 does not purport to create a new obligation, but rather, elaborates the existing obligation in Article 3.

5.120 Regarding Brazil's views on footnote 4, Canada argues that Brazil ignores the second sentence, which according to Canada was the sentence that finally made compromise on this provision possible. Canada submits that Brazil's claims should be examined in the light of the interpretation that Canada has put forward for Article 3, an interpretation that respects the second sentence of footnote 4. In Canada's view, this sentence renders Brazil's case null and void.

To continue with Arguments of Parties Regarding Canadian Measures Alleged by Brazil

¹⁴¹ TPC Annual Report 1996-1997, pg. 5 (Exh. BRA-5).

¹⁴² Canadian Taxpayers Federation study, "Corporate Welfare - A Report on Sixteen Years of Industry Canada Financial Assistance," 16 April 1998, Tab 4 (Exh. BRA-31). This statement of total DIPP funding to the aerospace sector represents aggregate DIPP funds released to four Canadian companies, i.e., Bombardier, de Havilland, Pratt & Whitney Canada Inc., and CAE Electronics Ltd.