GAAR, transaction, Federal Court of Appeal, subsection, transactions, economic substance, statutory scheme, Supreme Court, Justice Miller, Canadian Pacific, canadian tax journal, economic realities, Court of Appeal, provisions, legal form, Canadian tax law, legal relationships, OSFC, compound interest, contemplation, Canadian Tax Foundation, legal substance, Canada Trustco, The Court of Appeal, economic cost, statutory interpretation, the common law, Shell Canada, OSFC Holdings Ltd., the common law series, case
The Long, Slow, Steady Demise of the General Anti-Avoidance Rule Brian J. Arnold*
KEYWORDS: GENERAL ANTI-AVOIDANCE RULE
Misconceptions About the General Nature of the GAAR
The Series Aspect of the Definition of Avoidance Transaction
Analysis of Misuse and Abuse
Economic Realities, Economic Substance, and the GAAR
245 of the income tax
Act,1 the general anti-avoidance rule (GAAR), was introduced in 1987 and is generally applicable with respect to transactions occurring after 1987. The GAAR was adopted in response to the Supreme Court
of Canada's rejection of the business purpose test in the 1984 Stubart case.2 It was intended to provide the Canada Revenue Agency (CRA) and the courts with a balanced and effective statutory basis for combatting abusive tax avoidance. The first case involving the GAAR, McNichol,3 was heard by the Tax Court of Canada in 1997, almost 10 years after the rule was enacted. Since that time, there has been a steady stream of GAAR cases; as of early 2004, about 20 have been decided by the courts.4 Although the Supreme Court of Canada has yet to hear a GAAR * Of Goodmans LLP, Toronto. 1 RSC 1985, c. 1 (5th Supp.), as amended (herein referred to as "the Act"). Unless otherwise stated, statutory references in this article are to the Act. 2 Stubart Investments Ltd. v. The Queen,  CTC 294 (SCC). 3 McNichol v. The Queen,  2 CTC 2088 (TCC). 4 Cases decided by the Tax Court of Canada: RMM Canadian Enterprises Inc. v. The Queen,  1 CTC 2300; Nadeau v. The Queen,  3 CTC 2235; Fredette v. The Queen,  3 CTC 2468; Rousseau-Houle v. The Queen, 2001 DTC 250; McNichol, supra note 3; Husky Oil Ltd. v. The Queen,  4 CTC 2691; Jabs Construction Ltd. v. The Queen,  3 CTC 2556; 488 (2004) vol. 52, no 2
the long, slow, steady demise of the general anti-avoidance rule 489 appeal,5 the Federal Court of Appeal has decided 8 cases. The GAAR decisions to date, and in particular those of the Federal Court of Appeal, provide a sufficient basis for one to discern the fundamental attitude and approach of the courts to the interpretation and application of the GAAR. The purpose of this short article is to identify and analyze selected aspects of the case law dealing with the GAAR that, in my opinion, will inexorably render the rule largely ineffective, as the title of the article suggests. The article is not intended to be a comprehensive analysis of the GAAR and the related case law. Instead, it focuses on the core elements of the GAAR that are key to its effectiveness: the series of transactions aspect of the definition of "avoidance transaction," the test for whether an avoidance transaction constitutes a misuse or an abuse, and the role, if any, of economic substance in the application of the GAAR. The article commences with references to some prejudicial misconceptions that the courts have articulated concerning the general nature of the GAAR. It concludes with some simple suggestions for amending the GAAR to put the courts back on the right track. MISCONCEPTIONS ABOUT THE GENERAL NATURE OF THE GAAR Occasionally judges cannot resist the temptation to make general comments, often uncomplimentary, about the tax legislation they are called on to interpret. For example, they have complained about the prolix, turgid, impenetrable statutory drafting that characterizes the legislation.6 They have also made general comments about the GAAR. These comments provide insights into the judges' attitudes about the GAAR that could influence their decisions in GAAR cases. In my view, these Mathew v. The Queen,  1 CTC 2045; Geransky v. The Queen,  2 CTC 2147; Imperial Oil Ltd. v. The Queen,  2 CTC 2754; Hill v. The Queen,  4 CTC 2548; Loyens v. The Queen,  3 CTC 2381; and Canada Trustco Mortgage Co. v. The Queen,  4 CTC 2009. Cases decided by the Federal Court of Appeal: OSFC Holdings Ltd. v. The Queen,  4 CTC 82; Water's Edge Village Estates (sub nom. Duncan v. The Queen),  4 CTC 1; Kaulius et al. v. The Queen, 2003 DTC 5644; The Queen v. Canadian Pacific Ltd.,  2 CTC 197; The Queen v. Donohue Forest Products
Inc.,  3 CTC 160; The Queen v. Jabin Investments Ltd.,  2 CTC 25; The Queen v. Imperial Oil Limited, 2004 DTC 6044; and The Queen v. Canada Trustco Mortgage Company, 2004 DTC 6119. 5 The Supreme Court refused leave to appeal in OSFC Holdings Ltd., supra note 4, the only GAAR case to date in which leave to appeal has been requested. 6 See, for example, Citibank Canada v. The Queen,  2 CTC 2260, at paragraphs 29 and 33 (TCC) (aff 'd.  2 CTC 171 (FCA)), in which Judge Mogan referred to the statutory definition of "term preferred share" as follows: The definition of "term preferred share" is prolix in the extreme. The persons who drafted that definition did not practise any economy of words or language. One may well ask how many members of parliament understood the definition when it was made law by amendment to the Act. . . . It is so detailed; so particularized; so long and tedious and excessive in its use of language.
490 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
comments ref lect misconceptions about the GAAR that might prejudice its application by the courts. In the Jabs Construction case, Associate Chief Justice Bowman of the Tax Court referred to the GAAR as follows:
Section 245 is an extreme sanction. It should not be used routinely every time the Minister gets upset just because a taxpayer structures a transaction in a tax effective way, or does not structure it in a way that maximizes the tax.7
In the Hill case, Justice Miller referred to the GAAR as the government's "ultimate weapon":
The Appellant has most deliberately relied on the sanctity of legal relationships, not to be impugned by the economic realities of a situation, in achieving his goal. . . . [W]hile GAAR may be the ultimate weapon for the government to undo such legal relationships, in this instance the application of GAAR is simply ineffective.8
In the Canada Trustco case, Justice Miller warned that
GAAR is not to be imposed lightly. . . . [T]his is tax legislation to be applied with utmost caution.9
In the Fredette case, Justice Archambault wrote:
Section 245 is a powerful tool for discouraging and preventing flagrant abuses of the Act. It cannot serve as a tool for the Minister to force taxpayers to structure their transactions in the manner most favourable to the tax authorities.10
In the Canada Trustco case, Justice Miller referred to Justice Archambault's decision in the Rousseau-Houle case that the GAAR did not apply to a misuse or abuse of regulations under the Act in the following terms:
[I]t should only be Parliamentarians through legislative policy, not the executive through regulatory policy, that can yield [sic] as heavy a hammer as GAAR to deny a taxpayer a tax benefit.11
7 Jabs Construction Ltd., supra note 4, at paragraph 48. Also quoted with approval in the Tax Court decisions in Geransky and Loyens, supra note 4, and in Canadian Pacific Ltd. v. The Queen,  1 CTC 2190 (TCC). 8 Hill, supra note 4, at paragraph 63. 9 Canada Trustco Mortgage Co., supra note 4, at paragraphs 77 and 91 (TCC). 10 Fredette, supra note 4, at paragraph 76. See also the similar statement in Rousseau-Houle, supra note 4. 11 Canada Trustco Mortgage Co., supra note 4, at paragraph 58 (TCC).
the long, slow, steady demise of the general anti-avoidance rule 491 In Geransky, Associate Chief Justice Bowman wrote: The Income Tax Act is a statute that is remarkable for its specificity and replete with anti-avoidance provisions designed to counteract specific perceived abuses. Where a taxpayer applies those provisions and manages to avoid the pitfalls the Minister cannot say: "Because you have avoided the shoals and traps of the Act and have not carried out your commercial transaction in a manner that maximizes your tax, I will use GAAR to fill in any gaps not covered by the multitude of specific anti-avoidance provisions." That is not what GAAR is all about.12 The Federal Court of Appeal has been more circumspect to date in making general comments about the GAAR. In OSFC, Justice Rothstein indicated that [t]he Court will proceed cautiously in carrying out the unusual duty imposed upon it under subsection 245(4).13 These statements from the GAAR cases are selective. They are drawn largely from cases decided by the Tax Court, and they are attributable to only a few judges. As a result, they must be considered with caution. They do not necessarily represent the views of all judges. What these statements suggest clearly is that some judges consider the GAAR to be an extraordinary provision that must be applied cautiously. This is a predictable reaction. But the statements also suggest clearly that some judges regard the GAAR with considerable suspicion and that, because of its nature, they see their role in a fundamental sense as one of limiting its application. This view of the GAAR is not the only possible view. The GAAR can be seen as a necessary and measured response by Parliament, after trying all other possible responses, to the serious problem of abusive tax avoidance. Such tax avoidance undermines the fairness and integrity of the tax system. It is not inconceivable that the courts might view their role as assisting Parliament--or at least not frustrating the will of Parliament--in carrying out its clear intention to prevent abusive tax avoidance. Indeed, in the first GAAR case decided by the Tax Court, Justice Bonner expressed this general view of the GAAR: Section 245 itself must not be read in a disjointed way. It is not to be assumed that the legislature enacted subsection (4) based on some sort of consciousness that the scope of subsection (3) was far greater than is evident from its language. Tests suggested by counsel such as "extreme undermining of statutory purpose" and "ordinary tax planning" are of little assistance and are not justified by the language of section 245 read as a whole. To accept such tests would undermine the object and spirit of section 245 12 Geransky, supra note 4, at paragraphs 42-43. Geransky was also cited with approval in Loyens, supra note 4, at paragraph 116. 13 OSFC Holdings Ltd., supra note 4, at paragraph 69, quoted with approval in Water's Edge, supra note 4, at paragraph 52.
492 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
and run counter to the teleological approach mandated by the Supreme Court of Canada: Quйbec (Communautй urbaine) v. Corp. Notre-Dame de Bonsecours. The telos of section 245 is the thwarting of abusive tax avoidance transactions.14
I am curious as to why some judges see their role with respect to the GAAR as serving as a check or restraint on the authority of Parliament and the tax authorities to control abusive tax avoidance. It may be that this is how they see their role generally with respect to tax matters; however, the role of the courts in the tax system is a much larger issue that is well beyond the scope of this brief article. The views expressed in the foregoing quotations do not accurately reflect the reality of the GAAR. These statements are widely quoted and highly influential; but they are exaggerated and, what is worse, prejudicial to the application of the GAAR. The reality of the general nature of the GAAR is better understood in the following context:
1. Parliament had no other option for preventing abusive tax avoidance.15 The Supreme Court rejected the application of a business purpose test in the Stubart case. It has insisted on a literal or plain meaning approach to the interpretation of tax legislation. It has insisted on strict adherence to the legal form of transactions and consistently rejected any reference to economic realities in applying the provisions of the Act. Specific anti-avoidance rules are of limited effectiveness.16 Better-designed and better-drafted legislation is a desirable goal; but even if it is achievable to some degree, it cannot ever be an adequate response to the problem of tax avoidance. So what is Parliament to do? It did the only reasonable thing left for it to do--enact a general anti-avoidance rule. 2. The GAAR is an extraordinary provision. It is a provision of last resort, applicable only after all of the other relevant provisions of the Act have been applied. If the GAAR applies, it gives unusually broad discretion to the minister to determine the tax consequences as are reasonable in the circumstances.17 But the GAAR is not, as some judges have suggested, "an extreme sanction," "the ultimate weapon," or a "heavy hammer." The result of the application of the GAAR is that the taxpayer is deprived of the inappropriate tax benefit that would have resulted from the abusive tax-avoidance transaction. That is all. If the GAAR applies, no penalty is imposed to discourage taxpayers from taking the chance that they might get away with abusive taxavoidance transactions. The tax imposed is only the tax that would have
14 McNichol, supra note 3, at paragraph 23. 15 See generally Brian J. Arnold, "Reflections on the Relationship Between Statutory Interpretation and Tax Avoidance" (2001) vol. 49, no. 1 Canadian Tax Journal 1-39. 16 See Canada, Report of the Royal Commission
on Taxation (the Carter commission), vol. 3 (Ottawa: Queen's Printer, 1966), 554-56. 17 Subsection 245(5).
the long, slow, steady demise of the general anti-avoidance rule 493 resulted if the abusive tax-avoidance transaction had not been carried out. Small wonder, then, that tax-avoidance schemes have continued to flourish since the adoption of the GAAR. Abusive tax avoidance presents the possibility of significant tax savings, and little downside risk. 3. The CRA has applied the GAAR in a very restrained, responsible fashion. Any application of the GAAR as a basis of assessment must be approved by the GAAR Committee, a head office committee with representatives from the Department of Finance and the Department of Justice
, as well as various divisions of the CRA.18 According to the most recent statistics, as of March 31, 2004, the GAAR Committee has approved the application of the GAAR as either the primary or an alternative basis of assessment in only 374 cases out of a total of 570 cases considered since 1988. Moreover, the application of the GAAR in 374 cases over 15 years must be considered in the context of over 12,000 corporate audits performed annually by the CRA. These statistics indicate that the CRA is not imposing the GAAR lightly or indiscriminately, and certainly not whenever taxpayers structure their affairs in a tax-effective manner. THE SERIES ASPECT OF THE DEFINITION OF AVOIDANCE TRANSACTION The GAAR applies to avoidance transactions that involve a misuse or an abuse of the provisions of the Act. An "avoidance transaction" is defined to be a transaction that would result in a tax benefit unless the primary purpose of the transaction, objectively determined, was other than to obtain the tax benefit.19 But the definition also includes a transaction that is part of a series of transactions that would result in a tax benefit unless the primary purpose of the transaction--not the series--was other than to obtain the tax benefit.20 In other words, if a series of transactions results in a tax benefit, the GAAR will apply to each transaction in the series unless the transaction has a primary non-tax purpose. The series aspect of the definition of an avoidance transaction is a key element of the GAAR because most aggressive tax planning involves more than a single transaction. The series rule is intended to prevent taxpayers from inserting tax-motivated transactions without any commercial purpose into a series of transactions that has an overall legitimate business purpose. 18 See Wayne Adams, "The General Anti-Avoidance Rule (GAAR) Committee," in Report of Proceedings of the Forty-Seventh Tax Conference, 1995 Conference Report (Toronto: Canadian Tax Foundation, 1996), 54:1-9; and Patrick Boyle, Sharon Gulliver, Jerry Lalonde, Anne-Marie Lйvesque, and Paul Lynch, "The GAAR Committee: Myth and Reality," in Report of Proceedings of the Fifty-Fourth Tax Conference, 2002 Conference Report (Toronto: Canadian Tax Foundation, 2003), 10:1-20. Approval of the GAAR committee for the application of the GAAR as a basis of assessment in a particular case is not required if the committee has previously approved the application of the GAAR in a similar situation. 19 Paragraph 245(3)(a). 20 Paragraph 245(3)(b).
494 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
There is no definition of "series of transactions" for purposes of the GAAR or for purposes of the many other provisions of the Act that contain this phrase. However, subsection 248(10) provides that a series of transactions is deemed to include any related transactions carried out in contemplation of the series. In the leading GAAR case, OSFC Holdings Ltd., the Federal Court of Appeal held that the term "series of transactions" means a preordained series, so that at the time that the series commences, there must be "no practical likelihood that the subsequent transaction or transactions will not take place."21 The Court of Appeal based this restrictive interpretation of "series" on the UK case law and the administrative position adopted by the CRA. However, the court held that subsection 248(10) expanded the narrow case law meaning of a series of transactions:
As long as the transaction has some connection with the common law series, it will, if it was completed in contemplation of the common law series, be included in the series by reason of the deeming effect of subsection 248(10). Whether the related transaction is completed in contemplation of the common law series requires an assessment of whether the parties to the transaction knew of the common law series, such that it could be said that they took it into account when deciding to complete the transaction. If so, the transaction can be said to be completed in contemplation of the common law series.22
There are two problems with the case law concerning the series aspect of the GAAR. First, in the Canadian Pacific case,23 the Federal Court of Appeal indicated that the purpose of any particular transaction in a series could not differ from the overall purpose of the series. It quoted with approval the following statement from the Tax Court decision in the case:
The transactions . . . were, when viewed objectively, inextricably linked as elements of a process primarily intended to produce the borrowed capital which the Appellant required for business purposes. The capital was produced and it was so used. No transaction forming part of the series can be viewed as having been arranged for a purpose which differs from the overall purpose of the series.24
According to both courts, all of the steps involved in the weak-currency borrowing had the same purpose as the series as a whole, namely, the borrowing of money for use in the taxpayer's business.25 Neither the Tax Court nor the Federal
21 OSFC Holdings Ltd., supra note 4, at paragraph 24. 22 Ibid., at paragraph 36. 23 Canadian Pacific Ltd., supra note 4. 24 Canadian Pacific Ltd., supra note 7, at paragraph 15, quoted in the FCA decision in Canadian Pacific Ltd., supra note 4, at paragraph 27. 25 In the Canadian Pacific case, the series of transactions constituting the so-called Aussie-yen loan involved the following steps: the initial borrowing in Australian dollars; the swap of Australian
the long, slow, steady demise of the general anti-avoidance rule 495 Court of Appeal provided any analysis to support this interpretation of paragraph 245(3)(b). It cannot be justified by reference to the wording of the definition of "avoidance transaction," which clearly distinguishes between the series as a whole and the transactions that make it up. Paragraph 245(3)(b) clearly states that the purpose of each transaction in a series must be judged separately. Moreover, the court's interpretation flies in the face of very clear statements in the technical notes to the GAAR: [W]here a taxpayer, in carrying out a series of transactions, inserts a transaction that is not carried out primarily for bona fide non-tax purposes and the series results in a tax benefit, that tax benefit may be denied under subsection 245(2). . . . Thus, where a series of transactions would result in a tax benefit, that tax benefit will be denied unless the primary objective of each transaction in the series is to achieve some legitimate non-tax purposes. Therefore, in order not to fall within the definition of "avoidance transaction" in subsection 245(3), each step in such a series must be carried out primarily for bona fide non-tax purposes.26 On the facts of the Canadian Pacific case, the primary purpose for swapping Australian dollars for Japanese yen was not to borrow funds for use in the taxpayer's business, the overall commercial purpose of the series. That commercial purpose could have been accomplished more directly and less expensively by borrowing Canadian dollars or by swapping the Australian dollars into Canadian dollars. Instead, the primary purpose was clearly to get larger interest deductions. If the Federal Court of Appeal's approach to the determination of the purpose of transactions forming part of a series in the Canadian Pacific case is generally adopted by the courts, the GAAR will be rendered largely meaningless. It will not apply where a taxpayer inserts into a series of transactions with a legitimate commercial purpose steps that are designed solely to produce tax benefits. It will apply only to solitary transactions or to series of transactions whose primary overall purpose is to obtain a tax benefit. It will not apply to any transaction that is part of a series, such as a financing arrangement or an arm's-length sale of shares, whose overall purpose is a legitimate commercial one. It is not yet clear how this issue will play out in the Federal Court of Appeal. In the cases in which the court has applied the GAAR, the series of transactions have been ones whose primary overall purpose has arguably been to obtain tax benefits. For example, in the Water's Edge27 and OSFC cases, the series of transactions as a whole were primarily tax-driven transactions. dollars for Japanese yen, accompanied by a foreign exchange
forward contract; the swap of Japanese yen for Canadian dollars, accompanied by another foreign exchange forward contract; and the use of the Canadian dollars in the taxpayer's business, as well as the swap of payments, the payments of interest, and the repayment of the principal amount of the loan. 26 Canada, Department of Finance, Explanatory Notes to Legislation Relating to Income Tax (Ottawa: Department of Finance, June 1988), clause 186. 27 Water's Edge, supra note 4.
496 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
The second problem, which is less serious than the first, relates to the Federal Court of Appeal's application of subsection 248(10) in the OSFC case. According to that case, whether a transaction is completed in contemplation of the series depends on whether the party to the related transaction knew of the common law series and took it into account. This approach is backward looking. It focuses on the intention or contemplation of the party to the related transaction, and it assumes that the related transaction occurred after the common law series. This backward-looking approach is inappropriate in certain circumstances. For example, many tax-planning arrangements are structured as a series of transactions, such as that involved in the OSFC case, followed by an arm's-length sale of property. In some cases, the arm'slength purchaser may not know or care about the prior series of transactions carried out by the seller in order to generate tax benefits. The related transaction, which is the arm's-length sale, is necessary to crystallize the tax benefits for the seller. In this type of situation, it should not matter whether the purchaser knew about the prior series or took it into account in purchasing the property. The alternative to the approach adopted by the Federal Court of Appeal in the OSFC case is a forward-looking approach. Under a forward-looking approach, the issue of whether a transaction was completed in contemplation of a series would be assessed from the perspective of the parties at the commencement of the series, not the perspective of the person carrying out the related transaction. A related transaction would be considered to be part of the series if the implementation of the series made sense only if the related transaction were carried out. For example, in the OSFC case, the series of transactions whereby mortgages with accrued losses were transferred to a new partnership made sense only if interests in the partnership were sold.
ANALYSIS OF MISUSE AND ABUSE By virtue of subsection 245(4), the GAAR does not apply to an avoidance transaction unless that transaction may reasonably be considered to result in a misuse of the provisions of the Act or an abuse having regard to the provisions of the Act read as a whole. Subsection 245(4) is obviously a key element of the GAAR. Not all avoidance transactions are intended to be subject to the rule. Subsection 245(4) provides the basis for distinguishing between legitimate tax planning and abusive tax avoidance, and it requires that distinction to be made in terms of the concepts of misuse and abuse. Transactions that are in accordance with the object and spirit or scheme of the Act are not abusive and are not subject to the GAAR. On the other hand, transactions that contravene or frustrate the object and spirit or scheme of the Act are considered to be abusive and subject to the GAAR.28
28 For the purposes of this analysis, I am ignoring the difference between a misuse of the provisions of the Act and an abuse having regard to the provisions of the Act read as a whole. The distinction is analyzed in the case law--in particular, in OSFC Holdings Ltd., supra note 4. In my opinion, it seems likely that the concept of abuse probably includes the concept of the misuse of particular provisions of the Act.
the long, slow, steady demise of the general anti-avoidance rule 497 In the OSFC case, the Federal Court of Appeal held that the application of subsection 245(4) requires a two-stage analytical process: The first stage involves identifying the relevant policy of the provisions or the Act as a whole. The second is the assessment of the facts to determine whether the avoidance transaction constituted a misuse or abuse having regard to the identified policy.29 The court went on to indicate that the relevant policy must be "clear and unambiguous"30 and that it would proceed cautiously, because subsection 245(4) involves overriding the words that Parliament has used on the basis of the policy of the Act. Justice Rothstein, writing for himself and Justice Stone, justified the clear and unambiguous policy test of a misuse and abuse in the following terms: It is also necessary to bear in mind the context in which the misuse and abuse analysis is conducted. The avoidance transaction has complied with the letter of the applicable provisions of the Act. Nonetheless, the tax benefit will be denied if there has been a misuse or abuse. This is not an exercise of trying to divine Parliament's intention by using a purposive analysis where the words used in a statute are ambiguous. Rather, it is an invoking of a policy to override the words Parliament has used. I think, therefore, that to deny a tax benefit where there has been strict compliance with the Act, on the grounds that the avoidance transaction constitutes a misuse or abuse, requires that the relevant policy be clear and unambiguous. The Court will proceed cautiously in carrying out the unusual duty imposed upon it under subsection 245(4). The Court must be confident that although the words used by Parliament allow the avoidance transaction, the policy of relevant provisions or the Act as a whole is sufficiently clear that the Court may safely conclude that the use made of the provision or provisions by the taxpayer constituted a misuse or abuse. In answer to the argument that such an approach will make the GAAR difficult to apply, I would say that where the policy is clear, it will not be difficult to apply. Where the policy is ambiguous, it should be difficult to apply. This is because subsection 245(4) cannot be viewed as an abdication by Parliament of its role as lawmaker in favour of the subjective judgment of the Court or particular judges. In enacting subsection 245(4), Parliament has placed the duty on the Court to ascertain Parliament's policy, as the basis for denying a tax benefit from a transaction that otherwise would meet the requirements of the statute. Where Parliament has not been clear and unambiguous as to its intended policy, the Court cannot make a finding of misuse or abuse, and compliance with the statute must govern.31 29 OSFC Holdings Ltd., supra note 4, at paragraph 67. 30 Ibid., at paragraph 69. 31 Ibid., at paragraphs 69-70. Justice Lйtourneau concurred in the result in the case on the grounds that the transactions constituted both a misuse of subsection 18(13) and an abuse of the Act as a whole. However, he did not endorse the clear and unambiguous policy test for subsection 245(4) enunciated by Justice Rothstein.
498 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
The analytical approach to the determination of a misuse or an abuse adopted by the Federal Court of Appeal in the OSFC case has been accepted and applied by the court in subsequent GAAR cases. It has also been accepted and applied by the Tax Court of Canada. Accordingly, the court's approach deserves close analysis. First, in my view, it is appropriate for the courts to establish a high standard for the application of the GAAR. If a relatively clear statutory scheme
cannot be identified, it is impossible to justify the application of the GAAR to a transaction because it contravenes the statutory scheme. At the same time, it must be recognized that by phrasing the threshold in terms of a clear and unambiguous statutory policy, the Federal Court of Appeal has made it relatively easy for judges to rationalize not applying the GAAR. It may fairly be said that very little in the Income Tax Act is clear and unambiguous. Nevertheless, as long as the courts apply the clear and unambiguous threshold reasonably, it represents a sensible approach that causes any doubt to be resolved in the taxpayer's favour. Second, I agree with the Federal Court of Appeal that the policy, or as I prefer, the statutory scheme, must be derived from the words of the Act and from extrinsic material. Sometimes the statutory scheme will be difficult to identify, and sometimes it will be impossible to identify. This simply underlines the importance of the exercise. Further, occasionally the statutory scheme will be easy to identify. For example, as discussed below, the statutory scheme for the taxation of corporate distributions was readily determined by the Tax Court in the first two GAAR cases involving surplus-stripping transactions.32 The fundamental problem with the approach to the determination of misuse and abuse in the OSFC case lies in the wording used to express the test, namely, the reference to "the policy of relevant provisions or the Act as a whole." In fairness to Justice Rothstein, he made it clear at the outset of his analysis that he was using the term "policy" to refer to several things:
The approach to determine misuse or abuse has been variously described as purposive, object and spirit, scheme or policy. I will refer to these terms collectively as [the] policy of the provisions in question or of the Act read as a whole.33
However, this explanation of the meaning of the term "policy" has been forgotten or ignored by subsequent courts, and the use of "policy" as opposed to "statutory scheme" as the operative term is prejudicial to the application of the GAAR. As Justice Rothstein put it, the misuse and abuse analysis "is an invoking of a policy to override the words Parliament has used."34 With the analysis framed in this light, the GAAR will rarely apply because judges inevitably prefer statutory words over
32 Also, as discussed below in connection with the Hill case, the statutory scheme with respect to the timing of the deduction of simple and compound interest is clear. 33 OSFC Holdings Ltd., supra note 4, at paragraph 66. 34 Ibid., at paragraph 69.
the long, slow, steady demise of the general anti-avoidance rule 499 unexpressed notions of parliamentary policy. An avoidance transaction complies with the wording of the provisions of the Act other than the GAAR; otherwise, there is no need to apply the GAAR. Justice Rothstein's articulation of the test of a misuse or an abuse in terms of policy suggests that the court must find something under, behind, or beyond the words of the Act to justify ignoring those words. The Supreme Court of Canada has quoted with approval on several occasions the following statement from a well-known tax textbook: [I]t would introduce intolerable uncertainty into the Act if clear language in detailed provisions of the Act were to be routinely qualified by unexpressed exceptions derived from a court's view of the object and purpose of those provisions.35 Moreover, in the Shell Canada case, the Supreme Court explained the same idea in terms of statutory interpretation: The Act is a complex statute through which Parliament seeks to balance a myriad of principles. This Court has consistently held that the courts must therefore be cautious before finding within the clear provisions of the Act an unexpressed legislative intention. . . . Finding unexpressed legislative intentions under the guise of purposive interpretation runs the risk of upsetting the balance Parliament has attempted to strike in the Act.36 By setting up an opposition between the words used in the Act and the policy of the Act, the Federal Court of Appeal has inadvertently framed the determination of a misuse or an abuse in a prejudicial fashion in favour of taxpayers. The GAAR, section 245 of the Act, represents words Parliament has used in the same way as does any other provision of the Act. Therefore, it would have been more neutral and fair for the Federal Court of Appeal to refer to the misuse or abuse analysis as "an invoking of some words Parliament has used to override other words Parliament has used" rather than "an invoking of a policy to override the words Parliament has used." Framed in this more neutral way, the misuse and abuse analysis would not be much different from the analysis required whenever two or more provisions of the Act conflict. It would be essentially the same as the analysis required under former subsection 245(1). The prejudicial effect of the way in which the Federal Court of Appeal has framed the test of misuse and abuse can be readily seen by comparing the first two GAAR cases decided by the Tax Court and two more recent cases decided by the Tax 35 Peter W. Hogg, Joanne E. Magee, and Jinyan Li, Principles of Canadian Income Tax Law, 4th ed. (Toronto: Carswell, 2002), 563. This statement also appears in various earlier editions of the book; it was originally quoted by the Supreme Court of Canada in Friesen v. The Queen,  2 CTC 369, at 373, and again in 65302 British Columbia
Ltd. v. The Queen,  1 CTC 57, at paragraph 51. 36 Shell Canada Ltd. v. The Queen,  4 CTC 313, at paragraph 43 (SCC).
500 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
Court on the basis of the Federal Court of Appeal's clear and unambiguous Policy approach
. In the McNichol case, Justice Bonner of the Tax Court, in applying subsection 245(4) to a surplus-stripping scheme, indicated that
[t]he scheme of the Act calls for the treatment of distributions to shareholders of corporate property as income.37
Later, he referred to the legislative scheme
concerning corporate distributions. In RMM Canadian Enterprises Inc., Associate Chief Justice Bowman quoted the preceding excerpt from the McNichol case and added:
[T]he Income Tax Act, read as a whole, envisages that a distribution of corporate surplus to shareholders is to be taxed as a payment of dividends.38
In these first two GAAR cases, the Tax Court identified a clear and unambiguous statutory scheme by putting together paragraph 12(1)( j), sections 15 and 84, and former section 247, and analyzed whether the transactions in question contravened that scheme. In contrast, in the Hill case,39 Justice Miller was unable to identify the relevant policy of the Act with respect to paragraph 20(1)(d). The Hill case involved a circular transaction whereby the taxpayer sought to convert accrued interest into principal so that interest on the accrued and unpaid interest would be deductible when payable rather than when paid. The government argued that the policy of the Act was patently apparent from the words of paragraphs 20(1)(c) and (d). The former allows interest on borrowed money to be deducted when it becomes payable. However, interest on unpaid interest (compound interest) is not interest on borrowed money and therefore is not deductible under paragraph 20(1)(c). Such compound interest is clearly deductible under paragraph 20(1)(d), but only when it is paid. It is hard to imagine a clearer statutory scheme. Justice Miller's search for the relevant policy in accordance with the Federal Court of Appeal's test caused him to lose sight of that scheme. In response to the government's argument that the statutory scheme was apparent from the provisions of paragraphs 20(1)(c) and (d), he wrote:
I can glean no identifiable policy from this argument. It is simply a reiteration of what the Act itself says, that is, simple interest can be deducted on a paid or payable basis and compound interest must be paid to be deductible. That is not an underlying Policy Statement
, that is a summary of the legislation. I was not referred by the Respondent to any materials that would assist me in understanding why the government permitted the deduction of simple interest on a payable basis and only permits the deduction of
37 McNichol, supra note 3, at paragraph 24. 38 RMM Canadian Enterprises Inc., supra note 4, at paragraph 53. 39 Hill, supra note 4.
the long, slow, steady demise of the general anti-avoidance rule 501 compound interest on a paid basis. What is the policy? It is not my role to speculate; it is the Respondent's role to explain to me the clear and unambiguous policy. He has not done so.40 With respect, Justice Miller has misapplied the Federal Court of Appeal's test of misuse and abuse. As discussed earlier, the Court of Appeal in the OSFC case used the term "policy" to refer collectively to the purpose of the legislation, the object and spirit of the legislation, and the statutory scheme. In contrast, Justice Miller is searching for the "underlying policy," the policy underlying the statutory scheme. In effect, he is requiring the government to show not only the statutory scheme of the Act, but also the tax policy justification for that scheme. With respect, the policy underlying the statutory scheme is irrelevant to the courts. It is no concern of the courts why Parliament chose to distinguish between the timing of the deduction of simple interest and that of compound interest. The role of the courts is to interpret legislation in accordance with the intention of Parliament. Where the intention of Parliament is clear, as it is in the case of the timing of the deduction of compound interest, it is not the courts' role to inquire into the justification for Parliament's actions. Justice Miller's approach to the determination of a clear and unambiguous underlying policy for purposes of the application of the GAAR puts the government at a huge disadvantage. It will often be difficult for the government to show a clear and unambiguous statutory scheme. It will rarely, if ever, be possible for the government to establish a clear and unambiguous tax policy justification for a statutory scheme, even one as clear as the timing of the deduction of simple and compound interest. If Justice Miller's approach in the Hill case is correct, the GAAR will seldom, if ever, apply.41 Justice Miller applied the same approach used in Hill in the more recent Canada Trustco case.42 In that case, he suggested that "[f ]raming the policy is critical" and referred to the "clear policy behind" the capital cost
allowance (CCA) rules.43 Justice Miller criticized both parties for not providing him with an explanation of the policy behind the rules: What neither party has addressed in depth is the why behind their policy descriptions; or as Justice Noлl put it in Water's Edge, the "reason for being" for these provisions.44 Justice Miller understood the difficulties involved in ascertaining the policy underlying provisions of the Act, but he felt obliged to engage in such an exercise because of the Federal Court of Appeal decision in the OSFC case: 40 Ibid., at paragraph 62. 41 For this reason, I remain mystified as to why the government chose not to appeal the Hill case to the Federal Court of Appeal. 42 Canada Trustco Mortgage Co., supra note 4 (TCC). 43 Ibid., at paragraph 59. 44 Ibid., at paragraph 61.
502 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
What this analysis highlights is the difficulty and risk in determining tax issues based on policy. Certainly GAAR invites such an approach, and the Federal Court of Appeal has made it clear that the only way to determine if there has been a misuse or abuse is to start with the identification of a clear and unambiguous policy. No clear and unambiguous policy--no application of GAAR. But at what level do we seek policy? And, as previously mentioned, do "policy," "object and spirit" and "intended use" all mean the same thing? Is there a policy behind each particular provision, a policy behind a scheme involving several provisions, a policy behind the Act itself? Is the policy fiscal? Is the policy economic? Is the policy simply a regurgitation of the rules? Does the identification of policy require a deeper delving into the raison d'кtre of those rules? How deep do we dig? The success or failure of the application of GAAR left to the Court's finding of a clear and unambiguous policy inevitably invites uncertainty. That is simply the nature of the GAAR legislation in relying upon such terms as misuse and abuse. As many have stated before, this is tax legislation to be applied with utmost caution as it directs the Court to ascertain the Government's intention and then rely on that ascertainment to override legislation. This is quite a different kettle of fish from the accepted approach to statutory interpretation where policy might be sought to assist in understanding legislation. Under GAAR, policy can displace the legislation.45
In my opinion, Justice Miller is wrong when he says that the GAAR requires the court to ascertain the government's intention and then use that intention to override the legislation. As the series of questions posed by Justice Miller accurately indicates, the type of policy analysis he envisages involves so much speculation as to be virtually impossible. It is certainly not a workable basis for the application of the GAAR. Parliament did not intend the courts to engage in the type of speculative tax policy analysis that academics engage in. It is more reasonable for the misuse and abuse analysis to focus on the identification of a statutory scheme and then to determine whether the avoidance transaction in question contravenes that statutory scheme. The identification of a statutory scheme does not require the courts to go behind, beneath, or beyond the legislation itself to determine the rationale or justification for the statutory scheme. "Statutory scheme" is synonymous with the purpose of the legislation, the object and spirit of the legislation, and the intention of Parliament. These concepts should be familiar to the courts because they form part of the modern approach to statutory interpretation as mandated by the Supreme Court of Canada.46 As noted above, the Federal Court of Appeal in the OSFC case
45 Ibid., at paragraph 91. 46 Stubart Investments Ltd., supra note 2, at 316, per Justice Estey: While not directing his observations exclusively to taxing statutes, the learned author of Construction of Statutes, 2nd ed. (1983), at 87, E.A. Driedger, put the modern rule succinctly: Today there is only one principle or approach, namely, the words of an Act are to be read in their entire context and in their grammatical and ordinary sense harmoniously with the scheme of the Act, the object of the Act, and the intention of Parliament.
the long, slow, steady demise of the general anti-avoidance rule 503 recognized the equivalence of these terms but unfortunately settled on the term "policy" as a shorthand way of referring to them. It is noteworthy that the Federal Court of Appeal affirmed the Tax Court's decision in the Canada Trustco case summarily from the bench.47 Unfortunately, the Federal Court of Appeal reiterated that the misuse and abuse analysis for purposes of the GAAR involves a search for the underlying policy. The court said: [C]ounsel was unable to refer to any source that satisfied us that there is a clear and unambiguous policy underlying paragraph 20(1)(a), or the CCA scheme when read as a whole, that renders it a misuse or an abuse of those provisions for the taxpayer to claim CCA in this case.48 The technical notes to section 245 provide support for the argument that subsection 245(4) should be applied by reference to the statutory scheme of the Act rather than its underlying policy. With respect to subsection 245(4), the technical notes refer to "the object and spirit of the provisions of the Act read as a whole," "the overall purpose" of the provisions of the Act, and "the general scheme of the Act."49 The technical notes give an example of income-splitting transactions that might violate the statutory scheme of the attribution provisions of the Act. The notes also refer to the scheme of the Act with respect to estate-freezing transactions, and to the scheme of the Act and "the expressed object and spirit" of the corporate loss limitation rules. A 1988 article by David Dodge
, who was at the time senior assistant Deputy Minister
of finance, provides additional support for the view that the misuse and abuse analysis for purposes of subsection 245(4) should be made with reference to the scheme of the Act: [T]he words "misuse" and "abuse" are intended to have an objective rather than a subjective meaning. These words are meant to exclude transactions that do not involve a use of the Act that is contrary to its general scheme. . . . Therefore, the application of subsection 245(4) should involve an analysis of the object and spirit of the provisions of the Act read as a whole in the context of each particular case.50 This article was used by the Federal Court of Appeal in the OSFC case as authority for the court's interpretation of the term "series of transactions." It should be equally persuasive with respect to the proper interpretation of subsection 245(4). 47 The Queen v. Canada Trustco, supra note 4 (FCA). 48 Ibid., at paragraph 3. 49 Supra note 26, at clause 186. 50 David A.
Dodge, "A New and More Coherent Approach to Tax Avoidance" (1988) vol. 36, no. 1 Canadian Tax Journal 1-22, at 20-21.
504 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
ECONOMIC REALITIES, ECONOMIC SUBSTANCE, AND THE GAAR The Supreme Court of Canada has consistently confirmed the fundamental principle of the Duke of Westminster case51 that a taxpayer is entitled to arrange his or her affairs so as to minimize the amount of tax payable. The Duke of Westminster principle is based on adherence to the legal form of transactions. The provisions of the Act are applied to the actual legal rights
and obligations (the legal substance) created by the transactions, and not to the economic substance of the taxpayer's position. In short, there is no doctrine of economic substance over form in Canadian tax law as there is in US tax law.52 It follows from adherence to the legal form of transactions that Canadian tax law does not permit the recharacterization of a taxpayer's transactions for the purposes of determining the tax consequences of those transactions. The Supreme Court has been emphatic about its rejection of any role for economic realities in interpreting and applying tax legislation.53 In the Shell Canada case, the Supreme Court said: This Court has repeatedly held that courts must be sensitive to the economic realities of a particular transaction, rather than being bound to what first appears to be its legal form. . . . But there are at least two caveats to this rule. First, this Court has never held that the economic realities of a situation can be used to recharacterize a taxpayer's bona fide legal relationships. To the contrary, we have held that, absent a specific provision of the Act to the contrary or a finding that they are a sham, the taxpayer's legal relationships must be respected in tax cases. Recharacterization is only permissible when the label attached by the taxpayer to the particular transaction does not properly reflect its actual legal effect. Second, it is well established in this Court's tax jurisprudence that a searching inquiry for either the "economic realities" of a particular transaction or the general object and spirit of the provision at issue can never supplant a court's duty to apply an unambiguous provision of the Act to a taxpayer's transaction.54 As I have argued previously, the Shell Canada case involves "a total rejection of any role for economic or commercial realities"55 in applying tax legislation (other than the GAAR). If economic realities cannot be used to recharacterize legal relationships
51 Inland Revenue Commissioners v. Westminster (Duke),  AC 1 (HL). 52 See Brian A. Felesky and Sandra E. Jack, "Is There Substance to `Substance over Form' in Canada?" in Report of Proceedings of the Forty-Fourth Tax Conference, 1992 Conference Report (Toronto: Canadian Tax Foundation, 1993), 50:1-63; and Guy Masson and Shawn D. Porter, "Canada," in International Fiscal Association, Form and Substance in Tax Law, Cahiers de droit fiscal international, vol. 87a (The Hague: Kluwer Law International
, 2002), 187-210, at 188. 53 See generally Arnold, supra note 15. 54 Shell Canada Ltd., supra note 36, at paragraphs 39-40. 55 Arnold, supra note 15, at 21.
the long, slow, steady demise of the general anti-avoidance rule 505 except where the label attached to the relationship is inappropriate (that is, the transaction is a sham), then the legal form of the transaction governs and it is difficult to see any viable role for economic realities. This result is not surprising because the doctrine of economic substance over form is fundamentally inconsistent with the Duke of Westminster principle. In the Singleton case in 2001,56 Justices LeBel and Bastarache argued in dissent that the Shell Canada case did not amount to a complete rejection of the economic realities doctrine. In their view, the economic realities of a situation must be considered in order to determine whether the legal relationships are bona fide. Nevertheless, a clear majority of the Supreme Court remains firmly committed to a severely constrained role for economic substance. It may be relevant where the statutory provision at issue is not clear and unambiguous. Further, it may apply where the statutory provision at issue requires recourse to economic realities. For example, subsection 16(1), which requires the bifurcation of blended amounts payable under a contract or other arrangement into income and capital elements, applies "irrespective of . . . the form or legal effect" of the contract or arrangement.57 Economic realities may also be relevant under former subsection 245(1). Although the Supreme Court in the Shell Canada case refused to apply former subsection 245(1) to a weak-currency borrowing, the Federal Court of Appeal has recently struck down a blatant loss utilization scheme using that provision.58 Although the Federal Court of Appeal did not mention the economic realities doctrine specifically, it held that in order to determine whether an expense artificially reduces income, it is necessary to look at all of the relevant circumstances
. According to the court, because the series of transactions was "pre-ordained, circular and limited in time,"59 the deduction generated by the series was artificial. Former subsection 245(1) was repealed when the GAAR was adopted. However, the Federal Court of Appeal's approach to the application of former subsection 245(1) may provide some clues as to the court's application of the GAAR. The issue is whether economic substance or economic realities are relevant under the GAAR notwithstanding that the Supreme Court has consistently held that they are not generally relevant for income tax purposes. More precisely, the issue is whether economic realities may be taken into account by a court in determining whether an avoidance transaction constitutes an abuse under subsection 245(4).60 The technical 56 The Queen v. Singleton,  1 CTC 121 (SCC). 57 Similarly, former subsection 245(2) applied to transactions whereby a person conferred a benefit indirectly on another person "notwithstanding the legal form or effect of the transactions." Former subsection 245(2) became section 246 when the GAAR was introduced. At that time, it was revised to exclude the words "notwithstanding the legal form or effect of the transactions." 58 722540 Ontario Inc. [Novopharm Limited] v. The Queen,  3 CTC 1 (FCA). 59 Ibid., at paragraph 44. 60 There is also an issue about whether economic realities can be taken into account in determining the purpose of an avoidance transaction under subsection 245(3). In my view, the determination
506 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
notes to the GAAR provide some very useful commentary on this issue. The technical notes recognize that in determining whether a transaction is an avoidance transaction, the transaction cannot be recharacterized. In other words, the GAAR applies to what the taxpayer actually did, the legal form of the taxpayer's transactions, and not some other transaction that a court may think is economically equivalent. It is worthwhile to quote the technical notes in this regard:
Subsection 245(3) does not permit the "recharacterization" of a transaction for the purposes of determining whether or not it is an avoidance transaction. In other words, it does not permit a transaction to be considered to be an avoidance transaction because some alternative transaction that might have achieved an equivalent result would have resulted in higher taxes. It is recognized that tax planning--arranging one's affairs so as to attract the least amount of tax--is a legitimate and accepted part of Canadian tax law. If a taxpayer selects a transaction that minimizes his tax liability and this transaction is not carried out primarily to obtain the tax benefit, he should not be taxed as if he had engaged in other transactions that would have resulted in higher taxes.61
Thus, the technical notes recognize that Canadian tax law is based on the legal form of transactions and that the recharacterization of transactions in accordance with their economic substance is not permitted. It must be emphasized, however, that these comments in the technical notes are directed at the definition of an avoidance transaction in subsection 245(3). With respect to the concepts of misuse and abuse under subsection 245(4), the technical notes indicate that the economic substance of transactions is clearly relevant:
Subsection 245(4) recognizes that the provisions of the Act are intended to apply to transactions with real economic substance, not to transactions intended to exploit, misuse or frustrate the Act to avoid tax.62
The view expressed in the technical notes is supported by the wording of subsection 245(4). Subsection 245(4) provides an exemption from subsection 245(2), the charging provision, "where it may reasonably be considered that the transaction would not result directly or indirectly in a misuse of the provisions of this Act or an abuse having regard to the provisions of this Act, other than this section, read as a whole." It is the result of the transaction that must be assessed in terms of the
of a taxpayer's purpose requires an examination of all of the relevant circumstances, including the economic realities of the situation. However, this issue is the same as the determination of purpose under the interest deduction in paragraph 20(1)(c). As the decisions in Shell, supra note 36, Ludco (sub nom. Ludco v. MNR,  1 CTC 95 (SCC)), and Singleton, supra note 56, indicate, the Supreme Court has clearly rejected the consideration of economic realities for that purpose. 61 Supra note 26, at clause 186. 62 Ibid.
the long, slow, steady demise of the general anti-avoidance rule 507 scheme of the Act. This suggests that all the consequences of the transactions-- legal, financial, commercial, and economic--should be considered. A broader analysis of the GAAR in its historical context
, and in the context of the jurisprudence relating to tax avoidance generally and the economic substance doctrine in particular, confirms that economic realities must be relevant under subsection 245(4) if the GAAR is to be effective in preventing abusive tax avoidance. The GAAR applies as the provision of last resort only to transactions that have otherwise complied with all of the other relevant provisions of the Act. With a few exceptions, as noted above, the other provisions of the Act apply to the legal form of the transactions in question; economic substance is irrelevant in applying the provisions of the Act other than the GAAR. Accordingly, it is difficult to see how transactions could be considered to be abusive if the economic substance of what the taxpayer did cannot be considered. Although the early GAAR cases seem to recognize this fundamental point, more recent GAAR decisions have confused the prohibition against the recharacterization of transactions for the purposes of applying the GAAR and the consideration of economic substance for purposes of determining whether there has been a misuse or an abuse of the Act. The analytical approaches applied in the first two GAAR cases, McNichol and RMM Canadian Enterprises, and the more recent Canadian Pacific and Canada Trustco cases are compared below. In the McNichol and RMM Canadian Enterprises cases, the Tax Court considered the economic substance of surplus-stripping transactions in concluding that the transactions abused the statutory scheme for taxing corporate distributions. In McNichol, Justice Bonner wrote: The transaction in issue which was designed to effect, in everything but form, a distribution of Bec's surplus results in a misuse of sections 38 and 110.6 and an abuse of the provisions of the Act, read as a whole, which contemplate that distributions of corporate property to shareholders are to be treated as income in the hands of the shareholders. It is evident from section 245 as a whole and paragraph 245(5)(c) in particular that the section is intended inter alia to counteract transactions which do violence to the Act by taking advantage of a divergence between the effect of the transaction, viewed realistically, and what, having regard only to the legal form, appears to be the effect. For purposes of section 245, the characterization of a transaction cannot be taken to rest on form alone.63 Similarly, in RMM Canadian Enterprises, Justice Bowman quoted with approval the preceding excerpt from Justice Bonner's decision in McNichol and said: A form of transaction that is otherwise devoid of any commercial objective, and that has as its real purpose the extraction of corporate surplus and the avoidance of the ordinary consequences of such a distribution, is an abuse of the Act as a whole.64 63 McNichol, supra note 3, at paragraph 25. 64 RMM Canadian Enterprises Inc., supra note 4, at paragraph 53.
508 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
In both these cases, the Tax Court respected the legal form of the transactions and did not recharacterize them. The legal form of the transactions was a sale of shares. The Tax Court accepted the legal form, even though the sale was made to an accommodation party, for the purposes of determining whether there was an avoidance transaction. For purposes of determining whether there was a misuse or an abuse under subsection 245(4), however, the Tax Court realized that it was necessary to go beyond the legal form of the transactions and consider their economic effects. As discussed above, the Canadian Pacific case involved the application of the GAAR to a weak-currency borrowing. The government's argument with respect to the misuse and abuse analysis was that, in economic substance, part of the interest payments made in Australian dollars represented repayments of the principal amount of the loan. In economic terms, a weak-currency borrowing is simply a synthetic borrowing in the ultimate currency. For example, the Aussie-yen loan in Canadian Pacific was, in reality and in economic substance, a borrowing of Canadian dollars.65 Further, the government argued that there was a clear statutory scheme to the effect that interest payments are deductible and repayments of principal are not. There was an abuse of this statutory scheme because the taxpayer's transactions had the economic effect of converting non-deductible principal payments into deductible interest payments. The government's argument is consistent with the approach adopted by the Tax Court in both the McNichol and the RMM Canadian Enterprises cases. Nevertheless, the Federal Court of Appeal rejected the government's argument "because it depends on recharacterizing the interest payments as capital payments."66 The Court of Appeal referred to the terms of the debenture agreement as authority for its conclusion that the payments of interest in Australian dollars were in fact interest and not principal. According to the court,
[t]here is no authority which permits the Court to ignore the nature of the relationship between [Canadian Pacific] and the lender and, in effect, to rewrite the terms of their agreement.67
This statement is simply wrong with respect to the issue of a misuse or an abuse under subsection 245(4). The Court of Appeal is correct that the agreement between the parties cannot be recharacterized for the purpose of determining whether or not a transaction is an avoidance transaction under subsection 245(3). Unfortunately, the Court of Appeal confused the analysis of avoidance transaction under
65 If the foreign exchange gain recognized at repayment is amortized over the term of the loan and netted against the interest expense incurred, as required by generally accepted accounting principles
, the net interest expense is equal to the interest expense incurred with respect to a Canadian-dollar borrowing. 66 Canadian Pacific Ltd., supra note 4, at paragraph 34. 67 Ibid., at paragraph 30.
the long, slow, steady demise of the general anti-avoidance rule 509 subsection 245(3) and that of misuse or abuse under subsection 245(4) and, as a result, did not even consider the government's alleged statutory scheme with respect to interest and principal.68 A similar confusion with respect to the prohibition against the recharacterization of transactions for purposes of subsection 245(3) and the relevance of economic substance for purposes of the misuse and abuse analysis under subsection 245(4) occurred in the recent Canada Trustco case.69 Canada Trustco involved a tax-driven sale-leaseback transaction under which Canada Trustco had no economic investment and no risk in the leased assets but claimed CCA in respect of those assets against its other income. With respect to the question of a misuse or an abuse under subsection 245(4), the government argued that the transaction was abusive because the economic result of the transactions was that Canada Trustco claimed CCA even though it had not incurred any real economic cost
in the assets. Justice Miller of the Tax Court rejected the government's argument: To accept the Respondent's position would be to recharacterize the legal form and substance of the transaction, for the purposes of then determining whether there has been a misuse or abuse. The GAAR provisions cannot be applied in that way.70 Justice Miller referred to the Supreme Court's rejection of the economic realities doctrine and concluded that "[i]t is the legal cost which is determinative, not the real economic cost."71 He continued: GAAR is not to be imposed lightly. It should not permit a recharacterization of a transaction to find the transaction is abusive in its recharacterized form. The transaction must be viewed in its legal context and if found abusive, only then recharacterized to determine the real tax consequences. That is how the GAAR provisions are set out: is there a tax benefit, is there a primary purpose other than obtaining that benefit, and does that avoidance transaction result in an abuse or misuse? All those questions require a review of the transaction, which is otherwise acceptable under all other provisions of the Act; that is the legal transaction.72 In my opinion, Justice Miller has confused the prohibition against the recharacterization of transactions for purposes of determining whether they are avoidance transactions under subsection 245(3) and the consideration of economic realities or economic substance for purposes of determining whether there is an abuse of the Act under subsection 245(4). As discussed earlier, it does not follow that, 68 It is unclear to me why the government did not seek leave to appeal to the Supreme Court in the Canadian Pacific case. 69 Canada Trustco Mortgage Co., supra note 4 (TCC). 70 Ibid., at paragraph 69. 71 Ibid., at paragraph 73. 72 Ibid., at paragraph 77.
510 canadian tax journal / revue fiscale canadienne
(2004) vol. 52, no 2
because transactions cannot be recharacterized for purposes of subsection 245(3), the economic substance or results of the transactions cannot be considered under subsection 245(4). If Justice Miller's approach had been applied to the surplus-stripping transactions in the McNichol and RMM Canadian Enterprises cases, the GAAR would not have applied. Ignoring the economic realities in those cases, the transactions were in legal form sales of shares that resulted in capital gains. The scheme of the Act is that sales of shares generally result in capital gains. Considering only the legal form, the surplus-stripping transactions could not be considered abusive because they resulted in taxation in accordance with the statutory scheme. Excluding any consideration of economic substance and economic realities under subsection 245(4) in this way would render the GAAR incapable of ever applying. Such an interpretation of the GAAR is clearly contrary to Parliament's intention in enacting the GAAR.
CONCLUSION Unless the trends in the case law identified in this article are reversed, the GAAR will be rendered largely ineffective. It will apply only in those clear cases of egregious conduct where, arguably at least, the tax benefits of the schemes could be denied simply by applying the provisions of the Act in a reasonable way.73 Exposing the tax system to abusive tax avoidance without any meaningful restrictions is undesirable. The solution rests with Parliament. Section 245 should be amended in a targeted fashion to inform the courts that their interpretation of the existing GAAR is not in accordance with the intention of Parliament. The 2004 budget proposals to amend the GAAR to clarify that it applies to a misuse or an abuse of the Regulations to the Act,74 and to amend the Income Tax Conventions Interpretation Act to provide that tax treaties do not override the GAAR,75 provide clear evidence that the Department of Finance may be persuaded to act in this way in appropriate circumstances. The necessary amendments are not difficult and would not even be very controversial. First, a provision should be added to clarify that, in determining the primary purpose of a transaction for purposes of paragraph 245(3)(b) of the definition of an avoidance transaction, the overall purpose of the series as a whole is
73 See, for example, Water's Edge, supra note 4, at paragraph 48, in which Justice Noлl raised the possibility that the application of the GAAR was unnecessary: "Indeed, the object and spirit of the relevant provisions is so clear that I questioned during the hearing whether the Tax Court Judge properly concluded, and the Minister properly conceded, that the Act when construed without regard to section 245, allowed Klink to deduct the terminal loss." 74 Canada, Department of Finance, 2004 Budget, Budget Plan, Notice of Ways and Means Motion To Amend the Income Tax Act, March 23, 2004, resolution 18. 75 See the Notice of Ways and Means Motion To Amend the Income Tax Conventions Interpretation Act, ibid. I think this is the intention of the notice of ways and means motion, although what it says is quite different.
the long, slow, steady demise of the general anti-avoidance rule 511 irrelevant. Second, subsection 248(10) should be amended to clarify that whether a transaction is "completed in contemplation of the series" must be judged from the perspective of the persons carrying out the series at the time that the series commences. Third, a provision should be added to the GAAR to the effect that an avoidance transaction involves a misuse or an abuse under subsection 245(4) if it contravenes the relevant statutory scheme, not the underlying policy. Admittedly, this provision might be tricky to draft. Finally, a provision should be added to the GAAR to require the courts to consider the economic substance of a transaction or series of transactions in assessing whether an avoidance transaction constitutes a misuse or an abuse under subsection 245(4).76 Amendments along these lines would provide clear signals to the courts as to the proper interpretation of the GAAR. Undoubtedly, the judges would welcome the guidance provided by these amendments. 76 Giving the courts permission to consider economic realities would not be effective, in my opinion, because the courts would just follow the existing Supreme Court jurisprudence to the effect that economic realities are irrelevant.